UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

RBC treatment of a municipal landfill leachate : a pilot scale evaluation Peddie, Craig Cameron 1986

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-UBC_1986_A7 P32.pdf [ 20.75MB ]
JSON: 831-1.0062727.json
JSON-LD: 831-1.0062727-ld.json
RDF/XML (Pretty): 831-1.0062727-rdf.xml
RDF/JSON: 831-1.0062727-rdf.json
Turtle: 831-1.0062727-turtle.txt
N-Triples: 831-1.0062727-rdf-ntriples.txt
Original Record: 831-1.0062727-source.json
Full Text

Full Text

RBC TREATMENT OF A MUNICIPAL LANDFILL LEACHATE: A PILOT SCALE EVALUATION by CRAIG CAMERON PEDDIE A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DECREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Civil Engineering We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 14, 1986 ® Craig Cameron Peddie, 1986 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the THE UNIVERSITY OF BRITISH COLUMBIA, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Civil Engineering THE UNIVERSITY OF BRITISH COLUMBIA 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: October 14, 1986 ABSTRACT This study evaluated the on-site treatment of a moderately low strength municipal landfill leachate with a Rotating Biological Contactor (RBC), at pilot scale (0.9 m dia.). The leachate generally had COD and NH^-N concentrations of less than 1000 mg/L and 50 mg/L respectively. A high treatment efficiency for both carbon removal and nitrification was achieved despite variable and intermittent loading conditions. The effluent filtrable BOD^ was generally less than 10 mg/L and the effluent NH^-N concentration was usually less than 1.0 mg/L. This effluent quality was achieved at mass loading levels comparable to those for sewage treatment (10.0 g B O D 5 / m 2 * d for carbon removal and 0.8 g NH 3 -N/m 2 *d for nitrification). The results demonstrated that long hydraulic retention times (HRT >4 hrs.) can offset the effects of lower temperatures. Nitrification efficiency in particular was shown to be HRT dependent. Limited heavy metal data indicated that heavy metals were removed at efficiencies and relative affinities comparable to those observed in activated sludge studies. An aside to this study showed that trace organics, some of which are on the EPA priority pollutant list, were present in this leachate and were effectively removed during passage through the RBC. Keywords: Leachate treatment, Rotating Biological Contactor (RBC), carbon removal, nitrification, loading rates, hydraulic retention time (HRT) effects, heavy metal removal, priority pollutants. ii Table of Contents ABSTRACT ii ACKNOWLEDGEMENTS x 1. INTRODUCTION 1 2. SITE DESCRIPTION 3 3. RATIONALE 7 3.1 Purpose 7 3.2 Literature Review - Leachate Treatment 7 3.3 RBC Treatment 23 4. EXPERIMENTAL PROGRAM 45 4.1 Sampling And Analysis Program 47 4.2 Sampling Procedures 52 4.3 Analytical Procedures 54 5. LEACHATE QUALITY 58 5.1 Leachate Generation 58 5.2 Affect of Water Inputs on Leachate Quality 65 5.3 Premier Landfill Leachate 67 5.4 Organics 85 5.5 VFA's 89 5.6 Nitrogen 94 5.7 Total Solids and Specific Conductance 95 5.8 Metals 98 5.9 Specific Trace Organics 104 6. PILOT PLANT 106 7. RBC OPERATION 112 7.1 Start-Up 112 7.2 The Disruptions 116 iii 7.3 A New Beginning 119 8. TREATMENT RESULTS 131 8.1 Carbon Removal 132 8.2 Nitrification 140 8.3 Suspended Solids 147 8.4 Metals 147 8.5 Specific Trace Organics 150 9. DISCUSSION 156 9.1 Organic Removal 156 9.2 Nitrification 170 9.3 RBC Response to Variable and Intermittent Loading 183 9.4 Metals and Trace Organics 185 9.5 Toxicity 186 9.6 Implications for Full Scale Treatment 186 9.7 Experimental Program and RBC Operation 190 10. SUMMARY 193 11. CONCLUSIONS 195 12. RECOMMENDATIONS FOR FURTHER RESEARCH 197 13. REFERENCES 199 14. APPENDIX 1 206 15. APPENDIX 2 214 16. APPENDIX 3 220 iv List of Tables 3.1 Summary of Leachate Treatment Studies 12 4.1 Sampling Program 48 5.1 Variability of Leachate Composition 59 5.2 Premier Leachate Characteristics (Well #1) 69 5.3 Leachate Heavy Metal Levels (AA) 102 5.4 Leachate Metal Analyses (ICP) 103 5.5 Leachate Trace Organic Content 105 6.1 RBC Specifications 107 8.1 RBC Heavy Metal Removal (AA) 151 8.2 RBC Metal Removal (ICP) 152 8.3 RBC Biomass Metal Levels (AA) 153 8.4 RBC Biomass Metal Levels (ICP) 154 8.5 RBC Trace Organic Removal 155 9.1 Design Loadings for RBC Treatment of a Municipal Wastewater 188 v List of Figures 2.1 Photo of New Section of Premier Landfill looking North-West (June 1983) 3 2.2 Premier Street Landfill - North Vancouver, B.C. Site Plan and Location 5 2.3 Photo of New Section of Premier Landfill from Top of Old Landfill (June 1985) 6 5.1A Premier Leachate Characteristics vs. Time 82/83 70 5.IB Premier Leachate Characteristics vs. Time 83/84 71 5.1C Premier Leachate Characteristics vs. Time 84/85 72 5.2 Leachate Flow and Constituent Mass Release Premier Street Landfill 75 5.3A Leachate Carbon Content vs. Time 82/83 76 5.3B Leachate Carbon Content vs. Time 83/84 77 5.3C Leachate Carbon Content vs. Time 84/85 78 5.4A Leachate Nitrogen Content vs. Time 82/83 79 5.4B Leachate Nitrogen Content vs. Time 83/84 80 5.4C Leachate Nitrogen Content vs. Time 84/85 81 5.5A Leachate Total Solids and Sp. Conductance vs. Time 82/83 82 5.5B Leachate Total Solids and Sp. Conductance vs. Time 83/84 83 5.5C Leachate Total Solids and Sp. Conductance vs. Time 84/85 84 vi 5.6 TOC vs. COD 86 5.7 B O D 5 vs. COD 87 5.8 B O D 5 vs. BODr/COD Ratio 88 5.9 Leachate VFA conc'n vs. Time 90 5.10 COD(vfa) vs. Leachate COD and B O D 5 91 5.11 COD vs. VFA 92 5.12 B O D 5 vs. VFA 93 5.13 N H 3 vs. COD 96 5.14 N H 3 vs. Sp. Cond. 97 5.15 Tot. Solids vs. Sp. Cond. 99 5.16 COD vs. Sp. Cond. 100 5.17 B O D 5 vs. Sp. Cond. 101 6.1 Photo of RBC Prior to Start-up, Showing Disk Media and Influent Pump 106 6.2 Photo of RBC Installed Adjacent to the North Leachate Lift Station 107 6.3 Section of North Leachate Lift Station Showing RBC Connections 108 6.4 Photo of RBC Pump Inlet Screen 109 7.1 Photo of creamy, taupe coloured, initial bacterial growth (June 1983) 113 7.2 Photo of mature biomass growth during start-up (late June 1983) 114 vii 7.3 Photo of Pump tubing failure 115 7.4 Photo of Aftermath of 1 s t Flood in the RBC (November 1983) 117 7.5 Photo of RBC after being raised 1m to avoid flooding 118 7.6A RBC Operational History: Influent Flowrate and Loading 82/83 120 7.6B RBC Operational History: Influent Flowrate and Loading 83/84 121 7.6C RBC Operational History: Influent Flowrate and Loading 84/85 122 7.7 Photo of heavy dark growth on RBC during April-May 1984 123 7.8 Photo of single bellows leachate pump (165 rpm) and nutrient pump 126 7.9 Photo of twin bellows leachate pump (50 rpm) and nutrient pump 127 7.10 Photo of healthy bacterial growth 129 7.11 Photo showing average biomass thickness across the RBC 129 7.12 Photo showing leachate foaming in RBC first stage 130 8.1 RBC Effluent B O D 5 vs. Loading and Time 133 8.2 RBC Effluent B O D 5 vs. Loading Rate 135 8.3 1 s t and 4 t h Stage B O D 5 (settled) vs. Time 136 8.4 1 s t and 4 t h Stage B O D 5 (filtered) vs. Time 137 8.5 RBC Effluent C O D vs. Loading Rate 138 8.6 RBC Effluent C O D vs. Loading and Time 139 8.7A RBC Nitrification Performance - part 1 142 viii 8.7B RBC Nitrification Performance - part 2 143 8.8 Effluent NH^ and N 0 3 vs. Loading Rate 144 8.9A 1 s t and 4 t h Stage N H 3 and N 0 3 - part 1 145 8.9B 1 s t and 4 t h Stage NH^ and N 0 3 - part 2 146 8.10 1 s t and 4 t h Stage Suspended Solids 148 8.11 RBC Effluent Suspended Solids 149 9.1 B O D 5 Removal versus Loading Rate 159 9.2 COD Removal versus Loading Rate 160 9.3 B O D 5 Percent Removal versus Loading Rate 163 9.4 BODjj Removal versus Loading Rate Corrected for Temperature 164 9.5 B O D 5 Removal - Monod Kinetics Approach 168 9.6 N H 3 -N Removal versus Loading Rate 173 9.7 N H 3 -N Percent Removal versus Loading Rate 174 9.8 N H 3 -N Percent Removal versus Temperature 175 9.9 N H 3 -N Removal versus Loading Rate Corrected for Temperature 177 9.10 N H 3 -N Removal versus Hydraulic Retention Time (HRT) 178 9.11 NH-. -N Removal - Monod Kinetics Approach 180 ix ACKNOWLEDGEMENTS I dedicate this thesis, and give my greatest love and thanks, to my wife Joan, who patiently supported me in every way during the protracted process of producing it. I would also like to thank my parents and family, for their support and encouragement throughout my education. I also extend a special thanks to Professor Jim Atwater, for his guidance and support in overseeing this study. The greatest benefit which I received as a result of the extended duration of this study was that those people who might otherwise have been remembered as staff or aquaintances, have become valued friends. With this in mind, I would like to extend my sincere thanks to Sue Jasper, Environmental Lab Manager, Susan Liptak and Paula Parkinson, Research Technicians, and Tim Ma, GC/MS Technician, for their expert help and advice. During the course of this study, 1 also made considerable use of the facilities and expertise within the Civil Engineering Workshop. Again, with the above in mind, 1 would like to thank Dick Postgate, Head Technician, and his staff, especially Art Brooks, and Guy Kirsch. I would also like to thank the members of the Environmental Engineering faculty, Dr. Oldham, Dr. Mavinic, Dr. Hall, and Prof. Atwater, for their help and tutelage. Last, but not least, I would like to thank the many other members of the staff, and fellow students, whose help and interaction have been very valuable, especially, Fred Koch, Ann Davern, Kelly Lamb, Carolyn Foo, Bruce Anderson, Yves Comeau, Troy Vassos, Ken Johnson, Mano Ramarathan, Chris Town, and Paula Wentzell. I would also like to acknowledge the generous assistance and cooperation of the District of North Vancouver, the staff of the Premier Street Landfill, and CMS Equipment Ltd. of Mississauga, Ontario. Funding was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC). 1. INTRODUCTION Landfill leachates are wastewaters formed when water migrates through emplaced solid wastes and carries off or leaches, soluble matter, decomposition products, and fine solids. This water or leachate emerges from the solid wastes laden with organic and inorganic compounds, heavy metals, etc., with the potential to pollute the surface and groundwater environment unless control measures are taken. Efforts to control or prevent pollution caused by leachates have come relatively recently. Landfill leachates have only been acknowledged as environmentally significant for about the last 20 years and awareness at the regulatory and local operational levels has occured mainly over the last 10 years. Prior to this awareness, landfills were commonly established on cheap land, such as peat bogs, ravines, and abandoned gravel pits, with no particular site preparation. Today many of these sites are still in use. In retrospect, the choice of these types of sites to minimize disposal costs and to reclaim land was unfortunate, as their hydrological characteristics have often exacerbated the leachate problems at these landfills, necessitating expensive remedial action. Dykes, groundwater barriers, and leachate collection systems, have been installed at many of these landfills to prevent any further escape of leachate into the environment. Across North America large volumes of leachate are being collected from existing landfill sites. Many of these existing landfills are nearing capacity and so in several jurisdictions (such as Vancouver), an urgent evaluation of municipal waste disposal options is underway. Regardless of the types of solid waste handling and treatment methods employed, a residual component of the solid wastes remains for ultimate disposal, likely by landfilling. In addition, land disposal remains very competitive economically with other solid waste handling methods in many areas. 1 2 Thus landfilling will be an important component of most solid waste disposal schemes well into the foreseeable future. However, the recently developed public awareness of the pollution potential of landfills has made it very difficult to establish new landfill sites. One benefit of this awareness is the public insistance that new landfill sites include detailed measures to prevent pollution of the environment from leachate, either by in-situ attenuation in underlying soil layers, or by containment, collection, treatment, and disposal. The current trend in North America is towards full collection and treatment of landfill leachates before they are discharged into the environment. Therefore, increasing volumes of leachate can be expected from new landfills, carefully engineered to contain all the leachates they produce. Once these leachates have been collected, efficient methods of treatment and/or disposal must be found. As landfill leachate characteristics and site conditions (such as climate, proximity to sewers, etc.) are highly site specific, a variety of effective treatment and disposal schemes must be developed for these wastes. The Environmental Engineering Croup at the University of British Columbia has had, and continues to have, an extensive research interest in landfill leachate issues. Numerous studies have been conducted on leachate generation, characterization, toxicity, treatment methods, and treatment parameters. To further this effort, this study was initiated to evaluate the performance of a Rotating Biological Contactor (RBC) treating a relatively weak municipal landfill leachate. The leachate that was treated generally had a chemical oxygen demand (COD) of less than 1000 mg/L and a total Kjeldahl nitrogen (TKN) value of less than 50 mg/L. This study evaluated the capacity of the RBC to remove both the carbonaceous and nitrogenous oxygen demanding material from this leachate at pilot scale and under field conditions. 2. SITE DESCRIPTION The Premier Street landfill is situated on a natural bench on the east bank of the lower Lynn Creek ravine in the District of North Vancouver. This bench lies within a steep-sided bowl, immediately downstream of the lower boundary of Lynn Canyon Park. The bench is composed of fluvial sand and gravel up to approx. 6 m in depth, underlain by a dense glacial till which is contiguous with the walls of the bowl. The District of North Vancouver began using the property for a landfill in 1959, and by agreement began accepting wastes from the District of West Vancouver, and the City of North Vancouver, in 1969 and 1970 respectively. In 1981 the then active landfilling area of the property was nearing capacity and the decision was made to develop the final 10.5 ha section of the property for use (see Figures 2.1,2.2,2.3). Figure 2.1 Photo of New Section of Premier Landfill look ing North-West (June 1983) 4 -PREMIER STREEJJ y LANDFILI Burrard"""ir" Inlet V a n c o u v e r N.Vancouver S C A L E = 100 200m Figure 2.2 Premier Street Landfill • North Vancouver, B.C. Site Plan and Location 5 Figure 2.3 PhoIo _ _ bT^ew^eclion of Premier Landfill from Top of Old Landfill (June 1985) 6 This section is bounded by Lynn Creek to the north-west, the steep walls of the bowl to the east, and by a mountain of garbage, which was the previous active landfill area, to the south-west. A bentonite slurry trench and dyke were constructed along Lynn Creek, paralled by a perforated leachate collection pipe. An extension of the slurry trench runs along the base of the previously filled area to isolate the old and new sections. The slurry trench and underlying glacial till combine to form a relatively impermeable dish beneath the site to contain the leachate produced. A ditch at the base of the bowl slopes diverts storm water around the site into the creek. The leachate collection pipe terminates in the north lift station which transfers the leachate via another pipe to the municipal sewer system. These preparations were completed in January 1982 and this new site began receiving wastes shortly thereafter. Up till now the landfilling activity has been restricted to an area adjacent to the old site, which is about 40% of the 10.5 ha available. Since drainage from the whole section is recovered by the leachate collection system, rainwater from the unfilled area has a diluent effect on the characteristics of the leachate received in the pumpwell. Although the site is at a low elevation (<70 m), its postion at the base of the North Shore mountains attracts a relatively heavy annual rainfall of approximately 2000 mm. 3. RATIONALE 3.1 PURPOSE The purpose of this study was to evaluate the suitability and effectiveness of a Rotating Biological Contactor (RBC) for the treatment of a municipal landfill leachate. This evaluation was conducted at pilot scale (0.9 m dia.), and under field conditions at a landfill site in order to produce data which would closely approximate full scale expectations. The suitability of the RBC for the treatment of this leachate was evaluated primarily on the basis of three determinations: the capacity of the RBC to remove the carbonaceous component of the leachate; whether or not the leachate could be nitrified; and if so, the capacity of the RBC for nitrification. These capacities, for carbon removal and nitrification, would be defined as the maximum mass loading rate (g/m^*d of BOD^ or NH^) for which complete treatment was maintained (effluent BOD^s25 mg/L and/or NH^si .O mg/L). Loading rates for leachate treatment could then be compared to those established for domestic sewage treatment. 3.2 LITERATURE REVIEW - LEACHATE TREATMENT A search of the literature in early 1983, prior to the start of experiments, failed to find any references concerning RBC, or other aerobic fixed growth process, treatment of landfill leachates. In the absence of directly comparable results, some literature concerning leachate generation and composition, leachate treatment (primarily by aerobic suspended growth systems), and RBC treatment of other types of wastes, was collected to provide background information for this study. This discussion will focus on the later two topics as leachate quality is discussed adequately in Section 5.0. 7 8 Landfill leachates are a relatively recent topic of environmental concern. Chian and DeWalle (11) attribute one of the first studies to Merz, who investigated the leachate from incinerator ash dumps in 1952 and went on to later study leachates from municipal solid waste landfills. Over the next twenty years the extent of the pollution caused by leachates was documented and the emphasis of research shifted to studying the mechanisms of leachate generation and movement, and measures to control or treat the leachate after it has been produced. With respect to the treatment of leachates, two of the earliest, widely referenced studies are those by Boyle and Ham (5), and Cook and Foree (15), both published in the early 1970s. Boyle and Ham looked at the treatability of landfill leachates by biological processes, both anaerobic and aerobic. In their anaerobic studies, which received the greater emphasis, Boyle and Ham achieved greater than 90 percent removals of both COD and B O D 5 , from influent COD concentrations of 2,240 to 22,400 mg/L. Loading rates ranged from 0.43 to 2.16 kg COD/m^*d, and detention times were 5 - 20 days, at an average temperature of 23° C. Effluent quality was found to improve with decreased loading rates and/or longer detention times. They also found that anaerobic system performance was very temperature dependent, with COD removals dropping from 87.2 % at 23° C, to only 22 % at 10° C. Subsequent studies have repeatedly confirmed the capabilities of anaerobic treatment of landfill leachates. For example Bull et al. (7), realized a 96.8 percent BODjj removal from an influent concentration of 5700 mg/l BOD^ at a detention time of 30 days. Although these two, and many other, studies have demonstrated a high percentage BOD^ removal, the effluent BOD^ values are typically greater than 100 mg/L (13). Effluent ammonia levels are also usually very high because of the minimal nitrogen requirements of anaerobic bacteria and the efficient conversion of organic nitrogen to ammonia. Therefore in most instances, anaerobic treatment can not be regarded as a complete treatment process and further treatment or effluent 9 polishing is required. The aerobic treatment studies conducted by Boyle and Ham were considered less successful. Three fill and draw reactors with a 5 day detention time were relatively heavily loaded (0.3 - 1.4 kg BOD 5 /m 3 *d) with landfill leachate. Effluent BODj- levels ranged from 160 to 1400 mg/L, and the units were plagued by foaming and solids separation problems which increased in severity at the higher loadings. However, the results indicated that for loading levels less than 0.48 kg BOD[-/m3*d and warm temperatures (23° C), that BOD,, removals of greater than 90 percent could be achieved, ln another segment of this study, Boyle and Ham also demonstrated that a landfill leachate (COD = 10,000 mg/L) could be combined with domestic sewage up to 5 percent leachate by volume for co-treatment in an extended aeration process without impairing process performance. The most important result of this study however was the demonstration that biological treatment of landfill leachates was possible. Cook and Foree (15) expanded upon the aerobic biological treatment studies of Boyle and Ham, and also evaluated various physical-chemical treatment processes for landfill leachate treatment. Using fill and draw aerated batch reactors, with a 10 day detention time and loading rates between 1.58 and 7.9 kg COD/m^*d, they were able to achieve BODj- removals in the order of 99.7 percent, from an influent BOD^ of 7100 mg/L, to effluent values of less than 26 mg/L. Nutrient additions of nitrogen (N) and phosphorus (P) did not improve the treatment efficiency significantly despite the nutrient ratio of the leachate (100:2.33:0.23 BODj.:N:P) being far less than the generally accepted 100:5:1 ratio for healthy growth. Solids settleability was observed to be very good but foaming remained problematic requiring the periodic use of a defoaming agent. A theoretical minimum detention time of 5.3 days was calculated from kinetic considerations and confirmed by the failure of a 5 day detention time unit. This result indicated that the 1 0 detention time of the reactors used in the Boyle and Ham experiments were probably too short to achieve stable operation or efficient treatment. The evaluation of physical-chemical treatment involved chemical coagulation followed by activated carbon. Chemical coagulants were effective for suspended solids and colour removal but since the COD of leachate is mainly soluble, COD removal was minimal. Activated carbon proved fairly effective at removing the soluble COD from the leachate, but given the high concentrations of organic material in many leachates, this treatment method would not be economical. As a polishing step for biologically treated effluents, activated carbon proved very effective for residual COD and colour removal. Chian and DeWalle (11) later reviewed the experience with physical-chemical treatment of landfill leachates and came to a similar conclusion, that physical-chemical treatment is best suited for polishing biological treatment effluents, or treating old leachates, which have a low soluble organic content. Reverse osmosis was shown to be the most effective treatment method followed closely by activated carbon. It was also effective for treating raw leachate, except that rapid blinding of the membranes made such an application impractical. The studies by Boyle and Ham, and Cook and Foree, indicated that biological treatment, both anaerobic and aerobic, could effectively remove organic material from relatively strong leachates and that physical-chemical methods were much less effective except for suspended solids removal. Subsequent studies of leachate treatability have expanded upon these initial results and qualified the conditions under which the various treatment methods are applicable. The remainder of this discussion however will focus on aerobic leachate treatment as the RBC is primarily an aerobic treatment process. The articles concerning aerobic biological landfill leachate treatment reviewed for this study were intended to be a representative sampling of previous treatment 11 experience. Since all of the studies involved aerobic suspended growth systems (activated sludge or aerated lagoons), and therefore were not directly comparable to this study, a more comprehensive review was unwarranted. The papers reviewed cover a fairly wide variety of different leachates, treatment conditions, operational problems, and treatment related topics. Table 3.1 summarizes various results and parameters from the treatment studies reviewed. A number of points are readily apparent from this table. Column 1 shows that the C O D and BOD, , values of the leachates used in these studies are, with the exception of Palit and Quasim (56), who used diluted feed, moderate to high in comparison to those of the Premier leachate used in this study. However the B O D ^ / C O D ratios of these leachates are all relatively high (>0.5). As pointed out by Chian and DeWalle (11), the organic strength of a leachate, as well as its biodegradability, as exemplified by its B O D ^ / C O D ratio, reflects the degree of stabilization of the landfill, with high organic strength and degradability being associated with fairly new or young landfills. Therefore all of these treatment studies have dealt with high organic strength leachates from young landfills. The discrepancy between organic strengths, as well as the lack of comparable leachates encountered in the literature, indicates that the Premier leachate is somewhat unique to have such a low total organic strength coming from a young landfill. The reasons for the low strength and other characteristics of the Premier leachate are explored in Section 5.0. Column 2 shows what is perhaps the most significant point, that all the studies demonstrated very efficient carbon removal from different landfill leachates. The results indicate that aerobic treatment processes, operating within limiting conditions, are generally capable of complete treatment to effluent B O D ^ values of less than 25 mg/L, regardless of the initial leachate strength. However effluent C O D levels are usually much higher (100 - 900 mg/L), largely due to refractory humic Table 3.1 Summary of Leachate Treatment Studies (suspended growth) Column No. References #1 #2 #3 Influent Effluent Loading Rate COD mg/L COD % Rem kg / m 3 (BOD mg/L) BOD % Rem (F/M Ratio) #4 #5 #6 #7 Op. Prob. Temp. Foaming Heavy Nitri-Effects or Metal fication Comments [BOD/COD Ratio] tSRT] Settling Removal Robinson & Maris (65,66) (1983,1985) Keenan ct al. (43) (1984) Ehr ig , H . J . (22,24) (1984,1985) 5028 (3035) [0.60] 18488 12468 [0.67] BOD<20 (0.21 or less) @ SRT>10 d |>I0 d) 939 1 18 95 99 285-49900 (27-29975) [0.013-0.92 1] Mostly BOD<25 (0.12-0.32) 0.0005-1.128 « 0 . 1 ) [10-70 d] Minor Antifoam Good Added Bulking @ SRTs5d rising sludge Foam Poor Settling Very -SRTs 1-20 days, <5 erratic, >10 Poor worked well -very long SRTs required for nitrification Very -full scale plant, 0.144 Mgd Good -influent NH3 conc*n toxic (1072 mg/L) air stripped. Excel, -full scale Idg. <20 g /m 3 prod. eff. <25 mg/L -F /M <0.05 recommended to avoid filamentous bulking -complete nitrification when N/MLSSS0.03 W o n g & Mavinic (81) (1982) 13000 (8090) [0.62] 148-888 7-188 >93 >97 (0.1 1-0.405) (best @<0.16) Minor Poor Good Control -nutrients 100:3.2:1.1, temp down to Reactor 5" C Only -nutrient levels have little effect, F/M deter, settling Zapf-Gilje & Mavinic 19000 <900 (86) (13640) <97 1981) [0.71] >95 0.96,2.14,3.21 >99 (0.18-0.49) [6,9,20 d] Minor Good -temperatures 2 5,16,9' C Poor Column No. #1 #2 #3 #4 #5 #6 #7 Influent Effluent Loading Rate Op. Prob. COD mg/L COD % Rem kg / m 3 Temp. Foaming Heavy Nitri-References t< t 3 0 D5 m9^-) % &ern \?AA Ratio) Effects or Metal fication Comments [BOD/COD Ratio] [SRT] Settling Removal Stegmann & Ehrig (74) 4000-16660 20-3500 40-94 0.16-0.9 (1980) (750-11253) 20-60 95-99.9 (0.02-0.1 1) [0.19-0.67] Foam Excel, -activated sludge and aerated lagoon studies -treatment dependent on BOD/COD ratio -complete nitrification when <1 kg BOD/m 3 Paid & Quasim (56) (1977) Uloth & Mavinic (79) (1977) 365 48000 (36000) [0.75J 29-55 85-93 (0.226-0.436) [7-23.8 d] >98 (0.06-0.22) [>20 d] Minor Good Good -influent diluted 22-26 times -kinetic parameters evaluated SRTs 10,20,30,45,60 d, BOD inhibition obs. -best treatment at F/M <0.12 and SRT>20 d -mechanical mixing and low air controlled foaming Cook & Foree (15) (1974) Boyle & Ham (5) (1974) Premier Leachate 15800 (7 100) [0.45] 2700-9200 (1550-8000) [0.47-0.87] 86-4421 (44-3020) [0.25-0.75] 290-360 10-26 430-6720 160-7800 [10 d] 0.3-5.28 [5 d] Foam -F /M > 1.5 for failed unit 14 and fulvic acids (11,12). Further C O D removal has been demonstrated with physical-chemical effluent polishing (11,23,74), activated carbon being particularly effective (as mentioned previously), but as indicated by Stegmann and Ehrig (74), the necessity of removing this residual C O D is a subject for debate. Column 3 gives the loading rates and/or solids retention times (SRT) at which the various experiments were run. The conditions for which good treatment was achieved were very similar in all of the studies. It was generally concluded that a SRT S* 10 days and a F/M ratio of less than 0.1 - 0.15 kg BODj/kg MLVSS, or volumetric loading less than 0.1 - 0.15 kg BOD,-/m 3*d, was necessary for efficient and reliable treatment (43,66,79,81). These operating conditions correspond fairly well to those describing the extended aeration variant of activated sludge (51). The long SRTs and low organic loading rates necessary for efficient leachate treatment indicate low rates of growth of the process bacteria, which should be reflected in the process kinetics. Mavinic (49) summarized the kinetic parameters for carbon removal from a number of leachate treatment studies and demonstrated that the difference between the values of the kinetic coefficients determined for a leachate, and those typical of domestic sewage, increased with increasing leachate strength. Therefore, despite high proportions of readily biodegradable material in the leachates studied, inhibition of bacterial growth was indicated and increased with the strength of the leachates. It was concluded that the SRT required to effectively treat a leachate increases markedly with the strength of the leachate. Since the SRT is inversely proportional to the F/M ratio and loading rate, the maximum loading rate would decrease as the leachate strength increased. Kinetic considerations also indicated that old leachates, those with a low BOD^/COD ratio and thus low biodegradability, would also require longer SRTs and lower loading rates. Mavinic also found that the kinetic parameters were greatly influenced by cold temperatures. 15 The effects of the changes in kinetics with leachate strength can be seen in the results of Stegmann and Ehrig (74). They reported on activated sludge and aerated lagoon leachate treatment from bench, pilot, and full scale studies. The more extensive aerated lagoon results indicated that complete treatment to BODj. values less than 25 mg/L were possible at loading rates and detention times determined by the BODr/COD ratio of the leachate. As reported previously by Chian and DeWalle (11), the BOD,-/COD ratio is a useful characteristic with which to catagorize leachates and evaluate their treatment results. High ratio (high strength) leachates (>0.4 BODj/COD) could be treated at loading rates up to 0.05 kg BODr/m.3*d, but longer detention times were also required. For leachates with intermediate ratios of 0.1 - 0.4, loading rates of <0.01 kg BOD^/m^*d were necessary but the detention times required were relatively constant. Leachates with low BOD^/COD ratios <0.05 required very low organic loading rates, < 0.002 kg BODg/m^*d, but the detention times were also reduced reflecting the small fraction of degradable material. In addition to maintaining reasonable SRTs and loading rates, top treatment efficiency was generally dependent on achieving a proper nutrient (N + P) balance in the process. Landfill leachate is generally found to have sufficient nitrogen present in the form of ammonia (NH^) but levels of phosphorous are usually deficient. Therefore most leachate treatment studies have added nutrients, particularly phosphorous, to prevent nutrient deficiencies. Cook and Foree (15) found, as mentioned previously, that nutrient additions did not have a great effect on total treatment efficiency, but the effluent quality was improved slightly when nutrients were added. Stegmann and Ehrig (74) also found that P addition had only a minor effect on effluent quality. A lack of phosphorous inhibited growth in one activated sludge unit but the reduction in soluble effluent BOD,, with P addition in another unit was judged insignificant compared to the overall removal. Wong and Mavinic 1 6 (81), while investigating the effects of sludge age and temperature on leachate treatment, confirmed the previous work by Temion and Mavinic (see 81), which showed a nutrient ratio of 100:3.2:1.1 BOD[-:N:P was sufficient to satisfy the growth requirements of the biomass for leachate treatment. They also found that phosphorous deficiencies resulted in poor settling of the bacteria. Robinson and Maris (65) also found that nutrient requirements for leachate treatment were less than the commonly accepted ratio for sewage treatment when ammonia nitrogen in excess of 100:3.6 B O D ^ N remained in the process effluent. The reduced nutrient nitrogen requirements for leachate treatment may help explain the small effect that P addition had in the Cook and Foree, and Stegmann and Ehrig studies, but the lower effluent soluble BOD,- levels achieved with P addition indicate addition i s -beneficial to attain high levels of treatment and improve process reliability. Column 4 shows that despite the indications of temperature sensitivity from kinetic considerations (49), temperature effects on treatment efficiency were generally found to be minor. Zapf-Cilje and Mavinic (86) observed a minimal loss of treatment efficiency with decreasing temperature down to 9° C. Similarly Robinson and Maris (65), and Ehrig (22), reported insignificant effects of lower temperatures on B O D j removal. However all of these studies reported impaired solids settling at lower temperatures, so this appears to be the main adverse effect. Column 5 summarizes the operational problems encountered during the course of some of these studies. There were basically two types of problems; excessive foaming of the leachate, and poor settling of the bacteria. Excessive foam formation is a common concern of leachate treatment. The high aeration rates usually required for treatment and the general use of inefficient coarse bubble diffusers, to avoid plugging, further aggravates the problem. Surface aerators frequently cannot be used because of concerns about foaming (or in other instances, heat loss)(33). The treatment studies show a trend towards increased 17 foaming as the leachate strength increases. Foaming problems also increased with leachate loading levels. Uloth and Mavinic (79) linked the foaming tendency of leachate to metal concentrations and anticipated severe problems with the high strength leachate used in their study. However, the use of mechanical mixing and minimal aeration was very successful at avoiding excessive foam formation. In other studies, anti-foaming chemicals were used effectively. For full scale applications, the results of Uloth and Mavinic indicate that mechanical mixing greatly reduces foaming problems, such that the use of submerged turbine-sparger combinations could reduce the need for chemical or physical foam control measures. The settling problems reported in some of these studies can be associated with either low temperatures, as mentioned above, excessive loading levels, or nutrient deficiencies, except for Keenan ef al. (43) who had denitrification and turbulence occurring in their clarifier. Both the low temperature and excessive loading conditions led to filamentous bacterial growth and sludge bulking conditions which impaired solid-liquid separation. Under more favourable operating conditions, excellent solids settleability was generally observed. Good solids settleability was frequently attributed to the inclusion of inorganic precipitates and adsorbed metals in the bacterial floes. As shown in column 6, heavy metal removals from the leachate during aerobic treatment was very good. Heavy metals are removed by various mechanisms but the two main ones are as inorganic precipitates, usually hydroxides or phosphates, and adsorbed or complexed with the biological solids (6,9,76). However the various metal species are not removed equally well. Studies of metal removal by activated sludge during domestic sewage treatment have established a fairly consistent order of metal removal efficiency. Iron, Zinc, Copper, Chromium, and Lead, are removed best while Nickel, Manganese, Calcium, and Magnesium, are removed least. Brown and Lester (6) reported average removals of Fe (86%), Zn 1 8 (69%), Cr (66%), Cu (66%), Pb (64%), Hg (63%), Al (51%) Cd (46%), Ni (33%), Mn (20%), and Ca (6%). Although landfill leachates often have higher heavy metal concentrations and different relative concentrations between metal species than domestic sewage, the leachate treatment studies have demonstrated heavy metal removals with similar affinities for metal species and similar or better removal rates. Higher removal rates during leachate treatment are probably attributable in part to the generally longer sludge ages employed as metal removal is enhanced by longer sludge ages (76). Very efficient heavy metal removals have also been observed during anaerobic treatment of leachates (5,7). Under anaerobic conditions, many heavy metals precipitate as sulphides which, combined with the biological complexing, yields the high removal rates. The efficient removal of soluble heavy metals, by precipitation and biological inactivation, explains the general lack of observed toxic effects despite the often very high concentrations of heavy metals in leachates undergoing treatment. For example, Uloth and Mavinic (79) had the highest leachate metal concentrations of the studies reviewed here, Fe 960 mg/L, As 3.6 mg/L, Pb 1.44 mg/L, and Zn 223 mg/L, yet a stable process was maintained and achieved >98 percent C O D removal, and very high metal removal rates. None of the other studies reviewed indicated any metal toxicity problems, except possibly for Jasper et al. (42), who proposed metal accumulation in their sludge as one explaination for a deterioration of nitrification performance. Studies are ongoing at UBC to clarify heavy metal toxic effects, particularly of Zinc, on nitrification/denitrification of leachate (18). Although leachate metal levels are generally non-toxic, Mavinic (49) indicates that they probably contribute to inhibiting bacterial growth rates and thus affect the process kinetics. In most cases however, leachate metal concentrations are not considered to significantly impair biological leachate treatment, either aerobic or anaerobic (11,33). 19 In addition to the high levels of carbonaceous oxygen demanding material found in landfill leachates, particularly young leachates, relatively high levels of nitrogenous oxygen demanding material, mostly in the form of ammonia (NH^ - N), are also generally present. Since, as will be discussed in Section 5.0, the high ammonia concentrations usually persist long after the organic strength has been reduced by the maturation of the landfill, there is an increasing interest and emphasis on ammonia removal from landfill leachates. Ammonia is removed biologically by conversion to organic nitrogen during bacterial cell synthesis, or oxidized to nitrate (NO^) by nitrifying bacteria. Depending upon site specific effluent guidelines or goals, the nitrate could possibly be further treated by biological denitirfication to remove the nitrogen completely as nitrogen gas, although Henry (33) indicates that experience with denitrification of leachate is insufficient to predict its reliability. Research emphasizing nitrification or nitrogen removal from leachate is very recent; thus, most of the studies reviewed looked at nitrification as a secondary topic to carbon removal. Given that, as mentioned previously, all these studies dealt primarily with high organic strength leachates, with COD concentrations much greater than ammonia concentrations, this approach was not unreasonable. Fortunately, the nitrification results, where given, are sufficiently detailed to support a number of conclusions. The nitrification results from the various studies are summarized in column 7 of Table 3.1. The first point demonstrated by the results of these studies is that the occurrance of nitrification depends upon the BOD^NHg ratio of the leachate. Studies in which this ratio was greater than or equal to roughly 100:3.6 had no nitrification take place (65,66,81). The lack of nitrification under these conditions is attributed to the greater growth rate of the heterotrophic bacteria which convert ammonia to organic nitrogen required for cell sysnthesis (34,35,47,53). As mentioned 20 previously, the 100:3.6 B O D 5 : N H 3 ratio represents the minimum nutrient requirement of the heterotrophic bacteria for cell growth. This ratio of organic strength also ensures a high growth rate of the heterotrophic bacteria at practical loading rates, since the heterotrophs follow a first order kinetic response to substrate concentration. Civen the relatively high organic loading rates employed in most of the studies reviewed, the heterotrophs were likely growing at or near their maximum rate in most instances. O n the other hand, nitrifying bacteria follow a zero order response to substrate concentrations greater than about 0.5 mg/L and thus essentially grow at a constant slow rate. Therefore, under nutrient limiting conditions, the heterotrophs consume essentially all of the ammonia for their nutrient requirements by virtue of their much higher growth rate. Results from the studies in which excess ammonia was present such that the BOD^/NHj ratio was less than 100:3.6, generally indicate that efficient nitrification of landfill leachate is possible. Stegmann and Ehrig (74) achieved complete nitrification in both lab scale activated sludge, and pilot scale aerated lagoon studies, ln the later case at organic loadings of up to 1 kg BODg/m 3 *d. When denitrification was attempted in the activated sludge studies, 99% removal of the influent nitrogen was achieved as an influent ammonia concentration of 973 mg/L was reduced to 8.2 mg/L N H ^ and 25 mg/L N O ^ in the effluent. Keenan er al. (43) reported greater than 99% nitrification of leachate in a full scale activated sludge plant. Efficient nitrification was also observed in the control reactor of the Wong and Mavinic (81) study, to which excess nutrients had been added. More recently, Dedhar and Mavinic (18) maintained efficient nitrification of an old, low organic strength leachate. Good denitrification performance was also achieved but complete denitrification could not be maintained due to variable carbon loading to the anoxic zone. Therefore the results of these studies support the general view that efficient nitrification of landfill leachates is usually readily achieved (13,33). 21 However the experience with nitrification of landfill leachates has not been as overwhelmingly positive as that for carbon removal. Robinson and Maris (64,65,66) reported that for a study involving both a high and low organic strength leachate, no nitrification had occurred after a retention time of 20 days. In the case of the high strength leachate, influent ammonia levels were initially completely converted to organic nitrogen but continued detention beyond 20 days allowed for a gradual conversion of some of this organic nitrogen to nitrate, essentially by aerobic digestion of the sludge. The low organic strength leachate had only a small portion of the influent ammonia convert to organic nitrogen after 20 days retention, reflecting the low BODj. removals, and ammonia losses were largely accounted for by volatilization at the pH of approximately 9.3. Continued detention beyond 20 days resulted in some nitrification but a total retention period of 70 days was required to reduce ammonia levels to less than 1 mg/L. Very low suspended solids levels in these units probably contributed to the poor ammonia conversion (MLSS and MLVSS were less than 200 and 100 mg/L respectively). In a previous series of experiments, Robinson and Maris (65) subjected activated sludge units operating with a 10 day SRT at 10° C, to artificially elevated influent ammonia levels. No appreciable nitrification was observed in any of the units over 82 days of operation and influent ammonia in excess of the nutrient requirements of the micro-organisms remained in the effluent. This result was not entirely unexpected because of the relatively short SRT and lower temperature. Therefore the SRT was increased to 20 days and the units operated for a further 70 days. Most of the experimental units did not stabilize at the new SRT and while some nitrification occurred in units with excess influent ammonia, it was unstable and incomplete. The incomplete conversion of ammonia to nitrite (NC^), and the accumulation of nitrite in the reactors, reduces the pH, as was observed, and inhibits the re-establishment of stable nitrification (22). Robinson and Maris 22 concluded that their results did not show a fundamental reason why nitrification of leachates could not be achieved, but rather indicated that a greater degree of process control, particularly of pH, and much longer SRTs in some instances, are required to maintain a reliable nitrification process. Jasper et al. (42) also reported unsatisfactory nitrification/denitrification performance from their study. Initially, during the first eight weeks of the study, efficient nitrification was established in 10 and 15 day SRT units and ammonia removals in excess of 90% were achieved. However, as the study progressed, the nitrification performance deteriorated such that in the final three weeks (22 - 25), ammonia removals were just over 50% and nitrate levels in the aerobic zone were less than 10 mg/L. The denitrification performance of these units was also very poor. During the first eight week period, no denitrification took place despite the high levels of nitrate available. Following some operational changes at the end of week eight, denitrification was established and outperformed the nitrifiers, but it too deteriorated as the study progressed. On a percentage basis the denitrification rate increased towards the end of the study to over 90%, but this was due more to the reduction in nitrate levels than an increase in denitrification performance. The authors speculated that low effluent ammonia levels could be achieved at very long SRTs (>>30 days), as suggested by Robinson and Maris (65) above, but this would be due as much to other removal mechanisms (assimilation and stripping) as nitrification. It was also postulated, as mentioned previously, that the inability to establish a stable nitrification/denitrification process was due to toxic inhibition, possibly from accumulated heavy metals in the sludge. Not withstanding the difficulties with nitrification of landfill leachate experienced in a small number of studies, biological leachate treatment, especially by aerobic suspended growth systems, has proven to be a very effective and efficient treatment alternative for removing the majority of pollutants from landfill leachates. 23 Aerobic suspended growth treatment systems operated within limiting conditions have demonstrated complete removal of biodegradable substrates (BODj. <25 mg/L and N H 3 <1 mg/L), as well as efficient removal of suspended solids, heavy metals, odours, and colour. Henry (33) generalizes these limiting conditions as SRTs of twice, and loading rates of half, those used for domestic sewage treatment, but it is the leachate characteristics in each case which dictate specific operating limits. The discussion above has also underlined the need for careful and detailed process monitoring and control in order to maintain process reliability and high levels of treatment, especially for reliable nitrification/denitrification. 3.3 RBC TREATMENT Given the proven capabilities of aerobic suspended growth systems to effectively treat landfill leachates, one might question the need to evaluate other processes such as the Rotating Biological Contactor (RBC). However the aerobic fixed growth RBC process has operational characteristics and bacterial growth conditions which are distinct from other processes, particularly suspended growth systems, and claims to have various advantages over other processes, especially for nitrification. Given the wide variety of composition of landfill leachates, and the increasing emphasis on nitrifying leachates, the RBC could well prove advantageous for the treatment of at least some types of leachates. Therefore the evaluation of the RBC's performance with respect to landfill leachate treatment is worthwhile. The Rotating Biological Contactor is a very simple treatment device both mechanically and conceptually. Thin discs, or some other shaped media, which provide the surface area upon which the bacteria will grow, are mounted on a horizontal shaft over a trough containing the wastewater so that the disks or media are partially submerged in the wastewater. Generally the depth of submergence is such that 40 to 50 percent of the total surface area is underwater. The shaft is 24 then slowly rotated so that the media surface is alternately immersed in the wastewater and exposed to the atmosphere. A laminar film of water remains attached to the rotating media and it is within this film that the bacteria grow and affix themselves to the media surface. Once the media surface is covered by the bacterial growth, the surface roughness of this biomass helps determine the thickness of the water film which attaches to it. During the time the thin water film is exposed to the air, gas transfer occurs across the air-water interface with for example, oxygen diffusing into the film, and gaseous products of bacterial activity such as C O 2 , diffusing out. When the film of water is immersed in the wastewater, diffusion of soluble materials; substrates, products, and gases, occurs across the liquid-liquid interface. For example, this is considered the main mechanism by which the bulk liquid is aerated or stripped of dissolved gases. Some mixing also occurs between the bulk liquid and the attached water film but the extent of this exchange and the thickness of the laminar film are uncertain, and depend upon other factors such as the rotational speed of the media, and the surface roughness of the attached growth (as mentioned above). The rotation of the shaft and media also causes a gross mixing of the wastewater in the trough such that it is considered completely mixed, which enhances mass transfer at the liquid-liquid interface, and also contributes some mechanical aeration of the bulk liquid, although this is considered minor and frequently neglected (26). To complete the system, baffles are placed between sections of disks on a shaft, or a number of troughs are connected in series, to provide a number of stages which prevent hydraulic short circuiting, and enhance treatment efficiency. As RBC systems are generally staged, the process is analogous to a series of completely mixed reactors, with the important difference that the bacteria are retained in each reactor, and practically speaking, only the wastewater moves 25 between the different stages. The RBC thus combines many attributes of completely mixed and plug flow reactor systems. These attributes, along with characteristics unique to the design and operation of RBC systems, can be used to explain many of the advantages and disadvantages claimed for the RBC process. One of the most obvious advantages of the RBC system is the simplicity of the mechanics and process itself, which results in economical operation. In a mechanical drive RBC system, operation and maintenance costs are minimal as the drive components and bearings are simple and reliable, and the energy required to turn the shaft is relatively small. Air driven RBCs similarly use proven blower technology, and the low air pressures and volumes required to rotate the shafts result in even lower operating costs. Process control is also very simple as the biomass is self regulating. Control is limited to periodic monitoring of organic and hydraulic loading rates to ensure they are within design limits, as well as checking dissolved oxygen levels, pH, effluent quality, etc., to both monitor process performance and meet regulatory requirements. RBCs have been used widely for small package plants because of their economical and simple operation with minimal operator skill and attention. Smith and Bandy (71) in their state-of-the-art review of RBC technology (1983), found that hourly labour requirements for RBC plants ranged from 1 to 7 hours/week, averaging 2.6 hours/week, and that power requirements for 100,000 f t 2 (9290 m 2 ) of standard density media were 3.6 kw for mechanical drive units (1.6 rpm), and 2.93 kw for air driven units (1.2 rpm). Capital costs of RBC plants are higher than for activated sludge plants for capacities above approximately 1 M G D (3800 m 3/d) but landfill leachate volumes are generally much less than 1 M C D and are within the range where RBC plants are economically viable. One of the most important attributes claimed for the RBC process is its high performance for carbon removal and nitrification (84). RBCs operating within design limits have demonstrated their capability to produce high quality effluents 26 with soluble BOD, . (SBOD,.) levels of less than 10 mg/L and ammonia concentrations of less than 1 mg/L (54). This effluent quality is comparable to that achieved by other high efficiency aerobic systems and represents a practical limit for biological treatment. The mechanisms by which this treatment is achieved in the RBC, and the factors which affect it, are at present best explained in terms of the mass transfer/kinetic models of Famularo and Mueller, et al. (26,53). It has been found, and is now generally accepted, that the biofilm develops up to three layers or zones of biological activity of varying thicknesses, depending upon loading conditions (48,53). The outermost layer consists primarily of rapidly growing aerobic heterotrophs which utilize the carbonaceous substrates from the wastewater. When conditions permit, slow growing nitrifiers will predominate to form a second aerobic layer beneath the aerobic heterotrophs. A third innermost layer generally forms against the supporting media where anaerobic conditions can prevail as the oxygen is unable to penetrate to this depth in the biofilm. The activity in the anaerobic layer varies according to the penetration of other substrates but can include; acid fermentation and methane production if exogenous substrates are available, denitrification if both carbon and nitrate are present, and endogenous reduction of the biomass. Other autotrophs such as sulphur bacteria can also become active when conditions permit. In stages receiving high loadings of organic carbon, usually the first few stages in a multi-stage system, the heterotrophic bacteria in the aerobic zone grow very rapidly. High influent substrate concentrations provide a strong concentration gradient for the diffusion of soluble substrates into the biofilm. The availability of substrate supports a large, actively growing biomass which exerts a high oxygen demand and similarly results in a strong gradient for the diffusion of oxygen into the biofilm. Growth is usually limited by the mass transfer rate of oxygen under these circumstances, rather than growth kinetics or substrate availability, and therefore 27 oxygen does not penetrate beyond the first layer. This limits the maximum thickness of the aerobic layer as well as the rate of substrate utilization. Though given the higher mass transfer driving forces present, the majority of the BOD,, removal and the highest BOD^ removal rates are commonly observed in these initial stages (3,57,60,71). However, overloading conditions, as will be discussed later, interfere with transfer rates and reduce removal efficiency. Influent suspended solids, both organic and inorganic are also effectively removed by adsorption and impingement on the thick biofilm. The high growth rate in the early stages results in a high sloughing rate of excess biomass so that suspended solids levels are usually highest in these first stages. Development of a nitrifier layer in heavily loaded stages is generally inhibited by the growth rate of the heterotrophs which outstrips that of the nitrifiers to the extent that oxygen and/or ammonia are unavailable to them. Marsh e( al. (47) observed that nitrification does not begin until BOD,- levels in an RBC have been reduced to 60 mg/L, and a stable nitrifier population only becomes established when B O D 5 levels approach 30 mg/L. At B O D 5 levels less than 10 mg/L the nitrifiers become the dominant population. The other studies reviewed all reported similar findings and it is now generally accepted that BOD,- levels of less than 30 to 40 mg/L are required to establish a stable nitrifier population. In terms of loading rates, Kincannon ef al. (44) found that nitrification did not begin until the loading in a given stage was reduced to 4.15 g SBOD^/m 2*d or less. Therefore it is frequently observed that TKN or ammonia removal in the first stages of an RBC is limited to the nutrient requirements of the heterotrophs and nitrate production is minimal (34,39,44,53,58). The limited pentration of oxygen into the biofilm under the heavy loading conditions prevalent in the early stages allows the formation of the thickest anaerobic layers. Depending upon the diffusion of organic substrates there may be 28 some exogenous growth but most of the activity involves endogenous reduction of the biomass. Denitrification generally does not occur because of both the lack of nitrate production, mentioned above, and the lack of carbon penetration into the anaerobic layer due to its rate of utilization in the aerobic layer. Under very heavy loading, anaerobic conditions become stable enough to allow the growth of sulphur reducing bacteria. This generally leads to operational problems as will be discussed further below. In stages receiving moderate loadings of organic carbon, corresponding to intermediate stages of a multi-stage system, biological activity in all three layers generally interact in relation to the availability of the various substrates. The growth rate of the aerobic heterotrophs is usually substrate limited and reduced from that observed in the previous stages or at higher loading rates. This results in a thinner biomass and increased penetration of other substrates and oxygen into the biofilm, as mass transfer rates exceed the rate of utilization in the first layer, ln particular, oxygen and ammonia can diffuse into the biofilm to a depth at which the slow growing nitrifying bacteria can outgrow the heterotrophs to form their own layer. The greater diffusivity of ammonia over carbon substrates aids in this process (36). Nitrates produced in this layer then diffuse outwards in both directions in response to concentration gradients. Within the anaerobic layer, which underlies the nitrifier zone, considerable denitrification can occur with nitrates readily available and using carbon from residual substrate or endogenous respiration of the heterotrophs (1,48,53). The substrate limited growth rate of the aerobic heterotrophs in these intermediate stages derives from the reduced influent soluble BOD^ concentrations. Low B O D j concentrations in the bulk liquid result in even lower concentrations in the biofilm because of the gradient required to drive the mass transfer. The lower mass transfer rates and heterotrophic growth rates result in lower BOD - removal 29 rates and reduced removal efficiency being observed in the intermediate stages or at lower loading rates (57,60). A contributing factor to the lower heterotrophic growth rates is that the soluble fraction of the organic load entering these later stages contains a higher proportion of compounds resistant to degradation, the readily utilized substrates having been preferentially consumed in the initial stages. However in these later stages, many slower growing and/or specialized bacterial populations can compete more effectively to remove much of this material. A significant portion of the organic load to the intermediate stages comes from the suspended or sloughed solids from the previous stages. These solids generally become re-attached to the media in subsequent stages (43) and the extra retention time under substrate limited conditions encourages the degradation of the entrapped organic solids and endogenous reduction of the excess biomass. The endogenous reduction of the biomass reduces the overall sludge production of the RBC process, which is another advantage claimed for the system. Kincannon et al. (44) determined a sludge production rate of 0.37 kg solids/kg BOD,, removed, treating domestic sewage. However the sludge production for an RBC treating landfill leachate would increase somewhat due to the addition of inorganic precipitates as was observed for suspended growth systems. ln stages receiving very light organic loadings, corresponding to the last stage of a multi-stage system, nitrifier activity predominates and the biomass is very thin reflecting their low growth rate. Heterotrophic activity is limited primarily to endogenous respiration of re-attached biomass from previous stages. Denitrification is also limited by the lack of carbon and the penetration of oxygen. Removal efficiency of all substrates under these severely substrate limited conditions is further reduced because the low concentration gradients provide a very small driving force with which to overcome the mass transfer resistances. This explains why it is not possible to reduce soluble substrate concentrations to zero. 30 From the discussion above it follows that efficient BOD, , removal can be achieved with the RBC process provided that the sizing, and number, of the stages is adequate. Since B O D ^ removal rates increase with the applied loading, B O D ^ removal can be enhanced by increasing the proportion of surface area in the first stage compared to the subsequent stages. This factor is frequently incorporated into RBC plant design; for example the pilot plant used in this study had twice as much surface area in the first stage than in each subsequent stage. Conversely the decreasing BOD, , removal efficiency with successive stages, or diminishing return, means that the maximum number of stages for B O D ^ removal is generally taken to be four. Configuring an RBC plant to achieve maximum BOD,- removals in the early stages not only improves the B O D j removal efficiency but the nitrification performance as well. As discussed above, the onset of nitrification is dependent upon the prior removal of most of the carbonaceous substrates and therefore the earlier in the process this occurs, the more surface area is available for nitrification in the following stages. In cases of high influent BOD, , and/or ammonia concentrations, or the treatment of substrates which are difficult to degrade (such as phenols), additional stages may be required to achieve complete treatment or high effluent quality. As will be discussed in Section 5.0, landfill leachates frequently have one or more of these characteristics. The attached growth nature of the RBC provides several advantages for nitrification over suspended growth systems. Nitrifying bacteria are generally described as being sessile (22,24) and therefore the RBC provides a prefened environment for their growth. Crowing attached to the RBC media also results in indeterminately long solids retention times which allows the slow growing nitrifiers more than adequate time to develop. Since only two species of bacteria are involved in nitrification, Nitrosomonas and Nitrobacter, they have proven more sensitive to 31 various environmental conditions such as, temperature, pH, and inhibitory levels of substrates or other substances, than the more diverse and adaptable populations responsible for carbon removal or denitrification. Therefore, the relatively protected location of the nitrifiers, within an interior layer of the biofilm, usually in the later stages of the process, could significantly improve the stability of the nitrifier population. As will be discussed further below, the fact that the biomass is attached also greatly reduces the possibility of the slow growing nitrifiers being washed out of the system. When all these factors are considered, the fixed growth RBC system provides an environment far more conducive to the growth of a stable nitrifier population. Therefore a more reliable and efficient nitrification process should result. The attached nature of the biomass also greatly enhances the settling of the effluent suspended solids which are frequently mainly nitrifiers. It is generally reported that nitrifying bacteria from suspended growth systems are finely dispersed and settle very poorly (22,66), which aggravates the wash out problem in these systems. However in the RBC process, the effluent suspended solids are concentrated in chunks sloughed off of the media and these chunks generally settle well, even when they are composed mostly of nitrifiers. In the case of leachate treatment, effluent suspended solids settling is further improved by heavy inorganic precipitates. The mass transfer mechanisms can also be used to explain the observed reduction in BODj. removal efficiency as the size of the RBC is increased. Kincannon ef al. (44) found there were no scale-up effects for overall loadings of less than 4.9 to 7.3 g SBOD 5 /m 2 *d or first stage loadings of less than 12.2 g SBOD[-/m 2*d. However, at loading rates higher than these levels, full scale RBC units became oxygen limited and lost removal efficiency sooner than smaller scale units. Similarly Wilson, Murphy, and Stephenson (54,80) found that a 0.5 m diameter 32 RBC unit achieved 15 percent higher C O D removals than a 2.0 m diameter unit and assumed that a further 10 percent reduction in performance would occur in 3.5 m diameter units. Famularo et al. (26) expained that since the peripheral velocity of the media is kept constant in most studies, to limit shear forces, the rotational speed of the shaft decreases as the diameter of the media increases. Therefore the period of rotation increases with diameter and an element of surface area is either immersed or exposed to the air for increasing periods of time. When this period of rotation is too long for the rate of growth of the bacteria, the substrate concentration within the liquid film will be depleted and the biomass will become inactive for part of each cycle, reducing the removal efficiency. Presumably, oxygen would be similarly depleted for some portion of the immersion cycle, further reducing the activity of the biomass. As indicated by the results of Kincannon et al. (44), at lower loading rates the growth rate of the biomass is not sufficient to cause this effect. It follows from the above discussion that if rotational speed were kept constant, scale-up effects would be reduced or eliminated. Results such as those of Friedman et al. (29) indicate that to some extent this is true. They found that as the rotational speed, and thus peripheral velocity, of their RBC units were increased that the maximum removal rate increased. Since their units were of the same diameter (11.88 inches) this improvement would be primarily due to increased mass transfer rates caused by the increased mixing and turbulence at the biofilm interface, rather than overcoming the substrate depletion decribed above. In a full scale RBC, both of these factors would improve performance but rotational speed is limited by the need to keep the peripheral velocity within acceptable limits so that the hydraulic shear doesn't strip off the biomass. More or less out of tradition this limit has been set at 0.3 m/s or 1 ft/s, but Friedman et al. demonstrated that up to a 50 percent increase in this value may prove practical. However power costs 33 also increase with the rotational speed. It has been found that nitrification within the RBC proceeds at a constant, temperature dependent, rate per unit area of nitrifying bacteria and thus does not exhibit any scale-up effects (44,53,54,80). Kincannon et al. (44) determined a constant nitrate production rate of 0.73 g N O y m ^ d . Similarly Murphy and Wilson (54) calculated a constant TKN removal rate of 1.12 g TKN-N/m 2 *d at 20° C. The constant reaction rate is due to the slow growth rate of the nitrifying bacteria and their zero order response to substrate concentrations above about 0.5 mg/L. Given that substrate utilization is reaction rate limited, the other factor limiting nitrification is the number of nitrifying bacteria. It appears as though the nitrifier layer, in those stages in which it develops, grows to a fairly uniform thickness, limited by reaction rates, endogenous decay, and to a lesser extent predacity, and mass transfer limits. This would explain the strong areal relationship of nitrification in the RBC. Another important advantage claimed for the RBC process is excellent resistance to shock organic, hydraulic, or toxic loadings (54,81). Of these, the RBCs resistance to hydraulic loading is the easiest to explain. As the bacteria are attached to the fixed media they are much less prone to wash-out of the process than in a comparable suspended growth system and therefore major losses of biomass usually do not occur. Upsets in a suspended growth process typically result in reduced solids settleability and high hydraulic flows would only increase the losses of biomass into the effluent. Biomass losses are particularly serious in a nitrifying system. In a RBC system however, process performance can be impaired due to a reduction in the hydraulic retention time (HRT). RBCs designed for sewage treatment generally have short HRTs, in the order of 0.5 - 2.0 hours at design flowrates, which is considered another advantage of the process itself (84). Therefore under the conditions of a hydraulic surge, the system HRT could fall below some limiting value. Filion et al. (27) investigated the effects of variable hydraulic loading on an 34 RBC and found that varying the HRT between 0.44 and 0.94 hours did not significantly affect carbon removal, but nitrification performance did respond to HRT variations between 1.12 and 3.36 hours. However nitrification performance recovered much more quickly from hydraulic loading fluctuations, than from changes in TKN loading or influent concentration. Poon ef al. (60) found that for their system, B O D j removal was adversely affected by a reduction of the HRT from 0.73 to 0.42 hours, but low influent B O D j concentrations were cited as a contributing factor (lower reaction rates). Therefore, while the RBC provides good resistance to loss of biomass during high hydraulic loadings, process performance, especially for nitrification, may be reduced by lower hydraulic retention times. The RBC's resistance to shock organic or toxic loads stems from the large mass of highly concentrated micro-organisms resident on the fixed media. As noted previously the first stage of an RBC is typically larger or has more surface area than subsequent stages to maximize the mass of bacteria in contact with influent conditions. In the case of peak organic loadings, this large biomass provides the assimilative capacity necessary to absorb extra substrate over short term periods. With respect to toxic loadings, toxic effects become manifest when the levels of toxin exceed a critical ratio to the mass of bacteria, rather than reach a specific concentration. Therefore the greater the biomass present, the greater the concentration of toxin which can be tolerated. Theoretically, a completely mixed suspended growth system with the same total biomass as an RBC would be somewhat more effective at resisting shock loads because all of the bacteria would be available to moderate the shock conditions. By the same reasoning an RBC would be more effective than a plug flow suspended growth system due to a higher proportion of biomass near the influent end of the process. However, as pointed out above, the RBC system is much less prone to losses of biomass if the bacteria become stressed or upset. 35 The results of Filion ef al. (27) tend to confirm the greater responsiveness of a completely mixed suspended growth system; a side by side comparison of comparable activated sludge and RBC systems indicated that the effects of loading peaks on effluent quality were more pronounced and lasted longer in the RBC system. As might be expected, the impact of peak TKN loadings on nitrification performance were roughly three times greater than the effects of peak TOC loadings on carbon removal. Peak loading rates of 24 - 27 g TOC/m 2 *d and 6.0 -7.2 g TKN/m 2*d obviously exceeded the assimilative capacity of their RBC system. The recovery times of roughly one hour for carbon removal and three hours for nitrification indicate that longer HRTs could significantly increase the RBCs resistance to shock organic loads. Poon ef al. (60) found that for shock organic loadings from 4.3 to 15.4 g SBOD^/m 2*d, representing 124 to 444 percent of the normal applied load, that no adverse affect on the- RBC unit performance was observed. They conclude that these loadings were within the assimilative capacity of the RBC since other studies had demonstrated removals of up to 17.0 g SBOD,., or 28.3 g total BOD,., per m 2 *d . Therefore, the RBC has demonstrated good handling of shock organic loads within its assimilative capacity and reasonable response to even higher loads. Again the main advantage may be the RBCs resistance to biomass losses, which often plague suspended growth systems during shock conditions. Resistance to shock organic loadings is very important with respect to leachate treatment because as will be discussed fully in Section 5.0, landfill leachate composition and flowrates can be highly variable, and peak organic and hydraulic loadings frequently occur coincidently. Landfills also often receive various types and amounts of toxic material, knowingly or otherwise, which can end up in the leachate. One of the main disadvantages of the RBC process is its sensitivity to low temperatures. The same large surface area and very thin water film in contact with 36 the atmosphere which maximizes gas transfer also maximizes the potential for heat transfer. Since as a general rule of thumb the rate of biological reactions is reduced by half for each 10° C decrease in temperature, the rapid cooling of influent wastewater can significantly decrease process performance. During warm weather the heat transfer efficiency of the RBC is beneficial as influent wastewaters are generally cool. The heating effect of warm temperatures is moderated somewhat by evaporative cooling which Kincannon et al. (44) observed could reduce temperatures by 2 - 3 °C across an RBC unit. During cold weather conditions however, the situation is deleterious as the RBC will efficiently lose heat to the atmosphere and evaporative cooling further aggravates the problem. Therefore, RBCs installed in areas subject to low temperatures (less than 10° C) are usually fitted with insulated covers to reduce heat loses. Under very cold conditions they must be covered to prevent icing problems. The reduction in RBC performance as water temperatures decrease has been well documented (28,54,57,80,85). In most instances the effects of temperature on the reaction rates of the RBC have been expressed in terms of an Arrhenius equation coefficient 8. For carbon removal, coefficient values ranging from 1.03 to 1.11 have been determined for various temperature ranges. Wilson et al. (80) for example, used a value of 0 = 1.05 over the temperature range of 5.5 to 13° C. Many of these studies showed that the carbon removal rate does not increase above 10 to 15° C, so corrections are not usually applied at higher temperatures (28,54). Nitrification has been observed to be much more sensitive to temperature effects and over a wider range of temperatures. The temperature coefficients determined for nitrification are therefore generally larger than for carbon removal alone, typically 1.09 to 1.11. Murphy et al. (54) for example, determined that a factor of 0 = 1.09 applied for carbon removal with nitrification up to 20° C, above which no further correction was required. However, as pointed out by Forgie (28) 37 the use of Arrhenius coefficients to correct for temperature effects is not strictly correct. Although the Arrhenius equation and coefficients are widely used and accepted for correcting reaction rates for temperature effects, Forgie recalls the fact that 6 itself is a function of temperature and that the use of a constant 6 is only an approximation. Therefore the use of a constant 6 over a relatively wide temperature range could lead to a significant error. Secondly, he pointed out that the form of the equation is exponential, which implies that reaction rates increase continuously with temperature, and conversely, that temperature effects decrease as the temperature decreases. As indicated by Forgie, a number of studies have shown that, in fact, the reaction rates drop off more sharply as the temperature decreases; thus temperature effects increase with decreasing temperature, and also reaction rates level off at warm temperatures, rather than increase continuously. Therefore, he concludes that the Arrhenius coefficients provide a reasonable approximation of actual temperature effects only when used over small 4 to 5° C temperature ranges. Forgie presented an empirical curve fit equation from experimental data, which indicated a parabolic shape. Experiments conducted by Forgie also produced a couple of other interesting results. The first was that an established nitrifier population could continue to nitrify well at temperatures as low as 1° C. However, these low temperature runs were only maintained for short periods of one or two weeks so as Forgie conceeded, it is uncertain whether or not this performance could be maintained at this low temperature. The second point illustrated by his results was that hydraulic retention time had an influence on the temperature effects. Specifically, longer HRTs reduced the adverse effects of low temperatures and restored some of the process efficiency. This effect is atttributable to longer contact times between the wastewater and biomass, offsetting the reduced reaction rates. Wu ef al. (83,84) also found 38 HRT to be a factor in RBC performance and therefore included it as a parameter in their empirical models of the RBC process. Another disadvantage of the RBC process is a history of operational problems, poor performance, and mechanical failures, which has made engineers wary of RBCs. Most of the mechanical problems can be attributed to early RBC installations in which poor design and fabrication of the units, resulting from inexperience with the weight of biomass which could accumulate and the forces involved, lead to numerous shaft and media failures. Although RBC design and manufacture are greatly improved, mechanical failures still occur periodically, usually in units which have been continuously overloaded and suffer fatigue failures. Other operational problems and poor treatment performance have also generally resulted from overloading conditions, frequently from underdesign. Some early designs were based on the performance of pilot scale studies by designers not cognizant of the scale-up effects which reduce performance. In other instances, RBC manufacturers have used overly optimistic design factors for competitive reasons. This underscores another disadvantage of the RBC process; that RBC design is still proprietary, which makes comparison and evaluation of RBC units and designs from different manufacturers difficult (71). A variety of operational problems have been observed in RBC units overloaded hydraulically or organically. As discussed previously when response to shock loading was considered, hydraulic overloading results in incomplete treatment and thus poor system performance. Organic overloading on the other hand can lead to a number of unpleasant conditions. Excessive organic loads, which occur quite frequently in the initial stages of RBC plants, cause overgrowth of the biomass, both on the media and eventually in suspension. The excessive growth frequently results in the biofilm bridging the gaps between the media surfaces which reduces the active surface area by restricting the access, and thus the transport, of oxygen 39 and substrates (25). Bridging of the biomass also reduces the ability of hydraulic shear to control the biofilm thickness and the rate of biomass sloughing is greatly reduced. The excessive growth rates drastically reduce the dissolved oxygen levels in the bulk liquid which, combined with the bridging effects, allows most of the biomass to become deeply anaerobic, severely reducing reaction rates in most of the biomass and thus it contributes very little to the removal performance of the system. Within the anaerobic zone, sulphur and hydrogen sulphide is often produced and this encourages the growth of Beggiatoa bacteria which oxidize these products to produce energy. These micro-organisms store sulphur in their cells giving them a white milky appearance, which is characteristic of overloaded RBC stages (25,34). In addition to the reduced BOD,- removal indicated by the Beggiatoa growth, the sulphuric acid they produce can lower the system pH and adversely affect the nitrifying organisms. Anaerobic conditions in one or more stages can also cause severe odour problems. The excess biomass in overloaded stages can dramatically increase the weight of the shaft and media, thus greatly increasing the stress in these structures. Since the shaft is rotating, the stresses are cycled continuously and the effects of fatigue are multiplied, reducing the life expectancy of the shafts and media. The increased weight of the shaft also requires significantly more energy to turn it so energy costs are increased while performance decreases. Therefore, overloading conditions generally result in much higher operating, maintenance, and replacement costs. Evans (25) surveyed a number of RBC plants for operational problems and found that for plants in which first stage loading were less than 17.6 g BOD,-/m 2*d no problems were reported, but for plants with first stage loadings greater than 43 g BOD,./m 2*d, problems always occurred. At plants with first stage loadings between these two extremes, no clear pattern was observed and the 40 occurrence of overgrowth conditions was attributed to other factors such as; wastewater characteristics, rotational speed, temperature, tank design, and media configuration. It was statistically determined from the survey results that a first stage loading of 35.6 g BODj/m^*d had a 50 percent probability of operational problems. Various modifications of the standard mechanical drive RBC system have been employed to alleviate overloading problems associated with the initial stages and/or improve overall system performance. Step-feeding, de-staging, internal recycles, and supplementary air diffusers, have been used to reduce first stage loadings or prevent oxygen depletion. De-staging and step-feeding are two very similar methods of reducing the loading rate in the initial stages. They involve rearranging the process flow path so that the initial stages operate in parallel, or splitting the influent flow between the initial stages which still operate in series, respectively. Internal recycling of aerated wastewater from the last stage of the RBC back into the first stage both reduces the loading in the first stage by dilution of the influent and adds dissolved oxygen. However, internal recycles are only beneficial to process performance when used to alleviate an oxygen deficiency, otherwise the dilution of the influent reduces removal efficiency by decreasing mass transfer gradients and the HRT in each of the stages (3,55). Supplementary air diffusers placed in the initial stages both prevent oxygen deficiency, and produce additional turbulence in the wastewater which helps control biomass thickness and prevent bridging. One of the most effective developments in RBC technology has been the use of air-driven RBCs, which expand upon the benefits of supplemental aeration. Hynek and Chou (39) conducted a comparison study of air and mechanical drive RBC units and reported a number of advantages associated with air driven units. Diffused air introduced from the bottom of the tank bubbles up through the media and becomes trapped in cups on the periphery of the disks to cause the rotation. The diffused air both aerates the bulk liquid as well as causing increased mixing 41 and turbulence at the interface with the laminar water layer and biofilm. This increased turbulence and air contact allows aeration of the biofilm to occur during the submerged cycle and causes increased shear forces on the biofilm which prevents excessive buildup. The result is a thinner, more active biomass. Air drive RBCs have good resistance to oxygen depletion in the initial stages at high loading rates and Hynek and Chou observed greatly reduced growth of Beggiatoa and other filamentous micro-organisms. The air drive RBCs also permit easy regulation of rotational speed, rotate slower for a given removal rate, require less energy, and develop less stress in the shafts and media due to the thinner biomass. Nitrification was also enhanced because BOD^ removal was achieved in a fewer number of stages leaving more surface area available for nitrification in the remaining stages. Another benefit of air driven RBCs is that heat recovery from the blower air is possible in covered units. The one disadvantage of air drive RBCs for landfill leachate treatment could be that the diffused aeration would promote foaming; however, this may be controlled somewhat by the media. Having discussed the various properties and characteristics of the RBC system, the treatment capacity and performance of the RBC remains to be stated. Early design specifications for RBCs were in terms of hydraulic loading rates determined from manufacturers nomographs, using the waste strength and desired removal efficiency. This design approach did not easily adapt to differing waste types or permit simple comparisons of loading levels. Kincannon and Stover were thus prompted to introduce the total organic loading concept in the early 1970s and it has since gained wide acceptance (44,54). Therefore, RBC loading rates or treatment capacity are usually expressed as mass of substrate applied or removed, per unit area of media surface. Design specifications are also now generally expressed in terms of organic loading rates although from the previous discussions, maximum hydraulic loading rates, which are temperature dependent, should be specified to 42 ensure an adequate HRT for efficient treatment. These HRT limits would generally only apply to low strength wastewaters or low temperature nitrification. From the studies reviewed, the capacity of the RBC to remove carbonaceous substrate (BOD^) and achieve complete treatment (effluent BOD , ^ 2 5 mg/L) is in the order of 15 - 18 g B O D 5 / m 2 * d , at temperatures > 15 e C. Forgie (28) achieved a good effluent quality and removals in excess of 90% at loadings up to 15.2 g BOD,-/m 2*d and at 15° C. Kincannon ef al. (44) found that at loading rates less than 9.8 g SBOD 5 /m 2 * d , soluble effluent B O D 5 <10 mg/L were achieved, but at a loading of 18.3 g SBODj./m 2*d, which corresponds to a total BOD,- loading considerably higher, the removal efficiency was only 53%. Paolini and Variali (58) found that their effluent quality deteriorated at loading rates greater than 19 g BOD,-/m 2*d. Poon et al. (60) treating a clarified trickling filter effluent (tertiary treatment) achieved an average effluent SBOD^ of less than 15 mg/L at loadings up to 7.8 g SBODi j/m 2 *d . Murphy et al. (54) found that good treatment efficiency could be achieved up to a loading of 15 g BODg/m 2 *d after which some scatter in the results occurred. This led them to recommend a design loading rate of 15 g B O D j / m 2 * d for temperatures of 15° C or higher, which compares favourably with many other design factors and practical experience, as Evans (25) found that all the plants he surveyed had loading rates <19.5 g BODg/m 2 *d and most were operating at <12 g B O D 5 / m 2 * d . Some of the modifications or variations of the RBC system can increase the removal capacity of the RBC somewhat beyond these levels without reducing effluent quality. In particular, air driven RBCs and the use of oxygen enriched RBC systems have demonstrated higher capacities. Hynek and Chou (39) comparing air and mechanical drive RBCs found that air drive units were 18% more efficient for carbon removal and 25% more efficient for carbon removal with nitrification, but recommended designs with 7 and 5 percent higher loadings for each mode 43 respectively. Huang and Bates (38) investigated the potential benefits of enriched oxygen environments with pressurized air and pure oxygen RBC systems. As indicated by Famularo et al. (26) earlier, since the RBC is typically limited by the mass transfer of oxygen, the RBC could benefit appreciably from an oxygen enriched atmosphere as predicted by their model. Greatly increased biomass thicknesses resulted from the oxygen enrichment, particularly with pure oxygen under pressure, but increased COD removal was not consistently observed. The lack of improved COD removal in the first stage was blamed on severe bridging of the biomass, but COD removals in the second stage were increased by the oxygen enrichment. Nitrification was observed to be improved by pressurized air, but the use of pure oxygen, especially under pressure, resulted in inhibitory high dissolved oxygen levels. They concluded that oxygen enrichment would prove beneficial for COD removal if the RBC was modified to prevent bridging. The RBC system is also capable of higher removal rates if complete treatment is not required. Mikula et al. (52) found that an RBC treating dairy wastewater was capable of 71.1% COD removal at a loading rate of 38.5 g COD/m 2 *d (27.4 g COD/m 2 *d removed). At this high loading rate the fourth stage accounted for 18 - 30 percent of the total removal. Poon et al. (60) reported that BOD,, removals ranging from 17 to 28.3 g BOD^/m 2*d had been found in the literature. Higher removal rates are achieved at higher loadings by making more efficient use of the later stages for carbon removal. Since the BOD^ loadings to the later stages are increased, higher carbon removal rates are achieved at the expense of nitrification, which will be inhibited. As mentioned previously the capacity of the RBC for nitrification is determined by the surface area participating in nitrification. Nitrification then occurs at a fixed rate, established by Murphy et al. (54) to be approximately 1.2 g 44 TKN/m 2 *d at 20° C. Their design loading recommendations for both nitrification and carbon removal are presented in Table 9.1 of the Discussion. As mentioned at the beginning of this section, a review of the literature prior to the start of the experimental program failed to find any references concerning RBC treatment of landfill leachate. However a second review of the literature, conducted after the protracted experimental phase, did yield a couple of studies on this topic, by Ehrig (22,24) and Coulter (16). Ehrig reported on the treatment of three different old, or methanogenic phase, leachates and found that the RBC was capable of almost complete nitrification of these leachates at loading rates up to 2 g N/m 2*d. Coulter reported some results of a companion study to this one and found that efficient carbon removal was achieved at loading rates of 9.6 and 18.3 g COD/m 2 * d (6.2 and 11.6 g BOD 5 /m 2 *d ) . An interesting lack of nitrification was also observed during this study. These papers are discussed more fully, within the context of the results of this study, in the Discussion, Section 9.0. The scarcity of studies concerning RBC treatment of landfill leachates was confirmed by Chian et al. (13) when their review of the literature failed to find enough data to enable them to present ranges of expected treatment efficiency for aerobic fixed film processes. However, as reported by Ishiguro (40), and Masuda ef al. (48), the dearth of experience with RBC treatment of landfill leachates does not apply to the Japanese literature. Ishiguro noted that Japan has had extensive experience with RBCs, treating mostly industrial wastes, and as of 1983 had more than 1600 RBC plants installed. He also reported that there were 135 plants treating landfill leachate, the first having been installed in 1976. Therefore, it seems an effort should be made to benefit from their experience as translation is much more economical than research. 4. EXPERIMENTAL PROCRAM The experimental program proposed to fulfill the purpose of this study had three main component parts. First among these was the characterization of the Premier Landfill leachate, both to determine the constituents of the RBC influent for process evaluation, and to provide a basis for comparison of the treatment experience from this study to other leachates and waste treatment situations. Of primary interest was the carbonaceous and nitrogenous content of the leachate, as these are the main fractions removed by biological treatment; however, several other tests were conducted to determine typical levels of selected heavy metals and some specific trace organic compounds. In addition to the chemical analysis of the leachate, physical properties such as total solids, specific conductance, pH, and temperature were monitored regularly. The second part of the experimental program was concerned with the evaluation of the capacity of the RBC to effect carbonaceous removal from this leachate. For simplicity, the RBC would be operated under pseudo steady-state conditions, for which the flow rate would be set and the influent strength, temperature, etc., would be allowed to vary naturally. In order to determine the maximum capacity, or mass loading rate, of the RBC for carbonaceous removal it was proposed to begin operation at a low mass loading rate (and therefore low flowrate), and then increase the loading rate by increments until the effluent quality deteriorated, indicating an overloaded condition. The starting flow rate and size of the incremental flow increases would be determined by the leachate strength measured prior to each change in flow. After each increase in loading the RBC would be allowed to stabilize over a minimum period of three weeks. A nutrient 45 46 solution of phosphoric acid (H^PO^), and when necessary, ammonium chloride (NH^CI), would be added to the first stage to maintain a nutrient level in excess of 100:5:1 BOD,-:N:P, so that no nutrient deficiences would limit growth (4,7,18,66). The performance of the RBC would be monitored by twice weekly sampling of the RBC influent, effluent, and operating parameters. The third part of this study concerned the evaluation of the capacity of the RBC to nitrify this leachate. It was intended to set the influent flow rate such that the carbonaceous loading rate was approximately 25% of the maximum capacity determined previously, and then to vary the ammonia (NH^-N) loading rate with additions of ammonium chloride (NH^CI) . The ammonia loading rate would be doubled during each increase until the effluent ammonia levels indicated overloading ^.conditions. Then the loading rate would be adjusted downwards to find the maximum capacity. Again the RBC unit would be allowed to stabilize at each loading level before the next change was imposed. The later two parts of the experimental program outlined above would probably have worked well to provide the data necessary to evaluate the performance of the RBC if it had proceeded as planned. However, mechanical problems, natural calamity, and variable leachate strength, imposed numerous upsets and operational changes such that no orderly progression of loading rates could be maintained. In practice, the experimental program involved operating the RBC as steadily as possible during the periods between upsets. Variation of the loading rate was accomplished largely by the natural variation of the leachate strength, although changes in the influent pumping rate were also made. A drastic reduction in the BOD^:NHj ratio of the leachate during the course of the study made additions of NH 4CI unnecessary for both nutrient requirements and the evaluation of nitrification. Efforts to evaluate the carbonaceous removal capacity of the RBC were exasperated by a decline in the carbon content of the leachate. Complete details about the 47 variation of leachate strength and the operation of the RBC are given in Sections 5 and 7 respectively. In addition to the three main parts of this study, a number of smaller topics received a cursory investigation. These topics include: the generation and settleability of suspended solids, the removal and fate of some heavy metals, the presence of several trace organics, and the effects of variable and intermittent hydraulic and organic loading rates. Observations on these topics were in part based on specific test results and in part based on general operating data. 4.1 SAMPLING AND ANALYSIS PROGRAM An extensive sampling program was set up to characterize and monitor various raw leachate parameters as well as monitor the performance of the RBC (Table 4.1). The sampling program was based upon grab samples and field measurements taken during twice or thrice weekly visits to the landfill. Since the landfill is a 45-60 minute drive from the University, it was considered impractical to go to the site on a daily basis. Automated sampling was ruled out because of a lack of resources. This would have been an expensive alternative because the RBC installation was located beyond any supervision and therefore a secure enclosure for the sampling equipment would have been required to thwart vandalism, (for which there was precedent). Since the period of the study was expected to be many weeks, it was assumed that the twice weekly samples would provide sufficient data to evaluate the treatment efficiency of the RBC. The main parameters used to characterize the raw leachate were: chemical oxygen demand (COD), ammonia nitrogen (NH^-N), specific conductance (Sp. Cond.), total solids (TS), and pH. Sampling of these parameters had begun in Oct. 1982, when weekly grab samples were collected as part of another study. During 48 Table 4.1 Sampling Program - Samples to be collected twice weekly: (Tues. & Fri.) - Twice Weekly Procedures: (each site visit) - Check & Record: - influent flow rate - nutrient flow rate - influent temperature (raw leachate) - 1 s t & 4 t n stage water temperature - Sample: - influent (raw leachate) COD,TKN,NH3,Sp.Cond.,TS,pH - 1 s t & 4 t h stage (raw) COD,TSS/TVSS,pH (settled) COD,TSS,TVSS,TKN,N H 3 , N O B (filtered) COD - Once Weekly Procedures: (in addition to above) - Sample: - influent (raw leachate) BODj.,TOC,aIk. 1 s t , . 2 n d , (raw) B O D 5 & 3rd stage (settled) BOD 5 ,TOC (filtered) BOD^alk. - 4 t h stage (raw) BODj (settled) BOD 5 ,TOC (filtered) BOD 5 ,P0 4 ,a lk. - Measure: - D.O. levels in all stages - depth of biological growth on all stages - Before each change in loading: - scrape off areal sample from each stage for biomass determination - collect biomass samples from each stage for nutrient (N,P), and heavy metal analysis - Periodic Samples: - collect samples of raw leachate, 1 s t & 4 t h stage liquid for metal analysis, attempt to sample at high, medium, and low leachate production rates 49 the course of this study, additional tests were performed at various times for: biochemical oxygen demand (BOD^), total organic carbon (TOC), volatile fatty acids C1-C3 (VFA), total Kjeldah! nitrogen (TKN), and alkalinity (Alk.). These analyses, when combined with the previously mentioned heavy metal and trace organic analyses, constitute a fairly thorough characterization of this landfill leachate. While the sampling and analysis of the raw leachate proceeded as outlined in Table 4.1, the monitoring of the RBCs performance did not proceed entirely as planned. The twice weekly procedures of Table 4.1 were generally earned out as proposed; however, the remainder of the sampling procedures were either performed less frequently, or discontinued. These reductions in the sampling program were caused by the operational problems alluded to earlier. The irregular operation of the RBC reduced the significance of many of these extra tests and samples, and also reduced the time available to make these tests, as maintenance procedures often took precedence. However the quantites which were measured on the twice weekly basis were the most important with respect to evaluating the RBCs performance for carbon removal and nitrification. The other supporting data, while desirable under other circumstances, was not central to the goals of this study. Thus, the RBC process was monitored primarily by sampling liquid from the first and fourth stages of the unit. These samples whether raw, settled, or filtered, were analysed for most of the same parameters as the raw leachate. ln addition, they were analysed for: combined nitrate and nitrite nitrogen (NO^ + N O j -N), total suspended solids (TSS), and total volatile suspended solids (TVSS). Field measurements for the most part consisted of recording the liquid temperature of the first and fourth stages, and making notes on visible changes in such factors as; thickness and colour of the biomass, foaming within the RBC, settleability of the suspended solids, and effluent clarity. The only significant change made to the twice weekly routine of Table 4.1 was the twice weekly, rather than weekly, measurement 50 of BODj. during the later half of the study. This change was made after it became apparent that BOD was a better indicator of process performance than COD, because of the relatively high levels of refractory COD which persisted in the RBC effluent. This refractory COD could conceivably mask the breakthrough of degradeable COD when the process became overloaded. The rest of the sampling and analysis program was carried out to varying extents in response to changing conditions and priorities. Samples from the intermediate stages, stages two and three, were collected during the first three weeks of the study and then stopped when operational problems developed. These samples, in combination with those from the first and last stages, were intended to monitor the progression of treatment through the RBC unit. The results of the first three tests, and subsequent data from sampling the first and fourth stages, indicated that most of the treatment was occurring in the first stage and that there were only slight changes in the liquid quality between the first and fourth stages. Therefore, it was decided to discontinue the intermediate sampling until the data from the first and fourth stages indicated that measureable changes in liquid quality would occur between the individual stages. The necessity of the intermediate sampling was not indicated during the rest of the study. A similar re-evaluation of priorities took place with respect to field measurements of dissolved oxygen (DO) and pH. In the case of pH measurements, an initial test was made using a laboratory pH meter which was taken to the site, but the lack of shelter and possibility of damaging the meter, ruled out its regular use. Since a reliable portable pH meter was not available, it was decided to measure the pH of the samples back at the laboratory. An initial measurement of DO levels in the RBC stages, made with a Yellow Springs Instruments Ltd. (YSI) portable dissolved oxygen meter, indicated levels approaching saturation except for the first stage which was about 1 mg/L less. 51 These results were obtained under light organic loading conditions of approximately 4.1 g COD/m 2 -d. Measurements during subsequent weeks were interrupted by operational problems. The DO measurements were then discontinued as one step to streamline the sampling program until steady operation could be achieved. While DO levels indicate the extent to which the oxygen transfer capability of the RBC is being used at a given loading level, and indicates inhibitory or limiting conditions, this information was of secondary importance in this study as effluent quality was used as the prime indicator of process performance. Therefore, it was decided that the measurement of DO levels in the RBC would be discontinued until limiting conditions were approached as indicated by the effluent quality or changes in the colour of the biomass. In practice, steady-state limiting conditions were not indicated, and no further DO measurements were made during the course of this study. The sampling program for TOC, alkalinity, and effluent orthophosphate (P0 4), should also be elaborated upon. The raw leachate and RBC samples were analysed for TOC during the first half of the experiment in order to establish a correlation between this parameter and the COD and BOD results. Once sufficient data had been collected to show whether or not such a correlation existed, the TOC analysis was discontinued as originally planned. Alkalinity on the other hand was only monitored during the second half of the study when greater emphasis was placed on evaluating the performance of the RBC for nitrification. The weekly checks of the effluent orthophosphorus levels were done quantitatively during the early part of the study, but this was later reduced to a qualitative check, and finally the frequency of these checks was reduced to approximately monthly. This reduction in sampling was justified on the basis of the previous results, which indicated that an excess of P O . was consistently maintained in the system. 52 Sampling of the biomass attached to the RBC disks consisted of one sample from all four stages, and two samples of just the first and fourth stages. This small number of samples reflects the practical limitations which were imposed upon the significance of the analysis on those samples. Once the biomass became established, it was quickly realized that measurements of the biomass thickness, or determining the weight of areal samples, would not yield good estimates of the total biomass, particularly in the first stage, which is the most important. This was because samples could only be taken off of the external fibreglass disks of each stage, and the growth on these disks differed from that of the internal mesh disks. During light loading conditions, the biomass appeared to grow preferentially on the fibreglass disks. Under heavier loading, the thick growth on the first stage was patchy on the external disks, and considerable bridging of the biomass occurred between the internal disks. In addition, the rationale for determining the total biomass was removed because the variable loading conditions made it impossible to relate the amount of growth to the availability of substrate. Therefore biomass determination was discontinued after one sample. The other two samples were taken for analysis of nutrients and heavy metal accumulation, checks which, for the purposes of this study, did not warrant further samples. 4.2 SAMPLING PROCEDURES The sample collection and preservation procedures used during the course of this study generally followed those recommended by Standard Methods (72). During each sampling visit to the landfill site, grab samples of the raw leachate and liquid from various stages of the RBC were taken. A bucket and rope were used to hoist a quantity of the raw leachate from the bottom of the North lift station wet well. Usually the temperature of the 53 leachate was determined with a thermometer, and then two samples of the leachate were taken from the bucket. A 2 L sample was taken for COD, BOD^, Sp.Cond., pH, and TS analysis, and a 500 mL sample was preserved with 1 mL of concentrated r ^SO^ for TKN, NH^, and TOC analysis. During the second half of the experiment, a 60 mL sample was collected for volatile fatty acid (VFA) analysis. Periodic checks made during the course of this study showed there was no detectable difference between leachate samples from the wet well and samples taken from the influent line to the RBC. Samples of the RBC liquid were collected by dipping from each stage to be sampled with a 125 mL plastic beaker. First 500 mL of the stage liquid was collected for COD, BOD,-, TSS, TVSS, and pH measurements. Then 1 L was withdrawn to be settled for 30 minutes in a 1 L graduated cylinder. The supernatant was then carefully poured off to provide the settled samples: 500 mL for COD, B O D 5 , TSS, and TVSS tests; 125 mL preserved with H 2 S 0 4 for TKN, NH^, and TOC; and 50 mL were filtered with a test tube plunger type filter and preserved with phenylmercuric acetate (CH^COOHgCgH^), for N 0 2 + NO^ analysis. Filtrate of the raw and settled samples from the TSS tests was used for the determination of soluble C O D and BOD,, as well as residual orthophosphate (P0 4 ) . The temperature of the 1 s t and 4 t n stage liquid was also usually recorded. The biomass samples were taken by momentarily stopping the rotation of the disk and then scraping a square patch of biomass off one of the fibreglass endplates of each stage of interest. A 3 inch (7.6 cm) wide metal paint scraper was used to remove the biomass from a 3 inch square (58.1 cm 2 ) area. The biomass was scraped into a previously acid washed and tared glass jar (8 oz. wide mouth jar). Then the samples were dried at 104° C and reweighed to determine the dry weight of the biomass per unit disk area for each stage. Additional biomass sample for nutrient and metal analysis was collected and dried in a similar 54 manner, and then ground to a fine powder and stored for analysis. The acid preserved samples of leachate, and settled first and fourth stage RBC liquid, from several dates were saved until the end of the study for heavy metal analysis. Samples were selected to be representative of high and low leachate strength and leachate production rate. In addition, some samples were taken specifically for metal analysis and were preserved with H N O ^ as prescribed by Standard Methods. O n three occasions samples of leachate and RBC effluent (settled 4 t n stage) were collected for a CC/MS scan of trace organic compounds, as an aside to this study. These samples were collected in clean, oven-dried glass vials with teflon caps. The vials were filled completely leaving no head-space or bubbles. 4.3 ANALYTICAL PROCEDURES The analytical methods used for this study were except as noted below from Standard Methods 1 5 t h ed. C O D - The potassium dichromate reflux method as per the 1 3 t n ed. of Standard Methods was used as this method has been adopted as a lab standard at U.B.C.. TOC - Acidified samples (pH<2) were analysed using a Beckman 915A Carbon Analyser. 55 BODj. - The 5 day BOD was determined using the dissolved oxygen probe method. Probe calibration was by Winkler titration. Dilution water was seeded with settled RBC solids or unsettled RBC effluent. TKN - Samples preserved with acid (pH<2) were analysed using a Technicon AutoAnalyser II and Technicon industrial method no. 325-74W. NH^ - Samples preserved with H^SO^ (pH<2) were analysed using the automated phenate method on a Technicon AutoAnalyser II, a tentative standard method (15 t n ed.) N O 2 + N O 3 - Samples preserved with phenylmercuric acetate were analysed using the automated cadmium reduction method on a Technicon AutoAnalyser II, a tentative standard method (15 t n ed.) Total Solids (TS) - as per 15 t n ed., 80 mL leachate samples in triplicate Total Suspended - as per 15™ ed., RBC samples in duplicate, Whatman 934-AH Solids (TSS) glass microfibre filters Total Volatile Suspended Solids (TVSS) as per 15 t h ed. 56 Specific Conductance (Sp. Cond.) Alkalinity (Alk.) - measured using a Radiometer Model CDM3, Conductivity in /zS/cm - titration to pH = 4.5 as per 151*1 ed. pH - laboratory pH meter Metals - Total Metals samples prepared as per 15 t n ed. and analysed on a Jan-ell Ash #810 Atomic Absorption Spectrophotometer (AA), flame method used except for lead (Pb), (graphite furnace). - Selected samples were sent to the Environmental Protection Service laboratory for an Inductively Coupled Plasma (ICP) metal scan 57 Volatile Fatty Acids - The analyses for Volatile Fatty Acids C2-C4 were performed (VFA) on a Hewlett-Packard 5750 Gas Chromatograph equipped with a flame ionization detector and using helium as the carrier gas. A 6 ft. by 1/4" O.D. and 1/8" I.D. glass column packed with 0.3% Carbowax/0.1% H 3 P 0 4 on 60/80 Carbopack C (supplied by Supelco Inc.) was used. The column was conditioned as specified on the instructions supplied with the packing. Quantification was by the external standard method using reagent grade standards dissolved in 0.1% aqueous phophoric acid. Samples were stored in 60 mL plastic bottles and preserved by freezing. - The analyses for specific volatile and semi-volatile trace organics were performed on a Hewlett-Packard 5985B Gas Chromatograph/Mass Spectrometer. A purge and trap method was employed in which the samples were purged with an inert gas (helium) and the volatiles then trapped onto an adsorbtive material (Tenax-GC/Chromosorb-101). The trap was then backflushed into the gas chromatograph column and the GC/MS analysis started. Samples were collected in 40 mL glass vials without headspace, and stored at 4° C until analysis, which was within 5 days. Organic Compounds 5. LEACHATE QUALITY 5.1 LEACHATE GENERATION Landfill leachates are complex wastewaters which reflect the unique circumstances of their formation in their varied chemical and physical properties. One of the first realizations of leachate researchers was that the composition of leachates varied widely from landfill to landfill such that it was impossible to describe a typical leachate. Table 5.1 shows the ranges of observed values for some leachate characteristics assembled from the literature by Pohland (59). Similar tables have been compiled by many other authors and generally also show a wide variation in leachate composition between sites. It was soon recognized that each landfill site had a different combination of the many factors which were thought to affect the nature of the leachate produced. Climate, types of wastes and their relative amounts, landfilling methods, compaction density, soil types, hydrology, site dimensions, collection system layout; these are just a few of the many parameters involved. In addition to the site to site variation of leachate characteristics attributable to physical differences, early comparisons of leachate data showed clearly that the nature of a landfill leachate changed with the increasing age of the landfill (11). Leachates from relatively new landfills (receiving wastes for less than 5 years), and research lysimeters, usually had very high concentrations of degradable organics (BOD), and high levels of ammonia (NH^-N), and heavy metals (relative to domestic sewage). These were labelled young leachates. Landfills which had been in operation for more than 10 years generally produced leachates with very low BOD 58 59 Table 5.1 Variability of Leachate Composition*1* pH 4.9 8.4 Total hardness (mg/L as CaCO^) 30 13,100 Total alkalinity (mg/L as CaCO^) 100 20,805 Total iron (mg/L) 2 1000 Sodium (mg/L) 85 1805 Potassium (mg/L) 28 3770 Sulphate (mg/L) 24 1220 Nitrate nitrogen (mg/L as N) 5 196 Ammonia nitrogen (mg/L as N) 0.2 1106 Chemical oxygen demand (mg/L) 246 750,000 Biochemical oxygen demand (mg/L) 5.9 720,000 Total volatile acids (mg/L CHjCOOH) <100 10,000 Total dissolved solids (mg/L) 1740 11,254 (1) from table 1. of Pohland (ref. 59) concentrations, considerably higher ammonia levels, and variable heavy metal concentrations. These leachates were called old leachates. The reduced condition of leachate constituents coupled with the observed production of large volumes of methane gas (CH^), led to the conclusion that solid wastes within a landfill were stabilized over time primarily by anaerobic microbial processes. Therefore, landfills are now generally conceptualized as large anaerobic batch digesters in which the infiltrating precipitation provides both the transport phase for leaching and mobilizing contaminants, as well as the moisture necessary to promote biological activity. Numerous lysimeter studies have examined the nature of the decomposition process within landfills; the interaction of various physical parameters with the biological processes, and the resulting leachate quality (12,13,17,63,75,77,78). These controlled studies in conjunction with more detailed and long term observations of full scale landfills have resulted in a basic understanding of landfill evolution and some cause and effect relationships. It has been observed that a landfill generally progresses through 5 identifiable stages between first use and final stabilization (13). 60 The two most important and dominant phases with respect to leachate quality are the acid formation and methane fermentation phases. The acid formation phase of a landfill's life becomes established quite quickly after the field capacity of the fill, or a zone of the fill, is exceeded and moisture begins to move through the wastes. With the onset of water movement, conditions become ideal for the growth and spread of an anaerobic microbial culture. The first group of bacteria to establish themselves are the facultative acid formers. These bacteria degrade the larger organic compounds, found dissolved in the pore water, or hydrolysed from the wastes, down to simple organic acids, hence their name. Acetic acid (CH^COOH) is the main catabolic end product of these microorganisms during anaerobic fermentation. Some of the acetic acid then undergoes condensation reactions, or is combined by other bacteria, to produce the other volatile fatty acids (VFAs) of higher order which are commonly found in leachate (such as propionic C3, and butyric C4) (11). These acids are produced in large quantity and their concentration in the leachate draining out of the wastes can range to over 10,000 mg/L. At such high levels it is not surprising that the leachate generally achieves its highest organic strength during this phase, and that the VFAs normally account for a large proportion of the total organic strength of the leachate. In terms of total organic carbon (TOC), the VFAs often represent 80-95% of the total value (32,64). The remaining fraction of the TOC is usually made up largely of refractory humic and fulvic acids (11). Since the VFAs are readily biodegradable under aerobic conditions, they exert a strong oxygen demand and generally also account for almost all of the BOD of the leachate. Production of these acids also reduces the pH of the pore water, or leachate, which increases the solubility of most heavy metals. Therefore metal levels in the leachate are usually highest during this phase. Low pH conditions also inhibit the growth of other types of bacteria, notably the methanogens, and thus the acid 61 formation phase tends to be self propagating. Lysimeter studies have identified a number of management options or physical conditions which prolong the acid formation phase within a landfill. Placement of the wastes in thick layers, shredding of the wastes, high compaction densities, and low moisture inputs, have been shown to promote acid formation (63,75). All these factors tend to reduce the movement of leachate through the wastes and therefore would maintain the low pH conditions and inhibit the development of the methane bacteria. Stegmann ef al. (75) observed that in a limiting case in which there was no moisture movement (65% moisture content), that an acid conservation effect, such as occurs in silage, took place. Leachate recycle was also observed to prolong the acid formation phase by maintaining high acid levels in the leachate moving through the wastes. However, leachate recycling also increased the rate of waste stabilization and intensified the activity of the methanogenic phase which followed (13,64). Factors which hasten the end of the acid formation phase and the onset of the methane production phase are generally the converse of those mentioned above i.e., high moisture inputs, etc.. Stegmann ef al. also found that the placement of an uncompacted and/or aerobically composted bottom layer significantly accelerated the onset of methanogenesis. Similarly Robinson and Lucas (67) found that an unsaturated soil zone beneath the fill rapidly developed a population of methane bacteria such that VFAs produced in the wastes were not observed to penetrate through the layer. (This result was probably aided by the low leachate production rate and therefore long detention time in both the fill and the soil zone at this site.) Thus, the duration of the acid formation phase in a landfill is also a function of all the site specific conditions mentioned earlier, and has been observed to vary from less than one year to more than 10 years. In addition to the organic carbon and heavy metal content of the leachate, high concentrations of nitrogen compounds are usually present. During the anaerobic 62 degradation of the organic material in the wastes, the organic nitrogen component is rapidly reduced to the ammonia form. Since the growth rate and therefore nutrient requirements of the anaerobic bacteria are relatively low, very little nitrogen is assimilated by the bacteria. Therefore, the ammonia passes readily through the wastes in the leachate. As it is the same degradation process which produced the high concentrations of VFAs, high concentrations of ammonia (over 1000 mg/L) can also be produced. The dissolved solids levels in the leachate during this phase are also generally very high. Since the acid formation phase is established rapidly with the onset of leachate migration, this leachate contains the first flush of soluble inorganic material from the wastes in addition to the dissolved organic matter. Straub and Lynch (77) showed that the inorganic strength of the leachate decreases exponentially as the cumulative volume of water passing through the wastes increases. They found that the inorganic strength was stabilized after approximately four moisture changes through the wastes. Therefore, the inorganic material would generally be flushed from the wastes while they are in the acid formation phase, adding to the dissolved metals and organic compounds to increase the total dissolved solids observed during this phase. Although the duration of the acid formation phase may vary, the end of this phase is initiated by its very beginning. The simple acids produced by the acid forming bacteria are the prefered substrate for various other bacteria, most importantly the methanogenic bacteria. While growth of these bacteria may be inhibited by the low pH conditions produced by the acid formers, gradually the population of methane bacteria establishes itself and eventually balances the activity of the acid formers. When the balance point is reached, the methane forming bacteria consume most or all of the organic acids produced by the acid formers and thus the VFA content of the leachate is drastically reduced. This marks the 6 3 establishment of the methane fermentation phase. It is frequently observed that the transition between the acid formation and methanogenic phases is relatively abrupt, particularly where leachate recycling is practiced. This would tend to indicate that the methane bacteria population develops and spreads, but at reduced activity, until inhibitory conditions moderate enough to permit a rapid exploitation of the available substrate. From the previous discussion it follows that with the virtual removal of the organic acids from the leachate, the organic strength of the leachate is drastically reduced from that of the acid formation phase. Leachates from landfills with a well established methanogenic phase typically have a very low BODg concentration (<100 mg/L), although the C O D may remain significantly higher due to the refractory compounds. The establishment of the methane bacteria also affects the pH and ORP conditions within the landfill and leachate. As the acids are consumed, the pH rises to approach neutral values. The leachate shifts from a volatile-acid buffered system, to a predominately bicarbonate buffered system. ORP values generally decrease gradually prior to the rapid growth of the methane bacteria and reflects the development of conditions favourable to the growth of these obligate anaerobes. These two conditions combine to greatly reduce the mobility of most heavy metals during this phase. The higher pH levels reduce the solubility of the metals, while the low ORP conditions encourage metals which are dissolved to precipitate as sulphides. Therefore, heavy metal levels in the leachate are generally much lower during this phase. A possible exception to this trend is lead (Pb), which forms a stable complex with humic substances and thus remains mobile (32). The reduction in dissolved metals and organic acids is reflected in the reduced concentration of dissolved solids and value of the specific conductance. During the acid formation phase, heavy metals and organic acids constitute a major 64 portion of the dissolved material. Therefore, the immobilization and removal of these materials during the methanogenic phase results in significantly lower concentrations of both dissolved solids, and charged species which contribute to the conductance. While the concentration of almost every other constituent of the leachate is reduced markedly with the onset of the methanogenic phase, the concentration of ammonia generally remains constant or even increases slightly. This reflects the fact that degradation of the wastes is continuing at similar rates as occurred during the acid formation phase. The pathway over which the organic carbon leaves the landfill (as methane CH^) may have changed, but the fate of the ammonia produced remains the same. If the rate of water movement has decreased by this time due to the increased depth of the fill, or placement of the final cover, the concentration of the ammonia in the leachate may be observed to increase over time. This persistence of high ammonia concentrations in landfill leachates over very long periods of time (until the wastes are fully stabilized), has led increasing numbers of researchers to conclude that the ammonia content of leachate is a more serious and difficult problem than the organic carbon content (22,64). Unlike the acid production phase, the methanogenic phase does not end abruptly but rather fades out as the stabilization of degradable material is gradually completed. The methane fermentation phase is also less stable than the acid formation phase and subject to upset. Jasper ef al. (41) observed that for a landfill with a short hydraulic retention time, and subject to large water inputs, that wash-out of VFAs occurred periodically, coincident with major rainfall events, after the methanogenic phase had become established. This further supports the concept of a landfill as being a large anaerobic digester subject to similar constraints such as hydraulic overloading. However the literature indicates that at most landfills conditions are more moderate, and once the methanogenic phase is established, it is usually quite stable and the breakthrough of VFAs is not observed. 65 The discussion thus far has described the affects on leachate quality of a shift from the acid formation phase to the methane fermentation phase within a landfill undergoing stabilization. Due to the numerous factors which affect the stabilization process, there are no specific parameter values which define these two phases but rather they show to varying extents the characteristic changes mentioned above. 5.2 AFFECT OF WATER INPUTS ON LEACHATE QUALITY When the rainy season begins in the Fall, wastes placed during the Summer are rapidly soaked to their field capacity and the top layers of the landfill can become almost saturated with each new rainfall. Additional water inputs increase the hydraulic flux within these top layers, conceivably forcing the water to move faster through existing pathways in underlying unsaturated layers, as well as opening up new paths, saturating more wastes, and exposing more surface area to the water. In less dense wastes the former mechanism would probably predominate, leading to a heavy flush of pollutants, followed by reduced concentrations due to the reduced contact time with the wastes. Within dense wastes the later mechanism would dominate , leading to increased concentrations of pollutants as more wastes were exposed. Saturated zones below the watertable, or perched higher in the landfill, could also lead to higher concentrations of pollutants due to greater contact with the wastes. The residence, or contact time, of the water with the wastes affects the leachate strength by varying the length of time which chemical and biological processes have to concentrate soluble products in the passing water (dilution). Residence time can also affect the ability of other chemical and biological processes to remove soluble constituents from the leachate. Jasper ef al. (41) observed that 66 the concentrations of organic constituents, TOC, BOD, COD, VFA, and VSS increased with increasing leachate production or water inputs. It was theorized that these increased concentrations came about because the increased water contact with the wastes, combined with a shortened leachate retention time, overloaded the methane bacteria and resulted in the wash-out of organic material. They also observed that the nitrogen content, NH^ & TKN, as well as TIC, CI", Alk., and Sp. Cond. levels decreased during peak leachate flows. These parameters are generally less affected by biological activity and more affected by the exposure of wastes to the water and dilution. It was noted that the product of the leachate flow and parameter concentration or value, increased with increasing flow, supporting the notion that greater contact of water with the wastes was occurring. For the remaining tested parameters, metals, pH, TC, TP, TSS, and TDS, concentration was relatively independent of the rate of water input. Results from the monitoring of another landfill assumed to have a long leachate retention time (3 to 4 months), indicated that the levels of all parameters are relatively independent of water inputs. Other researchers have observed similar variability of leachate strength with water input. Bull (7) also indicates that heavy rainfall may cause an increase in leachate strength by reducing the residence time of leachate within the fill. However, Raveh ef al. (63) observed that for their lysimeter study, the concentration of pollutants in the leachate was independent of the level of water application up to 1100 mm of water per year. They speculate that retention time was not limiting in their case which allowed pollutants to concentrate to their saturation level. Therefore, while the variation of concentrations of pollutants may be variable with respect to water inputs, it is now generally held that the amount, or mass, of pollutants leached from a landfill increases with increasing water flow (17). Considerable effort has been applied towards formulating a mathematical model capable of simulating the production of landfill leachate. Such a model would 67 be invaluble to help explain the interaction of the many physical, chemical, biological, and hydraulic influences on the concentrations of the various leachate constituents and to aid in the design of leachate control measures. So far, these efforts have resulted mainly in hydraulic models to estimate leachate volumes, and simplified empirical models which can be made to fit observed data by varying coefficient values. Such models are useful tools for the analysis of historical data and can help identify which mechanisms are important in the leaching process (78). The work of Straub ef al. (78) is a case in point. Their model indicated that high moisture flow rates increased the relative importance of water movement and decreased the importance of microbial activity, which agrees with the previously discussed notions of leachate flow and residence time. While empirical models can yield useful insights into the leaching process, a mechanistic model would be more useful for predicting leachate quality. However, given the number of variables which affect leachate quality, the formulation of such a model seems an impossible task. 5.3 PREMIER LANDFILL LEACHATE A leachate sampling program was started in October 1982, just a few months after the new section of the landfill site was opened (recall Fig. 2.2), to provide data for this and other studies. The weekly samples, and later the leachate feed for the RBC, were taken from the lift station wet well and recall, were therefore already diluted roughly 50% by drainage from the unfilled portion of the site. This is one reason why this leachate would be described as weak compared to most others encountered in the literature. Column A of Table 5.2 shows the high and low values of various tested parameters for this leachate to date. A comparison of these values with the corresponding ranges of Table 5.1 shows clearly that this leachate has concentrations of the typical leachate constituents 68 nearer the low end of the given ranges. Average values were omitted from Table 5.2 because, due to the nature of the strength fluctuations described later, they are not meaningful. The dilution of the leachate by drainage from the unfilled portion of the site demonstrates how important physical site conditions such as the collection system layout are in determining leachate quality, ln this case the placement of the collection pipe within the sand and gravel underlying the site promotes rapid drainage and collection of the water from beneath both the filled and unfilled areas. Due to the lower hydraulic conductivity and extra thickness of the compacted wastes however, the drainage from the filled area would lag behind that of the unfilled area. Therefore the dilution ratio would be variable. Rapid drainage also probably means that the soil zone below the wastes is unsaturated most of the time, ln other cases, the collection system may affect leachate quality by collecting leachate from areas in one phase of stabilization rather than another, or by maintaining a saturated zone below or within the wastes. Suffice ft to say, the collection system design can greatly influence the quality of the leachate collected, as it does in this case. Another reason for the relatively low strength of the Premier leachate is the quite high moisture flux through this landfill. The wastes were placed over the fluvial gravel in comparatively thin lifts (<2 m), covering the whole area of the fill before the next lift was started. Moderate compaction densities were achieved using a small BOW-MAC compactor and/or a large bulldozer, and a thin layer of permeable cover material was placed over the wastes daily. The above method of placing the wastes increases their exposure to precipitation and promotes good drainage of water through the wastes. When subjected to the heavy annual rainfall normally received at this site, the field capacity of the wastes is rapidly exceeded and large volumes of water drain relatively quickly through the garbage. The large 69 Table 5.2 Premier Leachate Characteristics (Well #1) low - high B<2> low - high C W low - high COD mg/L 86 - 4421 263 - 1527 150 - 434 B O D 5 mg/L 44 - 3020 ( 4 ) 161 - 1035<4> 49 - 251 TKN-N mg/L 8.1 - 53.8 18.5 - 51.2 20.1 - 41.2 NH 3 -N mg/L 6.9 - 49.1 17.1 - 46.4 18.4 - 40.3 VFA mg/L (as acetic) 1 - 1470 48 - 888 5 - 108 T.S. mg/L 540 - 3595 764 - 2176 639 - 1238 Alk. mg/L (as CaC0 3 ) 288 - 782 428 - 750 350 - 673 Sp. Cond. /xS/cm 527 - 3567 1162 - 2594 1070 - 1890 pH 5.6 - 7.4 6.3 - 7.0 6.4 - 6.8 (1) Data Period A - October 22/82 to March 31/85. (2) Data Period B - April 10/84 to July 24/84. (3) Data Period C - January 18/85 to March 31/85. (4) BODr value estimated from COD. volume of the water and the resulting short contact or residence time within the wastes act to reduce the strength of the leachate produced at this site as previously discussed. Figures 5.1 A,B,C, show the variation in concentration of the primary leachate constituents from the start of monitoring in October 1982, until June 1985. These figures illustrate several interesting points about the variation of leachate strength at this landfill. First, note that the levels of all these constituents parallel each other very closely. This contrasts somewhat the results of Jasper et al. (41) as they found that the ammonia levels would decrease, and total solids levels would remain constant, during peak concentrations of organic strength and peak leachate flows. The reasons for these differences becomes clearer when one notes how the variation of pollutant concentration relates to the pattern of rainfall or water inputs. PREMIER LEACHATE CHARACTERISTICS VERSUS TIME AND PRECIPITATION 1 9 8 2 1 9 8 3 PREMIER LEACHATE CHARACTERISTICS VERSUS TIME AND PRECIPITATION z o < 20 O UJ D J F 1 9 8 4 M M 0 N 1 9 8 3 D J F 1 9 8 4 i L L J L h i J LL M M o Legend A COD X B0D5 • T. SOLIDS H Sp. Cond. X NH3-N PREMIER LEACHATE CHARACTERISTICS VERSUS TIME AND PRECIPITATION • A K O C T N O V D E C J A N F E B M A R A P R M A Y J U N J U L O C T N O V D E C J A N F E B M A R A P R M A Y J U N J U L 1 9 8 4 1 9 8 5 60 50 h40 30 D) E < 20 O Ho < 0 1 9 8 4 1 9 8 5 i II Ii • 1 i . 1 1 1 . • i 1 1 i • 1 1 . . i 1 | Legend A COD X BOD5 • T.SOLIDS H Sp. Cond. K NH3-N 73 Figures 5.1 A,B,C, clearly show that the pollutant concentrations are highest during the wet Winter and Spring months and then decrease over the dryer Summer and early Fall period. Upon closer examination it can be seen that sharp increases in leachate strength are generally preceeded by wet periods or major rainfall events. This is most noticeable during March 1983, December 1983, January 1984, and December 1984. In general, it can be observed that leachate volumes and strength responded quickly to rainfall events. The increases in concentration generally lag behind the rainfall peaks by a few days and so these rainfall peaks often correspond to sharp dips in the concentration values. This reflects the time lag between the drainage from the unfilled and filled portion of the site. Water from the unfilled portion of the site is collected more quickly than that from the filled area and therefore has an initial diluent effect. This time lag can also be seen in Figure 5.2, from Jasper ef al. (41), which shows the leachate production volumes and the mass of major pollutants released over the same period for this site. The pollutant discharges lag behind the peak leachate discharges by several days. This figure also clearly shows that the mass of pollutants discharged increases with the volume of leachate produced. Therefore, the data from this site indicates that the main mechanism governing the leachate strength is the area of contact between the wastes and the water. As noted earlier, increased water inputs increase the surface area or volume of waste in contact with the passing water. These figures also show quite well the evolution of this site and its leachate quality through the acid formation phase to the start of the methane fermentation phase. As the leachate sampling began just a few months after the first wastes were placed into the new landfill area, and near the start of the first wet season, it appears as though some of the first leachate to be produced from this section was collected. This is indicated by the very low concentrations of the first few samples. The leachate strength as exemplified by COD, rose rapidly from 64 mg/L 74 P H A S E I 450-1 r36 400 o 350 28 x m X 300 24« z t— o 250 •20 c JC 200 • 16 150 • 12 100 8 •~.5.0-• 4 OCT NOV 82 OEC JAN 83 FEB I8r MAR APR MAY JUN JUL P H A S E Z AUG SEP OCT 450T36 400 [ 32 350-LEGEND r 300 Leochott voL Mas*COO Most NH 3 Mo** T.S. 200 • 16 150-12 28 24 OT 20 OCT NOV 83 OEC JAN 84 FEB 18 150 100 50 OCT NOV 84 DEC JAN 83 FEB MAR APR MAY JUN JUL AUG SEP OCT from Jasper et al. (41), 12 8 4 0 Figure 5.2 Leachate Flow and Constituent Mass Release Premier Street Landfill 75 in October 1982 to a high of 4421 mg/L in April 1983, indicating the start of the acid formation phase. During the Summer and Fall of 1983 the leachate strength tapered off gradually to approximately 1500 mg/L C O D due to dryer conditions. Although the dryer conditions could be expected to increase leachate strength because of increased residence time and less dilution from the rest of the site, the opposite occurred, possibly due to a minimum groundwater flow beneath the site. Note that all the main leachate constituents decrease proportionally during this period. The leachate strength then rose slightly over the Winter of 1983 to about 2000 mg/L COD, which held steady through the January to March period of 1984. After that, the leachate strength decreased steadily like the previous year, except that the C O D decreased proportionately more than the other parameters. This indicates the establishment of the methane fermentation phase after less than two years. As mentioned previously, moderate VFA concentrations, pH, and high water inputs encourage the rapid development of the methanogenic bacteria. Therefore this period from March to October 1984 represents a transitional phase of leachate quality (which will be mentioned again in later discussions). Moderate rainfall during the Fall of 1984 caused the leachate strength to vary between 150 and 350 mg/L COD, with a slight increasing trend as the field capacity was re-established after a dry Summer. Then the leachate strength increased sharply in response to a heavy week of rain in December 1984, indicating a washout condition like that observed by Jasper ef al. (41). However, the landfill recovered very quickly once the normal hydraulic regime was resumed (Figure 5.1C). For convenience and clarity the data for the various major leachate parameters are presented separately in subsequent Figures (5.3 - 5.5 A,B,C). In addition, the raw data from the analyses of this leachate is included in Appendix 1. Z2/Z2 * S A l ua juo^ uoqjC3 a i e i p e a i v€'S ajnSiJ COD, B O D 5 , & TOC (mg/L) LEACHATE CARBON CONTENT vs. TIME and PRECIPITATION 2500 - i 2000 1500 H 1000 H 500 £ 30 u z o «C 20 CL O CC Q. 10 0 N 1983 0 N 1983 J J U D J F 1 9 8 4 hi J U M A M J J A S O Legend A C O D X BODp; • T O C LEACHATE CARBON CONTENT vs. TIME and PRECIPITATION 1500 - i e r a c n n o 3" 01 n t cr o 3 o O 3 re 3 < in H 3' 00 00 VI — 1000 LO Q O CQ 66 Q O CJ 5 0 0 H E 30 o z o < 20 UJ Legend A COD X BODfi 1 1 1 1 1 1 1 1 i i O C T N O V D E C J A N F E B M A R A P R M A Y J U N J U L 1 9 8 4 1 9 8 5 JL ± • i l l - , i I I i • I 1 . • • I -f O C T N O V D E C J A N F E B M A R A P R M A Y J U N J U L 1 9 8 4 1 9 8 5 00 LEACHATE NITROGEN CONTENT vs. TIME and PRECIPITATION 60 -i 1982 1 9 8 3 LEACHATE NITROGEN CONTENT vs. TIME and PRECIPITATION 60 -i CD E BOH ^ 40 H y£ 30 < 20 i ( H E 3 0 (J z o u a. IXI 1 0 H 1 r~ 0 N 1983 0 N 1983 i r D J F 1 9 8 4 M M 1 D J F 1 9 8 4 JUL 111 J LL M A M o 0 Legend X TKN LEACHATE NITROGEN CONTENT vs. TIME and PRECIPITATION Legend X Nr-h X TKN J ! ! ! , j ! , , , , O C T N O V D E C J A N F E B M A R A P R M A Y J U N J U L 1 1 9 8 4 1 9 8 5 1 r -1 1 | 1 • l| i . I I I r i I I I l i 1 1 . • lrl r -O C T N O V D E C J A N F E B M A R A P R M A Y J U N J U L 1 9 8 4 1 9 8 5 E 4500-] o co 4000-a 3600-3000--6 c 2500-o u 2000-Cu CO <3 1600-co T> 1000-o CO 600-" T O n t— u • z o <t 20 u UJ <r CL 10 Total Solids & Specific Conductance vs. TIME and PRECIPITATION i i i i i i : i i i i i i O N D J F M A M J J A S O 1982 1 9 8 3 n r 1 1 ^ 1. 11 -1 1. • • I I i O N D J F 1982 1 9 8 3 n— — r i — i M A M J J A S O Legend A Tot. Solids x Sp. Cond. 0 0 c In 00 to tu O sr -4 O l/l ai 3 Q. C/l 13 o o 3 Q. C n a) 3 o fD 3 CO OJ CO E o co 3 3000-1 e3 2500 CD S -6 c o U CL c3 CO T J — I O O O H o CO ro 2000H 1500 500 J 3 0 z o !< 20-1 UI UJ UJ 5 1 r-0 N 1983 0 N 1983 Total Solids & Specific Conductance vs. TIME and PRECIPITATION D J F 1 9 8 4 M M JjJ-4 i i J U L LLI D J F 1 9 8 4 M M Legend A Tot. Solids x Sp. Cond. i 0 o CO Total Solids & Specific Conductance vs. TIME and PRECIPITATION Legend A Tot. Solids X Sp. Cond. 85 5.4 ORGANICS Figures 5.3 A,B,C, present the leachate COD, BOD 5 , and TOC data. These figures show more clearly how these related parameters parallel each other. The close linear correlation between these values is further demonstrated in Figures 5.6 and 5.7 which show TOC and BOD^, plotted against COD respectively, along with their corresponding tables of linear regression results. The data was analysed in roughly six month intervals to indicate whether or not the relationship between the various parameters changed over the period of this study. In the case of the relationship between TOC and COD, Figure 5.6 shows that the strong linear relationship appears steady over the period of this data. The regression analysis reveals that the slope of this curve is only slightly less than would be predicted by stiochiometric considerations (0.3320 vs. 0.3750), indicating that oxidation of organic carbon accounts for a large proportion of the COD, as expected. A similar close correlation is apparent between BOD,- and COD (Fig. 5.7). The regression analyses show a slight trend toward a decreasing slope, which one would expect with increasing time, with the exception of the last interval when the slope increases markedly. This unexpected increase in slope is probably due to the contribution of the ammonia oxidation in the BOD^ test becoming more significant with respect to the low total BOD,, values. Since ammonia was quite likely oxidized in the BOD test due to the use of an acclimatized nitrifying seed, but is not oxidized in the COD test, this small difference can significantly affect the BOD/COD ratio. The interference of the ammonia oxidation can also be seen in Figure 5.8 which shows the BOD^ values plotted against the BOD/COD ratio a la Stegmann ef al. (74). A comparison of this data with that of Stegmann et al. (74), reveals that this data would lie below the results they found, but follows a similar trend of decreasing BOD/COD ratio with decreasing BOD- values. The primary reason for 86 TOC versus COD 1000-1 800-O) 600-E ^ 400-1 r -200-Legend A 10/82 to 6/83 X 7/83 to 12/83 • 1/84 to 6/84 i i i i 500 1000 1500 2000 COD (mg/L) 2500 3000 Figure 5.6 TOC vs. COD Linear Regression Results Data Group Slope Y intercept Correlation Coefficient No. of Data Points 10/82 to 6/83 7/83 to 12/83 1/84 to 6/84 10/82 to 6/84 0.2087 0.3614 0.3363 0.3320 336.5 24.08 70.90 70.36 0.9466 0.9855 0.9867 0.9823 5 40 42 87 87 B O D 5 versus COD 1 5 0 0 n 1 0 0 0 -CO E Q O CO 5 0 0 -A A A A A A A 5 0 0 1 0 0 0 1500 2 0 0 0 COD (mg/L) Figure 5.7 B O D . vs. C O D Linear Regression Results Legend A 7/83 to 12/83 i j X 1/84 to 6/84 I • 7/84 to 12/84 | B 1/85 to 6/85 2 5 0 0 Data Croup Slope Y intercept Correlation Coefficient No. of Data Points 7/83 to 12/83 1/84 to 6/84 7/84 to 12/84 1/85 to 6/85 0.6380 0.6374 0.6160 0.7504 -32.29 -4.966 8.558 -55.86 0.9180 0.9225 0.9043 0.9332 16 11 43 13 88 1 5 0 0 - . 1 0 0 0 -£ LO Q 2 5 0 0 CO B0D5 vs. BOD/COD Ratio A A A A A m A • • 0.25 0 .50 0.75 1 BOD/COD Ratio 1.25 Legend A 7/83 to 12 /83 X 1/84 to 6 / 8 4 • 6 / 8 4 to 12 /84 E 1/85 to 6 / 8 5 1.50 Figure 5.8 B O D . vs. B O D . / C O D Ratio 89 the difference between the two sets of data is the dilution of this leachate which reduces the BOD,- and C O D values by about 50%, but would not alter the BOD/COD ratio. A second difference is the higher BOD/COD ratios observed at low BOD concentrations. These are probably attributable to the ammonia oxidation mentioned above. The abnormally high ratios skew the plot to the right at the lower levels, which explains the otherwise unlikely results in which the BOD/COD ratio is >0.8, let alone >1.0. Once these two factors are considered, the BOD/COD data from this study compares favourably with the results of Stegmann et al. 5.5 VFA'S The VFA concentration is closely related to the C O D and BOD^ results and vice versa. Figure 5.9 shows the variation in the VFA concentration over the period for which they were monitored. This period covers the transistion to the methanogenic phase as indicated by the steady decline in concentration from March 1984 to October 1984. The wash-out of VFAs during the Fall and Winter of 1984-85 is also demonstrated. Figure 5.10 shows even more clearly the significant contribution that the VFAs make to the organic strength of the leachate and the reduced acid levels after the transition period, with the exception of wash-out events. Correlation plots of the concentration of VFAs versus C O D and BOD^ values (Figures 5.11 & 5.12), also show that high C O D and BOD^ levels are due in large part to the VFA contribution. The regression results indicate some scatter in the data (particularly for BOD,-, as might be expected), but still show a reasonably strong linearity. Therefore, this data conforms to the experience of other studies which show that the organic strength of a leachate is largely determined by the fate of the VFAs produced during the decomposition of the wastes (32). 10000q 100CH 10CH Volatile Fatty Acid Concentration vs. Time M A M J J A S 1984 D J F M A 1985 Legend A ACETIC X PROPIONIC • BUTYRIC B Total VFA VFA Theoretical COD vs. Leachate COD and BOD 5 2000 -1 1984 1985 92 2000 n COD versus VFA A A 500 1000 1500 VFA (mg/L) Figure 5.11 C O D vs. VFA Linear Regression Results Legend A 3/84 to 6/84 X 7/84 to 12/84 • 1/85 to 3/85 2000 Data Croup Slope Y intercept Correlation Coefficient No. of Data Points 3/84 to 6/84 0.9788 450.8 0.8672 24 7/84 to 12/84 1.8214 114.3 0.9907 48 1/85 to 3/85 1.6303 139.2 0.9643 22 93 600 n B 0 D 5 v e r s u s V F A A A A A 200 400 VFA (mg/L) A Legend A 3/84 to 6/84 X 7/84 to 12/84 • 1/85 to 3/85 600 Figure 5.12 B O D 5 vs. VFA Linear Regression Results Data Group Slope Y intercept Correlation No. of Data Coefficient Points 3/84 to 6/84 0.7158 172.6 0.8624 10 7/84 to 12/84 1.2234 75.43 0.8539 42 1/85 to 3/85 1.7396 20.19 0.9110 13 94 5.6 NITROGEN Figures 5.4 A,B,C, show the variation of ammonia -N and TKN -N over the course of monitoring period. These figures show that, with the exception of a few early values, virtually all of the leachate nitrogen is in the ammonia form, as indicated by the very small difference between the total Kjeldahl and ammonia values. It is also readily apparent that the ammonia level of this leachate is quite variable within the narrow range of values recorded thus far. The ammonia concentration was generally between 10 and 50 mg/L. From Figures 5.3 A,B,C, there are two points to note about the nitrogen strength of this leachate. Firstly, that the ammonia concentration parallels that of the other constituents very closely, and secondly, that the ammonia concentration is much lower than the values of the other parameters, particularly during the first year. Proportionally, however, the ammonia level increases with respect to the other constituents over time. During the first eight month interval, the average COD/NH^ ratio was 79.5:1, but during the final six months, the ratio was 7.8:1, roughly ten times less. This reduction is attributable to the decrease in the COD concentration from an average of 2619, to 183 mg/L over the same period, rather than an increase in the ammonia concentration. Figure 5.13 shows the changing relationship between ammonia nitrogen and COD levels graphically. This figure clearly shows how the ratio of N H ^ C O D shifts markedly during the 1/84 to 6/84 interval, which corresponds roughly to the transition phase between the acidification and methanogenic phases. A change of this magnitude in the N H ^ C O D ratio has important implications with respect to the treatment of such a leachate. Similarly, Figure 5.14 shows that the ammonia concentration is increasing with respect to the specific conductance, again due to a reduction in this later parameter. Therefore, the ammonia levels in this leachate are maintained over time, as has been the experience at most other 95 landfills. 5.7 TOTAL SOLIDS AND SPECIFIC CONDUCTANCE The results for Total Solids and Specific Conductance were closely related to each other and varied linearly with the other parameters (recall Fig. 5.1 A,B,C). Figure 5.15 and the associated linear regression results, show more clearly the correlation between these two parameters. The close correlation between total solids and specific conductance was due largely to the very low suspended solids content of the leachate, typically less than 5% (<75 mg/L), of which very little was volatile. Therefore, the total solids residue was primarily made up of previously dissolved material, including the ionic salts and organic acids which are indirectly measured by specific conductance. Periodically, in response to a sudden change in leachate flow, large chunks of biological solids would slough off of the collector pipe and be washed into the lift station wet well. These were the only incidents which increased the leachate suspended solids. The sandy soil layers beneath the wastes, through which the leachate must flow to reach to collector pipe, appear to filter most suspended solids out of the leachate. Figures 5.5 A,B,C, show quite clearly that a change takes place in the nature of the leachate with the onset of the methanogenic activity. Prior to March 1984, the numerical value of T.S. and Specific Conductance were almost identical. Beginning in March 1984, the T.S. value decreased with respect to the Sp. Cond. value until October 1984, when a new steady relationship is established. This is shown graphically in Figure 5.15, and numerically by the linear regression data. The figure and regression data show that the January to June period of 1984 was a transition period in which the slope of the relationship shifted downwards. A reduction in the total solids level can be attributed to the reduction in dissolved 96 N H 3 versus COD D) CO 1000 2000 COD 3000 (mg/L) Legend A 10/82 to 6/83 X 7/83 to 12/83 • 1/84 to 6/84 B 7/84 to 12/84 S 1/85 to 6/85 4000 5000 Figure 5.13 NH„ vs. C O D Linear Regression Results Data Croup Slope Y intercept Correlation Coefficient No. of Data Points 10/82 to 6/83 7/83 to 12/83 1/84 to 6/84 7/84 to 12/84 1/85 to 6/85 0.01264 0.01224 0.00685 0.01975 0.03363 0.217 3.703 26.50 16.26 19.63 0.9917 0.8906 0.5268 0.7082 0.4797 29 34 42 51 36 97 N H 3 vs. Specific Conductance 60 n 40-E CO x Z 20-1 Legend A 10/82 to 6/83 X 7/83 to 12/83 • 1/84 to 6/84 H 7/84 to 12/84 ffi 1/85 to 6/85 1000 2000 3000 Specific Conductance (nS/cm) i 4000 Figure 5.14 N H 3 vs. Sp. Cond. Linear Regression Results Data Croup Slope Y intercept Correlation Coefficient No. of Data Points 10/82 to 6/83 7/83 to 12/83 1/84 to 6/84 7/84 to 12/84 1/85 to 6/85 0.01711 0.01174 0.01835 0.02103 0.02464 -7.536 -0.185 -3.375 -6.174 -5.780 0.9827 0.8601 0.8828 0.9806 0.9523 30 32 36 46 35 98 species, both VFAs and heavy metals (the pH increased from 6 to 7 over this period). As shown in Figures 5.14 - 5.17, there were fairly steady relationships between the Specific Conductance and the other major leachate parameters. The relationships with the individual parameters changed during the transition phase as the landfill evolved, but apart from this brief period, the correlations were quite consistent. This raises the possibility that for situations where a similar correlation exists, the easily measured Specific Conductance values may be used for monitoring, and/or treatment process control, purposes (68). 5.8 METALS Tables 5.3 and 5.4 summarize the results of the heavy metal analyses performed on this leachate. It is readily apparent that the metal concentrations in this leachate are, like the other parameters, moderate to low in comparison with other leachates (11). Although the number of data points is small, it can be seen that the metal levels appear to parallel the organic strength of the leachate. Therefore, these results tend to confirm the reduction in metal mobility with the onset of the methanogenic phase. As has been observed frequently by others (11,73), filtered and unfiltered samples of leachate gradually changed colour, from clear or pale yellow, to a rust brown colour, as ferrous ions were oxidized to the ferric form which precipitates as a hydroxide. This colour change was pronounced despite the low concentrations of iron in the leachate. 99 Total Solids vs. Specif ic Conductance 4 0 0 0 -i 3 0 0 0 -E ^ 2 0 0 0 O if) "co n 1000 X A A Legend A 10/82 to 6 / 8 3 X 7/83 to 12 /83 • 1/84 to 6 / 8 4 Kl 7 /84 to 12 /84 ffi 1/85 to 6 / 8 5 1000 2 0 0 0 3 0 0 0 Specif ic Conductance (i|S/cm) i 4 0 0 0 Figure 5.15 Tot. Solids vs. Sp. Cond. Linear Regression Results Data Group Slope Y intercept Correlation Coefficient No. of Data Points 10/82 to 6/83 1.0847 7/83 to 12/83 0.9569 1/84 to 6/84 1.1241 7/84 to 12/84 0.7374 1/85 to 6/85 0.6314 -146.1 -58.05 -568.8 -137.7 -43.75 0.9965 0.9814 0.8747 0.9512 0.9572 34 25 31 46 35 1 0 0 COD vs. Specific Conductance 5 0 0 0-1 1 0 0 0 2 0 0 0 3 0 0 0 Specific Conductance (ujS/cm) Legend A 10 /82 to 6 / 8 3 X 7/83 to 12/83 • 1/84 to 6 / 8 4 R 7 /84 to 12/84 m 1/85 to 6 / 8 5 4 0 0 0 Figure 5.16 C O D vs. Sp. Cond. Linear Regression Results Data Croup Slope Y intercept Correlation Coefficient No. of Data Points 10/82 to 6/83 1.3281 7/83 to 12/83 0.9397 1/84 to 6/84 1.0658 7/84 to 12/84 0.4050 1/85 to 6/85 0.2408 -541.6 -270.9 -986.6 -286.4 -127.8 0.9886 0.9407 0.6712 0.6402 0.6587 33 35 36 46 35 101 BODg vs. Specific Conductance 1500-1 1 0 0 0 -E LO Q O 5 0 0 CQ A A A ^ A 5 0 0 1 0 0 0 1500 2 0 0 0 Specific Conductance diS/cm) A Legend A 7 /83 to 12 /83 X .1/84 to 6 / 8 4 • 7 /84 to 1 2 / 8 4 1/85 to 6 / 8 5 2 5 0 0 Figure 5.17 B O D . vs. Sp. Cond. Linear Regression Results Data Croup Slope Y intercept Correlation No. of Data Coefficient Points 7/83 to 12/83 0.6561 -295.2 0.9548 16 1/84 to 6/84 0.2573 -90.17 0.7294 11 7/84 to 12/84 0.1716 -81.49 0.5933 42 1/85 to 6/85 0.2128 -180.2 0.7291 12 Table 5.3 Leachate Heavy Metal Levels (AA) Leachate Samples from the North Leachate Lift Station (Well #1), Premier St. Landfill June*1) May 11 June 22 July 17 Oct 19 Nov 30 Dec 21 Jan 1 Feb 15 May 10 83 84 84 84 84 84 84 85 85 85 j . COD 3520 1527 460 377 138 264 1352 434 155 126 Ca 265 Cd 0.13 Cr <0.02 0.0068 Cu 0.4 0.026 0.013 0.023 0.057 0.019 0.033 0.047 0.010 0.01 1 Fe 185 60.7 63.2 53.5 31.6 18.6 27.7 31.8 22.4 23.9 Mg 49.0 37.3 27.0 27.9 16.2 34.1 28.4 15.2 20.3 Mn 5.98 5.18 4.28 3.54 1.77 3.68 3.30 2.24 2.30 Ni 0.01 13 Pb <0.02 0.0036 Zn 0.420 0.420 0.124 0.293 0.218 0.1 10 0.476 2.93 0.082 0.1 13 (1) from Raina, 1984, (62) note: analyses by Atomic Absorption Spectroscopy (AA), all results in mg/L. o 103 Table 5.4 Leachate Metal Analyses (ICP) Element (mg/L) .CP. Metal Scan of Premier Leachate Samples 5/11/84 7/17/84 1/21/85 5/10/85 As B Ba Be Cd Co Cr Cu Mn Mo Ni P Pb Sb Se Sn Sr Ti V Zn Al Fe Si Ca Mg Na Hardness Ca, Mg Total <0.05 0.704 0.159 <0.001 < 0.002 0.168 0.014 0.018 4.83 <0.005 <0.02 0.38 <0.02 <0.05 0.09 <0.01 1.23 0.093 0.007 0.33 I. 85 57.4 II. 7 291.0 37.4 109.0 880 1000 <0.05 0.454 0.036 <0.001 <0.002 0.141 0.008 0.015 3.7 < 0.005 <0.02 0.05 <0.02 <0.05 0.08 <0.01 0.777 0.028 0.006 0.236 0.08 48.5 8.2 138.0 28.0 84.4 460 556 <0.05 0.461 0.069 <0.001 < 0.002 0.096 0.01 0.039 2.86 <0.005 <0.02 0.64 <0.02 <0.05 0.07 <0.01 0.804 0.03 0.006 2.4 0.15 32.9 8.1 162.0 28.4 83.9 522 592 <0.05 0.342 0.138 < 0.001 <0.002 0.058 <0.005 < 0.005 2.01 < 0.005 <0.02 0.12 <0.02 <0.05 0.05 <0.01 0.528 0.026 < 0.005 0.092 <0.05 22.0 7.3 81.4 20.3 67.6 287 287 104 5.9 SPECIFIC TRACE ORGANICS An interesting adjunct to this study were the results of two samples of Premier leachate which were analysed for some volatile, and semi-volatile, organic compounds. Table 5.5 presents the results of these analyses. A number of these compounds as indicated, are on the EPA list of priority pollutants. Most of the compounds indentified are found in solvents and paint products which often find their way into landfills. Harmsen (32) conducted a more extensive analysis of organic compounds found in two leachates taken from landfills in different phases of stabilization (acidification and methanogenic). He identified many different aliphatic, aromatic, and polar compounds; some similar to those found in the Premier leachate, and in particular also observed a strong toluene peak. Harmsen also found that the concentrations of these compounds were lower in the old leachate, but not to the same extent as the volatile fatty acids (VFAs). Table 5.5 Leachate Trace Organic Content 105 Premier (Well #1) Compounds (ppb) 7/17/84 11/30/84 * Benzene 13.1 2.80 * Toluene 385.0 84.80 * Ethylbenzene 13.0 6.64 * Chlorobenzene 1.0 -* Dichlorobenzene - -m - Xylene 24.4 16.21 o & p Xylenes 20.8 18.99 1 - methylethyl benzene -n - propylbenzene Trace Trace 1,3,5 - Trimethyl benzene 1.6 n - Butylbenzene - Trace * Compounds on the EPA list of priority pollutants (-) = Not Detected Trace = <1 ppb 6. PILOT PLANT The RBC unit used for this study was a model S5 package plant manufactured by CMS Equipment Limited of Mississauga, Ontario (see Figure 6.1). Table 6.1 lists some specifications of this small unit which is rated for a maximum hydraulic load of 3400 L Id of domestic wastewater. The unit has three chambers: primary settlement, disk zone, and final settlement, of which only the disk zone section was used. The disk zone is divided into four stages, with the first having roughly twice the volume and disk area of the other three. Each set of disks consists of two outer fiberglass plates which support the interior disks made of thin plastic mesh (roughly 4 mm thick), with 10 mm square openning. The RBC unit was installed adjacent to the North leachate lift station at the Premier Street Landfill in May of 1983. A small excavation was made to sink the Figure 6.1 Photo of RBC Prior to Start-up, Showing Disk Media and Influent Pump I 106 107 Table 6.1 RBC Specifications Make & Model model S5, CMS Equipment Ltd. Disk Diameter 0.9 m No. of Stages 4 No. of Disks 36 (arranged 15,7,7,7) Disk Area 47 m 2 Disk Zone Volume (net) 245 L Surface to Volume Ratio 190:1 m 2 /m 3 Rotational Speed 6 rpm Peripheral Speed 0.29 m/s Figure 6.2 Photo of RBC Installed Adjacent to the North Leachate Lift Station 108 Finished Ground Elev. 28.50 M ink RBC Influent And Effluent Return Lines Lag Pump O n 25.60 Leachate Col lect ion Pipe Inv. 25.00 Manhole Lid Elev. 29.02 All Pumps Off 24.70 Sump Elev. 24.40 Check Valve Transfer Hose From Pipe Tee to Bucket (3/4 in. ID) Location ol Inlet Screen During Later Phase of Experiment (Inside 19 L Plastic Bucket Mounted Beside Pipe Tee) Elev. 27.00 Check Valve Location o( Inlet Screen During Initial Phase of Experiment Flygt Submersible Leachate Pumps Scale 1:31.6 Figure 6.3 Section of North Leachate Lift Station Showing RBC Connections 109 Figure 6.4 Photo of RBC Pump Inlet Screen RBC to ground level so that the plumbing and electrical connections were below ground and out of harms way (see Figure 6.2). The electrical power line was run from the power panel for the lift station to the RBC though metal conduit. This supplied power for the 1/4 HP disk drive motor, as well as the leachate (and later chemical) feed pumps. Both the influent and effluent lines for the RBC were run into the lift station through a small hole punched through the wall of the wet well. To start with, the influent line consisted of 3/8 in. (.95 cm) OD plastic tubing run inside metal electrical conduit. Effluent from the RBC flowed by gravity back into the wet well via 1.5 in. (3.8 cm) dia. ABS plastic drain pipe. The effluent 110 return flows had no significant effect on the influent leachate characteristics due to the relatively tiny volume pumped through the RBC, and the spacial separation of the inlet and return lines within the wet well (see Figure 6.3). During the previous few months of planning and preparation for this study, the leachate was quite strong (recall Figure 5.1A). The C O D varied between 2000 and 4000 mg/L. It was assumed that the C O D would not average less than 1500 mg/L over the next 6 - 8 months (the anticipated study period), so that a maximum flowrate of about 750 mL/min. (1080 L/day) would be adequate to overload the RBC. The corresponding loading rate is approximately 34.5 g COD/m 2 * d or 23.3 g BOD/m 2 *d. On this basis a Masterflex 1" peristaltic pump fitted with a no. 1717 pumphead, rated for up to 1680 mL/min. flow, was installed in the RBC to pump the leachate feed up from the lift station wet well (see Figure 6.1). The required suction lift was approximately 3.4 m (see Figure 6.3). An inlet screen was placed on the end of the leachate influent line to help prevent solids from plugging it (Figure 6.4). This leachate, like most others, is nutrient deficient (15,16), particularly of phosphorus (P). Therefore, a solution of ammonium chloride NH^Cl, and phosphoric acid H ^ P 0 4 , was added to the first stage of the RBC. The solution was initially added via a gravity fed drip system from a constant head reservoir, but later a pump was employed. Over the course of the study, the concentration and flowrate of the nutrient solution varied, (the NH 4 C I addition was later stopped), but the nutrient levels in the RBC were maintained in excess of the 100:5:1 ratio of BOD[-:N:P which is generally accepted as adequate for good bacterial growth. This level of nutrient addition is particularly generous with respect to nitrogen, in light of the findings of other studies conducted here at UBC (62,79,81,86), which found the minimum nutrient ratio for leachate treatment to be 100:3.2:1.1. A preliminary T Reg. TM, Cole-Parmer Instrument Company, 7425 North Oak Park Ave., Chicago, Illinois, 60648. 111 jar test determined a phosporous demand of about 20 mg/L P due to precipitation with dissolved metals; however, this demand was accounted for by maintaining an effluent orthophosphate concentration of generally >0.5 mg/L. The hook-up of the RBC and the mounting of its ancillary equipment was completed within three weeks and the RBC was ready for operation in early June of 1983. Various changes and modifications were made to the pilot plant and its support equipment during the course of this study, but these will be discussed in the following Section RBC Operation . 7. RBC OPERATION 7.1 START-UP The RBC was filled with leachate and went into operation in early June of 1983. Seeding of the RBC with bacteria was not considered necessary as a sample of Premier leachate examined for a microbiology laboratory course had shown a very high bacterial count. This was soon borne out by the development of a bacterial film on the disks within two weeks. Warm summer temperatures and the relatively high organic strength of the leachate during this period (recall Figure 5.1A), doubtlessly contributed to this rapid growth. As observed elsewhere (57), the initial growth on the RBC disks was quickly supplanted by a more diverse bacterial population. This transition is generally marked by a change in both the colour and texture of the biomass. Figure 7.1 shows the light taupe colour of the short lived initial growth. After a few more weeks, the growth on the first stage in particular, was much thicker, and the texture had changed from creamy smooth, to a spongy filamentous structure. As seen in Figure 7.2, the colour had also changed to a light rust colour which darkened with successive stages. The usual progressively darker brown colour of the biomass observed in sewage treatment (57), is generally augmented in the case of leachate treatment by the precipitation of iron oxides, which explains the red tinge. Thus, within about six weeks, the biomass had developed to the extent permitted by the applied loading. The rapid development of the biomass took place despite numerous interruptions of the leachate flow due to tubing failures in the Masterflex pump. During the start-up period, the affect of these interruptions was dampened because 112 113 Figure 7.1 Photo of creamy, taupe coloured, initial bacterial growth (June 1983) the first stage was connected to the large primary chamber, which acted as a reservoir. Tests showed there was considerable mixing between these two zones and the liquid was essentially homogenous. When the connection between the primary chamber and the first stage was closed off, and the leachate pumped directly into the first stage, then the flow stoppages became more problematic. This tubing problem was quite unexpected as this type of pump and silicone tubing has been used extensively at UBC without proir problems. The silicone tubing has a service life expectancy to 825 hrs. at 100 rpm according to the 114 I Figure 7.2 Photo of mature biomass growth during start-up (late June 1983) manufacturers specifications 1. Yet in this instance, at about 60 rpm, the tubing often failed to last the two to four day (48 - 96 hr.) period between site visits. Figure 7.3 shows a typical tubing failure (notice the dark leachate puddle below the pump head). Numerous adjustments such as changing the tubing completely each visit, using a different type of tubing, and using a new pumphead, were unsuccessful. Installing two pumpheads in parallel to reduce the rotational speed only doubled the frequency of tubing failures, and caused a second problem when the tubing became knotted up inside the pumphead and jammed the pump. The cause of these problems remains unclear. According to the tubing compatability data provided by the manufacturer, silicone tubing is sensitive to some substances found in the leachate such a toluene, but these materials are present in only trace amounts. This would also fail to explain the problems with other types T 1985 - 86 Catalog, Cole-Parmer Instrument Company, 7425 North Oak Park Ave., Chicago, Illinois, 60648. 115 I Figure 7.3 Photo of Pump tubing failure of tubing which have different sensitivities. Chemical compatibility also fails to explain why similar pumps used to pump the same leachate in lab scale experiments back at the university did not experience similar problems, unless some volatile component was responsible. Another possible cause is abrassion from particles and precipitates in the inlet line. Since the speed of the pump in this case was higher than most previous lab scale uses required, this may have pinched material beneath the rollers, which did not occur in previous experience. In any event, this experience suggests that the use of Masterflex pumps (or similar tubing pumps) for pumping leachate, particularly at speeds above 20 rpm, may be inappropriate in some instances. After a couple of months of trying to establish a reliable pumping regime using the Masterflex pump without success, it was decided to replace it with a Cormann Rupp Industries (CRI)'* bellows pump. This small positive displacement T Gormann-Rupp Industries, Bellville, Ohio, 44813 116 pump is rated for a maximum flow of 1730 mL/min. and was installed in the RBC on November 10, 1983. A much smaller no. 1713 pumphead was then mounted on the Masterflex pump, which was relegated to dispensing the nutrient solution. Initially the bellows pump also had problems with broken valves, and a collapsed inlet line due to suction. These problems were remedied by installing valve springs to relieve the strain on the elastic valve stems, and by installing a thicker walled tubing and check-valve on the inlet line. With these modifications the bellows pump performed very well. The valve springs in particular should be recommended for use whenever these pumps are used with the applicable poppet-valves. 7.2 THE DISRUPTIONS Scarcely a week after the new bellows pump was installed, the first mishap of what was to be a series of three major interruptions occurred. An unusually heavy rainstorm during the week of November 18, 1983, completely overwhelmed the leachate lift station and the resulting pond flooded out the RBC. Figure 7.4 shows part of the gooey aftermath of this flood. (Notice the high tide level of mud on the electrical cord). The high water level was about 16 in. (40.6 cm) above the normal water level in the RBC, and just short of the disk drive motor. As the drive motor did not stop, oil washed out of the oil bath for the drive chain was whipped up into a frothy grey emulsion, which along with the considerable amount of mud washed into the RBC, coated everything. Both the bellows leachate pump and the Masterflex chemical pump were stopped, but the Masterflex sustained the most serious damage as both the speed controller and the motor windings were shorted out. The one bright spot of this event was that the shaded-pole type motor of the bellows pump was not damaged by the dunking and only required a thorough cleaning. Therefore the bellows pump and a 117 borrowed Masterflex pump were reinstalled in the RBC just a week later. This time however, the pumps and electrical wiring were mounted on a platform above the high tide mark within the RBC. During the last six weeks of 1983, the RBC limped along, as minor problems such as the aforementioned broken valves, collapsing feed lines, and icing due to a December cold spell, caused interruptions. Then on New Years Day 1984, another unusually heavy rainstorm caused a second major flood, which again stopped the pumps, and this time also stopped the disk drive, although the motor was not damaged. Once again the bellows leachate pump only required a good cleaning, so a second bellows pump of lower capacity was ordered to replace the shorted-out Masterflex pump (the repaired original pump had been re-installed just 10 days earlier), for dispensing the nutrient solution. It was then decided to dig up the RBC and raise it 1 m, to avoid the possibility of further flooding. Figure 7.5 shows the RBC in its new position. The location of the inlet screen in the lift Figure 7.4 Photo of Aftermath of 1 s t F lood in the RBC (November 1983) I 118 station wet well was also raised to avoid increasing the suction lift (recall Figure 6.3). It took only three weeks to move and reconnect the RBC but, in the process, the RBC was drained and the biomass dried up. Therefore when the RBC was restarted on January 20, 1984, the biomass had to be re-established before the study could be continued. During February 1984, the previously mentioned poppet-valve springs and a check-valve for the inlet line arrived and were installed. To prevent further collapsing of the feed line due to the pump suction, the inlet tubing was replaced by a heavier walled 3/8 in. ID tubing. This was connected to the metal conduit such that the leachate now flowed through this conduit, and was therefore in contact with the metal. These modifications greatly increased the reliability of the leachate pumping. Also during this period, the biomass was regrowing quite rapidly despite the cool winter temperatures. However, during the first week of March 1984, the RBC was vandalized, which was the third major interruption to befall this study. Figure 7.5 Photo of RBC after being raised 1m to avoid f lood ing I 119 One of the vent covers was pried off and the drive chain derailed, which stopped the disk, and 60% of the biomass was partially dried out. The remaining biomass grew anaerobically into a thick shaggy black mass. This vandalizism sparked a string of mishaps over the following few weeks, resulting in a burnt out drive motor and the drying of the rest of the biomass. By the beginning of April however, the RBC was back in operation with a rapidly growing biomass and the disruptions were coming to an end. Figure 7.6A shows graphically the erratic operation of the RBC throughout the disruptions, (October 1983 through March 1984). The observed influent flow rate was that measured upon arrival at the site during each visit. This value was used to calculate the loading rates which prevailed at the time of sampling. The reset influent rate was that measured at the end of each site visit after maintenance procedures were performed, or the flow rate otherwise varied. An average rate for the preceeding period (between site visits), was calculated from the observed and previous reset rate values. With these terms explained, one can see that the influent flowrate was frequently interrupted during this period. 7.3 A NEW BEGINNING From April 10, 1984, to July 24, 1984, the RBC operated continuously except for one minor interruption of the leachate flow, caused by a fouled check-valve. During the first six weeks of this period the biomass re-established itself. The new biomass grew very rapidly over the dried mat of previous growth and in the first stage particularly, the new growth was very thick and shaggy. This heavy regrowth of the biomass was no doubt encouraged by the relatively high organic loading applied (averaging 14.5 g COD/m 2), and the warmer spring temperatures. The rough growth periodically sloughed off in large chunks, giving the RBC Operational History: Influent Rowrate and Loading 800-1 600 H 2 400 H 200 H a o o 3 10 17 24 31 7 14 OCTOBER NOVEMBER 1983 i ) < ! » < ) < X i ani 21 28 S 12 19 26 DECEMBER 2 9 JANUARY 1984 FEBRUARY rr 20 27 6 12 MARCH Legend X Observed Rate $ Reset Rate • Average Rate CM E »^ CD +-» (0 CC •o 36 30 25 20-16-10-5-0 i " i Legend • COD Ldg. r — r - i 1 RBC Operational History: Influent Flowrate and Loading 2000 1600-2 I O O O -500 4> X X X x- r u X A / # x x X 4 11 APRIL 1984 -1 1 1 r 18 28 2 0 MAY T 18 23 30 JUNE iX i 1 r 8 13 20 27 T— JULY X X X 18 28 1 8 AUGUST — i 1 r IB 22 29 8 12 SEPTEMBER .XX)00O< Legend X- Observed Rate $ Reset Rate • Average Rate 18 28 3 OCTOBER C M 20-16-® 10 co CC •a D EiB Legend • COD Ldg. E 2 BOD Ldg. •1 NH 3 Ldg. RBC Operational History: Influent Flowrate and Loading 1600 C E E 1000 H © CO o ^ 5 0 0 c CD 3 Legend X Observed Rate $ Reset Rate • Average Rate 2 9 16 OCTOBER 1984 23 30 0 13 20 27 4 11 18 28 1 8 18 22 29 8 12 NOVEMBER DECEMBER JANUARY FEBRUARY 1985 19 28 8 12 MARCH 19 28 APRIL CM .E CD CO cc ti> •a 35-30-25-20 H 15-10 5-0 - i r -i r Legend CD C O D Ldg. E2 BOD Ldg. •I N H 3 Ldg. 123 disks a patchy appearance, and the colour of the growth was observed to be much darker brown than usual. Figure 7.7 shows the heavy patchy growth on the first stage during this time. This type of heavy growth continued until mid May, when almost all the rough growth quite suddenly sloughed off the disks and was washed out of the RBC as suspended solids. It appears as though this sudden general loss of the biomass and its dark colour, were at least partially caused by the underlying mat of residual biomass left over from the previous vandalism episode. Since the disks had rotated intermittently during this problem period, the biomass had not dryed out completely. A dry surface layer formed which probably protected deeper layers from moisture loss. When the normal RBC operation resumed and the new growth started, it appears as though this old anaerobic layer was revitalized. This produced an anaerobic layer which was much thicker than is normally developed. The extra Figure 7.7 Photo of heavy dark growth on RBC during April-May 1984 I 124 thickness of anaerobic growth probably caused the large patchy sloughing of biomass and finally, the complete sloughing of the rough growth. It is generally viewed that one mechanism for biomass sloughing is reduced adhesion between cells in the anaerobic layer. In this case it appears as though the anaerobic layer gradually broke down until ultimately it came unglued completely. The dark colour of the biomass during this period was probably due in part to depressed dissolved oxygen levels, because of the heavy organic loading on the first stage and the greater oxygen demand of the thick biomass, and in part from the dark colour of the thick anaerobic layer showing through. The biomass which replaced the rough growth was much thinner, but of uniform thickness over the disk area, and small patches of distinct bacteria cultures could be seen. Within another week the first stage biomass had regained its light rust colour and the bacteria were more homogenous. The biomass continued to evolve during the following two and onehalf weeks from June 1 to June 19, 1984. This was indicated by poor floe settleability, and thus higher effluent suspended solids, due to the presence of fluffy filamentous floe particles. The settleability problem cleared itself up by June 19 and the RBC operated extremely well through July 24, 1984. This period of continous operation took place during the previously mentioned transition phase of the leachate quality and therefore the organic strength was decreasing. Column B of Table 5.2 shows the range of leachate composition over this period. Although the leachate pumping rate was increased, the organic loading rate decreased steadily from the 14.5 g COD/m 2 *d of April - May, to an average of 7.7 g COD/m 2 * d during July. This is shown graphically in Figure 7.6B. It can be seen that the influent flowrate was maintained much more consistently than during the previous period. The declining leachate strength, while frustrating the desired increases in loading, also gave rise to a new maintenance problem. It was 125 observed that biological growth oh the inlet screen, and within the inlet line, was increasing. Towards the end of July the fouling rate became unmanageable, such that on three successive site visits the inlet line was choked off completely. In response to this intermittent flow and loading, a large proportion of the biomass was ejected from the disks. Following this loss of solids the RBC operated fairly steadily from August 8 to September 4, 1984. Despite flow rates around 1 L/min., the loading rates were less than 5 g COD/m 2 * d and would only support a relatively thin, but healthy biomass. After September 4, another series of minor pump problems and periodic fouling caused numerous stoppages. This biological fouling problem, which did not appear during the previous year, was observed to increase from May onwards, as the leachate strength declined. It appears this problem arose because as the leachate strength declined, the leachate in the wet well, and particularly in the intermediate bucket, was able to become increasingly aerobic. Aerobic conditions, as well as increasing temperatures, greatly accelerated the rate of growth on the screen and in the lines. On one occassion in particular, the inlet screen was caked with a 0.5 in. (1.3 cm) layer of bacterial solids, which had closed off the screen to the extent that it had partially collapsed under the suction of the pump. This growth occurred within the three days since the previous visit, when both the inlet screen and inlet lines were thoroughly cleaned. Aside from the rate of growth, aerobic conditions were also indicated by the light rust brown colour of the growth on the screen, which appeared very similar to that of the first stage growth on the RBC. This colour indicates that metal precipitates (mostly iron oxides) were also adding to the fouling problem. A single sample of inlet line deposits analysed for metal content was found to be 38.9% iron on a dry weight basis. While constant cleaning of the inlet line was bothersome, the problem only became serious when the leachate 126 strength declined to very low levels (less than 250 mg/L COD). Once the leachate strength rose above 250 mg/L COD in the late fall, the fouling rate decreased sharply, and became manageable with regular maintenance. The original bellows pump which turned at 165 rpm, the highest speed available for this type of pump, wore out a crank bearing by September 14, after approximately 10 months of continuous operation. A new bearing was easily fabricated but it appeared that this was an inherent weakness of this pump. The teflon bushing could not stand up to the continuous wear at this speed. Therefore, a twin bellows pump which turned at 50 rpm was installed on November 23 . Figures 7.8 and 7.9 show the original single, and later twin, 1.5 in. leachate bellows pumps respectively, as well as the smaller 0.5 in. dia. pump used for nutrient addition. By the time the pump problems had been ironed out, the wet fall weather had restarted the leachate flows. During December there were three mini floods Figure 7.8 Photo of single bellows leachate pump (165 rpm) and nutrient pump I 127 Figure 7.9 Photo of twin bellows leachate pump (50 rpm) and nutrient pump during which the leachate level in the lift station wet well rose high enough above normal levels to float the reservoir bucket and tip out the inlet screen. When these flows receeded, the inlet screen was left high and dry, thus interrupting the leachate flow. However, these heavy leachate flows also caused the washout of organic material mentioned previously (Section 5.3), so the organic loading of the RBC increased dramatically between flow interruptions. The highest recorded loading occurred on December 21, 1984. At a loading of 32.7 g COD/m 2 * d , it was observed that there was considerable foaming in the first stage and a thin white growth covered the surface of the biomass in the first and second stages. Although the loading rate fell sharply during the following week, this event caused a noticeable increase in growth on the later stages, while the first stage growth became very thick, shaggy, and sloughed in large chunks. This growth gradually thinned out over the next two weeks, but it provided a thick, healthy, biomass for the start of the next period of relatively stable operation. 1 2 8 A second period of stable operation (two minor interruptions due to fouling) occurred between January 18, and the end of March, 1985. The leachate characteristics during this period are given in column C of Table 5.2. Figure 7.6c shows the influent flowrate variation over this period. Since the C O D of the leachate was generally less than 270 mg/L, the carbon loading rate was quite low (less than 10 g COD/m 2 * d or 5.0 g BODg/m 2*d). However, since the ammonia levels in the leachate and the influent flowrates remained high, the ammonia loading rates were highest during this period. The effective loading rates for nitrification were possibly even higher when temperature effects were taken into account. This aspect will be discussed further in Section 9.2. Therefore, the nitrification performance of the RBC during this period is of particular interest. The biomass was fairly thick and healthy looking during this period, as it had been for most of the study. Figures 7.10 and 7.11 show the colour colour gradation, and thickness, of typical healthy growth. Foaming was never a problem with this leachate. Figure 7.12 shows an above average foaming condition. The heavier foaming incidents during high loading had, at most, 6 in. of foam in the first stage, with much less in following stages. Collection of treatment data from the RBC was suspended on April 10, 1985. The end result of nearly two years of operating experience with this RBC unit was two periods of two or three months continuous operation, and numerous shorter periods of a few days or weeks. While this operational history is less than was hoped for, the data collected is interesting none-the-less. Appendix 2 contains a listing of the RBC operational history and field observations. For the most part, the RBC operated very well under difficult circumstances. The interruptions in leachate feed and loading fluctuations generally had only a minor effect on the biomass or effluent quality. Figure 7.10 Photo of healthy bacterial growth Figure 7.12 Photo showing leachate foaming in RBC first stage 8. TREATMENT RESULTS The RBC unit performed remarkably well in treating this leachate under difficult operating conditions. A good effluent quality was maintained throughout most of the fluctuations in hydraulic and organic loading. Recall (from Section 4) that the effluent for this study was taken as the supernatant from a sample of the fourth stage liquid settled for 30 minutes in a 1 L graduated cylinder. Effluent from the final clarifier zone of the RBC was not representative of the RBCs performance, as settled solids were not removed, and considerable resolubilization of organics occurred in this zone. The main sampling program occurred from May 25, 1984, to March, 1985. Samples of the influent leachate, and first and fourth stage liquid, were taken on a two or three times a week basis, except during December, when samples were taken only once a week. During the previous period of operation, from July 1, 1983, to May 25, 1984, the data is less complete because the operation of the RBC was not considered stable enough to warrant a complete sampling and analysis program (recall section 7 RBC Operation ). Appendix 3 contains the raw data from the analysis of the RBC process samples. Loading rates were calculated from the measured flow rate and leachate COD, or NHg-N, etc., concentration, at the time the sample was collected (observed rate). This assumes that the flowrate and leachate strength were constant over the previous detention time period of the RBC. At a flowrate of 1000 mL/min., the theoretical detention time is 245 minutes. In most cases, the leachate strength probably did not vary significantly over this period of time, particularly since most of the data was collected over the later half of the study when the short term variability of the leachate strength was reduced. As the flowrate of the pump 131 132 was consistent, when it was running, these assumptions seem reasonable. The main sampling program extended over both of the longer periods of steady operation, as well as many shorter periods of continuous operation. Results from the two longer periods of continuous operation demonstrate the abilities of the RBC for carbon removal and nitrification of this leachate at low to moderate loading rates. Some of the shorter periods of stable operation occurred during higher loading rates of up to 22 g B O D ^/m 2 * d , and their results are also significant. A study by Filion ef al. (27), found that a RBC would recover to steady-state conditions within about 1 hour for carbon removal alone, and 3 hours for carbon removal with nitrification, in response to a step increase in influent loading. Since these recovery times are shorter than the hydraulic retention time, or the time required for substantial changes in leachate quality, it would be expected that the RBC was essentially at steady-state during these short periods also. 8.1 CARBON REMOVAL The carbon removal efficiency of the RBC treating this leachate was very good. An effluent soluble BOD^ of less than 10 mg/L was maintained for all but a few samples and the settled effluent B O D ^ was generally less than 25 mg/L. Those few samples which did have a soluble BODj- greater than 10 mg/L occurred after the loading rate had more than doubled over the previous few days. However, not every sharp change in loading was followed by a significant reduction in treatment. On a percentage basis, the soluble BOD^ removal was usually above 95%. Most of the lower removal percentages were caused by influent BODj. values so low that the effluent BOD^ values were relatively large in comparison. Figure 8.1 shows the variation of effluent BOD^ over the main sampling period. This figure also presents the loading rates and percent removals calculated for this data. 1000q CD E LO Q O CO 100-o L U cc 10 Q O 03 10-100-, 90 80-70 RBC EFFLUENT B0D 5 VARIATION with LOADING RATE and TIME JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR 1984 1985 i r JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR 1984 1985 r10 * CN D) -5 U J !< DC CD Q < O Legend A INF BOD X SET BOD • FIL BOD 134 Figure 8.2 presents the same effluent BOD,- values versus the corresponding loading rates as a scattered plot. This figure shows that only 7 settled samples and four filtered samples exceed 25 and 10 mg/L, respectively, over this period. The high settled values generally correspond to higher suspended solids levels, while the high filtered values were preceeded by sharp loading increases or brief interruptions of the leachate flow. Figures 8.3 and 8.4 show that the BOD^ removal was occurring primarily in the first stage, and that generally less than 10 mg/L of additional BOD^ was removed across the remaining three stages. The C O D results paralleled those of the BOD^ test with a few minor differences. Effluent soluble C O D values generally ranged from 40 to 100 mg/L, as show in Figure 8.5, indicating a relatively consistent refractory component. The size of this refractory component did not appear to reflect the influent C O D values closely. This refractory component reduces the significance of the C O D removal percentages because it accounted for up to 60% of the influent C O D when the leachate was weak. Secondly, the range between the C O D values for the settled and filtered samples was, in some cases, considerably greater than the corresponding range of BOD^ values. These differences occurred when the effluent suspended solids were relatively high and the volatile component of the solids was low (less than 30%). This indicates that the settled effluent C O D contained a large refractory component from the highly endogenous suspended solids. Figure 8.6 shows the variation of effluent C O D with loading rate over time, and also presents the highly variable percent removal data. The C O D and B O D 5 data shows that the RBC could maintain efficient carbon removal at loading rates up to .roughly 15 g COD/m 2 *d, or 9 g B O D 5 / m 2 * d , treating this leachate. This rate of loading would be considered only moderate for sewage treatment. C O D results for a couple of samples indicated that the capacity of the RBC is probably considerably higher but there were no B O D 5 135 E F F L U E N T F 3 0 D 5 v s . L O A D I N G R A T E 40- i CD 30 E IO Q S 20 LU U J X X o o • X o X X X , UJ UJ X , -i cn v X £ U J L T w X 1—1 X Ul ; J «P u Legend Ei! 10J f j i ^ X 2 x 8 a ° • DATA PERIOD B U _ | X ^ D + B X J # + + ° • • X + OTHER DATA • • DATA PERIOD C 2 4 6 ft 10 B0D5 LOADING RATE (g/m^d) Figure 8.2 RBC Effluent B O D . vs. Loading Rate 1 S T & 4 ™ STAGE BOD5 1000 100H 1 }• LO Q O CQ 10 1 / v. r < 3 i f / 1-1 1 1 1 r JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR 1984 1985 APR Legend A INF BOD X #4 SETTLED B #1 SETTLED 1 s t & 4 ™ S T A G E B O D 5 138 E F F L U E N T C O D v s . L O A D I N G R A T E 200 - i X + —I 1 r-10 20 ^0 COD LOADING (g/m *d) o a Ul Ul - i cr o X • 40 Legend DATA PERIOD B OTHER DATA DATA PERIOD C Figure 8.5 RBC Effluent COD vs. Loading Rate RBC EFFLUENT COD VARIATION with LOADING RATE and TIME lOOOOq 100CH E Q O CJ 100d 10 J 100-. < > 80-o UJ cc 60-Q O O 40-N D J F 1985 140 results for comparison. The BODj. results proved to be a better indicator of the treatment efficiency because increases in effluent COD could not otherwise be attributed to an increase in biodegradeable, or non-biodegradeable, material. There was only one sample which indicated a possible overloading condition for the RBC. This sample occurred on December 21, 1984, during the organic wash-out event described in Section 5.3. The COD of the leachate rose briefly to about 1350 mg/L which produced a loading rate of 32.7 g COD/m^*d. This loading resulted in an effluent filtrable COD of 152 mg/L, about twice the normal level, (see Figure 8.5) which probably contained an increased biodegradable fraction. Unfortunately, there was no BODj. value for confirmation as this occurred over the Christmas holiday. The response of the RBC to this loading event will be discussed further in the following chapter. 8.2 NITRIFICATION The nitrification efficiency of the RBC when treating this leachate was also very good. Generally the effluent ammonia nitrogen (NH^-N) and total Kjeldhal nitrogen (TKN), were less than 1.0 and 10.0 mg/L respectively. A large portion of the effluent TKN was presumably from the suspended solids in the settled samples. Nitrification was established during the first week on August 1983, after 2 months of operation. A month later, NH^CI was added to the nutrient solution and the nitrification capacity of the RBC was exceeded, as evidenced by high effluent NHj-N levels. The effluent ammonia levels varied between 30 and 60 mg/L, while the corresponding nitrate levels were 70 to 90 mg/L. Unfortunately, the nitrification performance of the RBC could not be quantified over this period because the leachate flow was too intermittent, and the flow rate of the gravity fed nutrient solution was unsteady. 141 When the RBC was restarted at the end of March, 1984 (after the disruptions), it took another 2 months until May 22, 1984, to start nitrifying again. Complete nitrification was re-established by May 28, 1984. The NH^Cl addition had been stopped, as it was no longer required to maintain the B O D ^ N ratio below 20:1. Figures 8.7 A,B, show that nitrification of the ammonia nitrogen (NH^-N) was very efficient over the main sampling period except for a few samples. In most cases, an effluent NH^-N concentration of greater than 1.0 mg/L resulted from high loading rates, or system upset due to an interruption of the leachate feed. A number of samples from January to March, 1985, had effluent ammonia levels greater than 5.0 mg/L, which indicated that the nitrification capacity of the RBC was being exceeded. The loading rates were greater than 0.7 g NH.j-N/m 2*d at water temperatures below 10° C. Thus the effective loading rates, corrected for temperature, were probably somewhat greater. Nitrification was stopped briefly on a few days in December 1984, when the water temperature fell below 5° C. During these periods however, the effluent ammonia level was not observed to increase. This point will be addressed further in the Discussion. Figure 8.8 presents the effluent ammonia (NH^-N) and nitrate (NO^-N) levels versus the NH^-N loading rate as a scattered plot. This figure shows a trend towards increasing effluent ammonia levels with increasing loading rate. Effluent ammonia levels above 1.0 mg/L generally occurred at loading rates greater than 1.0 g NH^-N/m^d. Figures 8.9 A,B, show that as with BOD^ removal, most of the nitrification occurred in the first stage, with less than 10.0 mg/L of ammonia being removed across the last three stages, except during cold weather or heavy loading. This will be discussed further in Section 9.2. RBC NITRIFICATION PERFORMANCE E i CO O + CO 16 22 29 6 13 20 27 10 17 24 31 14 21 28 -o E CO x z Uj 0.6 < CC O z Q < APR MAY 1984 JUN JUL AUG SEP 1 \ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 • • ( • - > Y | 1 1 — f 13 20 27 4 11 18 25 1 8 15 22 29 6 13 20 27 3 10 17 24 31 7 14 21 28 APR MAY JUN JUL AUG SEP 1984 RBC NITRIFICATION PERFORMANCE 6 0 - i 5 12 19 26 2 9 16 23 30 7 14 21 28 4 11 18 25 1 8 15 22 1 8 15 22 29 5 12 OCT NOV DEC J A N FEB M A R 1984 1985 144 EFFLUENT NHo & NOq -N vs. LOADING RATE 40 - i — 30H X X X X o X m § ( xo< ° X X X>X0<p x*< 9x 3 • x D X D x m + • < o x • ~l ^ I 1 1.2 1.4 N H 3 -N LOADING RATE (g/m^d) z o 2 < Legend DATA PERIOD 6 OTHER DATA DATA PERIOD C Figure 8.8 Effluent NHL and N O - vs. Loading Rate APRIL MAY JUNE JULY AUGUST SEPTEMBER 1984 1 ° 1 & 4 1 n S T A G E N H 3 & N 0 3 -N 8.3 SUSPENDED SOLIDS In general, the suspended solids levels in the RBC were quite low, <200 mg/L. Figure 8.10 shows the variation of the suspended solids levels in the first and fourth stage. It can be seen that in most cases, suspended solids levels substantially above 200 mg/L followed loading, or leachate feed, interruptions. Effluent suspended solids, as shown in Figure 8.11, were less than 25 mg/L during stable operation, and usually less than 100 mg/L during upset. The suspended solids were generally concentrated in clumps of biomass which settled rapidly. Although the solids separation achieved over 30 minutes in the graduated cylinder was quite good, even better results could be expected from a properly designed clarifier. The suspended solids level in the RBC disk zone was observed to fluctuate more during the periods of low organic loading (<3 g BOD^/m 2*d). Under these conditions, much of the biomass, particularly in the later stages, was highly endogenous and easily sloughed off with the changes in organic and hydraulic loading. The volatile component of the solids decreased to less than 30% during these periods. In fact, the volatile proportion of the solids appeared to be a good indicator of the general health of the biomass, varying from a low of 20%, to a high of just over 80%, depending upon the organic loading rate. 8.4 METALS The results of the metal analyses, Tables 8.1 and 8.2, indicate that the RBC generally removed over 80% of the iron (Fe), manganese (Mn), and zinc (Zn), as well as 50% of the copper (Cu), and lead (Pb), and lesser amounts of other metals. Results from the analysis of a few samples of biomass scraped from the RBC disks, Tables 8.3 and 8.4, indicates that the removed metals are concentrated f 1 & 4 x n S T A G E S U S P E N D E D S O L I D S 100003 Q _ | n u n l l u r i H l l l l H u n n u i i n u | in m i nu nrj H II | n H nuiln f nil l in U ip i l H ( n n null nil I in u nil M J I I I M IIII nil lin |in nu , MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY 1984 1985 RBC EFFLUENT SUSPENDED SOLIDS _ I O O O O 3 0 J 1 1 1 1 1 1 1 1 1 1 1 1 r M J J A S O N D J F M A M 1984 1985 1 5 0 in the biomass. The result from the one sample of the inlet line deposits tends to show that, particularly for iron, precipitation is a major removal mechanism. These precipitated metals are presumably then adsorbed onto the biomass, while further removal is affected by other mechanisms such as absorption, and chelation (6,9). The results from these few samples are not conclusive as to the metal removal efficiency of the RBC, but the observed relative affinities of the metals for biological removal, and the removal rates, are similar to those found for other biological processes (6,9). 8.5 SPECIFIC TRACE ORGANICS An interesting adjunct to this study were the results of a few samples of Premier leachate and RBC effluent which were analysed for volatile and semi-volatile organics. Table 8.5 shows the results of these analyses. The results indicate that the RBC removed these compounds very effectively. However, it is not known how much of these compounds was removed by bacterial degradation, and how much was volatilized into the atmosphere. A number of the compounds indentified are on the EPA list of priority pollutants (see Table). 151 Table 8.1 RBC Heavy Metal Removal (AA) Sample (mg/L) Cu Mn Mg Fe Zn Well #1 May 11,1984 0.0263 5.98 37.24 60.7 0.420 1 s t Stage 0.0495 3.02 26.54 54.1 0.310 4 t h Stage 0.0727 2.76 29.21 60.0 0.333 % Removal -176.4 53.9 21.6 1.1 20.7 Well #1 June 22,1984 0.0130 5.18 27.0 63.2 0.1235 1 s t Stage 0.0256 2.19 23.93 3.65 0.0628 4 t h Stage 0.0088 1.61 23.68 1.56 0.0354 % Removal 32.3 68.9 14.0 97.5 71.3 Well #1 July 17,1984 0.0225 4.28 27.88 53.5 0.293 1 s t Stage 0.0085 1.29 22.49 6.25 0.392 4 t h Stage 0.0119 0.711 24.53 3.62 0.0398 % Removal 47.1 83.4 12.0 93.2 86.4 Well #1 Oct. 19,1984 0.0571 3.54 16.18 31.64 0.2184 1 s t Stage 0.0546 0.848 15.66 5.09 0.0531 4 t h Stage 0.0284 0.252 15.18 2.32 0.0514 % Removal 50.3 92.9 6.2 92.7 76.5 Well #1 Dec. 21,1984 0.0333 3.68 34.05 27.73 0.476 1 s t Stage (set) 0.0988 3.81 26.60 10.59 0.140 4 t h Stage (set) 0.1099 2.36 25.52 8.82 0.108 % Removal -230 35.9 25.1 68.2 77.3 Well #1 Feb. 15,1985 0.0104 2.24 15.18 22.35 0.0820 1 s t Stage (set) 0.0248 0.808 13.80 4.68 0.0568 4 t h Stage (set) 0.0263 0.244 11.55 2.03 0.0478 % Removal -153 89.1 23.9 90.9 41.7 Well #1 May 10,1985 0.0906 2.57 20.92 25.0 0.1662 1 s t Stage (set) 0.0131 0.75 12.44 5.07 0.024 4 t h Stage (set) 0.0126 1.13 12.64 5.92 0.0745 % Removal 86.1 56.0 39.6 76.3 55.2 Average of 4 best 53.9 83.6 27.5 93.6 77.9 % Removals note: Lead (Pb) levels for all sampli es <10. ppb 152 Table 8.2 RBC Metal Removal (ICP) I.CP. Metal Scan of Leachate and RBC Samples Element Well #1 1st 4th % Well #1 1st 4th % (mg/L) 7/17/84 Stage Stage Removal 5/10/85 Stage Stage Removal As <0.05 <0.05 <0.05 B 0.454 0.357 0.371 Ba 0.036 0.037 0.039 Be <0.001 <0.001 <0.001 Cd <0.002 < 0.002 <0.002 Co 0.141 0.022 0.011 Cr 0.008 < 0.005 <0.005 Cu 0.015 0.009 0.008 Mn 3.70 1.39 0.689 Mo <0.005 <0.005 <0.005 Ni <0.02 <0.02 <0.02 P 0.05 1.10 0.73 Pb <0.02 <0.02 <0.02 Sb <0.05 <0.05 <0.05 Se 0.08 <0.05 <0.05 Sn <0.01 <0.01 <0.01 Sr 0.777 0.61 0.598 Ti 0.028 0.022 0.024 V 0.006 < 0.005 < 0.005 Zn 0.236 0.02 0.02 Al 0.08 <0.05 0.14 Fe 48.5 6.96 3.51 Si 8.2 6.0 5.9 Ca 138.0 114.0 111.0 Mg 28.0 23.4 23.5 Na 84.4 72.9 75.2 Hardness Ca,Mg 460 382 374 Total 556 398 383 <0.05 <0.05 <0.05 18.3 0.342 0.255 0.258 24.6 0.138 0.183 0.098 29.0 < 0.001 < 0.001 < 0.001 <0.002 < 0.002 <0.002 92.2 0.058 0.023 0.027 53.5 < 0.005 < 0.005 <0.005 46.7 < 0.005 0.008 0.008 81.4 2.01 1.23 1.54 23.4 <0.005 < 0.005 < 0.005 <0.02 <0.02 <0.02 0.12 3.71 3.58 <0.02 <0.02 <0.02 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.01 <0.01 <0.01 23.0 0.528 0.35 0.348 34.1 0.026 0.028 0.030 <0.005 < 0.005 < 0.005 91.5 0.092 0.022 0.024 73.9 <0.05 0.15 0.17 92.8 22.0 7.35 8.44 61.6 28.1 7.3 6.3 6.5 11.0 19.6 81.4 58.4 59.6 26.8 16.1 20.3 12.1 12.3 39.4 10.9 67.6 51.3 51.3 23.4 287 196 199 331 212 219 153 Table 8.3 RBC Biomass Metal Levels (AA) Sample (Mg/g) Cu Mn Mg Fe Pb Zn <10.0 184.5 793.6 388780 11.3 96.5 27.9 1021.3 5230.8 255729 21.3 418.5 37.3 1824.9 14552.1 178821 42.6 1077.3 5.0 1316.7 22083.3 116667 24.0 596.7 13.2 1154.2 17121.9 57961 18.3 433.0 22.7 849.4 5696.2 159893 4.3 2901.4 10.0 869.7 19326.9 65478 5.7 4148.6 20.3 1584.2 20342.1 212554 24.6 302.2 45.7 1780.8 54206.8 108746 30.4 651.8 Inlet Line Deposits July 17/84 2nd 3rd 4 t h Stage Stage Stage Stage Oct. 5/84 1 s t Stage 4 t h Stage May 10/85 1 s t  1 4 t h Stage Stage 154 Table 8.4 RBC Biomass Metal Levels (ICP) I.CP. Scan of Selected Biomass Samples Element (Mg/g) 1st Stage 4th Stage 1st Stage 4th Stage 7/17/84 7/17/84 5/10/85 5/10/85 As Ba Be Cd Co <80. 767. <2. 8. 24. <80. 1350. <2. <3. <8. <80. 1260. <2. <3. 31. <80. 1380. <2. 5. 53. Cr Cu Mn Mo Ni 16. 18. 4940. <8. <30. 32. 32. 13000. <8. <30. 21. 24. 18000. <8. <30. 34. 30. 13300. <8. <30. P Pb Sn Sr Ti V Zn Al Fe Si Ca Mg Na 29200. 90. <20. 613. 45. 56. 403. 2060. 280000. <200. 43200. 2400. 1100. 41700. 80. <20. 885. 111. 50. 647. 4990. 186000. <200. 63900. 4100. 1500. 60300. 90. <20. 1020. 101. 61. 292. 3610. 289000. <200. 55600. 3700. 1300. 42200. 100. 20. 900. 114. 55. 655. 5070. 186000. <200. 63200. 4100. 1500. 155 Table 8.5 RBC Trace Organic Removal Premier (Well #1) RBC Effluent Compounds (ppb) 7/17/84 11/30/84 7/17/84 7/20/84 11/30/84 * Benzene * Toluene * Ethylbenzene * Chlorobenzene * Dichlorobenzene m - Xylene 0 & p Xylenes 1 - Methylethyl benzene n - Propylbenzene 1,3,5 - Trimetyl benzene n - Butylbenzene 13.1 2.80 - - -385.0 84.80 - - 1.24 13.0 6.64 1.0 . . . 24.4 16.21 20.8 18.99 Trace Trace - Trace 1.6 - -Trace * Compounds on the EPA list of priority pollutants (-) = Not Detected Trace = <1 ppb 9. DISCUSSION The results of this study, presented in the previous section, showed that an RBC could effectively treat this landfill leachate to remove degradable carbonaceous and nitrogenous material, heavy metals, and some trace organic compounds. This treatment was achieved despite difficult operating conditions of variable and intermittent loading. While these basic results are encouraging, there are a number of points or observations which warrant further discussion. Also, there are implications with respect to extrapolating this data to a full scale application. Finally, there are a number of practical aspects of the experimental program and pilot plant operation which require further comment. 9.1 ORGANIC REMOVAL As presented previously, the carbon removal efficiency, as represented by BODj, was very good. Settled and filtered effluent BOD^ values were generally less than 25 and 10 mg/L respectively. This effluent quality exceeds the most stringent provincial requirement of 30 mg/L BOD^ (19). These effluent values were achieved at organic loading rates ranging up to 8.6 g BODj/m^*d; however, the majority of the data relates to loading rates less than 6 g BODj/m^*d. At these relatively low organic loading rates, this effluent quality could reasonably be expected. The organic loading rates for sewage treatment are generally two to three times as much. Evans (25), surveyed RBC plants for operational problems and presents data showing most plants were operating at loadings between 10 and 15 g BOD^/m^*d, to a maximum of 19.5 g BOD^/m^*d. Paolini et al. (58), while investigating kinetic response of RBCs, tried loadings between 8.7 and 39.5 g BODr/m 2*d and found 156 157 the effluent quality started to deteriorate above 19 g BODg/m 2 *d. Murphy and Wilson (54), operated a 2.0 m diameter pilot scale plant at between 4 and 36 g BOD^/m 2 *d loading. They found a fairly linear response between BOD removal and loading up to a loading of about 15 g BODg/m 2 *d, after which the removal rate tended towards a maximum. Their recommendations were for a maximum design loading of 17 g BOD,j/m 2 *d, at temperatures above 15° C, to achieve an effluent B O D 5 <30 mg/L. In the only comparable leachate treatment study which could be found in the literature (with respect to carbon removal), Coulter (16) reports the results of a companion study to this one conducted in Montreal. They were treating a leachate with a BODg of 850 mg/L and ammonia levels of 60 mg/L. At their highest organic loading of 11.3 g B O D j / m 2 * d , they achieved an average effluent B O D j of 19 mg/L, or 97% removal. At this level, first stage loading was nearing capacity as dissolved oxygen (DO) levels averaged 3.0 mg/L with a 1:1 recycle from the fourth stage already being employed to reduce the oxygen depletion in the first stage. The profile of BODj- removal across the RBC indicated that extra capacity remained in the later stages, if the loading could be increased further without adversely affecting first stage removals. The relatively low organic loading rates achieved during most of this study were insufficient to accurately determine the maximum capacity of the RBC for carbonaceous removal. During the first half of the study, the leachate BOD^ was greater than 800 mg/L, which could have produced loading rates greater than 30 g BOD[j/m 2 *d at maximum flowrates. However, the operational problems recounted in Section 7, continually postponed the planned increases in loading, as it was undertaken to establish steady operation at moderate loading first. By the second half of the study, when a reliable mechanical configuration had been achieved and the other calamities overcome, the transition phase of leachate quality had begun 158 and the high influent BODj. levels were gone. At the lower BODj- values (OOO mg/L), the influent pump did not have the capacity to achieve the desired loadings. An exception occurred during the brief organic wash-out event. During this event, an estimated B O D 5 loading of 22 g BODj-/m 2*d was achieved. Although no BODj. results are available for this brief period, because the BOD testing had been suspended over the Christmas holidays, the COD results indicate effective removals at this rate, as will be discussed shortly. Figure 9.1 shows the plot of BOD^ removal rates versus the BOD^ loading rates. It is readily apparent that the BODj. removal rate was very linearly related to the loading rate over the range of this study. This follows the experience of many others (44,54). It can also be seen that temperature effects were apparently absent. This will be discussed further below. The regression results for this figure indicate that 98.4% of the applied BOD^ was very consistently removed. Figure 9.2 is a similar plot of COD removal versus loading. This figure also shows a high degree of linearity, and most interrestingly, this relationship appears to extend right up to the highest loading point, which occurred during the wash-out event. This later data point indicates a BOD^ mass removal estimated to be 18.2 g BODj7m 2*d, which is within the realm of possibility. A single point is not conclusive however, especially considering the large gap between it and the other data. The regression analyses indicates that an average of 91.5% of the applied COD was removed. Observations of the operation of the RBC during this period indicates that at this loading level the maximum capacity of the RBC was being approached. The darker colour of the biomass, and appearance of the white Beggiatoa sulphur-oxidizing bacteria (recall Section 7.3) are indicative of limiting conditions (25). Effluent COD values also indicated a somewhat reduced effluent quality. Unfortunately, this loading level did not persist long enough to assertain whether or not the results represent a steady-state condition. Since the loading rate jumped so 159 BOD 5 REMOVAL vs. UNCORR. LOADING Temperature Effects T3 * CM 2 O LLJ r r LO Q O CD 10-1 8-6-4-2 A A • X A D B O D 5 LOADING (g/m2*d) Legend A Temp. >12C X Temp. 8-12C • Temp. <8C "i 1 1 1 4 6 „ 8 10 Figure 9.1 B O D - Removal versus Loading Rate Linear Regression Results Data Croup Slope Y intercept Correlation No. of Data Coefficient Points Temp. >12°C 0.9974 -0.1078 0.9978 28 Temp. 8-12°C 0.9922 -0.1497 0.9988 9 Temp. <8°C 0.9507 -0.0223 0.9993 14 Overall 0.9842 -0.0886 0.9982 51 160 COD REMOVAL vs. LOADING Temperature Effects 30-i • 20 COD LOADING (g/m Z#d) Legend A Temp. >12C X Temp. 8-12C • Temp. <8C 40 Figure 9.2 COD Removal versus Loading Rate Linear Regression Results Data Croup Slope Y intercept Correlation Coefficient No. of Data Points Temp. > 1 2 ° C 0.9547 Temp. 8-12°C 0.8435 Temp. < 8 ° C 0.9119 -0.8687 -0.8937 -1.1268 0.9778 0.9661 0.9978 34 20 18 Overall 0.9145 -0.9258 0.9871 72 161 dramatically after a series of loading interruptions, it is likely that the RBC would require more than three days to increase its biomass a comparable amount, especially at temperatures around 6° C. Therefore, from the experience of this and other studies, the maximum capacity of this RBC unit would probably have been in the range of 15 - 18 g BOD^/m 2 *d. As pointed out by Murphy and Wilson (54,80), this would lead to design loading rates approximately 15% less to account for scale-up effects. The literature suggests that numerous modifications or variations of the RBC process can improve performance at higher loading rates. Coulter (16), reported the use of a 1:1 recycle from the fourth to the first stage, in order to reduce the influent concentrations in the first stage and thereby reduce oxygen depletion. In other cases, step feeding, destaging, and/or supplementary air diffusers, have been used in the first few or all stages to increase the performance of the RBC. Perhaps the best approach, especially for leachate treatment in which influent concentrations may be very high, is the use of air-driven RBCs. Studies have shown (39) improved perforemance for both organic carbon and ammonia removal with air-drive RBC units. These units have demonstrated greater resistance to oxygen depletion, a thinner, more active biomass, prevention of biomass overgrowth, and lower maintenance and energy costs. Figure 9.3 shows the BODj. % removal versus loading. This figure raises a number of interesting points. The first point is that the removal efficiency was generally very good, >95%. Of the results less than 95% however, all of these occurred at loading rates less than 5 g BODj./m 2*d, thus under light loading. This seemingly incongruous result has a number of possible explainations. Many of these low removal percentages result from very low influent BOD^ values, which make the effluent BOD- relatively more significant. This reason was given earlier and is especially true for the lowest % removal values. A contributing factor however, may 162 be an effect observed in other studies (57,60), in which the removal percentage decreased with decreasing influent concentration. Poon et al. (60), explain this effect in terms of mass transfer rates, which would be reduced because of the substrate limiting conditions (low driving force). The final point to note is the effect of temperature on the treatment efficiency. From Figure 9.3 it can be seen, despite the scatter, that the best removal efficiencies in each temperature group decrease slightly with decreasing temperature. This small effect is also indicated by the subtle decreasing trend of the slopes in the regression results of Figure 9.1. The COD data is more scattered and inconclusive. However, these slight temperature effects are much less than is normally observed for RBCs treating sewage. Murphy and Wilson (54,80) determined that an Arrhenius temperature coefficient of 0 = 1.05 applied to carbon removal for temperatures less than 13° C, for their sewage treatment study. Figure 9.4 illustrates the unnecessary distortion that this factor imparts to the data from this study. For this study, the linearity of Figure 9.1 indicates that essentially no temperature correction is required down to 5° C. This result is supported by a similar finding by Forgie (28), who observed a very slight temperature effect between 5 and 15° C. Coulter (16), observed a similar lack of temperature effects in that study. He found support in the literature for the notion that the reduced activity of the bacteria at lower temperatures can be offset by longer hydraulic retention times and/or high degrees of treatment, which increase the contact time of the wastewater. This notion seems eminently reasonable, and as will be discussed shortly, is especially apparent in the nitrification results. The absence of temperature effects on the treatment of concentrated wastes such as landfill leachates at long detention times, (>4 hrs.), has important design implications as required surface areas are frequently increased over 50% to account for low temperature effects. 163 B O D 5 % REMOVAL vs. LOADING o LU CC 100-. 95 A 90 A LO Q O OQ 85 80-^ 75 A X * ^XMAV A " • A Q AX • • • X X X 2 4 6 8 B0D5 L O A D I N G (g/m z«d) Legend A Temp. >12C X Temp. 8-12C • Temp. <8C 10 Figure 9.3 BOD_ Percent Removal versus Loading Rate 164 BOD 5 REMOVAL vs. CORR. LOADING Temperature Effects 10 -i •o CN* 8 O UJ 4-O 2 CO • 4 6 8 „ 10 B O D 5 LOADING (g/rri *d) X • Legend A Temp. >T2C X Temp. 8-12C • Temp. <8C i 12 Figure 9.4 BOD,. Removal versus Loading Rate Corrected for Temperature Linear Regression Results Data Croup Slope Y intercept Correlation No. of Data Coefficient Points Temp. 8-12°C 0.8876 -0.1498 0.9983 9 Temp. < 8 ° C 0.7156 0.0098 0.9981 14 165 One anticipated effect which did not occur during this study was the interference of effluent ammonia concentrations on the soluble BOD^ results. Since RBC settled solids, or fourth stage liquid, was used as seed for the BOD test, it was anticipated that effluent BOD^ values would reflect the effluent ammonia levels. However, except for one data point, there was no apparent correlation between effluent ammonia levels and effluent BODj.. Although the investigation of mathematical models of the RBC process was purposely excluded from the scope of this study, some brief comments on work in this area wouldn't hurt. Researchers have taken a variety of different approaches towards mathematically modelling RBC process performance. Some, like Wu ef al. (83,84), have developed empirical models which relate parameters such as flowrate, substrate concentration, temperature, number of stages, rotational speed, etc., to effluent quality using coefficients determined from fitting curves to historical data. A well developed empirical model can predict RBC performance very effectively for conditions similar to those used to evaluate the coefficients, but the range of parameter values for which it remains valid is likely quite narrow. This is particularly true for the effects of different waste types. Therefore, a small-scale test run should be carried out to re-establish the coefficient values whenever a new type of waste is encountered. This disadvantage is not unique to this type of model however, as all models require calibration to new situations. An advantage of this type of model is its ease of use. The parameters used are generally those normally monitored, (temperature, flowrate, influent and effluent concentrations), and fixed values of system geometry, (no. of stages, rotational speed). A disadvantage of this type of model is that it gives little insight into what is happening within the process, so that when the model fails, there is no indication of what caused the problem. In an effort to develop a more theoretical model which would reflect the RBC process dynamics, various researchers have applied several kinetic approaches. 166 The process kinetics employed are generally adapted from those used to describe suspended-growth cultures. One approach which has met with popular success is the application of Monod type kinetics (44,57,58). Equation 1 shows the RBC form of the relationship. The greatest difficulty encountered with the application of this and other kinetic approaches has been the estimation of the amount of active biomass for the substrate in question. Since the biofilm is generally conceived to be layered with the different types of bacteria (heterotrophs, autotrophs, etc.) concentrated in specific layers, the biomass is generally estimated by arriving at a thickness value and then multiplying by the disk area. Estimates of the active thickness (Z) for BOD removal have ranged from 21 to 200 /im (26,58), depending upon the substrate loading. Therefore, arriving at an estimate of the active biomass for a fixed-film process is even more uncertain than for a suspended-growth process. Kincannon et al. (44), used a simpler analogue of the Monod equation which avoids the evaluation of biomass amounts. Equation 2 presents this simplified relation. The simplicity of this relationship was compelling enough to result in Figure 9.5, from which the values of U and K D were determined to be 199.5 and max D 212.2 g BODj-/m 2 *d respectively. These results are equal to 40.9 and 43.3 lbs. BOD^/1000 f t 2 * d respectively, which are roughly 4 times the values determined by Kincannon ef al. for sewage treatment ( U m a x = 10.0 and K g = 10.4 lbs BODj/1000 f t 2 * d ) . The significantly higher values of U m a x and Kg indicate that this leachate is more readily degraded than the sewage used in the other study. This finding fits nicely with the observation that most of the BOD of the leachate was in the form of readily degradable volatile fatty acids. These results also indicate that there was no apparent inhibition of the heterotrophic activity by any characteristic of the leachate composition. 167 U A X. Q(S Q-S) Z = 4s J - + R S (D A = disk area (cm 2 ) S Q,S= substrate concentration (mg/L) X, = concentration of active biomass D i ( m g / c n r r ) U = specific substrate utilization rate (1/day) k= M max. rate substrate utilization per unit active weight of bugs (1/day) Y Q = flow rate (Ud) Z = active biomass thickness (cm) K = Monod 1/2 velocity coefficient (mg/L) — = — B — * — — + — — (2) U U QS/A IT max ^ max U = substrate utilization rate K R = Saturation Constant (g BOD 5 /m 2 * d ) (g BOD 5 /m 2 *d ) U = Maximum Removal Rate QS/A= Applied Loading (g BOD,-/m 2*d) (g BOD t ; /m 2 *d) 168 1/BOD REMOVAL vs. 1/BOD LOADING Monod Kinetics Approach UJ DC 3-i 2.5-$ 2 J O 1.5-Q O m H 0.5-A • 1/BOD LOADING • Legend A Temp. >12C P X Temp. 8-12C • Temp. <8C i 1 1 1 1 0.5 1 1.5 2 2.5 Figure 9.5 BOD_ Removal - Monod Kinetics Approach Linear Regression Results Data Croup Slope Y intercept Correlation No. of Data Coefficient Points Overall 1.0586 0.0050 0.9976 51 U m a x = 1 9 9 5 (8 B O D 5 / m2 * d ) K R = 211.2 (g B O D 5 / m 2 * d ) 169 While the kinetic models have proven to be fairly successful at modelling the RBC process, their application is limited to conditions under which the kinetics control the bacterial growth. As pointed out by Famularo ef al. (26), in suspended-growth systems, kinetics almost always governs because mass transfer resistances are negligible across the relatively small dimensions of a bacterial floe particle. This is one reason for the success of kinetic models in suspended-growth systems. However in fixed-growth systems, biofilms can be quite thick, and substrates and products must move in and out of the biofilm generally from only one direction. Under these conditions, mass transfer effects become much more important and often determine reaction rates. It is quite likely that in many instances where kinetic coefficients have been evaluated for RBCs, they are in essence macroscopic approximations of many mass transfer effects. Mass transfer models, such as developed by Famularo and Mueller ef al. (26,53), incorporate both mass transfer and biological reaction kinetics into a comprehensive, albeit complicated, process model. These models have the advantage of being applicable over a wider range of operating conditions, as well as having the flexibility to predict the interaction between different groups of bacteria; carbon oxidizers, nitrifiers, denitrifiers, etc. This later capability has yielded some valuble insights into the factors which affect the performance of the RBC system. There is however considerable work still to be done to further refine these models (31). The numerous kinetic and diffusion coefficients used require further verification, and the effects of temperature and hydraulic retention time observed in this study (and elsewhere) could be incorporated. As presented in these papers, the models accounted for temperature by simply applying an Arrhenius coefficient, which as discussed earlier, is not always appropriate. 170 9.2 NITRIFICATION The nitrification efficiency of the RBC treating this leachate was also very good. Recall that the effluent ammonia nitrogen (NH^ -N), and total Kjeldhal nitrogen (TKN -N), were usually less than 1.0 and 10.0 mg/L respectively. This effluent quality was maintained at loading levels ranging up to 1.0 g NH^ -N/m 2 *d. The literature generally agrees that nitrification proceeds at a zero order reaction rate and therefore depends only on the number of nitrifying organisms (36,54,80). As the nitrifiers are considered to be concentrated within a relatively discrete layer on the RBC disks (53), their number is proportional to the surface area of the disks. In situations where nitrification occurs in all the stages, the nitrifier population would be proportional to the total disk surface area. This explains the strong relationship between the nitrification rate and total surface area observed in various studies (22,80). However, in other studies nitrification only occurs in the later stages of the process (57). In these instances, there is no close relationship between the rate of nitrification and disk area. The deferred onset of nitrification is associated with higher organic loading rates, which result in residual soluble BOD^ levels in excess of 30 g/L in the early stages. It has been a general observation in both fixed and suspended growth systems that nitrificaton is inhibited when residual BOD,- concentrations of this order exist. The reasons for this inhibition however are not well understood. Presumably in the fixed-growth systems, the higher BOD^ concentrations cause higher growth rates in the heterotrophic bacteria, which then concentrate in the outer layers, and thus reduce or prevent the penetration of oxygen and/or ammonia into the nitrifier layer. It follows that the nitrifiers would concentrate in a discrete inner layer, where they can compete against substrate limited heterotrophs, rather than be dispersed evenly throughout the aerobic biomass. The situation in suspended-growth systems is 171 less clear. Since the bacteria are completely mixed and in intimate contact with the substrates, there should be no mass transfer limitations, and thus the nitrifiers should get a portion of the substrates available. Hockenbury (35), conducted a series of tests which tended to show there was no good reason for nitrification to be inhibited by high BOD concentrations. However in practice, the inhibition of nitrification is observed. Recall from Section 8.2, that for a few days in December 1984, nitrification was reduced to zero as indicated by the effluent nitrate levels. However, since the effluent ammonia levels remained essentially zero, the lack of nitrification was not due to the low water temperatures ( < 5 ° C). It was then realized that this apparent anomally corresponded with the organic wash-out event. When the organic loading rate of 22.0 g BODtj/m 2*d is compared to the ammonia loading rate of 1.08 g NH 3 -N/m 2 * d , the ratio of BOD 5 :N is 20.4:1, which matches that of the nutrient requirements of the heterotrophic bacteria. Therefore, despite the presence of an established nitrifier population, the hetertrophs apparently consumed all of the available nitrogen for their growth requirements, which underscores the inhibition discussed above. The issue is clouded somewhat by the low temperatures, which have been found to reduce the activity of nitrifiers more than heterotrophs; however, as shown by Forgie (28), an established nitrifier population can continue to nitrify down to temperatures as low as 1° C. Also, the long hydraulic retention times which occurred probably reduced the temperature effects, as observed for organic removal, and as will be discussed further below. In this study, considerable nitrification (generally >50%) occurred in the first stage. This nitrifier activity reflects the the low rates of organic loading and long hydraulic retention times, resulting in high first stage BOD removals and low residual BODj. concentrations. Therefore, it is anticipated that there was a strong relationship between surface area and nitrification rate for this study; however, the data is 172 insufficient to be conclusive in this area. Figure 9.6 shows the ammonia nitrogen removal rate versus loading rate relationship. It is readily apparent that the removal rates are linearly related to the loading up to the maximum loading achieved in this study, 1.3 g NH.j-N/m2*d. The regression results indicate an overall average removal of 80%, but when the outlying points are ignored (see groomed results), the majority of the data indicates a better than 9 5 % removal efficiency. From Figure 9.6, the causative effects of the outlying points are not clear. The scatter of the data at the top end of the graph (above loadings of 1.0 g NH^-N/m^d) could be due to temperature effects, the RBC nearing it's capacity for complete nitrification, or some other factors. Explanations for the few scattered points at lower loading rates were also not immediately apparent. To further investigate the factors affecting the nitrification performance of the RBC in this study, the percent nitrogen removal was plotted against loading rate and temperature. The results are Figures 9.7 and 9.8 respectively. Figure 9.7 indicates a trend towards reduced removal efficiencies at higher loading rates, as noted earlier in Figure 8.9. However, the data is again quite scattered and the effect of temperature is difficult to assess. Figure 9.8 doesn't serve to clairify the issue much as there is again considerable scatter of the data and the lowest removal efficiencies do not occur at the extremes of loading or temperature. This figure does tend to show a slight trend towards reduced removal efficiencies at temperatures less than 10° C. However, temperature effects on nitrification were much less than is commonly observed or assumed for RBC treatment, which parallels the organic removal results discussed earlier. Murphy and Wilson (54,80) determined that an Arrhenius temperature coefficient of 0 = 1.09 applied to nitrification for temperatures below 20° C. Coefficients of this magnitude have been determined or used in many RBC studies. Figure 9.9 shows that, as with the organic removal example, 173 N H 3 -N REMOVAL vs. UNCORR. LOADING Temperature Effects 1.2 CM LU < LU CC 0.8 < 0.6H > O 0.4 ' 0.2-CO X • • • x • • • Legend A Temp. >12C X Temp. 8-12C • Temp. <8C 0.2 0.4 0.6 0.8 1 J . 2 1.4 N H 3 -N LOADING RATE (g/rrT*d) Figure 9.6 NHL^ -N Removal versus Loading Rate Linear Regression Results Data Croup Slope Y intercept Correlation Coefficient No. of Data Points Temp. > 1 2 ° C 0.8006 0.0678 0.9235 29 Temp. 8-12°C 0.7999 0.0531 0.9694 17 Temp. < 8 ° C 0.7518 0.0849 0.9208 19 Overall 0.7718 0.0750 0.9433 65 Temp. > 1 2 ° 0.9822 0.0034 0.9994 26 (groomed) Temp. 8-12° 0.9461 0.0095 0.9945 41 (groomed) 174 NH 3 -N % REMOVAL vs. LOADING o 100-1 90 H 80 70 CO X 60 H 50 X • • • • • X • • 0.2 0 .4 0.6 0.8 1 J.2 N H 3 -N LOADING RATE (g/rri *d) L e g e n d A Temp. >12C X Temp. 8-12C • Temp. <8C -1 1.4 Figure 9.7 NH„ -N Percent Removal versus Loading Rate N H 3 - N % R E M O V A L v s . T E M P E R A T U R E - ^ x 1 0 0 - i o LU rr 9 0 H 8 0 Z 70H CO 6 0 5 0 x X X X X X X X X X X 5 1 0 1 5 2 0 Temperature C 2 5 Figure 9.8 NH~ -N Percent Removal versus Temperature 176 such a correction (to 20° C) is excessive. The reduced effect of temperature is probably due to the long hydraulic retention periods, as discussed previously for organic removal. It was then decided to check the influence of hydraulic retention time (HRT), given that it had been an important factor in organic removal and that the scatter at high loading rates corresponded with higher influent flow rates. Once checks were made, it was observed that the other outlying points from Figure 9.6 also corresponded to high influent flow rates. Figure 9.10 shows the rather definitive relationship between hydraulic retention time and nitrification efficiency for this study. This relationship appears to be relatively independent of temperature effects, although the slight scatter at the corner of the graph seems to be temperature related. These results indicate that the nitrification efficiency is reduced sharply at hydraulic retention times less than about four hours. There is little indication in the literature surveyed that the effects of hydraulic retention time on nitrification have been investigated. Aside from the few studies found by Coulter, which compared temperature and retention time effects, retention time has generally been discounted as an important process parameter in RBCs. As pointed out by Wu et al. (83), regression analyses have shown retention time to be much less important than other parameters, but this may be because RBCs have generally operated over a narrow range of retention times. Since the ammonia levels in sewage are about the same as they were in this leachate, RBCs which have been operated to achieve complete nitrification have probably maintained hydraulic detention times of over four hours and therefore this effect may have gone unnoticed. Similarly, RBCs operated primarily for organic carbon removal generally have short hydraulic retention times, in the order of 0.5 to 2.0 hours, which would also fail to exihibit this effect. However, there is some other evidence of this effect as Mikula et al. (52), attributed the loss of nitrification, while treating 177 N H 3 -N REMOVAL vs. CORR. LOADING Temperature Effects * CM LU 1.2 n H rr o UJ rr co x 0.6H 0.4 H 0.2 x • • Ox • X • • • LTJJ X * • • • • • Legend A Temp. >12C X Temp. 8-12C • Temp. <8C NH 3 -N LOADING RATE (g/m2#d) Figure 9.9 NH, -N Removal versus Loading Rate Corrected for Temperature 178 AMMONIA % REMOVAL vs. RETENTION TIME o LU rr 100-1 90-80 70-60 50 A A n A 5 10 15 HYDRAULIC RETENTION TIME hrs. 20 Legend A TEMP. >12 C X TEMP. 8-12 C • TEMP. <8 C Figure 9.10 N H , -N Removal versus Hydraul ic Retention Time (HRT) 179 cheese processing wastewater, to a drop in HRT from 16 to 9.5 hours. Therefore, further research should be encouraged to define the interrelationship of temperature and hydraulic retention time, especially with respect to nitrification. The relationship between hydraulic retention time and nitrification efficiency is probably rooted largely in the growth kinetics and mass transfer rates which control the nitrification process. However, retention time itself is rarely included as a parameter in mathematical models of the RBC nitrification process. The models for nitrification in RBCs parallel those discussed earlier for organic carbon removal, so there is no need to repeat those comments, except to present the results of the Kincannon ef al. (44) approach as applied to the nitrification performance. Figure 9.11 shows the plot of 1/NH^-N removal versus 1/NH^-N loading. From the regression analyses the parameters U m a x and Kg were evaluated to be 4.69 and 4.54 g N/m 2*d, or 0.96 and 0.93 lbs N/1000 f t 2 * d respectively. These results show that the activity of the nitrifiers is roughly 43 times less than the heterotrophs, which reflects the lower growth rate of the nitrifying organisms. The low growth rate is probably a major reason for the observed retention time effect. However, since ammonia removal rates were comparable with those achieved in many other studies, there is no evidence that the growth rate of the nitrfiers in this study was inhibited or lower than normally observed for sewage treatment. The nitrification performance of the RBC treating this leachate compares very favourably to results from RBC treatment of sewage and general aerobic treatment of other landfill leachates. Removal efficiencies and loading rates correspond very well to those established for complete nitrification in sewage treatment applications. For example, Murphy and Wilson (54,80), found that the maximum loading rate for complete nitrification was between 1 and 1.2 g TKN-N/m 2*d, which relates very well to the results of this study (recall Figure 9.6). The results of Murphy and Wilson are very typical for sewage treatment. Therefore the results of this study indicate 180 1/NH 3 REMOVAL vs. 1/NH 3 LOADING Monod Kinetics Approach O U J cc 10 - i 8-6-CO X 4-2 -• A kA Legend A Temp. >12C X Temp. 8-12C • Temp. <8C _, , ! ( ! 2 4 6 8 10 VNH3 -N LOADING Figure 9.11 NHL -N Removal - M o n o d Kinetics Approach Linear Regression Results Data Group Slope Y intercept Correlation No. of Data Coefficient Points Overall 0.967 0.213 0.991 65 U m a x = 4 6 9 <B NH 3 -N/m2 *d) K B = 4.54 (g NH 3 -N/m 2 *d) 181 that the design nitrogen loading rates used for sewage treatment are applicable to this landfill leachate (considering the prevailing organic loading rates). As indicated previously for organic removal however, the loading reductions recommended for low temperature conditions may not be necessary. The results from other landfill leachate studies generally indicate that nitrification is readily achieved and maintained. Chian et al. (13) summarized that aerobic treatment processes are usually capable of 90% N H ^ -N conversion, and typically produce effluents with less than 10 mg/L N H ^ -N. However, there have been problems encountered while trying to nitrify landfill leachates. Many of these problems have resulted from the greater sensitivity of the nitrification process to upset. For example, Keenan et al. (43), found it necessary to reduce influent ammonia levels of roughly 1000 mg/L, by 50 to 60 % with air-stripping, to avoid inhibition of the nitrifying organisms. Robinson and Maris (66), found that nitrate production did not occur until the solids retention time (SRT), was greater than 20 days while treating an old leachate, and that an SRT of 70 days was required to reduce effluent ammonia levels to less than 1 mg/L. Their lack of success may have been due in large part to their inability to maintain adequate solids concentrations. The MLVSS of their reactors were typically <100 mg/L. Jasper et al. (42), failed to maintain consistent rates of nitrification after it was initially established and they speculated that the fade in nitrification performance was due to toxic effects of accumulated metals, especially zinc (Zn). Therefore, while these cases may be exceptions to the rule, they demonstrate that nitrification of landfill leachates requires greater control and is less certain than organic carbon removal. Although one of the advantages ascribed to RBCs is that they provide a more stable environment for nitrification (22), the few results concerning leachate treatment are inconclusive. The results presented by Ehrig (22) support the results of this study, as he found efficient nitrification of three different leachates from 182 methanogenic phase (old) landfills. These leachates had much higher ammonia concentrations than the Premier leachate, ranging from 206 to 1346 mg/L. In contrast, the study reported on by Coulter (16), observed an almost complete lack of nitrification. Effluent ammonia levels were in the order of 38 mg/L, while effluent nitrate levels were limited to 0.5 - 1.0 mg/L. The reasons for the lack of nitrification in this case were not determined conclusively. Coulter speculates that if nitrification was established during the first run, which was at a light BODj. loading rate and coincident with warm water temperatures, (a fact that was not established analytically), it was then upset and lost because of the doubling of the loading rate at the start of run #2. Nitrification would have then been difficult to re-establish because of the low wastewater temperatures ( < 1 1 ° C) during run #2. Toxic inhibition of the nitrifiers , by something within the 10% of industrial waste accepted by the Montreal landfill, was proposed as a contributing factor. The results of this study would tend to support the theory that some toxic effect was responsible for the lack of nitrification in the Montreal study. Leachates used in the two studies were quite comparable except that the Montreal leachate had mercury (Hg) levels of 0.5 mg/L, which is much higher than levels observed in Vancouver area landfills. A more extensive analysis of the Montreal leachate's composition may have found other toxic and/or inhibitory compounds, both inorganic and organic. In the absence of toxic effects, given the experience of this study, it seems implausible that nitrification would not have become established during the three month period of run #1, under the prevailing light loading conditions and warm summer temperatures. Once established, the nitrifiers would not likely be totally upset by just a doubling of the loading rate. During this study, the loading rates were highly variable and doubled on various occasions without even a loss of nitrification efficiency, let alone loss of the process. Since hydraulic retention times were similarly long during the Montreal study, the loading and temperature effects 1 8 3 were likely moderated for nitrification just as they were for organic removal, and retention time would not be limiting. Therefore, in the absence of further information, toxic inhibition seems the most plausible reason for the lack of nitrification observed in the Montreal study. This would then underline the importance of leachate quality in determining treatment feasibility and performance. 9.3 RBC RESPONSE TO VARIABLE AND INTERMITTENT LOADING As presented in Section 7 RBC Operation, the RBC operated under difficult conditions of variable and intermittent hydraulic and organic loading at times during this study. Overall, these conditions did not impair the process performance or adversely affect the biomass. Organic carbon removal and nitrification were observed to be relatively unaffected by the variability of the substrate loading, consistently maintaining a good effluent quality. The resistance of the RBC to the effects of variable loading were no doubt enhanced by the relatively gradual nature of the changes and the long hydraulic retention periods within the unit, ln the case of carbon removal, the low range of the organic loading rates was also a contributing factor. As pointed out earlier, Filion et al. (27) found that an RBC recovered in less than three hours to an instantaneous increase in loading. Therefore, given the more gradual changes in loading, and hydraulic retention times significantly greater than three hours, the RBC appears quite capable of responding to the loading variability observed during this study. However, for both carbon and nitrogen removal, a four fold increase in the loading rate over a four day period resulted in slight increases in effluent BODj. and NH^ values. This indicates that larger, or more rapid increases in mass loading rates would probably exceed the RBCs capacity to respond, without at least a temporary loss of effluent quality. Temperature would also affect the RBCs response time to increases in mass loading 184 rates. The variable organic loading was generally observed to have only a minor affect on the biomass or suspended solids levels in the RBC. As implied in the above discussion the biomass growth, and thus process performance, was able to adjust to the changing loading conditions. Agian, the low range of organic loading generally avoided many problems such as oxygen depletion, Beggiatoa growth, and substrate inhibition, associated with heavy loading conditions. Suspended solids levels within the RBC stages were usually lower during periods of steady operation. It was observed in this study and elsewhere (54), that suspended solids accumulated in the RBC during interruptions of flow. During brief stoppages of one or two days, the RBC solids generally continue to slough at a normal rate, and then accumulate because of the lack of flow through the unit. Murphy ef al. (54), observed that a flow of 10% of average flow was sufficient to wash out the sloughed solids. This accumulation affect explains the higher suspended solids levels during periods of unsteady operation. Therefore, process performance immediately after an interruption of flow depends mainly upon the ability of the final clarifier to handle this additional solids loading. Although it is expected that total solids production would increase slightly with loading levels, the data from this study was too scattered to establish sludge production rates. On two occassions, an interruption in the leachate flow resulted in a general sloughing and major loss of the biomass. It is uncertain why these two interruptions caused such a large loss of biomass while many others did not. In the first instance, in August 1984, warm temperatures may have increased the rate of endogenous decay which would weaken the biomass. The second instance, in November 1984, may have been a culmination of the various effects of previous upsets and declining temperatures. A sharp decline in the loading rate over the previous two weeks may also have been a contributing factor. Aside from these 185 two events however, the biomass was retained on the disks. During periods of very low organic loading, the biomass became endogenous, (the volatile component dropped to below 30%), but was retained on the disks ready to assimilate higher loads. Therefore, this study was able to demonstrate the good resistance of the RBC to any adverse effects of variable loading. 9.4 METALS AND TRACE ORCANICS The determination of some heavy metal and trace organic concentrations in the Premier leachate and RBC effluent was supplementary to the main topic of this study. The small amount of data collected does not support conclusions beyond the general results presented in the previous section; however, some additional comment is possible; For metal removals, the removal rates and relative affinity of the various metal species for removal were very similar to results observed for activated sludge systems. The removed metals were concentrated in the biomass to levels comparable to, or higher, than observed in suspended-growth leachate studies (82,83), with no apparent adverse effects. The trace organic results indicate that an assortment of compounds are finding their way into municipal landfills and that these compounds are quite mobile and readily enter the leachate. This raises the question of whether or not greater control over the disposal of these types of materials is necessary. The RBC effluent samples indicated that these compounds were effectively removed during treatment but further research will be required to determine the fate of these compounds. If volatilization or stripping into the atmosphere is the major removal mechanism, there may be a potential for a localized health hazard where leachates are treated. 186 9.5 TOXICITY A number of attempts were made to determine the toxicity of the Premier leachate and RBC effluent using the Daphnia bioassay procedure outlined in Atwater ef al. (2). However, problems were encountered with the survival of the Daphnia in the dilution water blanks and therefore no reliable results were produced. It was observed qualitatively that the Premier leachate was fairly toxic, which one would expect given the ammonia concentrations alone. The RBC effluent samples on the other hand were apparently non-toxic. This was indicated by the Daphnia growing better in the effluent than either the stock culture or dilution water. Therefore, it was indicated, but not conclusively, that the RBC was capable of producing a non-toxic effluent. 9.6 IMPLICATIONS FOR FULL SCALE TREATMENT There are many factors to be considered when extrapolating from the encouraging results of this and other studies, to a full scale application for the treatment of the Premier or other landfill leachates. Of primary concern are the chemical and physical properties of the leachate to be treated. The Premier leachate used in this study was quite weak, despite coming from a young landfill; this aptly demonstrates the effects that specific site conditions, such as climate, drainage patterns, etc., can have on leachate quality. There were no indications that this leachate was inhibitory to either the heterotrophic or autotrophic bacterial growth on the RBC. Experience at this university and elsewhere (16,18,42,43,66), with other leachates, particularly strong leachates, have shown that biological inhibition due to substrate concentration, heavy metals, and other compounds, is quite common. Fortunately, in most cases the inhibition results in reduced reaction rates rather than 187 process failure. Nitrification has proven especially prone to inhibition. In many cases, some form of pretreatment of the leachate was required before a stable biological process could be established. Therefore, the loading levels and treatment efficiencies achieved in this study may not be as readily attainable with other leachates, especially much stronger ones. The results of this study, as well as those of Coulter (16), and Ehrig (22) for organic removal and nitrification respectively, tend to show that the design mass loading rates proposed by Murphy and Wilson (54,80) for RBC treatment of domestic sewage, apply equally well to the treatment of some landfill leachates. Their design loadings are presented in Table 9.1. The aforementioned studies indicate that these loadings levels may be applicable to relatively weak young leachates, as well as most old leachates for which nitrification governs the loading rate. Further research is required to both confirm the initial results of these few studies, as well as determine the ability of the RBC to treat high organic strength leachates. To reiterate, these loading levels are probably not universally applicable to leachate treatment, but only more experience will determine over what range of leachate quality they are valid. Therefore in the mean time, these design guidelines should be confirmed by pilot scale studies of the particular leachate to be treated. Aside from the site to site variation of landfill leachate quality, the changes which occur over time as a landfill stabilizes must also be accounted for in a treatment design. Organic carbon removal will usually govern the design of a treatment process when a landfill is young, or in the acid formation phase, but after the transition to the old, or methanogenic phase, nitrification will govern the design. Therefore, the treatment design should incorporate a high degree of flexibility of operation to permit adjustment to changing conditions. The modular design of RBCs has the potential to permit the movement of units between sites depending upon demand, as well as the simple rearrangement of the staging or 188 Table 9.1 Design Loadings for RBC Treatment of a Municipal Was tewater^ Design Loading RBC System Design Objective Parameters (g /m 2*d) (Ave. Value mg/L) 15°C 10°C 5°C BODr removal B O D 5 s 2 0 B O D 5 L o a d 9 7 7 6 6 0 TSS S 20 (total) B O D 5 removal B O D 5 ^ 3 0 B O D 5 Load 15 12 9.3 TSS S 30 (total) B O D 5 removal TKN «s 3 TKN Load 0.60 0.39 0.25 plus nitrification (filtrable) - assumes primary clarified wastewater feed to RBC with 180 mg/L BODr, and 30 mg/L filtrable TKN, - provides factors of 1.25 and 1.35 for BOD^ removal and combined BODj. plus TKN removal to correct for diurnal flow variation. (1) From Wilson e( al. (80). treatment flowpath to adjust for changing conditions. A given number of RBC stages in a flowpath also tends to be self-regulating with respect to allocating surface area to carbon removal or nitrification, although the later always defers to the former, which may reduce nitrification performance at high organic loading. Another difficulty with the application of biological treatment processes to leachate treatment is the large variation in hydraulic and organic loading which can occur over a short period of time at some landfills. Frequently, the mass of pollutants leached from a landfill increase with increasing hydraulic flow through the fill, so that the hydraulic and organic loading tend to increase together. During this study, the mass of C O D released from the landfill was observed to increase eight fold over four days, after a prolonged period of low flows, in a full scale plant, the biomass would likely be unable to assimilate so much additional substrate that quickly. The results of this study however did show that the RBC was very resistant to less severe variations in loading and interruptions in leachate flow. Recall that a 189 four fold increase in organic loading over a four day period resulted in only a slight increase in effluent BOD^. In most full scale applications, some form of equalization, to modulate the loading peaks would probably be necessary for any treatment scheme. Where possible, recirculation of some of the leachate back onto the landfill is an attractive method as it can both hold-over flows until dryer periods, and reduce the pollutant load due to in-situ stabilization of the leachate. The results of the RBC leachate treatment studies indicate that the RBC is particularly well suited for leachate treatment. Other studies indicate that air-driven RBCs may be even more so (recall Section 3). The fixed-growth of the RBC provides much better resistance to variable hydraulic, and to a lesser extent, organic loading than suspended-growth systems. Air-driven RBCs would provide a high degree of operational flexibility, as well as permit the RBC to accept organic loadings in the first stages which would be problematic in a mechanical-drive unit. Such air-drive units more closely approximate the ability of completely mixed, or tappered-aeration plug flow, suspended-growth units to accept peak organic loadings. The staging of an RBC process train potentially provides a protected environment in the later stages for the nitrifying organisms, which would be further protected by their location in an interior layer of the biomass. Given the predisposition of the nitrifying organisms to attached growth, these factors probably contribute to a more stable nitrification process in the RBC as opposed to suspended-growth systems. While RBCs are more sensitive to ambient temperatures than suspended-growth systems, air-driven RBCs in particular can use warm air from the blowers, retained within the insulated covers, to ameliorate temperature effects. It was observed during this study that there was relatively efficient heat transfer between the liquid and the surrounding air, as a temperature differential of up to 4° C was observed between the first and fourth stage liquid. 190 While the above discussion has proposed numerous advantages of the RBC for leachate treatment, and especially for the air-driven RBCs, there are few apparent disadvantages, there is no conclusive evidence that RBCs perform better than suspended-growth systems. Recall from Section 3.2 that Henry (33) generalized suspended-growth leachate treatment as requiring SRTs of twice, and loading rates of half, those used for domestic sewage, which roughly corresponds to an extended aeration mode of operation. While the results of this study indicate that an RBC can treat landfill leachates at loading levels which are the same as those used for sewage treatment, this does not necessarily indicate an advantage for RBCs since RBCs are generally considered to relate more closely to an extended aeration process. There have been few side by side tests of RBCs and suspended-growth systems, and these have been inconclusive. None have been conducted for leachate treatment. One problem with comparing the two systems is relating the loading rates in the two systems, as again, the estimates of active biomass are determined in different ways and are quite subjective. Further research and comparison studies, with particular emphasis on the nitrification performance of the two types of treatment systems is required to determine if there is a difference. Therefore, until one type of treatment system proves superior, designers will continue to chose on the basis of economics and personal experience. 9.7 EXPERIMENTAL PROCRAM AND RBC OPERATION The experimental program as proposed could not be evaluated because it was not carried out, but it still seems to be a valid approach. However, there are a couple of changes to the sampling and analysis procedures which would have been beneficial in hindsight. Firstly, it would have been helpful to have determined both a total and nitrification inhibited B O D r value for the raw leachate and filtered 191 effluent. This would have more clearly defined the residual carbonaceous and nitrogenous BOD, as the total results did not reflect the ammonia levels on a regular basis. Secondly, it would have been nice to have some dissolved oxygen values during December 1984, when the organic wash-out occurred, to confirm the other indications of oxygen depletion. Other changes which were not implemented in this study were the use of load cells on the shaft to give a measure of total biomass, and the use of an autosampler to collect process and leachate samples. Load cells at either end of the media shaft have been used in other studies with good success, to easily determine a relative measure of the total biomass (47). For reasons discussed earlier in Section 4, the use of areal biomass determinations was not satisfactory during this study. The use of load cells appears to becoming more popular, judging from more recent studies, and it certainly has the advantage of simplicity. Use of an autosampler during this study would have permitted a characterization of short term fluctuations in leachate quality, as well as the collection of more process data during the periods of stable operation. An autosampler may also have proved useful to more closely study the response of the RBC to loading fluctuations. However, autosamplers should be used judiciously to test specific notions; because, while the sample collection is relatively effortless, the analysis of those samples is not. With respect to the RBC pilot plant, its ancilliary equipment, and operation, the extensive and varied experience gained during this study gives rise to a number of recommendations. The pilot plant itself was adequate for the purposes of this study but one useful modification would see the top cover fit over the lip of the bottom section, rather than inside it; this would then prevent rainwater from entering the unit and affecting the hydraulic loading. Other modifications which are desirable are; stronger mounts for the shaft and disk drive motor, a sludge removal mechanism in the clarifier to permit its use as a clarifier. and uniform rigid media 192 to help prevent the biomass bridging which occurred between the flexible mesh media. The two biggest operational problems aside from the natural calamities were the pump failures, and the biological fouling. Over the course of this study, the pump problems were more or less sorted out, and the Cormann-Rupp bellows pumps, with the valve springs installed, proved to be adequately reliable and easily serviced. The biological fouling problem was never adequately resolved. Susequent consideration of this problem had led to the suggestion that a relatively high capacity submersible pump (approx. 20 L/m) should have been used to lift the leachate from the wet well into a short retention time reservoir mounted in the RBC. Feed for the RBC would then be pumped from this reservoir, through very short delivery lines, which would reduce fouling and facilitate easier cleaning. Excess flows would overflow the reservoir and return to the wet well. The line from the submersible pump to the reservoir would be much less likely to plug up because of high flow velocities and positive pump pressure conditions. Plugging of the pump inlet screen would also be less likely because of the higher flows and the use of a coarser screen. With these changes, hopefully many of the problems encountered in this study could be avoided. 10. SUMMARY The results from this pilot scale study of RBC treatment of a landfill leachate indicate that efficient treatment can be maintained even under difficult operating conditions. Settled and filtered effluent samples had BOD,- values generally less than 25 and 10 mg/L, respectively. This effluent quality was maintained despite variable loading and frequent interruptions of the leachate supply. Settled effluent suspended solids were less than 25 mg/L during periods of steady operation, and usually less than 100 mg/L during upsets. Sharp changes in loading or interruptions of the leachate flow were first reflected by increases in the suspended solids. Overall, the RBC demonstrated a remarkable resistance to fluctuations and interruptions in organic and hydraulic loading. The RBC operated under low carbon loading conditions for much of the test period, due to declining leachate strength and pump limitations. In most cases, the B O D j loading was less than 6 g BOD,-/m 2*d. However, a few samples had higher loading rates, ranging up to 18 g BOD j/m 2 * d , and still produced a high quality effluent. These results indicated that the carbon removal capacity of the RBC treating this leachate was comparable to its capacity to treat domestic sewage. Efficient nitrification of this leachate was also maintained throughout variable conditions. Effluent NH-j -N and TKN -N were usually less than 1.0 and 10.0 mg/L respectively. Nitrification was observed to stop under high organic loading conditions. The average nitrogen loading rate during the study was approximately 0.6 g N/m 2 *d. Results and loading rates for nitrification compare very well with those found for sewage treatment. Temperature effects for both carbon removal and nitrification were offset by long hydraulic retention times, and for nitrification in particular, retention time 193 194 appeared to be a controlling factor. This result indicates that for concentrated wastes like landfill leachates, for which hydraulic retention times exceed four hours, reductions in loading rates at lower temperatures will be much less than normally applied for sewage treatment. Therefore, the design loading rates for nitrification and carbon removal developed for sewage treatment at moderate temperatures could be applied to the treatment of some landfill leachates over a wider range of temperatures. This study also indicated, to varying extents, that the RBC was capable of removing heavy metals and specific organic compounds, and produce a non-toxic effluent when treating this leachate. Overall, this study showed that the RBC is a viable process choice for leachate treatment and possibly has advantages over other systems, especially for nitrification. 11. CONCLUSIONS 1. This study indicates, although not conclusively, that the capacity of an RBC for carbon removal from this and similar leachates is comparable to it's capacity to treat domestic sewage. The design loading rates recommended by Murphy ef al. (54), for BODj. removal from domestic sewage should therefore apply equally well for the treatment of many moderate to low strength landfill leachates. 2. This study showed more conclusively that the capacity of an RBC for nitrification of this and some other leachates, is comparable to its capacity to nitrify domestic sewage. The design loading rates recommended by Murphy ef al. (54), for complete nitrification of domestic sewage should therefore apply equally well for the treatment of landfill leachates, except possibly in instances of toxic effects. 3. This study demonstrated conclusively that hydraulic retention time is an important parameter with respect to RBC treatment efficiency. For nitrification especially, hydraulic retention time appeared to be a controlling factor. The results also showed that hydraulic retention times of greater than four hours could effectively offset the temperature effects which have been frequently observed at lower retention times, for both carbon removal and nitrification. This result indicates that, for situations where sufficiently long hydraulic retention times are maintained, that loading rates need not be reduced in response to lower temperatures. 4. This study showed that the RBC process is remarkably resistant to fluctuations and interruptions of organic and hydraulic loading. Process effluent quality was not impaired by these variations, likely due to the moderating effects of the 195 196 long hydraulic retention time. Sharp changes in loading or interruptions of the leachate flow were first reflected by increases in the suspended solids, indicating that solids separation may be the controlling factor for effluent quality in these instances. 5. This study indicated that heavy metals are removed from the leachate and concentrated in the RBC biomass at similar rates and affinities for metal species as observed in suspended-growth systems. 6. This study showed that various specific organic compounds are present in this leachate and are effectively removed during passage through the RBC. Some of these compounds are on the EPA list of priority pollutants. The mechanism of their removal was not determined however. 12. RECOMMENDATIONS FOR FURTHER RESEARCH 1. Given the general lack of experience with RBC treatment of landfill leachates, further studies should be undertaken to conclusively establish the capacity of the RBC to treat landfill leachates of varying strengths and compositions, with special emphasis on nitrification. 2. Side by side comparison studies of RBCs and Activated Sludge treatment of leachate should be undertaken to evaluate advantages or disadvantages of the two systems, especially for nitrification. 3. The relationship between temperature and hydraulic retention time effects should be investigated more fully. Possibly the volume to surface area ratios of RBC design could also be used as a factor to change hydraulic retention time and reduce temperature effects. 4. The types and concentrations of trace organic compounds in leachate should be investigated in more cases, and the major mechanisms of their removal during treatment determined. 5. The denitrification of landfill leachate could be investigated using a submerged RBC. If, in fact, the nitrification process is more stable in the RBC, the nitrite accumulation observed by Ehrig (49) may permit the short-circuit denitrification investigated by Sam Turk (PhD thesis, UBC, 1986), to occur more reliably. 6. Ishiguro (56) indicated in his paper that Japan has had considerable experience treating landfill leachates with RBCs since 1976. When this paper was presented in 1983, there were apparently 135 RBC plants treating landfill wastes. This indicates that a review of the Japanese literature may provide the answers to many questions concerning RBC treatment of leachate. 197 Although unrelated to this study, it would be interesting to investigate feasibility and performance of a sequencing batch RBC for biological phosphorus removal. 13. REFERENCES 1. Atwater, J.W., and Bradshaw, G., "Operational Performance of an RBC Unit in Conjunction With a Septic Tank", Can. /. Civ. Eng., Vol. 8, No. 4, 1981, pp. 477 - 483. 2. Atwater, J.W., Jasper, S.E., Mavinic, D.S., and Koch, F., "Experiments Using Daphinia to Measure Landfill Leachate Toxicity", Water Research, Vol. 17, No. 12, 1983, pp. 1855-1861. 3. Ball, R.O., "The Effects of Hydraulic Variation on Fixed-Film Reactor Performance", "Fixed-Film Biological Processes for Wastewater Treatment", Ed. Wu, Y.C., and Smith, E.D., New Jersey, Noyes Data Corporation, 1983, pp. 174 - 205. 4. Benefield, L.D., and Randall, C.W., "Biological Process Design for Wastewater Treatment", Prentice-Hall, Inc., New Jersey, 1980. 5. Boyle, W . C , and Ham, R.K., "Biological Treatability of Landfill Leachate", J. Water Pollut. Control Fed., Vol. 46, No. 5, May 1974, pp. 860-872. 6. Brown, M.J., and Lester, J.N., "Metal Removal in Activated Sludge: The Role of Bacterial Extracellular Polymers", Water Res., Vol. 13, 1979, pp. 817 - 837. 7. Bull, P.S., Evans, J.V., Wechsler, R.M., and Cleland, K.J., "Biological Technology of the Treatment of Leachate from Sanitary Landfills", Water Res., Vol. 17, No. 11, 1983, pp. 1473 - 1481. 8. Cameron, R.D., and Koch, F.A., "Toxicity of Landfill Leachates", J. Water Pollut. Control Fed., Vol. 52, No. 4, April, 1980, pp. 760 - 769. 9. Cheng, M.H. , Patterson, J.W., and Minear, R.A., "Heavy Metal Uptake by Activated Sludge", J. Water Pollut. Control Fed., Vol. 47, No. 2, February, 1975, pp. 362 - 376. 10. Chian, E.S.K., and DeWalle, F.B., "Sequential Substrate Removal in Activated Sludge Systems Fed with Naturally Occuring Wastewater", Prog. Water Technol., Vol. 17, No. 2, 1975, pp. 235 - 241. 11. Chian, E.S.K., and DeWalle, F.B., "Sanitary Landfill Leachates and Their Treatment", /. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 102, No. EE2, April, 1976, pp. 4 1 1 - 4 3 1 . 12. Chian, E.S.K., "Stability of Organic Matter in Landfill Leachates", Water Res., Vol. 11, 1977, pp. 225 - 232. 13. Chian, E.S.K., Pohland, F.G., Chang, K.C., and Harper, S.R., "Leachate Generation and Control at Landfill Disposal Sites", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 14 - 30. 14. Clark, J.H., Moseng, E.M., and Asano, T., "Performance of a Rotating Biological Contactor Under Varying Wastewater Flow", /. Water Pollut. Control Fed., Vol. 50, No. 5, May, 1978, pp. 896 - 911. 199 200 15. Cook, E.N., and Foree, E.G., "Aerobic Biostabilization of Sanitary Landfill Leachate", J. Water Pollut. Control Fed., Vol. 46, No. 2, February, 1974, pp. 380 - 392. 16. Coulter, R.G., "Pilot Plant Study of Landfill Leachate Treatability by Rotating Biological Contactors", Proc. Second International Conf. on Fixed-Film Biological Processes, Arlington, Virginia, July 10 - 12, 1984, pp. 777 - 813. 17. Dass, P., Tamke, G.R., and Stoffel, C M . , "Leachate Production at Sanitary Landfill Sites", J. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 103, No. EE6, December, 1977, pp. 981 - 988. 18. Dedhar, S., and Mavinic, D.S., "Ammonia Removal From a Landfill Leachate by Nitrification and Denitrification", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 502 - 516. 19. Department of Lands, Forests, and Water Resources, "Report on Pollution Control Objectives for Municipal Type Waste Discharges in British Columbia", Water Resources Service, Victoria, British Columbia, September, 1975. 20. DeWalle, F.B., and Chian, E.S.K., "Kinetics of Formation of Humic Substances in Activated Sludge Systems and Their Effect on Flocculation", Biotechnology and Bioengineering, Vol. 17, 1974, pp. 739 - 755. 21. Duggan, J.B., and Cleasby, J.L., "Effect of Variable Loading on Oxygen Uptake", J. Water Pollut. Control Fed., Vol. 48, No. 3, March, 1976, pp. 540 - 550. 22. Ehrig, H.J., "Treatment of Sanitary Landfill Leachate: Biological Treatment", Waste Management and Research, Vol. 2, 1984, pp. 131 - 152. 23. Ehrig, H.J., "Laboratory and Full-Scale Experiments on Physical-Chemical Treatment of Sanitary Landfill Leachate", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 200 - 214. 24. Ehrig, H.J., "Biological Treatment of Sanitary Landfill Leachate with Special Aspects on the High Ammonia Concentration", Proceedings of the New Directions and Research in Wasfe Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 232 - 248. 25. Evans, F.L., "Consideration of First-Stage Organic Overloading in Rotating Biological Contactor Design", /. Water Pollut. Control Fed., Vol. 57, No. 11, November, 1985, pp. 1094 - 1098. 26. Famularo, J., Mueller, J.A., and Mulligan, T., "Application of Mass Transfer to Rotating Biological Contactors", }. Water Pollut Control Fed., Vol. 50, No. 4, April, 1978, pp. 653 - 671. 27. Filion, M.P., Murphy, K.L., and Stephenson, J.P., "Performance of a Rotating Biological Contactor Under Transient Loading Conditions", j. Water Pollut. Control Fed., Vol. 51, No. 7, July, 1979, pp. 1924 - 1933. 28. Forgie, D.J.L., "Effect of Temperature on Rotating Biological Contactor Treatment Efficiency", Proc. Second International Conf. on Fixed-Film Biological Processes, Arlington, Virginia, July 10 - 12, 1984, pp. 730 - 753. 201 29. Friedman, A.A., Robbins, L.E., and Woods, R.C, "Effect of Disk Rotational Speed on Biological Contactor Efficiency", J. Water Pollut. Control Fed., Vol. 51, No. 11, November, 1979, pp. 2678 - 2690. 30. Gibs, )., Schoenberger, R.J., and Suffet, I.H., "A Simplified Buffer Capacity Model for Sanitary Landfill Leachate", Water Res., Vol. 16, 1982, pp. 699 -705. 31. Grady, C.P.LJr., "Model ing of Biological Fixed-Films A State-of-the-Art Review", "Fixed-Film Biological Processes for Wastewater Treatment", Ed. Wu, Y.C, and Smith, E.D., New Jersey, Noyes Data Corporation, 1983, pp. 75 - 134. 32. Harmsen, J., "Identification of Organic Compounds in Leachate From a Waste Tip", Wafer Res., Vol. 17, No. 6, 1983, pp. 699 - 705. 33. Henry, J.G., "New Developments in Landfill Leachate Treatment", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 139 - 149. 34. Hitdlebaugh, J.A., and Miller, R.D., "Operational Problems with Rotating Biological Contactors", /. Water Pollut. Control Fed., Vol. 53, No. 8, August, 1981, pp. 1283 - 1293. 35. Hockenbury, M.R., Daigger, G.T., and Grady, C.P.L., "Factors Affecting Nitrification", J. Environ. Eng. Div., Proc'Am. Soc. Civ. Eng., Vol. 103, No. EE1, February, 1977, pp. 9 - 1 9 . 36. Huang, C.S., "Nitrification Kinetics and its RBC Application", /. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 108, No. EE3, June, 1982, pp. 473 -487. 37. Huang, C.S., "Some Experiences with Fixed-Film Biological Processes", "Fixed-Film Biological Processes for Wastewater Treatment", Ed. Wu, Y.C, and Smith, E.D., New Jersey, Noyes Data Corporation, 1983, pp. 422 - 449. 38. Huang, Ju-Chang, and Bates, V.T., "Comparative Performance of Rotating Biological Contactors Using Air and Pure Oxygen", J. Water Pollut. Control Fed., Vol. 52, No. 11, November, 1980, pp. 2686 - 2703. 39. Hynek, R.J., and Chou, C C , "Development and Performance of Air-Driven Rotating Biological Contactors", Proc. 34th Industrial Waste Conf. Purdue University, Lafayette, Indiana, May 8 - 10, 1979, Ann Arbor Science Pub. Inc., Ann Arbor, Mich., pp. 805 - 815. 40. Ishiguro, M., "Current Status and Future Trends of Rotating Biological Contactors in Japan", "Fixed-Film Biological Processes for Wastewater Treatment", Ed. Wu, Y.C, and Smith, E.D., New Jersey, Noyes Data Corporation, 1983, pp. 466 - 484. 41. Jasper, S.E., Atwater, J.W., and Mavinic, D.S., "Leachate Production and Characteristics as a Function of Water Input and Landfill Configuration", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 64 - 83. 202 42. Jasper, S.E., Mavinic, D.S., and Atwater, J.W., "Influent Constraints on Treatment and Biological Nitrification of Municipal Landfill Leachate", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 843 - 863. 43. Keenan, J.D., Steiner, R.L, and Fungaroli, A.A., "Landfill Leachate Treatment", J. Water Pollut. Control Fed., Vol. 56, No. 1, January, 1984, pp. 27 - 33. 44. Kincannon, D.F., Stover, E.L., Emrie, D., and Jankovsky, M., "Comparison of Pilot Scale and Full Scale RBC Design and Operation", Proc. Second International Conf. on Fixed-Film Biological Processes, Arlington, Virginia, July 10 - 12, 1984, pp. 1840 - 1870. 45. Knoetze, C , Davies, T.R., and Wiechers, S.G., "Chemical Inhibition of Biological Nutrient Removal Processes", Water S.A., Vol. 6, No. 4, October, 1980, pp. 171 - 179. 46. Maris, P.J., Harrington, D.W., and Mosey, F.E., "Treatment of Landfill Leachate; Management Options", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 280 - 296. 47. Marsh, D., Benefield, L., Bennett, E., Linstedt, D., and Hartman, R., "Coupled Trickling Filter-Rotating Biological Contactor Nitrification Process", /. Water Pollut. Control Fed., Vol. 53, No. 10, October, 1981, pp. 1469 - 1480. 48. Masuda, S., Watanabe, Y., and Ishiguro, M., "Simultaneous Nitrification and Denitrification in a Rotating Biological Contactor", "Fixed-Film Biological Processes for Wastewater Treatment", Ed. Wu, Y.C., and Smith, E.D., New Jersey, Noyes Data Corporation, 1983, pp. 227 - 246. 49. Mavinic, D.S., "Kinetics of Carbon Utilization in Treatment of Leachate", Water Res., Vol. 18, No. 10, 1984, pp. 1279 - 1284. 50. McBean, E.A., Poland, R., Rovers, F.A., and Crutcher, A.J., "Leachate Collection Design for Containment Landfills", /. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 108, No. EE1, February, 1982, pp. 204 - 209. 51. Metcalf and Eddy, Inc., "Wastewater Engineering: Collection, Treatment, Reuse", 2nd. Ed., McGraw-Hill, New York, N.Y., 1979. 52. Mikula, W.J., Reynolds, J.H., Middlebrooks, E.J., and George, D., "Performance Characteristics of a Rotating Biological Contactor (RBC) Treating a Cheeze Processing Wastewater", Proc. 34th Industrial Waste Conf. Purdue University, Lafayette, Indiana, May 8 - 10, 1979, Ann Arbor Science Pub. Inc., Ann Arbor, Mich., pp. 335 - 353. 53. Mueller, J.A., Paquin, P., and Famularo, J., "Nitrification in Rotating Biological Contactors", J. Water Pollut. Control Fed., Vol. 52, No. 4, April, 1980, pp. 688 - 710. 54. Murphy, K.L., and Wilson, R.W., "Pilot Plant Studies of Rotating Biological Contactors Treating Municipal Wastewater", Report SCAT 2, Central Mortgage and Housing Corporation, Ottawa, Ont., 1980. 203 55. Oleszkiewicz, J.A., "Efficiency of Plastic Media Trickling Filters Operating Under Extreme Organic Loading", Progress Water Technol. Vol. 7, No. 3/4, 1975, pp. 607 - 615. 56. Palit, T., and Qasim, S.R., "Biological Treatment Kinetics of Landfill Leachate", /. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 103, No. EE2, April, 1977, pp. 353 - 366. 57. Pano, A., and Middlebrooks, E.J., "Kinetics of Carbon and Ammonia Nitrogen Removal in RBCs" /. Water Pollut. Control Fed., Vol. 55, No. 7, July, 1983, pp. 956 - 965. 58. Paolini, A.E., and Variali, G., "Kinetic Considerations on the Performance of Activated Sludge Reactors and Rotating Biological Contactors", Water Res., Vol. 16, 1982, pp. 155 - 160. 59. Pohland, F.G., "Accelerated Solid Waste Stabilization and Leachate Treatment by Leachate Recycle Through Sanitary Landfills", Progress Water Technol., Vol. 7, No. 3/4, 1975, pp. 753 - 765. 60. Poon, C P . C , Chin, H.K., Smith, E.D., and Mikucki, W.J., "Upgrading With Rotating Biological Contactors for BOD Removal", J. Water Pollut. Control Fed., Vol. 53, No. 4, April, 1981, pp. 474 - 481. 61. Pretorius, W.A., "Nitrification on the Rotating Disc Unit", Progress Water Technol., Vol. 7, No. 2, 1975, pp. 243 - 249. 62. Raina, S., "Combined Treatment of Landfill Leachate and Domestic Sewage", Masters Thesis, University of British Columbia, Vancouver, B.C., 1984. 63. Raveh, A., and Avnimelech, Y., "Leaching of Pollutants From Sanitary Landfill Models", /. Water Pollut. Control Fed., Vol. 51, No. 11, November, 1979, pp. 2705 - 2716. 64. Robinson, H.D., Barber, C , and Maris, P.J., "Generation and Treatment of Leachate from Domestic Wastes in Landfills", j. Water Pollut. Control, Vol. 81, 1982, pp. 465 - 478. 65. Robinson, H.D., and Maris, P.J., "The Treatment of Leachates From Domestic Wastes in Landfills - 1. Aerobic Biological Treatment of a Medium-Strength Leachate", Water Res., Vol. 17, No. 11, 1983, pp. 1537 - 1548. 66. Robinson, H.D., and Maris, P.J., "The Treatment of Leachates From Domestic Waste in Landfill Sites", /. Water Pollut. Control Fed., Vol. 57, No. 1, January, 1985, pp. 30 - 38. 67. Robinson, H.D., and Lucas, J.L., "The Behaviour and Attenuation of Leachate from Domestic Wastes in a Landfill with a Designed and Engineered Unsaturated Zone, instrumentation and Research at Stangate East Landfill, in Kent, England", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 31 - 49. 68. Robinson, H.D., "Treatment of Domestic Waste Leachate in a Full-Scale Automated Plant", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference, University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 166 - 184. 204 69. Salvato, J.A., Wilkie, W . C , and Mead, B.E., "Sanitary Landfill Leaching Prevention and Control ", /. Water Pollut. Control Fed., Vol. 43, No. 10, October, 1971, pp. 2084 - 2100. 70. Sawyer, C.N., and McCarty, P.L., "Chemistry for Environmental Engineering", 3rd. Ed., McGraw-Hill, New York, N.Y., 1978. 71. Smith, E.D., and Bandy, J.T., "State of Knowledge of Rotating Biological Contactor Technology", "Fixed-Film Biological Processes for Wastewater Treatment", Ed. Wu, Y.C., and Smith, E.D., New Jersey, Noyes Data Corporation, 1983, pp. 450 - 465. 72. "Standard Methods for Examination of Water and Wastewater", American Public Health Association Inc., 13th, 14th, & 15th editions, Washington, D.C, 1971, 1975, 1980. 73. Stanforth, R., Ham, R., Anderson, M., and Stegmann, R., "Development of a Synthetic Municipal Landfill Leachate", /. Water Pollut. Control Fed., Vol. 51, No. 7, July, 1979, pp. 1965 - 1975. 74. Stegmann, R., and Ehrig, H.J., "Operation and Design of Biological Leachate Treatment Plants", Progress Water Technol., Vol. 12, 1980, pp. 919 - 947. 75. Stegmann, R., and Spendlin, H., "Research Activities on Enhancement of Biochemical Processes in Sanitary Landfills", Proceedings of the New Directions and Research in Waste Treatment and Residuals Management Conference; ; -University of British Columbia, Vancouver, B.C., June 23 - 28, 1985, pp. 84 -107. 76. Sterritt, R.M., and Lester, J.N., "The Influence of Sludge Age on Heavy Metal Removal in the Activated Sludge Process", Water Res., Vol. 15, 1981, pp. 59 -65. 77. Straub, W.A., and Lynch, D.R., "Models of Landfill Leaching: Moisture Flow and Inorganic Strength", J. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 108, No. EE2, April, 1982, pp. 231 - 250. 78. Straub, W.A., and Lynch, D.R., "Models of Landfill Leaching: Organic Strength", J. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 108, No. EE2, April, 1982, pp. 251 - 268. 79. Uloth, V . C , and Mavinic, D.S., "Aerobic Bio-Treatment of a High-Strength Leachate", J. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 103, No. EE4, August, 1977, pp. 647 - 661. 80. Wilson, R.W., and Murphy, K.L., and Stephenson, J.P., "Scaleup in Rotating Biological Contactor Design", /. Water Pollut. Control Fed., Vol. 52, No. 3, March, 1980 , pp. 610 - 621. 81. Wong, P.T., and Mavinic, D.S., "Treatment of a Municipal Leachate Under Multi-Variable Conditions", Water Pollut. Res. /. Canada, Vol. 17, 1982, pp. 135 - 148. 82. Wood, J.M., and Wang, H.K., "Microbial Resistance to Heavy Metals", ). Environ. Sci. Technol., Vol. 17, No. 12, 1983, pp. 582A - 590A. 205 83. Wu, Y . C , Smith, E.D., and Cratz, J., "Prediction of RBC Performance for Nitrification", /. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 107, No. EE4, August, 1981, pp. 635 - 652. 84. Wu, Y . C , and Smith, E.D., "Rotating Biological Contactor System Design", J. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 108, No. EE3, June, 1982, pp. 578 - 588. 85. Wu, Y . C , and Smith, E.D., "Temperature Effect on RBC Scale-Up", "Fixed-Film Biological Processes for Wastewater Treatment", Ed. Wu, Y . C , and Smith, E.D., New Jersey, Noyes Data Corporation, 1983, pp. 287 - 304. 86. Zapf-Cilje, R., and Mavinic, D.S., "Temperature Effects on Biostabilization of Leachate", J. Environ. Eng. Div., Proc. Am. Soc. Civ. Eng., Vol. 107, No. EE4, August, 1981, pp. 653 - 663. 14. APPENDIX 1 Raw Data of Premier Leachate Analyses 2 0 6 u b c e n g i n e e r i n g 207 o_ r | o V<! O i — In Oo! o-i tV O Lni lfi' ^ 1 v? V'l A • ll'l l/> r-l/i 0-u\ vS> l/> r- r -ir, Ln (X lfi 0 v9 | E r c> LO Li f r i') cr v5 vS> - i q c < U J I n > o "3-O 0c v9 on r r-r-<> <si In i o (V) r.1 r - r— T r - i - j ol "J" vs O cr (N r~ v5> V5 ol r -rx VTl r~ T VP rvi 1— (^ 1 01 e~ OJ 1 ro (V r— rO ^~ r i tr-"3" fVl <r-r-(M f>, o" CXl ol > [Y| 1* (VI r-V9 V> t « r-TP tYl VA O-J tyi rvi u (N U\ (VI \ t vSJon o|o <ri ui O CV) ol (3 0 rvi v9 \ c / ) Q _ J 0 in ? z LU o (-—r ^. h !Y1 (S ffl o l/l H •3-l/I r~ VA (VI <T VS o l o r~ o 6 l/> r -v9 Oi rs o r-rs Do rflj — t>'| <^  r»j r-1—| VA H > o| v9 fN c r 5= V\ 0-v» ?r vS V9 r-ol Vl rs r-01 <r Ol v9 o| mpji 0.1! r-olj lJ1 O ro (T in N Vl > tr 0) > d Cr J r0 ol > v9 (Yl O (VI 1-6 Vi tvi vc > tr v» (Yl (So I~ r k IO V9 V> ty> 0 6 fl— (V\ <r r-01 (Yl In ol T vfl > r~ i i 1 VP h t 01 O Z i i I i i | .i L X Z z r -i in vS ui 1 O r" LA lA \n 6* ri t r VA v9 CVl fr H Ol o~ ro 00 0 v9 ci 1A i - ! • ? IA'| 6 T Ui 1 ol •3-fVl 0Q rv> O c -tr ro ^~ U O ui r-6^ ZT Lr, ui r 6 Vi c r In v5 ^ \r\ ol CM i i 1 i i Ol £ v^ Z o CD ac cr u o o h-i i i 1 i 8 v9 in O o CD a 0 u CN v9 cr to o-0 rS q r~ Oo i/i i — to Lfi 1 1 O cr / \ •3-r-00 tr T~ l/l !— L O cr v9 cr rv cr fv 0 (VI O r-r lo v-0 N 0 er r -(V) (« U r -K> (^ 0 ol Ln ol r l m rs ra 0 tr r rs Ul M 0 3-tr (V) 0 ° V9 cr (Y) r-<r m ^ (V> r-CX) 0 O rr M L« v9 H OJ a. LO \ d - 4 -a of oi rs C — r-~t) '0 U\ 0 "0 ti o Z rr" t> o V? 6 o z (VI J Cx 0 w r~ <j c ; ^ (Yl 0 r i -T r< s: Cr=> i d r 3-£> U-Q) LL. ^ , - j »< Li. ID J> s 6 I V 0 /_ 6 l? c (Yl V 6 If. . i ol CN Cr fN w r ^ v9 (3 2 52 c T 0 (S jD t~ H M •J ; O (V > r rj a) i s u b c e n g i n e e r i n g 208 z r ' o_ o O vS q v'j u\ 0 ; _ . | >v\ Oj C*.; <?j r~ ^ c , v9i U'\j U l I— o _ O q vi 0 v9 v5 q q "I v3 v5 In v i 0 vr, O v9 v i <r 0 ^r iw r t (VI zr r>l v9 cy> <w ti r" (VI' i . - i 0' f> <r c> 0 vP O r J c-J r~ ZJ 0 r-rsi (5~ ol 0 r i tn in cr cr o -<r O r— cv c r 0 L9 0-o--r< O i s O 0 m i.'j r -r (.••• 1 r-r~ zr z r 0 r r c>" fi~ t>-r -ZT O /— "H \ i in Q _ i 0 CP £ Z O! \3 O a: I-—7 *£-</> r-ro C>i O ro H rS O \f< rN K i <Y| vi> (VI rs 3-o" r KI f» v.9 o~ c<> v_o O TT 6 (<n r~ o r--> O r^ l 0-0 Kn 0 bo O — r -<r r -cr 0 0 CV^  v5 O i'-. r ( Lo f ri Lo zr r~ (V) vS O OO > r-^n! O ! m u> h 1 ro O Z r— r- r-\r> 1 In! j v«i -cr | T i j i i c<> X Z z .Y t-6 zr O H 0 c p i r~ fi-ts In q ff" q 6 ts rn V? cr ol \* r \ i t 1 r ; cr> rs Do rv< 3-rs q cs q TM C» r«i rx rs q zr 6 rS o< cs 0 a1 cr ZT Lo r -Lri 1.1 • c r cr zr K l tr r-rl er (VO *zr K> q r" c> q K I K. e>-ej cV zr e» (S r-r / zr zr 0 Kl 00 (vi r-In rs i zr! o° LA! \ $ rSl rS I v_9 1^ rv r In rs L0 \fi rs cr ro rS Zr M rs (NI LO N (VI rs rS vP rs r-; CN CS v5 c>> — in 50 zr cr in cr \ CP \ i ro cs O *C \ £ v_^  Z o CD Ct (£ U 0 0 k bo cN cr r>t v9 cr i ' & CP c° Lo r \ t>.. O ZT L O Ln In N m 0 O v9 o o In vS r-o z r Lr> O r~ to >--) v5 vi> rs n i VP in 0 c> n l A r-(>Q In (/< LO c< L"J rs in LO 0 ° U0 cr LO rs \f, r— L O m r~ zr Lrv z r rs rS O 6 rS o CO H M O ' (*> <J> H O i~ 0 (T r— c c J 0 cr r~ r " r— <r rr^ rs Lf> 0 in cr L I cr r 0 zr ? Q 0 u zr v9 v3 <S r -CT H r l >-Tl V Kl r j m (VI Ln C N rsl r H rs vS v3 o cr vS zr In rj <v\ L.-i ^9 vf zr Co \s\ ol v3 er CV) v3 zr zr £-= r -v9 Cx-(•'J fi-ts 1/) cn 0 0 zr Vrj 0 (VI t/i zr O vS O 0 I T in zr L0 cr 3" 0 rs rS lo 0° r4 Ln 0" 1)" 0 c> cr -c-O lr> r s vS QJ CL E \ -+-tf O H <U CN 0' h <s 1 O p"> a> ; m Ln i ? h i ? h in ' < n e~ i 5 h , s •> —• r~ j Ln V" •2 cr i 6 3 X H ) j sT • 6 s Z3 <Z 'JI ) 6 <? f J U-0> 1-1 f™ (V —«-a/ t ' i - r ai - f 0 10 0 r "5 0 H 0 zr 0 1— 0 cr - r 0 ,-s 0 00 0 6 0 — O 0 u b c e n g i n e e r i n g 209 tr v i tr-; lo rS v9 \S v i —\ • 1ft _ ! VJ> H vO Lo r o V9 v i rs V S on V 9 LA cO v 9 v i ' v i tr v i i rr v 9 v i r>-» v i I'll L / rvj ro v i ! v i CO v i O l v i c 0 (J 0 " F -v5 Lf i c* r fx/ 0 rr CO 0 0 vS lo f r o~ 1 C T l f T -v y -H v<D V ? Ln r J rvi rj-r>) fv' cy cr cr (V) o l f M O t> t>^  Ln cr r/» o( rvi O rs rS tr tr I S Lo O Lo CN O tr o l ro rs r~ cs tr r -Oa O Lo cs vS r-ro cs Co L, Cyo 0 0 cr v9 Lr cr ffl £ t/7 a _ J 0 LP ? ffl f •—' z L U \ 0 0 ca t-—7 ^~ CO h r-tn cr 0 Lo o | co 0 rr In tr r-0 T r > J f> r> v9 v9 O tr v<? (>-H rs 0 o l 0 ov c r LA r s r~ c» rs H n -1 o i l > tr o< 0 LA H rs LO O tr o4 v3 ro O cs <S i i L3 tr CXD (Yl tr to tn > • i i { u. > i 01 0 z ! i 1 j i 0 r-rr rr v 9 tr cr O ro vj> Vn rj-Ln C~ Vn LT) 0-rr 10 ' i I ( j 1 1 tp rr z Z Y-! Q i <s jLo joc er O CM (vi Oo .-s r~ rr o l r— rs 6 r s rA rr r o c-l 0 o l co 6 o l r r rs cr rv> rs LT. v i CO v9 CA ( O rsl OA 0* r s 0= c^  rO O ea r o Ln c~ tr-£ S r S <3 r— o l rA S3 Pt j i WA Lo" 3-\i\ 0 •3-u \ CVJ r o tr-IJT rs • S cr 3-LO _ r o rr (VA rs LA LA Ln ? r o Lf) £ 0 0 o< 0-1 }a: LO LO! cr OSI CS O c s ; ro LO \ i vS> r~i <N " 3 -6 _cs rs 6 Ol cr CS vS> Cr-CS. c S Lf) rf cs °? rs| vA 0* N 5 ! vS N Q 5r rvi 0-r A ui t>i 00 •3-r A 0 LA <vi -3-00 •3-i i -3-cr i o r-•3-(> o i LO 0 0 6 lo ct- 5? q 5-r o rr -cvS ro 0 h3 ro Civ9 K i — r s j i O l LO C»lo<" o(| f O °? r-(S 6 cS O ZJ-r o •3-Lo - j 1 N cn £ z 0 CO or cr u O O O r 0 to O'l L i ' . r~ o l tN v9 cr m Lo r~ rr er-rs rr Lf) r S LO r -r r O rv> L A (VA V v 0 " ) r S L A cr r l t~ c r r-v9 O vS to r~ Lf) rS -J r v 5 rr r— rr Ln r~ rO v S LO r— r S V9 v9 t r V9 P-v9 vS cS rr vS CT rr O Lo, _V5 f O v P i <""irr > S | c r Lnlco 1 i vS; v s — ; c r rr.rr i/i O O CO j i 1 O 0 0 O t r Oo O rr of r 0 T v9 r r r r O rf) ri O c s v5 r r vf) to ro tyi v5> Lf i 0 ° 0 vP CV) O t - (.,-. c~ rv'l ~T" cy> 0 ty> C ~ cr O <v> v 9 r r O cr r r r~ v9 tA L O v9 r~ v9 r~ 0 cr 0 vS O 0= v9 <r o l O VA c r — rr O 1— rr H O c x E tf -+-0') •3-0 r 3 r _o L'i o< 6 0 z f S 6 0 OJ O cr J Ql O O a v9 J a' O 0 r s rvi N (J » 0 L> Q) O r v3 ( f> -• V 0 s h r -> (VI _D a ' L i -O S> U -r -(U ol _ D <u U -r-l _ D u. C o r s j5 (!) LL. cs D r vS 0 2. H •> r 0 E 0 0 1 0 r rs 0 X O (V) 6 X L5 ' > 0 SZ O >. <3Z *>• r~ - ^  s-1 > u b c e n g i n e e r i n g -* 210 re x i -J v i v9 I 1 1 rr - • "5" v i - • vS v5 rr VI rs v i r -v9 VI r~ v9 r-v i cr v9 r-v9 q c r v i cr vs 60 v i Cr v5> v i 0 r-' LA cr V Q cr v9 V5 rx, vs L1 i A v9 vS u ° t VS J ConoJ. l/l In r l r l T cT CN rvi cr O rr rS rs rr tr LO F < cr vJ3 V/> vS OO v5 r-r-In rr \ S •J' r 0 r< LA r 0 0 ° r r r 1 cr vJl 0° cr vS rs cr vS cr tr LA 7\ r s v i IT C r~ rr v5» Cr rr v9 r-vS 0o r-V S O O o l (y L A CS tr rr r-v9 ~> rA vS tr tr VS vS O tr-•*v tn £ CO Q - J 0 to 1/) h i* rA cr ro LT) Ln In vS vJ r 0° CN r~ --cyi r4 CT f> \A 0 lA r~ £2 O cr (VI iA 1A rvi cr Q ^^  H lA tr vS r-r -cr O cr 0 r- L i cs 0 LA L A U I cr cr P tr cr vS O tr — -CT V.P • - ... — - —- 6 r - (vi 0° Ln O r~ cv*i CP vS rS r— O v9 r-cx O r-O O r-( S 0° r-00 tr tr tr \S O vS oO CS v5 VA cS vS VA V\ v5 O c« v9 Cr* vn Lo In 0 CO Ln VA OA Ln CO rr LTl 0 Ln a u. > rs lA v9 LA L A tr v i V9, rj-cr v9 rr 0 0 vi) • r ~ 0= rj~ O •rr rvi Lf) O tr r ^ rv> t f cv rH rA N. r ~ In rs v9 lA rs: vS CA vA > > ol cS ol O iA 0 0 tr 0 0 rr 1— Cr-rS rs rs CA 0 ! ? <*> £ 2 Ui tD O aa h-Z i « O 2 .... ._. -.._ - -2 i <o X Z o rvi VS CO Vn r f rn 0' Vri CM m O V-A _ J O in rr rO 0 0 IN'l "in rr & V ? po O !•> v i t r rs cS rr r-rA rr oo to CO **v? C N CM cr-r-J O d rfl •a-s c-m -3 r v rO P-rw 00 r — rs r-J rO C*> O CO c-<_ CS rf 2 0 0 to 0 LO* po CO ro vT (S_ rs I N I CA CO -3-V " r* rr v£ LA r~ Ln C O cv in cs ""0 v5 c s Ln q tr r— "T cs. CO 0 ro rS ("A O CO f A 0 Q r s ro •3" f A rA rf rA r-l C X 0 i > cS Vo ro" r s O rf cs V A 1— -3" rs 17" rf o< VA r f cs vS rf cs •or v-r rf cs Lo 3-r-s LA cs cs. CS C v j c s 2 r-tr-rv> o •o tr rO « r~ ,0 -3-m lO zr ra rr O vS •a* «"A "3 rA tv\ (J* cn tr "o CO* zr c-s in Cr rs C° rs rr 0. 00 r r f V \ vS> rn O r o •3-Vi <yi vj> rs rr-oi rS. CS CVl 0° CP ro 0 O (V* ro 0 rf LCl r> CS (« o l O ro O <5 C N A V ? (VA t>° rf A A 0) v9 r r O CS. o l Vri o l t » rf cs c J VA O l rr-LTJ CS 0o C N Os " H VA C S en Z o co CC cr u o 0 k a~ r -*A cn 0 0 tr tr rr r-*3-§ l—1 r S L O v i )-T) 8 m cr 3" v5 VA on LA r r rO L*n rvA LA rA CO V? r s VA O ro LA VA rr rs O rs rA C/A VA -u> O co rd m rr o< v>. Ln CT U7 m &o rr £ a rr r" ro 0 cn m vS> cr vn tr-rs-in XT CS & r-m ro tr m 1— v9 r< rS vi> C S rA cS V i C v9 cr 0 tr CY) r - CT* a 0 u > vS» m rS r~ r r a* co *r cr rs1 cr r tr, L/l rs cx O rvi O rr 0 0 rs 1— r-cx> vS CO c*» cr tn r r 0° v ^ <n rs v9 vn Ln Lf> O V_9 rj-OQ rA LTi rA Vn 8 rS Vo Ln rs vS cr r-r rfi L ) ;-\ ) (v'l V 9 C S -v3 H r S rA H H C H O rr \ — CL £ tf to -4-tf o r j -C « ~c~j 3 * 3 -H v. Q r-H .4 3 Z C5 X tj-T> c 0 0 _T 0 X to 0 0 cs-r( O r LA CS a X Lf) h t» c h 5 Hi £ Hi rH c cr OJ H H eJ % 3 h v5 r4 o> er-rs a1 S r "i h v5 O 1 i h 5 fV> I 3 r-6 O rv) oJ ' 5 "3 vJ> o J h M ! 5" h CSl cv CT - | O 6\ u b c e n g i n e e r i n g * CXT o_ .,J» \S OQ vS r-v i u r-vS l/> °? 00 v i r-00 00 -v9 v9 r vi °? vS 1 v S v i r vS Cr \ 9 r -v i r v i q r-rS cr v i rJ r-0 O u CXi v i v i VA cr vs ~r r~ 00 v i v9 0° ~s • c o vj ^? cr v9 T f> 0 ° LA o tr bo O 0° e In CO r to rs" VN r-O t r H Nj> 0 CO r-lA O in 0 <r v5 rs cS M CT O O tv\ r-jbo Wi cr 1 O r tN 0 O O Cr O O vi tS 1A r r Vn v5 cs LA v5 O Ul cn 0 0 r r LO tn I in a _i 0 Ul I- 0 cr ol M 0~-tfi Csl Oo O vS r0 Oo \S o <y oo r-cr v i r-O r O r— 6^ S5 «S v S I A 09 <s9 in v5 Q v3 S 1— r~ &° r* Cs \S cs: t r If) Ln LO co rs" v^ 0° v9 v3 oi r— in r-t r in CN 0° r-r-vS on O v J o( CM O ff * J O t r m VA LO O -M ks m 0° Oa tj-\i\ or t r 0 zr 0 -rs O zr 0 t r t r 60 t r I" Oo 0Q 0b c° > rr tj-ttr t r C tr t r V ? r C c t r vS cn fA S O c% v5 ol tr t r r-t r O ti- t r O CT-vS Do oC LD O 00 ro 0 r-rn v5» LO Ln a > O A m — t r ro t r on m cs vS — 6 Ln O T m C S t r v5 CS r r r— er O v9 CO =r vi> tr 0- in v5 £ 2 Ul \D O at z z 1 O z -• - . . . . . . . --2 i •a a: Z V A CS <-< c4 o r-vS CS o c^ cs cS o C-xl INT. cs cs cs q rs I — erf rf rs of 0 r~ Vv r1 \ A "tf v i rs 0 irv 0 vc 0 v9 V A <d O Cr ro v i q cr cs v5 rri rn CCi cr v i q rr cs cs 07 cs CS vS rs \ . ' \ rt r-0 r * r— in r-cs VA rs c^ v; cs r-' CO in cn t-ri tr rf In Lfj V A r-r— va 1-7 c c-v Oo z P- t r rS v j cs o t f cs °°. rn rn t r c \ rs t r cs cs o cs r< 2S t r fr 0 rs r>l Cr O of v_9 v^ t r C P 00 v9 cr O tr cs cr O c-v cs r £ 0? e~ \A O t r 0 Q in t r =r tr K } rf cr H cs O J cvi O rs H O cs Vn CP O CO ~Q 0 en e z o 00 CL <x u O O 1--10 a O CO tr-CP t r t r vS O tr-ot O V 0° I-o s 0" LO <r 0 S t r cr cr TT" v9 *-rA t r o~ ro v9 r- 0 r r - r-to t r cr r- CS t r ol vS CP 0 t r CA o l fA CS r~ o l Lf) a 0 u v3 Ln r-r 0° CP C P cr cr cs; % O cr> 1— LO cr m cf cs-cr cs 0 if" vS ty> r-0 ° H IV) c< 0 ° m cr rN >j\ rs v-C cs CS rr cn t r ot vS oi r -tr 0 ° v5 rn CS tr v3 CS O v S r r <u CX £ \ o t r t r 6 a 3 'i N >-cj t r rH »tf . 3 <E bo ol rG 3 <JE •3- r-ox - t i v/i 3-% •/> • s Si • V Lrv rs S>. ft! rs QJ cs1 -> • 0 O Lo -f C cr -r 1 0 •rJ -V 0 0 c r 0 O CS 0 v^ rs O O O 0 ol _& O 6 0 "2 cr jO 0 72 v3 —• _ c 0 c 2 M 0 r-H O 0 T2 0 •3 c -! 1 t r ii u b c e n g i n e e r i n g — ir , : 212 or cr r— v i vi r-vi ZT vi vS vi v i vS v i tr vi zr "vS vi vJJ v i v i SS s i r-vi v i lA v i VA vi cv vi tr vi ej v3 zr * TS C tf o TT 00 In r.0 tr c< in zr ro t r in r-O cx 0 ° 0 cs X O cr-^^  O l^ 1-1 0 IA CA 0 N O r-O 0 r-0 e"> Ln in I zr v3 oC O 0-0 0 0 Lo Vn O In 0 sS O > O zr »A O rr 0 cr O VA O 10 io cr 0 r-X i \J tfi Q _J 0 CP CO h O r v3 ry> ZT cr rA m (P O T cT cn zr in O c r v5 r r 0 O CT" r-01 r~ 00 cn cs. 0 r-C» cs CP LA cr r~ zr 0~ r~ LA r [~ O Cx> v9 cr rv) v3 cs" sS D« O r 1 CA CXJ v3 sS 0 zr cr r0 LT) v9 r -cr bo LO 00 Cr t» vS tr cr-C tr vS r~ tr O v3 00 s9 s9 cr 00 6o cN Oo Oo m m r~ vj M O in O r--J Cio CV ZT vS in Ln u~-vS zr O cn P Ln Ln 00 zr Ln 0 ( CP Zr zr cn r-ro 00 Lri cy> ol Ln a: LL > LO oi rA r -ei 0° 00 Ln cx s9 O fA r r rs-v3 ro s9 D° O in r— rA cr in zr OA ZT vS on O zr v^ 1 O r- zr zr Ln 0 ° hi 0 N r- v3 m o< 0 et rO ol ^V *—' z O \-Z z 1 m O z zz I co X z Z .y r-vn rs cx in tr cs w> rz v5> 0 cr cn r--=r rs tr-r s U-) r — CS N S CS. tr cs <-s m in rs. x/\ cs r-^  ex rH 0 c s 8? c s Ln r-.5 0 CS m e; 0 0 lA IV) O t r r^i i n cs Cxi rx 0^ 8 r-rs 0-t r rs s? cr cr 0 ° Ln c-s v.» cs' f^ vi (S> VA d rS rn v9 tr fsA tr cs e( rs 0 O { £ LA. cvi CH 8 & 88 0^ C' r — r s CO (JO ct> r s VA c ^ r s c n of rTi cr1 c s r <-n O rv-> c s cvi tr o 1 J-_ ox-en 0O r-cS -r vS cs c s 06 C-S Oo c s cx CS tr O rr> ir rr rA rS FA cr 0' rS O W\ tr rz rS rs. rA rS O cv c s rsj rA rs O ; r~ r~ rs ^A m cr PA vn rs rS c£ Oj rvA N «c \ cn £ Z o CD ac cr u o 0 \- 00 LO in O O CO 00 v? r~ vJ zr cs. Ln O cr rr rA 0 -zr LA 1 cr Ln oC zr r-cr s trr es r~ a 0 u er-rs C" ei 0-Ln 0° Ln ot 0. LO tr 1— vS CA vi> <A cr tr tr tr e s <s zr m rr tr cs-O 0 CH vS m cs1 Ln O in cs-0 CS Ln t r co o( r-N r~-v3 CN r~ v9 ol \9 ( P A l 0 on zr >n H r~ O l L0 l-\ v5 v5 Lo rc Co v5 O n ol ~CL £ \ - f -tf Q tr 0° r a 15 J a tr 0 CD u 0. a 1 O in I r> h vj> i 6 h > r2 o< b K1 \* 0 CS cS v -_o a/ u_ tr Q) L i . CX S i u-6 "0 L u 5 cn QJ U-cr J> Q) U_ H r l c" rs j6 a; )j-c s L i -rs sj 0 X '^) Lf\ X vi> C -> 0 c V/l D v3 In 0 0 rr •> y 0 H e l 1^ D X V? d -^D r~ r~ cr 0 6 X 0 c^  ol V CJ = 1 1 In > r— u b c e n g i n e e r i n g 1 2 1 3 or Cu CY) vi vi ro vi Lo vS v3 V5» vi v i vi vS vi v9 vi N vS tr-v3 rA vi CX) S O O vS c 0 U c/» > N O cr lo 03 CN IT c r 0" t j -fV-> r-rv i rs 0 zr O r- v> VA CM \A Cr rr 8 N V) cr t r O cr 0 0 CO. r- O rr \n cr cr V) cr cr cr CT tr cr LA r-iA vS r O vS cr tr rS 0 r-v9 O 0-60 0 cr r-r-cr tr CT v. en £ CO Q _ J O If) LO h t r r-r-LA c\n t r r-Vn In Ln r-v5 r-t— O v5 v3 tr v3 er r t r O r t r 1A in S i A iA rA r~ I A tr r-in vS r-VA C)0 vS 0° cA iA tr iA tr Lo LA r~ ty O cT rr O OA VS O tr vs cv) CA r tr cr io ro tr LO r cr fA rs tr > vS> $ vs to LO > r--cr > v9 T V O N \J 1 1 1 1 j at I z Ui \ D O CY Z z m O Z i 1 i I-1 -i en ar Z rs O vS V\ LA t r cr vS vi rs 0 cr rs i-1 sr' cS LA r~ & rs 0 rS d r s O LA CN CA \A rv c~ r-— t r r-vS vS cr rz (V) Oo H rr cs LA rS cS t r r 00 r 00 r~ CA 6 cs vn rS rS Oo cr rr •r cs CS ev-VA r-3^ 2* O LA rs tf O rr Lr csl vS z 1-rs LA ro rS «f \ CT) E Z o CD a: cr o o 0 r -u> O O C O Q 0 u CM m o cs cs v9 O <rs vS CN Cr W= • \ A cr rr r-rs O O O r V) "3- r r~ cr 1— O C^ A tf t r v3 r-r- t r v9 rA tr r- O t<> er CP O r-rs 0 0 CM Lf) rr CD C L . £ \ -4-c? £ T & ce F 3 CS . 0 a : cr c*r V - V) <2 o CJ vi! vi <s. (• 0 cs •5 • •a CH 0 ) 5 C3 r r-> - , _ C 0-1. 1-i ! 5 0 2: 0 ° ! i r tA •3-r V Q J 8 a' h' rs 0' 1 O fS a t n 1 r er i r 1 . v9 /-Nl r^  rS a 1 1 cr 0 cr 1 O 61 5 <Z. (VI (Nl CJ! -5 O rxi CJ1 r 1 L/v e m -t— Ci 0) CA c TS r~ cs. V 4) L A 0 0 "0 O CA O rv -r 0 I OO J L« 3 15. APPENDIX 2 Listing of RBCs Operational History 214 Appendix 2 RBC Operational History Date Inf. Q Reset Q Ave. Q #1 T. #4 T. COD Ldg. BOD Ldg. Observations/Comments mL/min mL/min mL/min "C ' C g /m 2 2 g / m z Oct. 3 /83 52 90 • 2.346 Foaming 1 s t Observered Oct. 5 88 150 89.0 14.0 14.5 4.002 2n t* pumphead added tor influent Oct. 7 56 50 103.0 13.5 13.8 2.352 tubing split, cont. with one head Oct. 12 33 150 41.5 13.0 13.6 1.460 Oct. 14 0 160 75.0 13.0 13.2 0.000 pump tubing split Oct. 17 52 150 106.0 12.5 12.5 1.903 pump tubing split Oct. 19 0 140 75.0 1 1.8 1 1.8 0.000 pump tubing jammed in pump Oct. 2 1 0 0 70.0 12.5 12.5 0.000 pump tubing jammed, no fuse Oct. 28 0 0 0.0 1 1.0 1 1.5 0.000 restarted with one pumphead Nov. 4 0 0 0.0 0.000 Nov. 10 0 200 0.0 9.0 9.0 0.000 GRI bellows pump installed Nov. 14 320 320 260.0 10.0 10.0 3.850 Nov. 18 RBC FLOODED pumps knocked out, disc motor OK Nov. 25 0 510 0.000 restarted disc, covered in mud and oi Nov. 30 0 0 255.0 0.000 poppet valve broke, no replacement Dec. 2 0 405 0.000 rain guage frozen, snowing lightly Dec. 6 0 520 202.0 0.000 loss of suction, inst. 3/8in inlet line Dec. 9 345 345 432.5 13.559 feed line collapsed, growth reappearing Dec. 13 0 370 172.5 0.000 poppet valve broke, discharge side Dec. 16 310 310 340.5 13.820 Dec. 20 125 125 2 17.5 5.085 line partly collapsed, rebuilt Masterflex inst. inlet almost frozen solid Dec. 23 0 0 62.5 0.000 Dec. 30 RBC PARTLY FLOODED Jan. 3 /84 RBC FLOODED flooded Jan. 1, both pumps and disc stopped Jan. 20 0 475 Jan. 24 0 0 Jan. 2 7 0 560 Jan. 31 0 0 Feb. 3 0 0 Feb. 10 0 300 Feb. 17 0 465 150.0 Feb. 2 1 0 275 232.5 Feb. 24 0 640 137.5 Feb. 28 610 610 625.0 Mar. 2 0 500 305.0 9.0 9.3 Mar. 6 RBC VANDALIZED Mar. 7 0 0 Mar. 9 MOTOR BURNT OUT Mar. 16 0 600 Mar. 20 565 610 582.5 Mar. 23 565 550 587.5 Mar. 2 7 SPROCKET FELL OFF Apr. 4 0 505 11.2 11.2 Apr. 6 0 0 202.0 13.2 13.2 Apr. 10 0 470 Apr. 13 470 470 470.0 Apr. 17 460 460 465.0 Apr. 20 450 460 455.0 Apr. 24 430 430 445.0 Apr. 27 425 425 427.5 14.0 14.0 May 1 420 420 422.5 May 4 385 385 402.5 14.1 14.4 May 8 385 385 385.0 14.2 14.5 May 11 250 385 317.5 May 15 355 380 370.0 14.0 14.0 May 18 340 385 360.0 14.0 14.0 May 25 325 400 355.0 May 28 380 415 390.0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 30.396 0.000 0.000 0.000 18.476 22.306 0.000 0.000 0.000 14.678 17.209 18.414 14.487 12.457 1 1.945 13.987 16.078 1 1.453 8.786 10.506 7.527 10.123 surface leachate black, foam in pumpwell inf. pump lost prime poppet valve broke, growth reappearing poppet valve broke, no replacement installed new valves and valve springs pump lost prime pump lost prime, inf. checkvalve inst. checkvalve fouled growth coming along check valve and inlet line plugged drive chain knocked off restarted unbalanced disc drive motor pulled off mounts and jammed new motor and elec. breaker installed rapid regrowth cleaned inlet screen and checkvalve disc left stopped bellows nutrient pump inst. Mar. 30 feed pump stopped, removed for servicing rapid regrowth cleaned inlet screen feed line to bucket partly plugged new line to bucket sump installed inlet screen plugged, very high SS very heavy suspended solids solids washed out, thinner growth remains ammonia addition slopped Jun. 1 400 420 407.5 Jun. 5 440 405 430.0 15.0 Jun. 8 0 410 202.5 16.0 Jun. 12 360 400 385.0 19.0 Jun. 15 340 410 370.0 21.0 Jun. 19 410 600 410.0 19.5 Jun. 22 590 610 595.0 Jun. 26 570 600 590.0 17.0 Jun. 29 450 595 525.0 15.0 Jul. 3 460 600 527.5 20.0 Jul. 6 610 610 605.0 16.0 Jul. 10 660 650 635.0 17.0 Jul. 13 620 650 635.0 18.0 Jul. 17 650 990 650.0 21.0 Jul. 20 830 1030 910.0 17.0 Jul. 24 820 990 925.0 19.0 Jul. 26 10 1070 500.0 Jul. 31 0 1050 535.0 25.0 Aug. 3 0 940 525.0 21.5 Aug. 8 520 925 730.0 20.0 Aug. 10 870 930 897.5 18.0 Aug. 14 920 920 925.0 18.0 Aug. 17 930 1220 925.0 17.0 Aug. 2 1 1060 1 120 1 140.0 16.5 Aug. 24 260 1500 690.0 18.0 Aug. 28 2 10 1430 855.0 17.0 Aug. 3 1 385 1480 907.5 17.0 Sept. 4 1520 1520 1500 14.0 Sept. 7 0 1 150 760.0 17.0 Sept. 1 1 0 860 575.0 15.5 10.032 6.732 SS have filamentous floes, normal colouration returns 15.0 7.894 4.739 SS settle poorly, inlet line cleaned 16.0 0.000 0.000 checkvalve plugged, inlet lines cleaned 19.0 6.793 4.039 fluffy filamentous floe remains 21.0 6.273 3.366 settlability improving 19.5 6.777 3.641 settlability vastly improved 8.142 5.222 settlability good 17.5 9.200 7.097 solids increased but settle well 15.0 4.293 2.862 settlability good, inlet screen very plugged 20.5 6.969 4.375 inlet screen and valves cleaned 16.0 8.272 6.277 cleaned inlet screen 17.0 8.4 15 5.287 18.0 7.366 4.873 inlet screen cleaned 21.0 7.352 4.739 inlet screen lost down pumpwell 17.0 8.914 4.880 inlet valves cleaned 20.0 6.470 3.961 inlet valves cleaned 0.068 0.059 inlet valves cleaned 25.0 0.000 0.000 cleaned checkvalve, discs lost solids SS 21.5 0.000 0.000 checkvalve plugged, installed new inlet screen 21.0 2.964 2.699 inlet screen very plugged, SS mostly washed out, 1 s t stage lost most interior growth 19.0 4.541 2.480 19.0 4.306 2.318 cleaned inlet screen, slow regrowth 19.0 4.938 1.228 inlet screen cleaned 18.0 4.706 3.371 inlet screen cleaned 19.0 1.552 1.747 inlet screen cleaned 18.0 0.939 0.517 inlet screen cleaned 18.0 1.444 1.086 inlet screen cleaned 15.0 5.472 inlet screen cleaned 17.0 0.000 0.000 replaced valves 15.5 0.000 0.000 inlet screen plugged, pump bearings Sept. 14 0 0 430.0 17.0 Sept. 15 0 500 Sepl. 18 0 850 250.0 20.0 Sepi. 2 1 0 690 425.0 17.0 Sept. 25 0 1060 345.0 16.0 Sepl. 28 610 750 835.0 13.5 Oct. 2 0 900 375.0 17.0 Oct. 5 465 625 682.5 14.5 Oct. 9 230 0 427.5. 15.0 Oct. 10 0 1280 0.0 Oct. 12 0 1310 640.0 12.0 Oct. 16 1250 0 1280.0 12.0 Oct. 17 0 1300 0.0 Oct. 19 1280 1280 1290.0 10.5 Oct. 23 1 130 0 1205.0 10.0 Oct. 25 0 1200 0.0 Oct. 26 1 100 1020 1 150.0 10.0 Oct. 30 PUMP BROKE 5.0 Nov. 2 0 1320 0.0 8.0 Nov. 6 1 160 1200 1240.0 10.0 Nov. 9 810 1 1 10 1005.0 10.0 Nov. 13 770 840 940.0 9.0 Nov. 16 385 460 612.5 9.5 Nov. 20 0 0 230.0 9.0 Nov. 23 0 980 0.0 Nov. 27 973 500 976.5 6.0 Nov. 30 7 10 980 605.0 8.0 Dec. 4 950 980 965.0 6.5 Dec. 7 725 725 852.5 5.0 Dec. 1 t 0 960 362.5 6.0 Dec. 14 0 950 480.0 6.0 Dec. 18 0 960 475.0 1.5 Dec. 21 805 920 882.5 6.0 17.0 0.000 0.000 pump would not start, removed for servicing 0.000 0.000 reinstalled pump 20.0 0.000 0.000 pump stopped, restarted 17.5 0.000 0.000 pump lost prime, crud in the poppet valve 17.0 0.000 0.000 changed inlet screens and pump cam 14.0 1.867 0.641 inlet screen very plugged 18.0 0.000 0.000 inlet screen fully plugged 15.5 1.200 0.879 inlet screen changed 16.0 1.290 0.945 inlet screen changed, fully plugged pump restarted 12.0 0.000 0.000 pump prime lost, gummed up valves 13.0 6.825 4.388 assistant couldn't restart pump pump restarted 11.0 5.299 2.342 growth reappearing on 2 n d stage 1 1.0 4.034 1.254 pump wouldn't restart pump restarted 10.0 3.993 1.551 changed inlet screen 5.0 pump removed for repairs 8.0 0.000 0.000 pump reinstalled 10.0 13.607 8.387 10.0 5.856 2.819 inlet bucket sump not staying full 9.0 6.029 4.343 heavy foaming in 1 s t stage 9.5 2.853 1.617 growth getting thicker on 2 n d stage 9.0 0.000 0.000 pump stopped new twin bellows pump installed 7.0 7.998 4.991 8.0 5.623 3.238 7.0 13.110 «8.664 5.0 7.156 *4.676 pumpwell flooded, couldn't clean screen 6.0 0.000 0.000 inlet screen left high+dry after flood 6.0 0.000 0.000 inlet bucket had again been flooded out 1.5 0.000 0.000 bucket tipped again by high water levels 6.0 32.651 '22.049 lots of foam, white growth on 1 s t + 2 n d stages ^ CO Dec 28 350 875 635.0 7.077 •4.725 heavy growth in 1 s l stage, growth spreading, high SS Jan. 4 /85 0 820 437.5 4.5 5.0 0.000 0.000 Jan. 8 570 770 695.0 7.0 7.5 0.000 0.000 Jan. 1 1 250 250 510.0 6.0 6.0 1.868 1.260 Jan. 16 0 890 125.0 6.5 7.5 0.000 0.000 pump throughly cleaned Jan. 18 1 15 1 145 502.5 8.0 8.5 0.731 0.393 installed full stroke on front pump Jan. 22 1000 1 120 1072.5 6.0 6.0 13.020 7.530 Jan. 25 670 1 170 895.0 7.0 7.0 4.844 2.111 heavy, healthy growth on 1 s t stage Jan. 28 1093 1 160 1 13 1.5 6.5 7.0 6.558 3.246 Feb. 1 750 1 180 955.0 6.0 7.0 3.510 0.743 effluent line plugged, almost flooded out RBC Feb. 4 1 100 1 120 1 140.0 5.0 5.0 5.016 1.617 good growth on all stages Feb. 8 635 1 145 877.5 6.0 6.0 2.858 0.991 inlet line silted up, cleaned Feb. 12 1 145 1 175 1 145.0 4.0 4.0 6.939 Feb. 15 775 1015 975.0 7.0 8.0 3.604 1.372 overhauled pump Feb. 19 545 580 780.0 6.0 7.0 3.826 2.322 Feb. 22 225 0 402.5 8.5 9.0 1.465 1.330 inlet line plugged Feb. 23 0 1200 0.0 cleaned inlet line thoroughly Feb. 26 1 1 70 1 180 1 185.0 6.0 6.5 9.372 •6.037 Mar. 2 1 170 1 180 1 185.0 7.5 7.5 6.529 •4.107 Mar. 5 1 160 1 185 1 175.0 7.0 7.0 10.544 Mar. 8 1 105 1 125 1 163.5 8.0 8.5 8.420 Mar. 12 1000 1 100 1062.5 6.5 6.5 6.510 Mar. 15 630 0 865.0 9.0 10.0 2.986 pump and inlet line gummed up Mar. 16 0 1 160 0.0 cleaned and rebuilt pump and inlet Ii Mar. 19 1200 1220 1 180.0 8.0 9.0 5.940 Mar. 22 1 160 1 180 1 190.0 7.5 8.0 5.429 good growth on 1 s l + 2 n d stages Mar. 26 1 130 1 180 1 155.0 8.0 8.0 4.170 Mar. 29 840 1010 985.0 8.0 8.5 4.234 Apr. 2 700 1 160 855.0 9.5 10.0 3.150 Apr. 5 1000 1 160 1080.0 10.0 1 1.0 3.960 Apr. 9 11 10 1 130 1 150.0 1 1.0 12.5 5.062 Apr. 12 1 130 1 130 1 150.0 10.0 1 i.o 2.712 Apr. 19 0 1260 565.0 1 1.0 1 1.0 0.000 Apr. 26 50 1230 655.0 0.159 • BOD estimated from COD values 16. APPENDIX 3 Raw Data of RBC Process Sample Analyses 220 u b c e n g i n e e r i n g 221 o_ r-m r r-i— r-r r-l A i ^ M r ! rr! v_P r^  s9 r rvi r— LA r-tr r r r-c r r- r- Cr-'. r r r-Ln r 0° r— rs r-o 3 VJ J O rs rs c" C o tS (-• 0 ' Ln H r vi> v^ (>< (T Ln l Tl' o r? C! n'i 1.,. c v> 0 (•'• Ln O lo T r N O t) cr-r-rvl Ln cx> rr /^J m 01 H rs Ln oi r~ r cs-<n $ h (VI vS c r LO r" r . re O 1 ! 1 1 i-i cn r l o i I; H v5 X r co Q _J 0 LO £ Z LL! VD O c£ J -z. c/> r~ i r • rs X rs r-t r rs r O" rs rs — -r-rS ! \ i O ; rs T r r cn rr s5 Oo Ln rJ \ \ o | r J VP H i If) O z i I i 1 | rJ 0 ° LO rs Ln <-l <v r~ rvi Ln cn Lo rr i — tO a~ m Ln' rs r>l rr oo'rr rs; c l rs! (vl -| i i . z> i c rr Z z r-1 i V.0 VJ? rr rs CS r>'' 6 VC q e i i i i — i — 6* Lei rs vS rs cr ... er r<"> o i~ rS Ln r r r-rs Lo vi CO Lo cr-oc" m VB fA CY) rr Al r--£ Z o CO or cr u o o r-Ln Ln Lo m rS rs m (V) rr r~ r cr o rs Lo CP rr cr v9 vS cr cr 0 cr r rs O cn X v.? v J r-LA rv i i — ir)o 1 ! [V) u> O o CO t— O Ln v5> ry> v.9 rr u v.? rs cr cr- r4 f\ O vn r~ rsl r>'> O v9 LA Or. v9 r-rr Q 0 u o vS Ln rs vS cr r-CT r-er-rs Oo i n o vS O cr rs r< O <r •a-r-rS Lo rv") O rr O O rr rs ^= cr r~ rs r r (A 0 rs r 6° 10 \S v5 cr r- VJ> - r VJ v9 P Lo N LA Q . £ —f-Q tr o rr OJ i. ? UJ V-&_ 5 I ) 3 "5 rS , 5 cl Ln ) 3 h Lo I 5 r s. <j oJ. QC 10 5 al if 0 ~Z u . W cr i r> CU. La It* ) 5 h 0; tS 6 - r 1 r* rs ' J r> W VJ nt ) y "i 1-cu. _£ VJ rS rs i r r^  rV <f 5s . 5 5 r> ex. i 3 v> r r eS rf 6 Ln ' (5 o» a -iii o> Ln 5 % -V r 6 d. a~ UJ U r " V cf o -V V 6' u. O rJ 6 5 <£. a) c cr; ><5 5 s2 a -L V,.' 7 ^ <r . r: ii-' C d -r <J> 6 -2 zx-& V? rS , ej1 3 <z o> cS 6 -V VA 1" rA 6 <j w c ' c <z cr c5 - r i'' z. rs "5 VA o~ a>-LU a? —1 rs "5 a/ Vr I -r — r ~5 V/i a * cy a? r— »' VT cX 6 LA r —«' Vi 0 V') 3 -JJ-u> cT <y VA cl u b c e n g i n e e r i n g 222 tr Cu 0 r" v9i v i tritHrHoo v i ! v i ! v i ' r-I i O r-vS v9 r~ v i LA v i V5 r- cr vsj v i rO v i rA v9 cr vS c<q !"" C v ' l * i r O r-i c" 0 f V v9 v9 r-l ro rl Ln! (Vi rJ H LA 0-ol rv oil ^ ! tvij rsi I cr A1 oJ rr cx j C M LA rr f N tn L0 H r-T v9 O L n Lo O l ro v9 LO rS J CS rl OJ cs rO (VI LO rs r~ >jl v9 r~ f vv CO D _ i 0 LT) /o ? - — ' Z LU v£) O CO. Z LO h o-CO rA cr LA T cv ol r 0" r to LO > (A rr-r-rA r 1 q vs \ o. vi> O rr co c-i O in" 6 in Lo Ln v i LO H o ro r~ r lm o rr \ Lo fO Ln r-io i- : Z i ro O Z q rvi O* rv r-7t~ rr O v9 cr r o4 LA 0° oo c oJ 0" q 00 r~ ol vr> r~ ol VA r~ rr v i i — CO cr r -ro t> |-q CO o= q VA rr q do 0°; ro v i £ .*'• j 6 ° i rr r~ v9 r rr ol r~ VS r v5 I -rr 6 r-Lo LV Lo 6 rs ro 6 ZZ I <n X Z Z .y r-oJ rr r-r r L I r v9 Lo" q cr V. o rr cr r-N CA rv rr Ol CO 0° ol 6 Ln CA Cr 0c LA o t> Vn rr cr vo oi rr v9 d rr rr ol ol 0o rr rr ryi VO o °. co IS v9 cr UI c< r~ O r-O LO O rvi ro «i CP v9 o r-6 rn rr o! co C^  Ol rr-rr cs oi ro q rr Ln q LO LTl LO r -b° LA v i V? r-r-LO co Ln" o! N ~ k rr! C A i i q LA ol q rr vS ?2 v5 .•Yl o (VI oi rs rs rr cr LO" r-; L-) oi \ CJ) E z o CO cn cr u 0 O V-r~ In Cw o N ol c>o (Vc ol rr ol ro ro r— ni rr ol OA oJ ro VA rv| LA CP v5 — — i r o— ol of rV) r>i CV) LT) r~ LA vS rr o : cs Lo O CO K) VA r rvi ol rr rr ol Q 0 <j Ln VA rA rr r— 0 v9 (V) cr rr —-0 q o ol ol O rv> o| VA rr I A ol O o q r- rr o- ( S o» O r~ r r iS r- ro rs vf1 L'» O (VI o cr c° ol e£ CV) ot cs rs LA Q i " Q . £ \ -t-Q d a £ v/> e rA Ol -5 VO 0> 0 f -rvi " € a' \A cv 6 -r 1-0 o'' o> i LA A J . a* u. 1U 0 (V'i 3 LA ft' O O V" cv -*~-. Si VA 1) , CJ e rr 0 0' c -v v. •L A ' T3 0 Cf C - f LA 0 6 -r n 8 0 0 LA rr lo ! O' 0 "£ H VA -/** 0 »; r "? 0" r -ti G s r -<r <s ^ _ O l V c C - r r oJ 'l o c c —V I" -1-0 1 f-1. KJ 0 a< eS 6 rr ij 0 0 t —V o cy c 8 r 3 C5 To n r -f O i> ' c cs rr r~ o I" O 6 i. o» <r •V o •b ro cr O o l ~T> 0 ?• oJ - f u O rr rr _c rr cy D O 2 rr cr u cy O rA u IV O 7. / s ample C A R B O N O 9 / - O N I T R 0 6 E N ! S O L I D S O s A O C o no!. pH Ucvre C O D B O D s T O C T k N N O 3 - N T S S T V S S T S V * Dec. \i 32.7 2 G . 3 . 0 . 5 6 . Cec. IC q 5- 11.3 ! H . l 0 . 2</ ii D e c ic T l n.7 17 T O. 3 2 I 9 1 - Dec. 20 \12-H ? 2 26.S 22- / If b e e 2 0 \\Uo TM 2S.9 •2 .? V D « . 2 3 120. C 1 2 0 / I " MO. S T . ? j * l» Dec. 2 3 102.1 501 / "10 HO. /HI o.n o.n ?i FeW. n 720 Feb. 21 27? /zn 0. 32 • * FeW. 21 270/201 ?Xo O.ST 230 /1?0 O. IS-ii few. 2/y 112 >03 C ' / •• FeV>. 25 230 <o. 1 F-'T. 2^ : T . 6 H<oO 52 <o.i 1 * " » M a / . 2. n.s Hov. 1 IS? 20s- 23.0 I5' U Ma*'. 23. 313 / 20I IG5 30.4 /31.0 2^.3/22.15 l» He / . 23 IrJO 2JL2_ to 20.5/20.5 "I" M fipv'.l C 212. G l fa/.o 230 S'5 S3-I tl 250 101 II..5 IO1?.! '» 2C0 1 H 117.0 in.o 101.«/ 10? 1 n 0.07. KJfy.'i 13 113 103 52.0 55.2 O . 0 2 l« ns n.sliro/0 VI V1'1 n mi ! 13 31.8 35.? 21.0 /25.0 11 » W M tl : 2 0 1 2T ! ' I T ? 2^./ l\ 196 3t9.3 0.03 II \H5 36.6 0.02 ro u b c e n g i n e e r i n g 224 CL. PA r- r-t IA VA r— ... — ..... .... r r ; r^ lA C > r~ 0 D~ VA r< 03 in r-TJ ' C U° «/> o-vs Oo Oo 0 vS (~ — ... -£ CO Q _J 0 • - - - .... ••- — ... ... CS Ln O In OA Ln t r r-cX t r CO CO > - — ... 0 cr ol tr r-r-v i >« LA i tr SS rs Lr. 0 LA O fA cs £2 L/> L/> h-- — - . . . . .... -0 vS ^ S v S <9 vs in 0 6 rs r— "tr 0 rr rS. rs r-c^ sS tr \s rs q 0 CA 00 rA c< xr cn cs ? £ z LU \D O ol h-Z Z i A O z rs) O o H O 9 Ln 0 • O o -6 - O 6 O 6 -•• 0 m O . . cr co O rS 6 t r rA cs t r Ln * r-- c r 6 cs In fA rs - • • 2 i <» zr Z 0<=> S9 t r rJ C» t r O t r xr vj" r-(VI x_r> ts-tr Vv cr xr VA S — xr t r t r V\ rv\ ^~ m xr o rs Vv o t r 0 A X r-1 OA c r r-V \ 5-O xr r-•5 b° CA 0 C-"> i— ro v9 — z V r -cs S S LO o U i cs r-o cr t r CS° t r vn XT O LA t r tj» o r-vS s9 i— c£ VA vS o xr oo xf ba O o=> LA vi Ln bo c5 r— •xr r^ r— rs X t r xr O •xr r r i r 1 r^ r— rA X t f r-' vS r-t C» ro vn r— xr vP CT £ Z o co err tr u o o r -OO LO v5 oA Lfl t r r»"> rr r5 Ln r — O r~ o-Oc r -6° La V. La f> H vS vj> LA tr r -LA r-0 rA 10 o CO -• — » -LL -o T> m «J Ld CS Lo rA 0 0 cs: a 0 u cs v3 cO * c< vS t r r^ & rJ <r vj> xn cs" SS r~ = cs <r-a~ <r. in E; r y \ OA rs v5 o LA r H lA cr CA 6 cr nr. 0° 0° rs X vi» CV) _1. cr O L0 0 rs H m \? (V3 t r r~ r J AA cr rs Ln 1— 0° LA H CL £ tf to \ *> • t f O r rs ' £ a: «' d r rS ... T •3-1 ?1 t r t r r T r t r Cy > 5 i * 0° x 15 i r sr' XX r f © - r •£ t r LQ 'i X Ln < 8 TL 3 tx 1 s L0 i zr > 5 fSJ i 5 X v 5 r c* H H » 5 X X ri ri • 5 6 2Z -o _at xc in t_ rs i ? _c T c xr a; (/> i •' X LA H ) S t> r c5 6 7 LA r s )5 6 6 th •i T LA r s 2: Lr> rt 1 5 a r V . ' 1 t r t r r* p ^ 6 0' c 0 Cr *-v' C cr. ^ _ cr 1 r JT> /^ s i/> 1 1 u b c e n g i n e e r i n g 225 zr cu L A r- -rr-« r-U r-" cs C>5 VA CS Oo .... -i A rT 60' cr in bo V) cr 60* u Oo (V> 0 ° L A <y r-t DQ --5 ' c tf ^< Sv £ co a _J 0 s> .... -CX -r-C» rr 1— V.0 tr -OA O-Tr CO 0 ° rr . . . . O rr rjr 1 O r tr r-rr rr CN 10 tr s9 OA tr v5 La rj-O Lr> tr «v5 Ln tr rr rr C-N O LO O rr CO CO > O O rv> rO r-j O t r Lr> rr rS S9 — ' . . . Oft 1/5 OA L A rr s VA v9 rs 0 0 rsi LT> CS H rr rS v>° \ sS CP O csri O rr OS 0 rs r-OA OA r-r rS ry> Ln VS> O) Ln \ L A \> <"S ^v£ r-rz liv r~ rr r-OA CV1 Cv-1 OA r cS O i-s" rsl r— S9 rr v3 b » OA 0 S3 on o° cr 0 0 0 <s 0 v/> vV> r-tr <N X T rS ~ l 0 -<-i f— W. tr VA CN vS of. CN O T-cs r s r>J O A LrV .V9 O °i SJ> s CS 0 s9 rS O O CS <S 0° O 'CN OQ O 5 s | rr cr <y> rr CA O O rs t r r— CN r rS r-r 0 r~ r-0 ( A O r-vi I — rs s fA — S9 r I A v i OA vS rn O rA VA rr rs 0 r > q rs 8 c s tr O cs t r OO oj s 9 r S £ Z Ui VD O ad {-Z Z 1 to O Z Oo CN -VA OA CVi O xr "fA vn .... : Lfi OA V A rr s i L A L A s~ LA s5 CS t r s3 O r< CA r-O r r r~ • - 0 LA CS Lri VI vS vS v i v i cr - Lo cvi r r t r JZ 1 t» x t Z v9 O <5 V r 0 • 3 -0 6 . vn r— cri xr O O Vr, F-O" 0 b 6 0 tn XT vn .CD vS d • » O cr O •6 bo vS 6 Ol O 0— ti-i r; rr, rs rS 6 fs <5 yo 1— Oo O cS 0 vn 0 0 i* r-OA, oi o f OO O sS r O tr Of tr q rr cr -2 0-rr OA cN CO.* r-L-i v ^ cS v5 tf r— r S rS r— 0 r-' cr c-i vJ9 ci cn E Z O CO cc cr 0 O O f-rr O S S rr Ln rr r — LT> s 9 s9 OA H rr r— s9 oa r-0-1 t r4 r J vS v5 v5 IT) u> O O CO VT) L O rv [NO cs oA r O a ~ r 1 LLT o~ t r rr CV) Ln rr —D XI (U Xr-cn r-H r-in s S 00 O tr cr rs v J O i£ SJ? tr 0 r~ O In r -lS s ^ 1— r s c r VJ9 0 0 0 v_!> cS In cS. r~ cr rr1 FX H r r~ •O 0 •rA r 1 cr 0 in (V\ 0 cs" v5 L A O r- 0 cr LO cr r~ 1 vti r— s? "I cr~ S 3 cn rr 0-) Vi c~ v5 g cr~ Lo r S r cr xr vs v9 r S V 5 cs (V. o l O Ln r-cr s ^ r r r cn rr rs <u E : tf \ " t f n 0/ ! xr 7> •i -XL LA a) i i > ° • J v/» 7> 2~ xr T> xr VA y *--p x r -0) rr t> r> cV 6 6 ft" rr Q-x r V A in "6 _y x s v/i rr r l A rJ 1 x^ . ... £-rr XB, 0> V^ *1 T> 3* a?-V iy r r ' x O 0/ s. <y cf C5 5> rr - 0 a) vy> 1^ Q' XT' V 5" xr cr % 6 h cv 11 ^' tr ~r> »• xz <D VA a' XJ? 6) V t r CS CS > 1 L* IA * x r H a* -s »' l n J x z XT v5 cs CV b h CV L * tr -t> » 5. t o t> a. 3 ? cu V? i -xr er-rs e rJ -r" +- • r ^ rr u b c e n g i n e e r i n g 226 SA c/V c4 Oo 0 s 0> rr t r t r c s cr (VI tr (VI CO Q _ i O CA vS CA cr tA is rA Cr V9 o v9 cvi r-tr. O r o r s ; vs Ln cH "3 -rs e r r <s VQ "cvi -LC\ V5 s S (-1 o . cvi I CS 0 0 0 / Z Ui O at lV> r? cs cr Ln LD o o" v9 tr cr cs 00 LO cr rvi cr errs r r q 0-1 tr CA C>A \ CT) E Z O CD CH cr u V? tr O v 5 > cr vs 1A v9 KJ PA t r r r C>° tr Vi^ tr in cs; O rA in iA r -CyO iO in VTJS-/ v£> Cj rvi Vi1 Ln IA LA Q. £ V -+-tf l A rA i A ft h he a/ 3 : a) VA 0) ,3 t r CP LO c r (A H - iy) 3) h 1 "OJ LA 95 tr cN C V.A Lo 5> h J UA Si rz 0 LA a/ £ 6 tri t r cs c? rA C/Ci 1 far VA O I S S LA OA 5 3 h CD vA l / 1 0 — rs. i ~6 •X LA ^ ft; ^ u b c e n g i n e e r i n g 227 tc o_ - - cr r~ Vi c r r— O O o° . . . o-r~ (X) c r 6o CT r rJ cy 0° So -~s • c o U 7^ t o — /— \ £ to Q _J 0 0) • CC -v5 Cr O o tr o cr 0° V A v i vS tr o tr -O cS tr vS r-(V) VS3 O r r tr v9 cvi rr C S v^ CV) O tr o° V5 ro rr C A oC cr cS t o > r-pl o-r— c s i— CS3 tr cs CN rr o r r .o <V> s o r V3 tr t r r T r rr-bo to v S v9 cs tr o < r w r-r— T r r-V A o O f> o "Tr-vn C A Ln rr cy O r" vS r r r C O cr cs rr cr rr r C O Vn v5 rs c S cs O 0 o-O \ rr Cr r-o cr b O C N O S i/> k § CS o< CS rr-r? 1 vS> r-vS> r-v2 •3-Vr> rr rr o rr r S o) CS t r o4 o VO ra-v9 VA tr O iV/v V A cS cr O cr CA O S . v5 C S r \ A O •o o . o c s r-0° r" rr rj r -1A c S CJ^ o l o l O V9 O vs C A o r S O i — o -oo A A |— V ? r— (VI Ln' C A r A P -o r-o r A Vn* C A rr l A vS ro V A cr r v i oJ v5 ( S A Of r z Ui vO O Z Z i (0 O z V I rr V ) t r u\ <o r r vh r i rr . . . o C Cr r A r S Ln cr 6 C N rr L0 cr e< r-co" O cs-0 c r-i — LO oJ t~ <s q cr-cs I «» rr Z o in <Q cr" O 6 r r O d cr c s . O. rA d v9 v5 O P vii d r~ r-(T r~ c s r" 6 O o cs o l c^  s S cvi tr Ln r— r r J — cr O ro 6 Z r-- cs C O tr rr - cn r~ r A r~ Ln r~ CA rr rr cr rvS o r v i r S CP r v i & rr; r s * CV) rr1 r o £ Z o CD CL cr u o o k in £> O co c s O rr O cs s o l vS v9 \ r fo CN r-vo o L O cs ^ i r s r^-r r C O O r rr 00 r—-cr cH ' 3 .—' L ' ) O rA cr r S r • L ' l v.. D s rr (V) fA r-'i C A ' cs & R v9 r s r s V . ' LA o4 rr i CO r r r S 'H V y v3 t o o 0 u zr cr O fA •-.9 r.' ?"-} r- CL C S CT> A O cs <-~s i— LO v./ r CA v S cr rA P A cr r > N V > rr-cr >^ cr r-tr i— rr ^ s o^, i A O O ' 0" er-r s -o a; i— o s r— o ro ol CO Lo r C O r-t ~ LO r rvi / " ^ I— vt5 cr v5 CS> rr 0 ° *—\ vn v5 rs C-° —^1 v3 cr r~ 4) Q. £ \ V -+-tf O r A S $ -r (/> I T> te vo CA n CJ C5 d t O 7f i _ . rr to 75 3?" ty> cs rs a ti d 1.1 V \ " rl-• o ri I r s i C -D V^  f " f r r cr c f rr ~CJ a; 3 t a) V A if' rr r~ t> a1 c^  L 1 (V \ A v> *• rv' S ^ _ 3> ry Li / tr o < 6 a: C J cf 0 l o rL >-tr ~U -J£ o) VA 1h_ £| rr. u b c e n g i n e e r i n g 228 rXT - 60 -• . . . r-cS o v> r-o Cr r~ CsJ Oo 0o r~ o r-r q r- -TS ' C tf t/> ...... — ..... f CO £ Sy CO Q _l 0 tn « cr -• s3 0r cS 0° C S cs CO O m - -vS CS r r Cxi r-rn O s9 LA <N Ln CS" O r -cs Ln Co o o 6o O bo sS L A L A CO CO > 1 -Xr \ 0° Cr 0° \n O -0= o oA CN) rr rr rr r f rA O v9 CD OA rS I— vS q ss q s i o cn eO LA CV) O O A cr OA xr rA XT o <6 X3-r-rf r f o o 0o LA on in ~r r s5 LA r-S 9 O <r oo er-rs Cr o o o r r t r O «r-r cr T— s oi cn r-cr xr CA r— *n Xn tr OA O <S O rs O T-OA S 9 cr •r-r r vS 0" vS 0° vS o s i i A o= O rs O r_ 6 bo r S •c^ 00 OA r-s rA r~ cr> L A s9 OA LA" Vx <o rr fr i r~ o° OA CA r - t on | s9 oo V> r s cs rQ — o c« OA 0O r-X-VA oo cs rr OA <rs r>-C V ) "o vS AA r-r— 6 o rr vs. rv\ cr L \ to o xr cr CvO on r — bo" 0° r~ Oo ty> •s s ^ Z i (O O z - : in rS Oo OJ O O rA -TS q an cs --q v i q S S X3-r— q in L A CS Ln r- LA - Lh cS cr CS - LA r r Cr ot CA <0 OA <> CA 0° r-c S r-ol Z Ul O QC Z 2. 4 X Z LA OS xr < V\ H 6 . . . . ... J— U\ Lr\ r-•xr" r— 0° fS o o vi) in 6 q o o O cn O o r— O d LA O d Cr O c Ln 0 d In 6 Ln q d O z-r --O cn • ^ rJ CP cS* -• s9 Ln xr cr q xf rJ TT rn r OA of cs cs vS CS o m' cs rA q r r a on r r OO ^ v £ Z o CO or: cr u 0 o V-- ...... --•-10 CD O CO in x r 2 X3^ r r in rr XT V cs cS r r 'LA C cr xr • . . . . . v3 s5 s9 S Lr> rs v5 CS ol CYl 37 rs r r in 01 Ln c* O 3 cr rA £N) LO v_ • o T- sS> OA xr v^ tA o 0 u S> rs s9 ot | cr v3 LA r r r* cn e>s i— 6S . • ~ \ V9 s9 ->^ / s9 / •—s 0° in ol v5> rr-c>r r— /•-—> xr v9 s5i s5> r i q xr CP LO -xr oi in o-s9 r r xr 00 LA r r rs S ? OA xr ' o xr it "o xr r 5 LA OA cn Ln m O XT Cy> LA xr xr O LO <r-m r-r rr CP s9 4) CL £ tf \ "tf e sz a> D \n r "V .0) XT j?" xr-rA i <zz i. LA L ZT ~T> a) v> s-5" TT xrr ai VA • Q> 8 xr • y s -r-i (5 © rL rr t> XZ 0) LA 3.- if r r - i a) L A lV CT 6 -tr LA 7 1 £" xr LA i/i £~ X l-"i 0' cn C5 IA & r r ~T> 4--a? 1 l A T> of 0' 5 -i tf & CA L^  u b c e n g i n e e r i n g 229 a: ir •- I 4 -- ... r-; q do r 1 cr r 1 r; r ~ °T r Vi r~ fS 03 -' C tf «/> s \ £ s v tn Q _J 0 cn 0 8 I V : cs cr rr — vS In bo rr ol rr-O vS tr cr Cs/ O Ov o( C CS C A rr O ry) v3 on csl Cr rS C O to > r-rr t>-tj-CO rf x> ro r ~ xr i — TP r -rr O A r-2 o vi r A ro C A r-q w" ti-er rr o -5 o r¥ r -o" vs O o cs C A V) r ~ v 5 evj tA CN cv> rs ... O o o Q of) f ry o cs cs r-r~ vi r A L f l O co O cS C A r xf r -rf r-6 OA <v> Ln oo L A ry> IO Ln O 6 q 6 • O . 00 C S rs" cs f A in O c> r xr r xf v/> vo f -b° S9 ro r— r— rr L A xr ln o L n C O rr O xr .ss r~ tf r-d O C A L A r>0 O cn C A cr O 0 ° O C O S3 cs s S C A O A ... o C S ro "o rf rv\ s r— ? <VJ r ~ & Q PA V A C» r-r-Si r~ r s cv> r -q xr r ; rs O A rS CD C P OA 1A rs X3-cS -O A C A CV) rS rr a CO rr r t £ |SJ i — t*l r -r S r r^  Z U l VD O r-Z Z i fO O z r A • cS in _ « est r-• C S . in : -L P In Cf _ . Ln cS cn In r J .rr O • » C A cs -Ln O s9 q rr Ln « r-- cr rf s S O 2 I «o a: Z L A 6 b 6 O o J 0° O a H e O v9 0 6 xr • CP 0 6 CP Ol 6. X3-xr 6 9 cr\ s5 rs ^ cs o o d Ln O rr rr cf Lri Ln s9 rr sS> cf rr Lri r r L A Z r-o~ rn rf rr I -• -r< cs c cs n' in cs cr cs r- r-cs &> rr cs s3 V cr CP r-• -01 £ Z o CO cc cr u O O V--10 O O CO rO rs N I>A s * cr cr C A rr CP XV 3 ol 8 r- .xr 'in r s tyi rr rs o rS V/v In v ; CS O r cr Ln in lA ' N • s , ; cs r-vS LO C A r-r s rr rs Ln 73 O 8 o 0 u CV) 0^ CO t*> i n V? Cv> LA I— rr LO rr C S L7) r r rr Ln 5-rr cs CP oO rs vE) rs s ? t— xr <r4 s S r-L-v v^ r — XT", o= SS cr x r , IS) u\ r-r- r r-LO I— a> LA LD t T ) s9 CA . — L I Ln Ln Ln v^> r-I i i \9 r-xr> vi> y^ cr-s S m t— x^r r-r- ol r~ ^ s O J CS 0Q § s S s S V CL E tf CO -+-tf Q L A cs ts Q) V/1 LV g -r i • •f rr J ft' Vy Jf xx oV -s net) W 0 0 C - r L ^ •V~ v. -rr a) ^= rr cr-a | ar rr 1 5_ i r -o cv 0 V'1 *t 7 3 rr -D x±-i J cr •i o "5j_ •£ T> a> <J) L^ rr rJ V O $ 6 v/> ^ . » -o a ' tt^ rr vS ~v o s LA ,cP 0 V L-A xl *' rr <r o | V ~b t» xz V / rr j " =1 —j -J u b c e n g i n e e r i n g — 230 Q_ IA cr r - r -O ro . . . q V A CM 0Q . . . r~ iA - • Mr r r t Cr r1 CP r~ O (¥> °? r -0 cy5 --rs - c o U ..... — £ t/i Q _J 0 f l • <E 0 m 0» vs b* tr-ot in CN --o 1ft VJ — rr OA r s rA r- >A 0 CA CP r* rs ~T~ rA sS> 0 ° c M O LA r s rr rr rs CO CO > r-T— CS «A on er-r s r-* rA q rr-r c x r-rrj r— cs CO rA CA N\ to JA r-7 t£ c s 0 <D O r-Tr r-ro Vo vr> CA rr 0 •3- O \A 0 v9 1— CjO c s 0 rr CS r? r-r s 0 cy= (vA r 1 PA O & r-r-S9 0 V9 O VS O cs OA r~ CN <N. r-V9 r--1— vi LA <s O $ (VI cr CH r- Q r-oo (vA 0 0 6 rs O CN 0 d CS VP to © rr rs r-O S rr H CA T< rj-t-r <N CH O r s c s rvi •q r— O O 0~_ O • OA 1— 0 c s rs ---0 0 O PA rr O v i r c CA VA v-c O O r s rr r ~ v9 c s AA r -c s — OA CO bo rr vs iA CA c v i rr <3 r-r-rvj tvA O rr-r -rA CA VS rA iA VA sO >• r-rjj rv» CA rr c < LA c^ rA O c s (VJ r-|V> •0 rr-r 0 rA I--rf CA r-v9 £ z U I vO O oC z Z t rt O z K \A r-0? -OO rA rr r-& O cr . . . c s VA cs --r O v< rs O r i rs — v9 CA VA \ S - r-U r-0 . . . VA CO 0 b o Vo c— r f \A r^ •rr C« lA •S; I <a X z <H Vc r 1 co c s cs IA r r LQ rA O c s O CA O O O cs c s d in O O rr r^ O L"> cr ro s5 O ra-tr-VA •6 V^  > T77 CS O Vn in CA CS. d rs O CN 6 0Q CO* rA Cr O CA t — CA' d -A tr r>i O VS co" z 1-- <f (P IA - r -i« r-c n rr?. r-\n CX r-VA O rcr rr c s r-o \N (V) c s r r -c< . . . . rr" s i O rf in cr cr VA ro b o r i VA z o 00 cc cr u o o r --_ to rD O CO —-rA 3 r s c4 v„' CN cs lA rv T3-c x CA o CS r s t — CN v5> CN s & s r r -"Tr ra rs I i ! i CNl -J r-r< ro c s CO s po \ S L"> A O CA rA rr 8 CA CN •=r cr Q 0 U 0 VA v9 3 r-10 cs? 6 s5 c y -In LA v9 v? r r r r v_9 tn t> Lf) r-rt £ 0 s9 "c? r-r rr vs r r cS in rr 'r-8 " r n cr r O O r-LA 1— s S c<v rs rr cr rr-cr s9 IP r~ r O CP •c srS cS s9 <u cx E 1/1 \ "tf Q ro 31 8 6 > • rr 3 VA 5^ rr 1 VA -2 •3? V V~ V s s" 3 IT O fA t $ •r J rZ vi rr _o CS VA -y v> rr 6 c cu J t. 3 0; v^  *-cr _o a; c? 0 rr cv vo y rr OA 1) c5l 0 -r-\P 1. CC Ql IA> m _c V IT O -v Kr 7 it rr «./ VA Dale /Sample C A R B O N 0^9/-O N I T R O G E N S O L I D S i>9/.0 Conol, C O D B O D 5 T O C T Vi N i NH*-N N 0 3 - N T S S T V S S U3U.20M 3 7S" <l7 t p T i r u — / - L "2,44 ' : ! 2 2 i c /24</ S 3 (551 ' ? , ; o 2.3 0. 1 /0./5" 47.0 1-22 7.7 <«7 C60 2.0 0.0<3 4 7 . o 4G / F4 2-0 26 i-zo 7.7 1** S k x c * N o u . 2 3 1026 /qos 470 I020/H6«, 3.W/3C2 <•! (5"S-) 2.4 /5.Z O.I5/0./5 34-O 7 / £ 103(55) (3) 21 2.7 /5.7 0.015/0.11 I up I I ? 7.6<f 1'* H o o f rJo-J- 21 *?5 31 61-3/76.0 20.0/2^.7 qrt. „ - Z 5 13.3 16.7 4 4 C60) OO 23 3.15/2.4$ 10.s 2 ^ 0 IM Go C60) ( ? ) 'Co 2.9 l/.f> i<2.. I14.3 5.3 16.0 2 7 ^ 12 24 4/. zlq o.i 17.3/17.3 5*? 2 0 14.1/14.7 5c; C4G; CO IS 3.2 4.5 qn i. (4) 7 2.1 0.13/0.1? n.is 7.3 /6.0 in SIO^P pe<:.'-l q?.i/M4.o 23.5/2^.0 4 " L - $.7 / i l . 1 1" (35, n.*. /(o.O (b.l 5 iM bT.l A S O 10.0 / S . l 2 ? V 7 . 7 i sc no / K G 56 /s?,7 lg.l/2%0 11 &3) 1.1 /2.I5 5.S 3 2 0 7.4 7 4 (.ci) 5.0 lQ/c% 7 .1 21.Z/2V.1 I4.0/M.7 ? 2 0 7 . 7 T + Stc^ e Dec XI I°I3 /262 HI. MnS I' II N G / l i o io4 / | 4 0 243 (14,0/ 12.2/12.0 l.vr/\.<is o . o IO.fi 0.06/0.025" o . o z 7 6 . 0 2 . 0 6 G. 0 / ^ . 0 4 * S b e « Dec .2?, 2 G 2 4 ^ p SeWed " 123 c cr n CD D CQ CD CD 2 3 CD ro u b c e n g i n e e r i n Cu r-r r f - r-' IA r C r-Vv r-rA r — I/! r-c 0 U ^ 7 to r\ *H £ i/) Q _ l 0 05 • c r CSJ r r r S rrr r r o4 O cr JvA v j vS (VI oo r -N sS r s m O r-rs vS> o) OA O r -rs t o t o > r -o v9 _ C N O CS. t r t r v/> sS LA rs lo r r O l o l Cr rr r— rA cf ?^ cf to a-1 t r o r r r o <? r s S O «r Cv> o r5s o rs ty> to «r-Cs o rS r r s OA <rS r~ r~-vS o •Tr q c s r cs q cs cS. J— cf Co in CA in OA r O i f O r~ O 0 v i r-fO Ln CA to t o h v5> C~ tr K. rr tr t r (V v-< O 0° O l O <r r r CP t r co OS rs cS CS c> cr o vS r r sS r-4 cr <— 6 rr r-r>o IA O l r-s> c>-o cf cx O to r 1 VS o CN o t o LA S3 o -S r 6Q cS v— t r o P Lo r-rS tr ro r f r r o r f o t f CO tv\ J A r s ^ r r r-ISA "io VA' (— o r s r s r CS l 1— v i r s q rf (O n o rO z Ul O ccl h; Z Z ro O z 6 t r r r r f <S v3 vS oc oC cO LO L0 in LO in cs cf o cn -Ln cs In Ln cS d cs in q LA rS vS LO J— m Ln r cr zz c" 3C Z z r-iA CO Do {>£. v_9 Oo O o< ? o r o tr o r o O S 3 O Cy= o o oS o o cj r r i — >cS VA c » tr-r f v5 Cn Oc. i n cs rr, o o~ O o-o q t r tr CO rS c f r o t i -cs CN •XT 10 1 — L i Lo CO r f tn CX CX r~, r r o t r r o o CP l/v cr LA \A vS rs Ln LA' L n cc o f f c s m t ~ /-> *C \ 01 z o CD or cr o O 0 h ir> O O CO L") rS j — r — C4 0° r — u") ol s cr rr o f ro v3 r s LA s > cr O Ln t r o-i — rS cS Ln CA r , cr - — V r-n r cs m O m ' r -v./ CA r-s S3 3 o 0 u cs-c r cy> r d r r o CS r r Cx t r r r I * cv vS c r 1A r -r rS rA oJ O O cs rr rS 0° r-v J 0° D° o l in r — v9 Co r~ vS do r xr AO cr b s9 0 ° ol o r -o 'rf CS CP "3 0 r-r r CP IV) CP vf) on i— cr (JO CA rj; r SS Ln O 4) cx E c f LT) \ Q o-cr •O L TT T> o \A r r a' tr 0 —f V 1 if P V * r r v5> 0' c5 0 ~v V * r r "T> rre r r :> c5 cy I/ 1 fL •± r r ~t' rr: a' ni U o ] ^ _ •£ ». rrr* c L^ s r r l» c h° c c x l2— * r r T a1 "ti V-±. r r 60 o 5 6 s rr ft-l v r r u-0' 6 D - r L^> H rr T) a-r5 y V.A t r u b c e n g i n e e r i n 233 CL. CT v9 PO r— rs r-Cr vS> r~ r OA r~ r-rr r bo r-rr r r iA r 1 r TJ C O vj ^7 </> S - < /-~ \ £ tn a _J 0 01 ^£ E O N V-'"' O r-rs 0=> 00 O 6 0 r< O 0o CX rr PA rs v3 r s rA CN v9 cr cx oo Ln r< rr cs O rr H CN i A fsA rr 0 (vA O CY1 rr O CNA LA > o r=9 r-s3 o -rr-r o r-s9 r-v i © rA cr r-s i r— rA r~ o r r q o r-s3 r-r s r-6 rs r— rr r— o r-\9 r— s i q O O rr r q rA r— o rs r-0 ° r s i r - A l PA r r s (V) rf o rr- rv-> L A iyi VA CS O r rr o o r rs r-i iA & r r> rvi vi LA Ui i A o •rr rr rA ~~A q d r s r-rs rs o r s rs rA cr q s9 r-s9 OA rs rr -S rr O I--8 rA of rA O r s o •S O s q CP O rs fA cr r-r— Pjo" jrr-(Vl L" r-rr rA vA ISA rr. -o r-_ r—; VA' rr q q rs r—i v i <sl o„ cs r-r>, q rs ro O r f CO iA tr o Cxi M r-rr z Ui O a: r-z z 1 m O z Vn O LA s S cr s i rr rA rA O P A s i L/? r~ O o> s i rr r s LA Cr-cr LO r -Ln LA cs rA IS) Do 01 O '2 1 c" X z z 1-"rr rr rs I — rs rA rr r~ rr on r ~ i-n r-i s9 O L 1 Ln LA LD Cw ss> •o rs! LA q q V cr LT rr oo r-r i o-rr s9 LA t r r 1 V i r* rs VA s5 cr r i c -C^  rA rA IT) r~ r rC r" t~ rs VA PA PA rr fS <? VA rf rS cr 01 E z o CO a: cr u O 0 r -ex a O CD A j 5 LO O CO 0 tA rs rs rJ vn s_/ r; po r~ O 1 1 1 I 5 s ro ro rs; rn cs rr rr I n (.-rj- rr-hA LO cr o LA CA rl? CS, cs Ln a 0 u rr s9 s9 *s_ s5> V." Lf) LA o r-rr r r Cr§ r ^ O vA s^ rs s 9 rr si> O s9 "s •Hi V A s3 s LA r -s3 sS LA sj» rs LT\ s9 La O c° rs V3 O r~ vn s 5 03 rS r* 3 m o c-vn 0 ° r— cr s i Cr-sS m O t-m <r •ZJ LA <u a . E tf L/~) tf Q \ "Tr _o CV U -0 VA *• 0> U_ 4 ~b r 5 0* V>-i Ix LV 6 d • if 0 .£ rr 0' 5 : IA L u 0' 6 C - r " L / J rr V j c i iSL ai 6 r_ 1 rr -D CD V ^ f V •i zr r o o V rr" tl •3' rr rS ^\ il-a> (5 6 LA ? L rr a» _Z v/» if rr v5 r s _ Q (i> C D -r" V •£ rr "T-cu v/^  rr rs •> C5 r a> (? 0 - f i.n r : rr c 4) V A 7_ rr u b c e n g i n e e r i n g *Zl . _ _ _ 234 Cu r-LA r~ Ln. r-^  CP t r r-lo r-vO r-> 1— c U° <v N £ Ol a _ J 0 cn to h rrr rr s9 on O t r rr O cr 00 to to > on cT co PA ryi tr o o o j o o t r q vi rS o r f r \ o vi r-t f CO cf-kA r-; vj> O rf r-M on CM CA r~-rr v i (V) r-o >o rn r f fO ro 00 _^ A OA VA q rf o •o O tf Q CO r-r s r r s o <5 o Cr o s S r rf q r o O O LA r-to r o r-r f q rvj r— cf r o r f o ~r tr t f Co (vo to h-p° LA CS on 0° o rv) 6 t o c» O $ o rv) r; •o r~r rv, /^-\ rS> fA iA O r-r f rO cf CO cf-(O r>A O rf CO q OA Cr c~ t f o r; vi rs cvi r~ r~ t — r f rs (O O O f^^  JV\ rsi i f CS q "to CO i i r~ r f ro CS rA c f o C»a r o r-CO i f o r f fro c r o r~ r s r s O rf q rv ir» ro to OA q r s r s r; S3 r. r s ryi (VA f z Ui to O al r-Z Z i fO O z CA If) — -o r r in O cs cr Oo 0o LO r r r « r r r S Ln c f T~ LA cf q in c r r r L0 q s9 6 :cS i K> X z z r-rS LA-cs LT3 Lo tf q vi 0 ° cr LO c r LA ro cr cf V? rs O cf 0^ q cf CX> r/} cr t r o 6 rs cc? cS cy? r S rr v3 ty rs t r cf cf </; r \ cS ro (f q r r to °? d r S oi E Z o CO a CL U S CL V-q l i O cr 6 6 H r-<6 rs! r d c-6 v5 d — ^s C L'l r~ C in 6 in O 0 CO a 0 u r-O 0" H r s — C P s_> o rv) r-LA 1 <-s q r— I~ rr C r Cr cr ^rx / c r 3 r r -La > tr-vf3 cr r? v3 O C« o r LA O in rr r o b LA 0~ r-v3 cr La ^1 vVJl v5 'cr • cr w cr r~ vS r r f -r-v3 <u "a. E tf to \ -t-Q n _C a c C X" z t r ~V ai QJ LA •3-0 r 0 i _ tr •£ d ~> X o; 0" 0 a> 6 i / if r r J rE Lo '7 s uv o X r l i <£_ t> <v 32 a) 5- r r 0 H i Xn Si. s •£ r r -Tj i f 5 r r ri o X tr O io rr *a c> s o rS 21 ai § r r -c aJ 5-c r r v S _D gy CV 0" & V* 7 0) L/1 • ZT u b c e n g i n e e r i n g 235 — =c ex. • c U° £ m Q _ J 0 to I/) h 0 0 C/7 r -r-(>•>' o-r-s i O r> r\ri o s i o r-c>o o 0/> r r r A rA C A CA CA r~-o C A v; i — t r o V) <s t A q s i On w i O v i • A V/) \-r~_ v s v> cs o r-o PA r y i r--VA flA r-cs q r s cs r-C £ r-vs r CP Cvi CsA Cvi r-1 cs r ~ ; v i cs o cv q v i o ve PA o t> r S rs Cvi i A r v i r r / " S £ •~—' Z tu VO O h -Z z 1 ro O z 1 1" 3 T z Vi rV\ r C P tr v3 O tr- tr-0° vS VA O <5 l A O 6 V z 1-• j - tr C S 72} A/J t r r r 0 r r Cr o> r~ CN CV) rs 60 rr v9 r-6 cr O Q CN CT) £ S > z O CD Ct cr u 0 o h rA cr o r -v9 co r r (V &o fA °? rl vS 00 r s rr v9 0" O r to CV) r r 0 CO o x o 0 u cs r~ L A L A • — o-LA r r r— in u i L A O 0° v9 v_/ S<5 r-P ~ v—. r-(SO r~ V/i r-— SS, ' A r-sl V 9 r — v9 rA \3 -CA v3 v5 "a 2 5 ; r A in CD CL E tf on -+-tf o C S s i % 1 >.' : t> t»r o> v/> P T "& 0 ~Y l/> 0' t r -Ot \f *r r :£ t r c r V Si. <JX cy <? t> ~*c \P tr r r V>A 7 rr H s 8 t T T T> a) V> to o b T o> V i . 0. •3= 6 sT t r i '1 ' c ^. j r a -<-0 a» r*r (i 0 0 -V lA I-3" r~ i 6 -r \ ' A i d t> i. ri 0) 0" 0 7) P A e) Cf 6 tr i 3) 0; 0 o> A" O zr i 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items