CIVIL ENGINEERING STUDIES SANITARY ENGINEERING SERIES NO. 36 p tO AN COPT STUDIES ON THE REMOVAL OF NEMATODES BY RAPID SAND FILTRATION By IRVINE WEN-TUNG WEI Supported By FEDERAL WATER POLLUTION CONTROL ADMINISTRATION RESEARCH PROJECT WP-00047 Property of COLLEGE OF ENGINEERING DOCUMENTS CENTER UNIVERSITY OF ILLINOIS 157 GRAINGER LIBRARY 1301 WEST SPRINGFIELD AVENUE URBANA, LLINOfS 61601 USA DEPARTMENT OF CIVIL ENGINEERING UNIVERSITY OF ILLINOIS URBANA, ILLINOIS JUNE, 1966 STUDIES ON THE REMOVAL OF NEMATODES BY RAPID SAND FILTRATION by IRVINE WEN -TUNG WEI Supported by Division of Water Supply and Pollution Contro U. S. Public Health Service Research Project WP-000^7 Department of Civil Engineering University of Illinois Urbana , 1 1 1 i no i s May, 1966 Digitized by the Internet Archive in 2013 http://archive.org/details/studiesonremoval36weii I I I ACKNOWLEDGMENTS This report was submitted in partial fulfillment of the requirements for the degree of Master of Science in Sanitary Engineering under the direction of Dr. John H. Austin, Assistant Professor of Sanitary Engineering, University of II 1 i noi s . The author wishes to express his sincere gratitude to those persons who, in any way, assisted in the completion of this work. Especially he wishes to thank the following: Dr. R. S. Engelbrecht, for his valuable guidance and for his suggestions in writing this report. Dr. John H. Austin, for his valuable counsel, constant encouragement and guidance throughout the entire study, and for the many hours he spent in reading and discussing this report. Mr. Ronald L. Peterson, for his valuable suggestions in the inception of this study. Mr. Gerald Steiner, for his assistance in conducting the laboratory expe r iments . The Division of Water Supply and Pollution Control, U. S. Public Health Service, for the funds which made this study possible. IV STUDIES ON THE REMOVAL OF NEMATODES BY RAPID SAND FILTRATION ABSTRACT This study was designed to evaluate the removal of free-living nematodes by rapid sand filters. The addition of chemicals to the filter influent was also investigated in an attempt to increase the removal of nematodes by rapid sand f i 1 tration . D i ploqasteroides sp ., one of the predominant species found in the effluent at the Urbana -Champa i gn Sanitary District waste treatment plant, was used throughout this study. Tap water at 20°C was passed through a 2^-inch plexiglass filter column at a filtration rate of 3 gpm/sq ft. Ten mg/1 of alum solution and preformed alum floe, respectively, was applied directly to the filter influent. The percent removal of nematodes was correlated with various sand sizes, filtration times, and conditioning chemicals. In experiments using all dead nematodes, the filter sand was examined to determine the pene- tration of nematodes into the sand bed. The results of a study on sand size indicated that, at a low percent motility, the removal of nematodes increased with decrease in sand size. As the percent motility increased, the removal of nematodes decreased. Also, as the percent motility increased, the percent removal of nematodes by a 1 1 sand sizes studied approached one another. The removal of nematodes, having low motility, by sand of different sizes was found to be related to the initial The percent motility is defined as follows: ^ . , . ^ ,. 0/ \ motile nematode concentration in influent x 100 Percent Motility ( i n %) = ; ■; : : — t, total nematode concentration in influent headloss, i.e. high initial headloss yielded high nematode removal. When dead nematodes were used, the removal efficiency varied over a very narrow r ange, namely Sk to 98 percent, regardless of sand size. Filtration periods of 16 hours duration indicated that motile nematode removal decreased significantly with increased filtration time until an equilibrium was reached. Experiments with nonmotile nematodes gave relatively constant removals for the entire 16-hour filtration period. An evaluation of motile nematodes in the influent and effluent showed that most motile nematodes could eventually pass through the sand bed even though some of them could be temporarily retained in the sand during the early stages of the filtration process. From a balance of nematodes entering and leaving the system, it also appeared that motile nematodes could dislodge the nematodes which were initially retained in the sand bed. Thus, these two phenomena ""e suited in an overall decrease in removal of nematodes. In studies using alum and preformed alum floe, the results indicated that the increase of headloss did not increase nematode removal with time; on the contrary, nematode removal decreased as in those experiments without chemical conditioning. Obviously, the effect of motility, which tended to decrease nematode removal, was more significant than the expected benefits of the alum. However, in comparison with experiments without chemical con- ditioning, the use of alum appeared to be able to reduce the dislodgment of nonmotile nematodes, and thus increased nematode removal when the motility was less than 30 percent. This increase of removal was minimized with increase of motility. V I TABLE OF CONTENTS Page ACKNOWLEDGMENTS I ■ i ABSTRACT Iv LIST OF TABLES viii LIST OF FIGURES ix CHAPTER 1 . INTRODUCTION 1 1.1. The Problem 1 1 .2. Purpose of Study 2 1 .3 . Scope of Study 3 2. EXPERIMENTAL EQUIPMENT 7 2.1. Preparation of Sand 7 2.2. Preparation of Nematodes 7 2.3. Filter 8 2.3.1. Indicating Liquid of Flow Manometer 8 2.3.2. Chemical Feed Device 13 2. l.k. Air Sparger 13 3. EXPERIMENTAL PROCEDURES 15 3.1. Preparation of Sand 15 3.2. Preparation of Nematodes 15 3.2.1. Pretreatment of Nematodes 15 3.2.2. Estimation of Nematodes Required 17 3.2.3. Preparation of Nematode Suspension 18 3.3- Preparation of Chemical Feed 20 3.4. Backwashing 20 V I Page 3.5- Filtration 21 3.6. Nematode Sampling 21 3.7. Nematode Staining 22 3.8. Nematode Counting 23 3 .9. Sand Sampl ing 25 4. RESULTS AND DISCUSSIONS 26 4.1. Studies on Sand Size 26 4.1.2. Headloss 28 4.1.3. Nematode Distribution in Sand 30 4.2. Effect of Motile Nematodes on Removal Efficiency 32 4.2.1. Motile Nematodes in Influent and Effluent 32 4.2.2. Nematode Removal in 4-Hour Periods 42 4.2.3. Sixteen-Hour Experiments 48 4.2.4. Dislodging Effect of Motile Nematodes 50 4.3. Studies on Chemical Conditioning of Filters 56 4.3.1. Headloss 56 4.3.2. Nematode Removal 58 5. CONCLUSIONS 62 6. SUGGESTIONS FOR FUTURE STUDY 63 REFERENCES 64 APPENDIX 66 V I I I LIST OF TABLES Table Page 1 .1 VARIABLES STUDIED k 3.1 STAINING EFFECT OF 70 PERCENT ETHANOL AND 1.5 PERCENT EOSIN-Y AT DIFFERENT CONTACT TIMES 2k k.] HEADLOSS OF FILTER SANDS 29 k.2 CALCULATION OF NEMATODE BALANCE 5^ k.3 EFFECT OF SAND SIZE AND CHEMICAL CONDITIONING UPON THE DISLODGMENT OF NEMATODES 57 k.k HEADLOSS OF CLEAN SAND AND CHEMICALLY CONDITIONED SAND; SIZE 0.5 (1 .2) 58 A.l ABBREVIATIONS 67 A. 2 DATA SUMMARY OF 8-HOUR EXPERIMENTS 68 A .3 DATA SUMMARY OF 16-HOUR EXPERIMENTS 71 IX LIST OF FIGURES Figure Page 2.1 BAERMANN FUNNEL APPARATUS 9 2.2 FILTER APPARATUS 10 2.3 FLOW MANOMETER 11 2.4 CHEMICAL FEED DEVICE 14 3.1 SAND SIZE DISTRIBUTION 16 4.1 PERCENT NEMATODE REMOVAL vs. PERCENT MOTILITY FOR. ALL SAND SIZES STUDIED 27 4.2 DISTRIBUTION OF NEMATODES IN DIFFERENT SANDS 31 4.3 NEMATODE CONCENTRATION vs. TIME, EXP. 1, SAND - 0.5 (1.2) 33 4.4 NEMATODE CONCENTRATION vs. TIME, EXP. 15, SAND - 0.65 (1.6) 34 4.5 NEMATODE CONCENTRATION vs. TIME, EXP. 26, SAND - 0.5 (1 .2), ALUM FLOC USED 35 4.6 NEMATODE CONCENTRATION vs. TIME, FLOW RATE - 2 gpm/ sq ft (Peterson, 1965) 36 4.7 NEMATODE CONCENTRATION vs. TIME, FLOW RATE - 4 gpm/ sq ft (Peterson, 1965) 37 4.8 NEMATODE CONCENTRATION vs. TIME, FLOW RATE - 6 gpm/ sq ft (Peterson, 1 965) 38 4.9 TNIN, TNR, MNIN, AND MNR vs. FILTRATION TIME - A STATISTICAL SUMMARY 39 4.10 TYPICAL MODEL OF NEMATODE CONCENTRATION vs. TIME 40 4.11 PERCENT NEMATODE REMOVAL OF 4-HOUR PERIODS vs. PERCENT MOTILITY, SAND - 0.65 (1.6) 43 4.12 PERCENT NEMATODE REMOVAL OF 4-HOUR PERIODS vs. PERCENT MOTILITY, SAND - 0.5 (1.6) 44 4.13 PERCENT NEMATODE REMOVAL OF 4-HOUR PERIODS vs. PERCENT MOTILITY, SAND - 0.5 (1.5) ^5 LIST OF FIGURES (Continued) Figure Page 4.14 PERCENT NEMATODE REMOVAL OF 4-HOUR PERIODS vs. PERCENT MOTILITY, SAND - 0.5 (1.2) 46 4.15 PERCENT NEMATODE REMOVAL OF 4-HOUR PERIODS vs. PERCENT MOTILITY, SAND - 0.35 (1.5) *+7 4.16 PERCENT MOTILE NEMATODE REMOVAL OF DIFFERENT SANDS IN TWO 4-HOUR PERIODS 49 4.17 PERCENT NEMATODE REMOVAL IN FOUR 4-HOUR PERIODS OF 16-HOUR EXPERIMENTS, SAND - 0.5 (1.2) 51 4.18 TMNIN vs. TMNOUT IN FOUR 4-HOUR PERIODS OF 16-HOUR EXPERIMENTS 52 4.19 COMPARISON BETWEEN HYPOTHETICAL AND ACTUAL NEMATODE BALANCE 55 4.20 PERCENT NEMATODE REMOVAL vs. PERCENT MOTILITY FOR CLEAN SAND AND CHEMICALLY CONDITIONED SAND WITH THE SAME SIZE OF 0.5 (1 .2) 59 4.21 PERCENT NEMATODE REMOVAL OF 4-HOUR PERIODS vs. PERCENT MOTILITY FOR CLEAN SAND AND CHEMICALLY CONDITIONED SAND WITH THE SAME SIZE OF 0.5 (1.2) 60 1 . INTRODUCTION 1 . 1. The Problem Public health officials have speculated that nematodes found in public water supplies might carry and protect pathogenic bacteria and viruses through the water treatment processes, even chlor i nation , and thus might constitute a potential threat to public health (Chang, 1 960) . There is also some evidence that nematodes might be a contributing factor to taste and odor problems of water supplies (Cobb, 1918). In the past few years, two important studies demonstrated the seriousness of the nematode problem from the viewpoint of water treatment. First, it was found that nematodes were highly resistant to chlorination (Kelly, 1953). The resistance of nematodes to chlorination differs with different stages in their life cycle. The most resistant form is the egg, followed by sheathed larvae, adult, and nonsheathed larvae. Even the non- sheathed larvae were not entirely killed by a 3 to 5 mg/ 1 free chlorine residual at 24-25*C and pH 7.0-7.2 until the contact time exceeded 3 hours. Second, nematodes were found in the finished water in 16 of 22 municipal water treatment plants, employing rapid sand filtration of surface water supplies (Chang et_ aj_. , I960). These sixteen municipal water treatment plants were located in fifteen different states of the country, and the data from these plants showed an average total nematode removal of k6 percent by all the treatment processes employed, including coagulation, sedimentation, filtration, and chlorination. The most effective treatment process used for control of living organisms is chlorination. The ineffectiveness of chlorination was probably due to the tough cuticle associated with nematodes (Chang, I960), which is not effectively penetrated at the free chlorine residuals commonly used in pract ice . In view of the size of nematodes, about 0.5 to 2.5 mm in length and 20 to 50 u in diameter (Chang, 1961), the poor removal by filtration is unexpected. The rapid sand filter has long been considered an effective featment process for the removal of large particulate matter; thus, the penefation of filter media by nematodes presents a challenge to the water featment plant designer and operator. Although nematodes were found mostly in those water supplies using surface water source, some well supplies have shown the presence of nematodes (Maffitt, 1 966) . It was believed that these nematodes came from the soil, and gained access to the ground water via the aquifer. Peterson, in 1 965 j found that the motility of the nematodes seemed to be the most important single factor in nematode removal by rapid sand filtration and that the removal of nematodes was independent of the filtration r ates studied, i.e. 2, k, and 6 gpm/sq ft. 1.2. Purpose of Study The general purpose of this study was to obtain more understanding of the mechanism of nematode removal by rapid sand filtration, especially the role of motility. For this purpose, the variables of sand size and filtration time were studied; in addition, the penetration of nematodes into different types of sand beds was also investigated. Some chemical conditioning methods were studied, in the hope of finding methods of increasing the removal of nematodes. In this category, 10 mg/1 of alum solution and 10 mg/1 of preformed alum floe were added directly to the filter influent. 1.3. Scope of Study The variables investigated in this study and their ranges are shown in Table 1.1. Previous studies on the characteristics of filtering materials have used a uniform sand size. The possible reason for this may be the ease of preparing a sand of uniform size. In order to conform to existing practice, effective sizes (E.S.) and uniformity coefficients (U.C.) were selected as the criterion for the preparation of sand. Their ranges were also selected from those commonly used in practice (Anon., 1 9^+9) - A filtration time of 8 hours was selected for most study. It was believed that 8 hours should be long enough to reveal the results of most experiments. Four experiments with a filtration time of 16 hours were de- signed to evaluate the effect of filtration time upon the removal of ^e^atodes. The result of the 16-hour experiments also justified the use of 8 hours as the filtration time for most experiments. The filtration rate of 2 gpm/sq ft has long been adopted as the standard rate for rapid sand filters. Basically, there should not be any standard filtration rate; several factors should be considered in selecting the optimum filtration rate for practical design (Cleasby and Baumann, 1962) A detailed discussion of the selection of filtration rates is beyond the scope of this study. However, in recent years, more and more water plants have begun using filtration rates higher than 2 gpm/sq ft. The experience with a filtration rate of 3 gpm/sq ft in water treatment plants throughout the country has indicated that it may be used in many circumstances (Baylis, 1956). Therefore, a filtration rate of 3 gpm/sq ft was adopted and used throughout this study. Table 1.1 VARIABLES STUDIED Variable Range of Study Sand Characteristics Fl ltration Time Filtration Rate Sand Depth Influent Solution Influent Nematodes Genus Concentrat Ion Motility Life Stage Water Temperature E.S/ 0.35 0.5 0.5 0.5 0.65 U.C. 1.5 1.2 1.5 1.6 1.6 8 hours and 16 hours 7 .k cu.m./sq.m./hr. (3 gpm/sq. ft.) 61 cm. {2k inches) Tap Water Tap Water Containing 10 mg/1 of Alum Solution Tap Water Containing 10 mg/1 of Preformed Alum Floe DIploqasteroides 10 - 50/l!ter - 60% Larvae and Adul ts 20°C E.S. = Effective Size U.C. - Uniformity Coefficient From the viewpoint of nematode removal, it has been shown that sand depth, beyond a minimum depth, was not an important factor (Peterson, 1965)- Therefore, a sand depth of 2k inches was used, which is common i n pract ice . The presence of turbidity and the pretreatment of the raw water could conceivably have some effect upon the filtration, and will be studied in the future. For this study, tap water was used as the filter influent. The chemical conditioning of filters has been developed to improve the quality of filter effluents, especially for high-rate filtration (Conley and Pitman, i960); and it has been recently adopted for the design of some new water treatment plants (Culp, 1964). In this study, alum solution and preformed alum floe were used as conditioning chemicals, since they have been found to have different effects on improving the removal of algae by rapid sand filtration (Davis and Borchardt, 1965). A concentration of 10 mg/1 was selected as typical of coagulant dosages commonly used in practice. D i ploqasteroides sp . was used exclusively in this study since it was found to be one of the predominant species in the effluent from the U!"bana -Champa ign Sanitary District waste treatment plant. Peterson, in 1965, found that the total nematode concentration in the influent had little effect upon the total nematode concentration in the effluent. An influent total nematode concentration of about 25/1 iter was maintained for all the experi- ments except for experiment 23 in which a concentration of about 50/1 iter was purposely used to evaluate the "breakthrough" of dead nematodes. In the survey of nematodes in municipal water supplies (Chang, e_t a_j_. , I960), the average concentration of total nematodes found in the influent was 3 nematodes per liter and no influent concentrations greater than 36 nematodes per liter were found in the literature (George, 196*+). The constant concentration of about 25/ liter was selected because it was high enough to give a realistic and statistically meaningful result without being unduly difficult to be counted when a sample size of four liters was used. A nearly constant total nematode concentration in the influent facilitated the evaluation of dead nematode penetration in the sand bed. Extensive efforts were made to increase the percent motility in the influent, but it was found difficult to obtain percentages of motility g r eater than 60 percent. The cultures we r e concentrated by the Baermann funnel technique (Goodey, 1963). The reduction of motility was probably due to the loss of motility by the nematodes in passing through the apparatus before reaching the filter. A water temperature of 20®C was used throughout all the experiments This temperature was selected as a means to maintain as high motility in the influent as possible, since it was found that 20"C yielded maximum motility with D 1 ploqasteroides sp . (Chaudhuri, 196^). 2. EXPERIMENTAL EQUIPMENT 2.1. Preparation of Sand The granular materials employed in filtration differ in size and size distribution, in shape and shape variation, and in density and chemical composition. In this study, since the main interest was on the removal of nematodes, sand size was considered the most important factor of a 1 1 the sand characteristics. In addition, since the sand used was prepared from a commercially available filter sand, the physical and chemical properties of the sand were believed to be consistent with those used in actual water treatment plants. Therefore, in the preparation of sand, only the sand size was controlled by sieve analysis. For this purpose, a set of new sieves was used. ASTM sieve numbers 16, 18, 20, 25, 30, 35, 40, 45, and 50 were used. An electrically-driven sieve shaker was used in making the sieve ana 1 y s • s. 2.2. Preparation of Nematodes All the nematodes used were grown in T-30 culture flasks using the same medium as Chaudhuri in 1964. In order to remove extraneous suspended material found in the nematode cultures, the Baermann funnel technique (Goodey, 1963) was employed. This apparatus utilized a funnel with a piece of rubber tubing attached to the stem and closed by two spring clamps. The funnel was placed in a support and was filled with distilled water to about two-thirds of its capacity. The nematode cultures were transferred to a beaker of Red Flint Filter Sand, Eau Claire Sand and Gravel Company, Eau Claire, Wisconsin, 8 appropriate size (150 to 250 ml), and the beaker was almost filled with distilled water, covered with two layers of tissue and held in place by rubber bands. Then the beaker containing the nematode cultures was turned upside down and gently submerged into the water in the funnel. By this technique, the motile nematodes, regardless of size, could penetrate the tissue and sink to the bottom of the funnel stem. After three to four hours, a small quantity of water containing motile nematodes was removed from the funnel stem by adjusting the spring clamps. This apparatus is schematically shown in Figure 2.1. 2.3. Filter The filter built by a previous investigator for the study of the effect of filtration rate upon the removal of nematodes (Peterson, 1965), was utilized for this study. However, some improvements and changes in the apparatus were made for the convenience of operation and the precision of results. These changes and improvements will be explained in the following paragraphs. A schematic diagram of the filter is shown in Figure 2.2. 2.3.1. Indicating Liquid of Flow Manometer The original purpose of the "orifice meter," shown in Figure 2.2 was to generate a headloss between point 1 and 2 by means of the orifice at C. Thus, when any small fluctuation of flow rate occurred, a change of headloss between point 1 and 2 was generated and this change of headloss could be quickly detected by the fluctuation of the indicating liquid in the flow manometer so that proper adjustment of flow rate could be made. K. imwipes - product of Kimberly-Clark Company, Inc BEAKER RUBB£R TUBIftG STAND FUNNEL KIMWIPES SPRJNG CLAMPS D FIGURE 2.1 BAEKMANAI FUNNEL APPARATUS 10 C\J 11 The original indicating liquid was made with chloroform, methylene blue, and liquid detergent (ABS) to satisfy the following requirements: 1. The specific gravity of the indicating liquid should be greater than one so that it would be heavier than water. 2. The indicating liquid should be immiscible with water. 3. The indicating liquid should have a distinct color to be easily detectab le . The chloroform used had a specific gravity of about 1.58 and satisfied The first and second requirements. The methylene blue provided a distinct blue color but since this color could be extracted from chloroform by water, the detergent was added to chelate the blue color. This indicating liquid had a specific gravity of about 1.48. Difficulty was encountered in the control of flow rate with this indicating liquid. Usually the response to the changing flow rate was slow and overadjusting of the flow rate often occurred. Apparently, this indicating liquid was not sufficiently sensitive. VALVBC •?f \*- / fa r h. r 1 INDICATING LIQUID FIGURE 2. 3 FLOW MANOMETER 12 The function of the flow manometer is shown in Figure 2.3. Ac- cording to hydraulic principles, it is apparent that P. = P + 7 h_ + -vh k 2 w 2 ' P. = P. + 7 h, 3 1 w 1 and P 3 = P^ therefore P. + 7 h_ + 7h = P. + 7 h 2 w 2 ' w P. - P_ = 7 (h.-h.) + 7h = h (7-7 ) 2 w 2 1 w' thus, -—^ h tf- - - h(S-l) (1) 'w w where P. = hydrostatic pressure at point i (i = 1, 2, 3, 4) 7 = unit weight of water ' w 3 7 = unit weight of indicating liguid S = specific gravity of indicating liquid. From equation (1), it is apparent that the sensitivity of the indicating liquid depends upon the difference between its specific gravity and that of water; the smaller this difference, the greater will be its sensi t ivi ty . Accordingly, for this study, a new indicating liquid was devised for the flow manometer. It was prepared with D-7878 Merian Indicating Fluid and a red dye. The specific gravity of the liquid was found to be about 1.2 Merian Instrument Company, St. Louis, Missouri Dupont Red Powder Lot 791 . 13 The control of the flow rate was improved by using this new indicating 1 iquid . 2.3.2. Chemical Feed Device The electrolytic pump (Symons, 1 963 ) which was used for nematode feeding proved to be a dependable device. The rate of flow is directly proportional to the electric current produced from the D .C . source, and can be accurately regulated by the ammeter. Therefore, it was also used fc r the chemical feed. The schematic arrangement of this device is shown in Fi gure 2 M . A specific water level, L, in the gas pressure manometer, yielded a certain feed rate. Therefore, before the feeding operation was started, clamp B was closed and clamp A opened, and air was blown in through clamp A to raise the water level to L. With this arrangement, as soon as the desired current was applied and the clamp B opened, the whole system could be maintained at equilibrium, and the water level, L, would serve as an ■ndicator of the pressure in the system. 2.3.^. Air Sparger The air sparger Sn the filter column (Figure 2.2) was originally designed to reduce the effect of air binding. It was submerged in the water to produce large air bubbles which scrubbed out the dissolved gas in the water In the study of chemical conditioning, alum floe was used. In order to pre- vent the floe from being broken by the rising air bubbles, the sparger was removed from the water. However, it was kept in operation above the water surface in order to maintain proper pressure in the system. )k Q 15 3. EXPERIMENTAL PROCEDURES 3.1. Preparation of Sand A sand of specific size, controlled by effective size (E.S.) and uniformity coefficient (U.C.), was prepared on the basis of a sieve analysis. Af^er a se r ies of experiments, the sand was removed from the filter column, and another sieve analysis was made on a representative sand samples of one k logram to check for the E.S. and U.C. Figure 3-1 indicates that there was little change in the sand during a series of filter runs. For every sand with a different E.S. and U.C., the same procedure of preparation and verifi- cation was followed. It was felt that this procedure was quite satisfactory for the control of sand size. 3.2 Preparation of Nematodes 3.2.1. Pretreatment of Nematodes Before this study was started, it was observed that the nematode cultures were badly contaminated; many motile nematodes were found to be trapped in suspended particles and consequently resulted in a significant increase in nonmotile nematodes. The majority of the suspended particles were identified as a fungus, Zoopagales, which could trap, feed on, and kill nematodes (Alexopoulos, 1962). The other particles were suspected to be precipitates of the culture medium and/or excreta of nematodes. A decontam- ination study was started, and it was decided to utilize the Baermann funnel technique (see Figure 2.2) for the pretreatment of nematode cultures. The use of pretreatment was based upon the following considerations: 1. The contaminants reduced the percentage of motile nematodes which was an important factor to be studied. 16 *0 1 ->4 g N ^ < <0 1 3 5: &3A//^/ o/ o 17 2, The contaminants would increase the effective size of the nematodes appreciably, and 9 therefore, interfere with the results of f i 1 trat ion . 3. Usually, an incubation period of three weeks was used to build up a sufficient population of nematodes. Within three weeks, an appreciable number of old, dead nematodes were found. These dead nematodes were generally large, and this might interfere with the filtration results. h. From past experience, these large, old, dead nematodes might undergo disintegration in the culture flasks and/or staining process, which would interfere with the results of counting. The Baermann funnel technique maximized the motile nematodes that could be used in the experiments, and minimized the quantity of extraneous material . 3,2.2. Estimation of Nematodes Required The nematodes required for each exper iment were ca leu la ted from the flow rate, nematode concentration in the influent, and filtration time. The filtration rate of 3 gpm/sq ft was used throughout all the experiments. Since the flow was split in ha'lf above the filter, one half for the influent sample and the other half for the filter, twice the above rate was required. The nematode concentration of 25 nematodes/1 i ter was used in the calculation for all experiments except for experiment 23, where the nematode concentration was doubled to study the possibility of breakthrough during a 16 hour filtration period. A filtration time of 8 hours was used for all experiments except for experiments 21 through 2k, where 16 hours was used. 18 From experience, it was found that 3 T-30 culture flasks after 3 weeks of incubation at 20*C were sufficient to provide the 10,000 nematodes required for an experiment. In order to evaluate the effect of motility, usually a series of rhree or four experiments were conducted successively. Therefore, the total nematodes required was equal to 3 or k times the nematodes required for one single experiment. For the estimation of the nematode population obtained from the Baermann funnel, three small fractions from a completely mixed nematode suspension were pipetted into counting dishes, diluted properly for ease of counting, and counted for the total population in each fraction. The average of the three was used. 3.2.3. Preparation of Nematode Suspension Since the percentage of motile nematodes was found to have a signif- icant effect upon the removal of nematodes by rapid sand filtration, when the factors, such as sand size, filtration time, etc., we^e studied, the motility percentage was also considered an independent variable in every experiment. In order to obtain variable fractions of motile nematodes with minimal variations In total concentration, nematode size, or size distribution, a method used in the previous work was improved and the following procedures were used throughout this study. After the required number of nematodes was obtained from the Baermann funnel, all the nematodes were kept in a beaker, completely mixed with a magnetic stirrer. A number of aliquots was taken from this beaker equal to the number of experiments to be made on a specific sand size. These equal aliquots of nematodes were all contained in beakers. One of these was used for the 19 immediate experiment. The other beakers were placed in a refrigerator for the reduction of motility, and the second one was used after two or three days storage. The third one was used after four or five days storage and so on. In this way, a series of experiments using nematodes exhibiting decreasing motility was possible. The last experiment utilized all dead nematodes. For the preparation of dead nematodes, a previous investigator (Peterson, 1 965) applied 2 percent eosin-Y dve to the nematodes,, and the stained nematodes were used after four to five days contact with the dye. Before use, the dead nematodes were filtered to r emove the excess eosin-Y dye. Since the nematode samples had to be stained by the same dye for three more days before they were counted, the total contact time with the dye was seven to eight days. It was felt that there might be some disintegration of nematodes in such a long contact time, which might lead to incorrect counting results. Therefore, ethyl alcohol and formaldehyde were studied as possible substitutes for eosin-Y dye. It was found that one percent solution of formaldehyde was sufficient to completely kill the nematodes in 1.5 hours. Ethyl alcohol was not as effective. Consequently, forrna ldehyde was used, based upon the following considerations: 1. Formaldehyde has long been used for the preservation of nematodes In the field of Hematology. There is no danger of disintegration. 2. One percent solution of formaldehyde could kill nematodes more rapidly than the two percent aqueous solution of eosin-Y dye. 3. No additional filtration was required because formaldehyde was colorless; this would reduce the handling and eliminate the possibility of losing nematodes in the filtration process. 20 k. Formaldehyde can also increase the staining process just as ethyl alcohol does. Usually, the formaldehyde was added the night before the. experiment for convenience and to reduce the time required for preparation A slightly different procedure was used for experiments 25 through 30. These experiments were designed to evaluate the effect of chemical con- ditioning upon the removal of nematodes. Since these six experiments were designed to be carried out on successive days, it was decided to use alum solution and preformed alum floe alternately. The two experiments in the middle (experiment 27 and 28) used all dead nematodes while the first two had high motility and the last two had low motility. To achieve the maximum difference in motility the nematodes for the last two experiments were kept in the refrigerator for k and 5 days. 3.3. Preparation of Chemical Feed Chemical conditioning of the filter can be achieved by applying various kinds of coagulants, coagulant aids, or filter aids to the filter Influent. In this study, an appropriate amount of alum solution or pre- formed alum floe was supplied from the chemical feed bottle shown in Figure 2.2 to give a filter influent with 10 mg/1 of alum solution or 10 mg/1 of preformed alum floe. The dosages required for the chemical feed bottle were calculated from the final concentration, 10 mg/1, filtration rate, fil- tration time, and the capacity of the bottle. The preformed alum floe was prepared by the reaction between alum and calcium hydroxide. The chemical feed was prepared about one hour before filtration was started. 3.4. Backwashing In this study, 30 minutes of backwashing at 50 percent expansion was 21 used as a general criteria for sand cleansing (Peterson, 1965). In addition, before the nematode and chemical feed were started, an effluent sample of k liters was taken for microscopic examination. If it disclosed no nematodes, the experiment was started, otherwise backwashing was resumed until no nematode was found in the effluent. 3.5- Filtration A brief description of the operation on the filter will be given be 1 ow . The reservoir (Figure 2.2) was continuously supplied with tap water at a temperature of 20°C , controlled by thermomete r s 1 and 2. As soon as the nematode concentrate and chemical feed, if used, bottles were prepared and the filter sand was cleansed by backwashing, valve A was opened and both influent and effluent flow rates were measured, and the appropriate adjustments were made at valve B to supply a sufficient total flow rate, as monitored by the flow manometer. An equal division of the total flow rate was made by the adjustments of screw clamp D, and continuously monitored by the flow division manometer. When the hydraulic conditions reached equilibrium, the nematode and chemical feed were started by electrolytic pumps. The water level in the filter column was maintained 20 inches above the sand surface, with any fluctu- ations controlled within a + one inch range. When the headless increased, this water level rose, and the constant filtration rate was controlled by gradually opening valve E to resume the constant water level. Both flow rate measurements and headloss readings were taken when the nematode samples from influent and effluent were collected. 3.6. Nematode Sampling Samples from the influent and effluent were taken every forty minutes 22 by use of 5-gallon carboys. After the 5-gallon carboy containing the composite was vigorously mixed by a magnetic stirrer, a sample volume of four liters was withdrawn by applying vacuum to the vacuum flask. According to previous work (Chaudhuri, 1964), all the nematodes in the k liters will be completely retained on a 5 u. membrane filter. This membrane filter was carefully washed into a counting dish and counted for motile nematodes. After this counting, all the contents of the counting dish were filtered through the same membrane filter. The membrane was then stained by placing it in a test tube for the counting of total nematodes. In the previous work, two 4-liter aliquots were filtered and one was used for the counting of motile nematodes. Another was used for total nematodes. Statistically, the method used in the present study was subjected to less variation in the counting results, and therefore was used throughout this study. 3.7. Nematode Staining The staining technique gave the nematode a distinct red color, and therefore greatly facilitated the counting of total nematodes. In this study, two kinds of stains were used, depending upon convenience. One of the stains used was a 2 percent aqueous solution of eosin-Y dye. It had been shown that 100 percent staining of nematodes could be achieved after 72 hours contact time (Chaudhuri, 1964); and It was used for the experiments associated with sand size and chemical conditioning. For the purpose of quick staining, another modified stain was developed. It consisted of 70 percent ethanol and 1.5 percent of eosin-Y dye. It was found that 100 percent staining of nematodes could be achieved in fifteen minutes when the nematodes were counted in the counting dish. From past experience, counting total nematodes in the counting dish was found inconvenient and easily subjected to error. Therefore, after 23 72 hours contact with the stain, the contents of the test tube were washed and filtered again. This membrane was then placed with the retained material against the grid marks of a plexiglass counting dish. A drop of water was used to hold the membrane in place. The dish was then inverted and the stained total nematodes we'e counted on the membrane. Before the ethanol -and-eos in-Y method was adopted, the following experiment was carried out. to evaluate the contact time required. Sixteen test tubes were set up with 10 motile nematodes in each, and covered with the ethanol -and-eos in -Y stain. Every test tube was covered with an aluminum cap to prevent the evaporation of ethanol. All the nematodes were carefully picked up by a micro pipette so that the starting number of nematodes was assured to be ten in each tube. At different contact times, two or three test tubes were filtered and the stained nematodes were counted on the membrane. The result is shown in Table 3 • 1 • From the result in Table 3-1, it was shown that 100 percent staining on the pad could be achieved by a contact time as short as 11 hours. The results with a contact time of 2 days or longer were erratic at the first glance. A possible explanation was that some eggs might have hatched during the 2 to 5 day period. The decrease in nematode count after 7 days contact ^ight be attributed to the disintegration of nematodes. Therefore 3 when rapid staining was desired, such as with experiments with a 16-hour filtration time, this ethanol -and-eos In -Y stain was used and the counting was conducted after a contact time between 11 hours and 24 hours. 3.8. Nematode Counting All nematode counts, including motile and total nematodes, were conducted with a dissecting microscope at a magnification of 2k X. 2k Table 3-1 STAINING EFFECT OF 70 PERCENT ETHANOL AND 5 PERCENT EOSIN-Y AT DIFFERENT CONTACT TIMES STARTING NEMATODE COUNT = 10 PER TUBE Contact Time Nematode Count on the Membrane Tube I Tube II Tube III 1 1 hours 1 day 2 days 3 days k days 5 days 7 days 10 10 10 10 10 10 13 11 17 27 16 2k 10 16 10 9 25 3 .9- Sand Sampl ing In this study, the filter sand was sampled and examined for pene- tration of nematodes only in those experiments using all dead nematodes. This was necessary because the distribution of nematodes in the sand bed was affected by the total nematodes removed in the sand. Since the total nematodes in the influent were maintained as constant as possible, and the removal efficiency had little variation for all dead nematodes (about 96 percent), the total nematodes removed were then almost constant. Consequently, these sand samples served as a good measure of the effect of sand size upon the distribution of nematodes . The sand sampling procedures used by previous investigators (Baliga, 1964, Peterson, 1965)were adopted in this study. All the samples were taken immediately after draining the sand bed. From layers at different depths below the sand surface, an accurately weighed aliquot of sand approximating 3 grams was taken and transferred to a centrifuge tube. The nematodes were then separated from the sand by using centrifugal flotation at 2900 x g for 3 minutes in a sugar solution with a specific gravity of 1.10, and the nematode concentration was expressed as number of nematodes per gram of sand from which they were separated. 26 4. RESULTS AND DISCUSSIONS 4.1. Studies on Sand Size 4.1.1. Percent Motility and Percent Total Nematode Removal Five different sand sizes were selected for this study: the effective sizes (in mm) and uniformity coefficients being 0.35 (1-5), 0.5 (1.2), 0.5 (1.5), 0.5 (1.6), and 0.65 (1.6). The sand preparations of 0.35 (1-5) and 0.65 (1.6) represented two sands having extreme character- istics and used in existing practice while the other three were within the range commonly used. A compilation of calculated experimental results is shown in the Appendix. The relation between percent total nematode removal (PTNR) and percent motile nematodes in influent (PMNIN), both based upon the average of 8-hour filtration periods, is shown in Figure 4.1. The correlation between these two variables appeared to be significant for every sand size studied. This is shown by the fact that the higher the percent motile nematodes in influent, the lower was the percent of nematode removal. It can be seen from Figure 4.1 that when the motile nematodes in the influent were between 10 and 50 percent, most experiments showed a nematode removal of 40 to 60 percent. In the range of low motility, the use of a fine sand, such as 0.35 (1.5), increased nematode removal significantly, but as the motility increased above 40 percent, the motility appeared to play a much more important role than the sand size, and thus the finer sand did not lead to an increase in nematode removal. Since it was impossible to obtain a nematode population with 100 percent motility, the percent nematode removal at 100 percent motility was 27 - D ON 4- X k. 1 ills fc 7WW » 3O0MVW3AJ 1V10± J.N33Z/3C, 28 estimated from the percent motile nematode removal in those experiments with the highest motility for each sand size. In these experiments, the percent motile nematode removal ranged from 15 to 20 percent, a relatively narrow range. In other words, it might be hypothesized that when the nematodes in the influent were all motile, various sand sizes made little difference in nematode removal, as shown in Figure 4.1. However, the most interesting and most important result was obtained in those experiments with zero motility, i.e. all dead nematodes in the influent, The nematode removal was found to be almost constant, ranging from 93-9 to 97-6 percent, regardless of the sand size used. This demonstrates the importance of nematode motility, and it appears that dead, or nonmotile, nematodes are similar to other inanimate particles, and that rapid sand filters should have no problem in removing them effectively. In view of the size of nematodes, the above con- clusion was evidentally reasonable. 4.1.2. Headloss In order to minimize the effect of turbidity on nematode removal, a clear tap water supplied by the Urbana -Champa i gn water treatment plant was employed as the raw water. Thus the particulates in the filter influent were mostly the nematodes which were injected into the tap water. In this way the effect of sand size was clearly disclosed. Consequently, the headloss in an 8-hour filtration period did not increase significantly. The observed head- losses for different sand sizes studied are shown in Table 4.1. An investigation of Figure 4.1 and Table 4.1 reveals that the removal of nematodes of low motility was generally related to the initial headloss, i.e. high initial headloss yielded high nematode removal. This observation did not apply to nematodes with zero motility, and was obscured when the motility was increased. In other words, when all of the nematodes were nonmotile, the Table k.\ HEADLOSS OF FILTER SANDS 29 Sand Head 1 oss" in inches Initial F i na 1 0.65 (1 .6) 7.8 0.5 (1.6) 10.7 0.5 (1.5) 11 .2 0.5 (1.2) 12.8 0.35 (1.5) \9.k 8.1 11 ■ 9 12 .4 13 .5 22 • 7 These are average values calculated from the 8-hour experiments for the same sand. 30 hyd r aulic resistance of all sands studied was sufficient to effectively remove nematodes by approximately 96 percent. When all the nematodes were motile, the effect of motility would be the predominating factor, i.e. most motile nematodes could penetrate through the sand filter, and no significant effect of sand size could be expected. In the range between these two extreme situations, the inter- action between motility and sand size yielded a variation in nematode removal. It was suspected that the motile nematodes could dislodge the nonmotile nematodes retained in the sand bed and thus decrease the nematode removal. This hypothesis will be further discussed in section 4.2. The ability of a motile nematode to cause dislodgment is a function of headloss (i.e. sand size). Thus the sand size not only affects the primary nematode removal, but also the ability of motile nematodes to dislodge previously retained motile and nonmotile nematodes. 4.1.3- Nematode Distribution in Sand As mentioned in section 3-9, sand samples were taken at different depths of the sand bed at the conclusion of an 8-hour filtration period. These samples were carefully examined for nematode concentration,, i.e. nematodes per gram of sand. In Figure 4.2, the nematode concentration, expressed as a percent of that in the top layer (about 0.1 to 0.2 cm thickness) of the sand, was plotted against the sand depth. From these data, it can be seen that most dead nematodes removed by the sand filter were concentrated in the top layers, with much smaller percentages occurring at the lower layers. Even with the sand size of 0.65 (1.6), the nematode concentration decreased to less than 10 percent of the top layer at only 10 cm depth. It was evident that even though the effective size was large, if the filter sand followed a normal distribution, the top layer of the sand after backwashing still possessed a significant capacity for the removal of dead 20 30 40 50 SAND DEPTH CM. F/GUPE 4.Z D/STR/BUT/ON OF NEMATODES IN DIFFERENT SANDS 60 32 nematodes. The penetration of nematodes in different sand beds appeared to be related to the initial headloss of the sand, i.e. the higher the resistance the less was the penetration. 4,2. Effect of Motile Nematodes on Removal Efficiency From previous discussions, the following question naturally arises: What effect, does the influent motile nematode percentage have upon the nematode r emoval by a rapid sand filter?" The following sections will be devoted to an extensive discussion of the effect of motile nematodes on the sand filter operation. 4.2.1. Motile Nematodes in Influent and Effluent In the previous discussions, all values used, such as PMNIN, PTNR, and those shown in the Appendix were calculated averages, which were considered to be r he only reliable basis for making comparisons. However, an evaluation of indlvldua primary data was found to be rewarding. In the course of the experiments, it was often observed that the motile nematode concentration in the influent, sample was less than or close to that in the effluent sample. Generally this phenomenon was observed mostly in the latte r part of 8-hour filtration period. Since it seemed unreasonable, a study was made. Ir Figures 4.3, 4.4, and 4.5, the individual primary data of three typical experi- ments, including the number of motile nematodes and total nematodes per liter of influent and effluent, respectively, was plotted against filtration time, and also th r ee typical experiments reported by Peterson In 1965 were plotted in the sarne way and shown in Figure 4.6, 4.7, and 4.8. A complete evaluation of all of these data is given by the general trends in Figure 4.9- The trends were then simplified and shown in Figure 4.10, and the result discussed in the following paragraphs. 33 7///1 231/7 23c/ Q01VW1N 1V10± 7/AIW 23111 23d 3QQ1VMN 11I10W ?/MjL S On CQ N vo ^ ^ ~ * ^ £ 1/NJ. 7/zvw 36 J/NJL 1/NW 37 7/N1 7/NW ?/nj_ ?//vw 39 CTi vj y) vj lO v\) ^5 CM N O <^ s: ° ^. ^ in I ^ SAND - 0.6 5 (/-6) kk 0. F/RST 4 HOURS 20 20 40 eo PNfNIN FIGURE A.IZ PERCENT NEMATODE REMOVAL OF 4-HOUR PER fOD S l/S. PERCENT MOT/LJTy J SAND -0.5 ( A a) IO0V j-V — r h5 F//Z5T 4 UOUZ.S S£CO/VO 4 HOUJZ: 10 10 -I v PMN//V io so 35 F/CV&E 4-13 PERCENT NEMATODE REMOVAL OF 4 -HOUR PER/ODS VS. PERCENT SAND -0-5 (/-5) kS 100 80 60 1 40 20 FIRST A HOURS SECOND <4 HOURS i — f 10 2o 30 PMNIN FIGURE 4J4 PERCENT NEMATODE: REMOVAL Of 4-HOUK PERIODS VS. PERCENT MOTILITY SAND -0. 5 ( /.2 ) hi 60,- f/RST 4 HOURS o o SECOND <4 HOUZS A \o 20 30 PMN/N A0 50 60 FIGURE 4-15 PERCE AST NEMATODE REMOVAL OF 4-HOUZ. PER/ODS VS. PERCENT MOTILITY J SAND- 0-35 (1-5) 48 No consistent relationship was found between percent motile nematode removal (PMNR) and percent motile nematodes in the influent (PMNIN); therefore, the PMNR of both 4-hour periods of all the experiments were plotted in Figure 4.16, with solid lines indicating the tendency in each experiment. The consistent decrease In PMNR as filtration time increased was evident despite the variation in specific deceases. Also, it should be noted that most values of PMNR during the second 4 hours approached zero. Negative PMNR was not uncommon. A comparison among different sands shown in Figure 4.16 revealed that the PMNR, i.e. the capacity of retaining motile nematodes, In the first 4 hours appeared to be proportional to the fineness of the sand. The use of chemical conditioning methods, wh ich will be discussed in section 4.3, did not show signif- icant improvement in the capacity of the filter to retain motile nematodes. 4.2.3- SIxteen-Hour Experiments From the previous discussion the following facts were quite clear: first, the rapid sand filter could achieve an almost constant nematode removal of about 96 percent over the 8-hour filtration period regardless of sand size studied if all the nematodes in influent were dead or nonmotile; second, most motile nematodes could penetrate through the sand bed even though they might be temporarily retained In the sand bed. Based upon these results, the following questions were naturally raised: "Is It possible for the dead nematodes to break through the filter after a longer filtration time? or with a higher nematode concentration In the influent?" "Is there any equilibrium between the concentration of MNIN and MNOUT?" In order to answer these questions, the filtration time was extended to 16 hours when experiments were made on the sand size of 0.5 (1.5). In experi- ment 23, where dead nematodes were used, the influent nematode concentration was hs -50 -60 O A AVERAGE VALUE IN FIRST 4 -HOURS AVERAGE VALUE IN SECOND 4 - HOURS J_ _L J_ 0.(,S{1.6>) 0.5C/.6) 0.5(1.5) 0.5(1.2) 0-15(1.5) 0.5(7.2) 0.5(1-1) /0 "9/ 1 10 "3 /I ALUM FLOC ALUM solution TYPES OF SAND F/GURE 4./6 PERCENT MOTILE NEMATODE REMOVAL OF DIFFERENT SANDS IN TWO A- HOUR PERIODS 50 doubled, i.e. about 50 nematodes per liter, to evaluate the possibility of breakthrough . The removals of nematodes were calculated for every 4-hour period and are also listed in the Appendix. In Figure 4.17, the PTNR in consecutive 4-hour periods is shown. Figure 4.18 shows the total motile nematodes in the influent and the effluent in each 4-hour period. From Figure 4.18, it is evident that the TMNOUT values are almost equal to the TMNIN values after the first four hours, i.e. an equilibrium did e* st between TMNIN and TMNOUT. From Figure 4.17, the PTNR after the first four hours also showed insignificant difference from consecutive 4-hour periods. Both of these results would confirm the establishment of an equilibrium and the ability of motile nematodes to eventually penetrate through the filter bed. Both experiments 23 and 24, in which dead nematodes were employed, showed n o significant change in TNR over the entire period of 16 hours, in spite of the doubled nematode concentration in experiment 23. This result indicated that breakthrough did not occur and that there should be no problem for rapid sand filte r s to effectively remove nematodes if all the nematodes were nonmotile or dead . 4.2.4. Dislodging Effect of Motile Nematodes It was stated previously that, in the filtration experiments using all dead nematodes, the percentage removal of nonmotile nematodes fell within a v/e ry narrow range, about 94 to 98 percent, with an average of about 96 percent. Filtra- tion experiments involving motile nematodes, however, invariably produced percentages of nonmotile nematode removal less than 96 percent. This may be seen from the data listed in the Appendix. 100 90 so 70 60 50 40 30 - 25 EXPERIMENT 24 EXPERIMENT 2 i 51 EXPERIMENT 23 EXPERIMENT 22 J_ 1st. 2nd. 3rd. -4 th. FILTRATION TIME (4- HOUR PERIOD) FIGURE 4.11 PERCENT NEMATODE REMOVAL IN FOUR 4 -HOUR PERIODS OF IS-HOUR EXPERIMENTSj SAND-O-5 (A 2 ) 52 5IO 4A0 420 _ 390- 360- ? O 240 1st. 2nd. 3rd. 4 th. FIL TRA T/ON TIME ( 4 -HOUR PERIOD ) FIGURE 418 TMN/N VS. TMMOUT IN FOU/Z 4-HOUR PERIODS OF /6-HOUR EXPERIMENTS 53 In 1965, Peterson explained this phenomenon by assuming some loss in motility of motile nematodes in the filter effluent, thus reducing the apparent nonmotile nematode removal percentages, but some further analysis based upon this assumption produced extremely erratic results and was not successful. In this study, a different approach was employed to evaluate this phenomenon. The evidence strongly indicated the possible dislodgment of non- motile nematodes by motile nematodes as they penetrated through the sand bed. The approach employed was actually a nematode balance, based upon the primary data in each experiment. In Table k.2, columns (3) and (7) were obtained from the Appendix and show the percentage of motile nematodes in the influent and the percentage of total nematode removal, respectively, for all the experiments involving motile nematodes. The percentage of nonmotile nematodes in the influent was calculated in column (4). Since 96 percent of nonmotile nematode removal could be achieved by all the sands studied, then 4 percent of the nonmotile nema- todes in the influent should be expected to be detected in the effluent. Column (5) which represented the percent of nonmotile nematodes in effluent was thus calculated by multiplying column (k) by k percent. Then the assumption was made that all of the motile nematodes could penetrate through the sand bed and appear in the effluent. Theoretically, the maximum obtainable percent of total nematodes in effluent would thus be equal to the sum of columns (3) and (5) , and is shown in column (6). The actual per- cent of total nematodes in effluent was calculated from column (7) and shown in column (8). It should be pointed out that all the percentages shown in Table k.2 were based upon the total nematodes in the influent; therefore, a quantitative comparison could be made between columns (6) and (8), and the result was plotted in Figure 4.19- Without exception, the actual PTNOUT in column (8) was found to Table 4.2 CALCULATION OF NEMATODE BALANCE 3 54 (0 (2) (3) (4) (5) (6) (7) (8) _ % °100%-(3) = (4)x4% °(5) + (3) % -)OQ%- (7) Sand Experiment pNMNIN pNMN()UT Expected pTNR Actual Number Maximum PTNOUT PTNOUT 0.65 (1.6) 15 38.4 61.6 2.4 40.8 34.0 66.0 16 24.9 75.1 3.0 27.9 37.9 62.1 17 14.7 85.3 3.4 18.1 49.3 50.7 Average 28.9 Average 59.6 0.5 (1.6) 5 49.1 50.9 2.0 51.1 38.5 61.5 6 23.7 76.3 3.0 26.7 47.4 52.6 7 13.3 86.7 3.4 16.7 51.8 48.2 Ave. 31.5 Ave. 54.1 0.5 (1.5) 21 34.2 65.8 2.6 36.8 41.2 58.8 22 16.7 83.3 3.3 20.0 51.1 48.9 Ave. 28.4 Ave. 53.8 0.5 (1.2) 1 25.7 74.3 2.9 28.6 51.5 48.5 2 25.4 74.6 2.9 28.3 49.6 50.4 3 3.6 96.4 3.8 7.4 61.0 39.0 Ave. 21 .4 Ave. 45.9 0.35 (1.5) 9 58.1 41.9 1.6 59.7 37.5 62.5 10 44.5 55.5 2.2 46.7 38.4 61.6 11 33.9 66.1 2.6 36.5 57.0 43.0 19 H.O 89.0 3.5 14.5 80.3 19.7 Ave. 39.3 Ave. 46.7 0.5 (1.2) 26 31.3 68.7 2.7 34.0 43.0 57.0 and 10 mg/1 of Alum 29 10.1 89.9 3-5 13.6 73.3 26.7 Floe Used Ave. 23.8 Ave. 41.8 0.5 (1.2) 25 37.0 63.0 2.5 39.5 40.5 59.5 and 10 mg/1 of Alum 30 14.3 85.7 3.4 17.7 60.5 39.5 Solution Used Ave. 28.6 Ave. 49.5 a Based upon TNIN 55 o ACTUAL PTNOUT A EXPECTED MAXIMUM PTNOUT X X X X X X 0.65(1.6) 0.3(1.6) 0.5(1.5) 0.5(1.2) 0.35U?) 0.50-2) 0.50-2) io**9/t of tO">a/l of ALUM SOLUTfOH ALUM FLOC 5 A NO COMB I NAT ION F/GUZE 4.19 COMPARISON BETWEEN HYPOTHET/CAL AND ACTUAL NEMA TODE BALANCE 56 be mere than the maximum expectable PTNOUT in column (6); In other words, there' had to be some mechanism other than the penetration of motile nematodes to account for the difference between columns (6) and (8). The only reasonable explanation would be that most motile nematodes not only could penetrate through the filter bed by themselves, but also could dislodge a significant fraction of nonmotile nematodes. This resulted in the poor nematode removal in the experiments involving motile nematodes. The assumption that all the motile nematodes in the influent could penetrate the filter bed was actually the worst situation. Therefore, the explan- ation stated above was actually on the safe side. In order to make a comparison between different sand sizes, the average value of columns (6) and (8) were calculated and shown in Table 4.3. The difference in these two average values are shown in the last column of Table 4.3, and it may be considered as an indicator of the degree of dislodgment in these different sands. The comparison showed the highest degree of dislodgment in the coarsest sand and lowest for the finest sand, and little difference among sands with different uniformity coefficients but the same effective size. Also, the chemical conditioning method appeared to be somewhat helpful in reducing the dislodgment. 4.3- Studies on Chemical Conditioning of Filters As stated in section 1.3, a chemical conditioning method, employing alum solution and preformed alum floe, was studied In the hope of improving nema- tode removal. The sand size used was 0.5 (1.2), and the filtration time was 8 hours . 4.3.1. Headloss A comparison of headloss development is shown in Table 4.4. 57 Table 4.3 EFFECT OF SAND SIZE AND CHEMICAL CONDITIONING UPON THE DISLOOGMENT r OF NEMATODES (DATA FROM TABLE 4.2 ) Sand (6) Average (8) Averaqe (8) -(6) Averaqe 0.65 (1.6) 28.9 59.6 30.7 0.5 (1.6) 31.5 54.1 22.6 0.5 (1.5) 28.4 53.8 25.4 0.5 (1.2) 21.4 45.9 24.5 0.35 (1.5) 39.3 46.7 7.4 0.5 (1.2) and 10 mg/1 of Floe Used At urn 23.8 41.8 18.0 0.5 (1.2) and 10 mg/1 of Solution Usee Al 1 um 28.6 49.5 20.9 58 Table k.k HEAD LOSS OF CLEAN SAND AND CHEMICALLY CONDITIONED SAND, SIZE 0„5 (1.2) Sand Conditioning Head loss (in inches) Ini t ia 1 F i na 1 0.5 (1 .2) None 12.8 13-5 0.5 (1.2) 10 mg/1 Preformed Alum Floe 1.3.1 16.1 0.5 (1.2) 10 mg/1 Alum Solution 13-3 22.5 Apparently, the use of alum solution increased the headloss more than did the preformed alum floe, while there was almost no significant increase with the clean sand. 4.3-2. Nematode Removal All of the calculated values are listed in the Appendix. The nematode removal was calculated for the two 4-hour and the entire 8-hour filtration period, respectively. A comparison with the clean sand is shown in Figures 4.20 and 4.21. It can be seen from Figures 4.20 and 4.21 that the use of both chemicals appeared to increase the percent total nematode removal when the percent of motile nematodes in influent was small (less than 20 percent), but this effect appeared to be minimized as the motility increased. The possible explanation for the im- proved nematode removal at low motility might be that the chemicals used could cause the nonmotile nematodes to adhere more tenaciously to the sand grains and thus reduced the dislodging effect of motile nematodes. Obviously, this result would substantiate the discussion in section 4.2.4 and the analysis made in Table 4.3. From Figure 4.21 and Table 4.4, the increase of headloss did not appear to increase the percent total nematode removal as time proceeded. On the contrary. 59 100 CLEAN SAND 10 mg// OF ALUM SOL UT/ON USED 10 rr\3/t OF ALUM FLOC USED _L 10 20 30 40 SO 60 PERCENT MOTILE NEMATODE IN INFLUENT (pMA//N) FIGURE 4.20 PERCENT NEMATODE REMOVAL VS. PERCENT MOTILITY FOR CLEAN SAND AND CHEMICALLY CONDITIONED SAND WITH THE SAME SIZE OF 0.5 (1.2) 60 FIGURE 4. 2 I ZO PMNIN PERCENT NEMATODE REMOVAL OF 4-HOUZ PER 100 S VS. PERCENT MOTlL/TV FOR CLEAN SAND AND CHE MICA LLY CO NO I T/ONEO SA NO WITH THE SAME SHE OF aSO-Z) 61 the nematode removal decreased as filtration time increased just as in the situation found in clean sands. Therefore, the logical conclusion which may be drawn from this study was that penetration and the dislodging effect of motile nematodes were much more significant than the use of chemical condi- tioning, which might reduce the dislodging effect to a certain extent. 62 5. CONCLUSIONS 1. The removal of nematodes was about 96 percent regardless of sand size when all the nematodes in influent were dead or nonmotile. 2. When all the nematodes in influent were dead or nonmotile, the percent removal of nematodes was constant over an entire filtration period of 16 hours with a nematode concentration in the influent as high as 50 nema- todes per 1 i ter . 3- T h e possibility of breakthrough of nonmotile nematodes is remote at concentrations less than 50 per liter. k. The penetration of nonmotile nematodes increased as the hydraulic resistance of the filter sand decreased. 5. T he motility of nematodes has a much more significant effect than sand size upon the percent removal of nematodes by rapid sand filtration. 6. Most motile nematodes can eventually penetrate the filter bed although they might be temporarily retained in the filter bed during the early stages of f i 1 1 rat ion. 7. Motile nematodes appear to cause dislodgment of a significant fraction of the nonmotile nematodes. 8. The higher hydraulic resistance of the sand medium and the use of chemical conditioning methods may reduce the dislodging effect of motile nematodes in the range of low motility. 9. As the motility increases, the motility of nematodes was the predominant factor in determining the percent of nematode removal, regardless of sand size. 10. The nematode removal decreased as filtration time increased until an equil- ibrium was reached. This was due to the penetration of motile nematodes and dislodging of nonmotile nematodes by motile nematodes. 63 6. SUGGESTIONS FOR FUTURE STUDY 1. Investigation of the removal of nematodes by ether filter medium, such as anthracite or mixed bed of sand and anthracite. 2. Investigation of various means of immobi 1 iz i ng nematodes in influent, thus increasing their removal. 3. Investigation of the removal of nematodes by up-flow filtration. h. Investigation of the effect of turbidity in raw water upon nematode removal by rapid sand filtration. 64 REFERENCES Alexopoulos, C. J., "Introductory Mycology." John Wiley and Sons, New York (1962). Anon., "Tentative Specifications for Filtering Material." J. AWWA (March 1949). Baliga, K. Y., "Benthic Sampling, Analysis, and Ecological Studies of Nematodes." M.S. Thesis, University of Illinois (1964). Bavlis, J. R., "Seven Years of High-Rate Filtration." J. AWWA, p. 585-595 (May 1956). C^ang, S. L., Woodward, R. L., and Kabler, P. W., "Survey of Free-Living Nematodes and Amebas in Municipal Supplies." J. AWWA, 52, p. 613 (I960). C h 3nc, S. L., "Survival and Protection Against Chlori nation of Human Enteric Pathogens in Free-Living Nematodes Isolated from Water Supplies." Am. J. of Trop. Med. Hyg., 9, p. 136 (I960). Chang, S. L., "Operator's Guide to Significance of Viruses, Amebas, and Nematodes in Public Water Supplies." J. AWWA, 53, p. 288 (March 1961). Chaudhuri, N., "Occurrence and Controlled Environmental Studies of Nematodes in Surface Waters." Ph.D. Thesis, University of Illinois (1964). Cleasby, J. L., and Baumann, E. R., "Selection of Sand Filtration Rates." J. AWWA (May 1962). Cobb, N. A., "Filter Bed Nemas: Nematodes of the Slow Sand Filter Beds of American Cities." Contributions to Science Neonatology, 7 S p. 189 (1918). Conley, W. R., and Pitman, R. W., "Innovations in Water Clarification." J. AWWA, 52, p. 1319 (I960). Culp, R. L., "New Water Treatment Methods Serve Richland. !! Public Works (July 1 964) . Davis, E., and Bo'chardt, J. A., "Filtration of Flocculant and Non-Flocculant Matter." A Paper at the National Symposium on Sanitary Engineering Research Development and Design (July 1965)= George, M. G. s and Kanshik, N. K., ''Infestation of Surface Water Supplies bv Nematodes." Environmental Health, 6, p. 229 (1964). Goodey, J. B., "Laboratory Methods for Work with Plant and Soil Nematodes." Technical Bulletin No. 2, London: Her Majesty's Stationery Office ( 1 96 3 ) <. Kelly, S. N., "Infestation of the Norwich, England, Water System." J. AWWA, 47 s p. 330 (1953). 65 Maffitt, H. C. s "Elimination of Nematodes, Crenothrix, and Other Organisms J. AWWA, 58, p. 119 (1966) . Peterson, R. L.., "Removal of Free-Living Nematodes by Rapid Sand Filters." M.S. Thesis. University of I! line's (196.5). 66 APPENDIX CALCULATED DATA OF ALL THE EXPERIMENTS 67 Table A.l ABBREVIATIONS Exp. No PMNIN: PTNR: PMNR: PNMNR: 0_: H: ATNIN: 7 MN IN : TMNOUT: XYZ Here X = the xth 4-hour period YZ = designated experiment number, also representing the average of 8-hour period percent of motile nematodes in influent percent of total nematode removal percent of motile nematode removal percent of nonmoti le nematode removal average filtration rate average headloss average total nematodes in influent total motile nematodes in influent total motile nematodes in effluent Table A. 2 DATA SUMMARY OF 8-HOUR EXPERIMENTS 68 Sand Exp. No. PMNIN Percent PTNR Percent PMNR Percent PNMNR Percent gpm/sq ft H ft ATNIN N/1 iter 0.65 (1.6) 115 215 15 35.8 41.4 38.4 40.2 26.7 34.0 31.4 0.2 16.7 45.2 44.2 44.8 2.99 2.98 2.98 .66 .67 .67 27 20 24 116 216 16 23.8 26.1 24.9 42.1 30.2 37-9 14.6 -4.0 6.3 50.7 42.2 48.4 3.01 3.00 3.00 .68 .69 .69 28 29 29 117 217 17 15.3 13.9 14.7 56.5 36.8 49.3 29.6 -1.4 18.2 61.4 42.9 54.6 3.00 3.00 3.00 .66 .67 .67 28 22 25 118 218 18 98.3 96.1 97-2 98.3 96.1 97.2 3.00 3.00 3.00 .63 .65 .64 22 20 21 0.5 (1.6) 105 205 5 55-2 44.2 49.1 40.0 34.9 38.5 29.4 8.1 17.3 53.1 63.3 58.8 3.02 3.00 3.01 0.97 1 .03 0.99 25 28 26 106 206 6 28.0 18.4 23.7 66.4 31 .2 47.4 42.7 -41.3 4.1 75.6 47.6 60.9 2.99 2.98 2.98 0.89 0.93 0.91 25 21 23 107 207 7 13.8 12.9 13.3 69.1 38.6 51.8 48.7 -15.4 12.4 72.4 46.6 57.8 3.00 2.99 2.99 0.89 0.92 0.91 36 37 36 113 213 13 96.7 94.3 95.6 96.7 94.3 95.6 3.01 3.01 3.01 0.91 0.95 0.93 26 22 24 114 214 14 97.0 94.3 95.8 97.0 94.3 95.8 3.01 2.99 3.00 0.91 0.95 0.93 25 22 23 0.5 (1.5) 0.5 (1.2) See 16-hour experiments 101 201 1 26.7 24.3 25.7 59.6 44.0 51.5 42.6 10.4 20.4 65.8 54.7 62.3 2.97 2.99 2.98 1.05 1.11 1 .08 22 19 20 Table A. 2 (Continued) 69 Sand Exp. PMNIN PTNR PMNR PNMNR Q H ATNIN No. 1 J ercent Percent Percent Percent gpm/sq ft ft N/1 iter 0.5 (1 .2) (Continued) 102 24.5 54.0 53.6 59.4 3.01 .09 21 202 26.7 46.1 1 .6 67.7 2.98 1 .09 20 2 25.4 49.6 23 .7 58.5 2.99 1 .10 20 103 4.7 67.6 60.8 67-9 3.01 .09 15 203 2.2 57-0 20.6 57.8 2.99 1 .09 20 3 3.6 61 .0 50.3 61.4 3.00 .09 18 104 98.8 98.8 2.96 .08 22 204 96.5 96.5 3.01 .13 13 4 97.6 97.6 2.98 .11 17 0.5 (1 .2) + 10 mg/ 1 Alum Solut ion 125 32.1 62.4 50.0 68.3 3.01 .34 18 225 43.5 10.7 19-0 4.3 3.01 .78 14 25 37-0 40.5 34.5 43-9 3.01 .56 16 130 19.8 66.9 57.2 69.3 3.02 .26 12 230 8.5 53.7 -52.1 63.5 3.01 .57 12 30 14.3 60.5 25.6 66.3 3.01 .42 12 127 97.1 97.1 3.01 .15 17 227 98.2 98.2 3.01 .25 13 27 97.6 97.6 3.01 .20 15 0.5 (1 .2) + 10 mg/ 1 Prefo rmed Floe 126 31.1 47.6 48.9 47.0 3.03 1 .20 17 226 31.6 38.4 16.7 48.9 3.00 .27 17 26 31.3 43-0 32.2 48.0 3.01 .23 17 129 10.8 74.7 43.9 78.4 3.01 1.14 17 229 9-4 71.8 17.7 77.4 3.00 1.29 16 29 10.1 73.3 32.1 77.9 3.00 1 .22 16 128 98.9 98.9 3.01 1.15 17 228 97-0 97.0 3.01 1.25 13 28 98.1 98.1 3.01 ! .20 15 0.35 ( 1.5) 109 65.1 54.9 50.0 64.2 3.01 1.68 14 209 5K2 32.1 6.2 59.2 3.00 1.84 14 9 58.1 37.5 19.7 62.2 3.00 1.75 14 Table A. 2 (Continued) 70 Sand Exp. PMNIN PTNR PMNR PNMNR H ATNIN No. Percent Percent Percent Percent gpm/sq ft ft N/liter 0.35 (1 .5) (Continued) 1 10 58.7 41.6 28.6 60.2 3.02 .65 14 210 31 .0 35.4 12.1 45.8 3.00 1.82 14 10 44.5 38.4 22.7 51.1 3.01 • 73 14 111 24.3 57-3 21.6 68.8 2.99 .69 14 211 43.2 56.8 16.0 87.8 3.00 .83 15 11 33.9 57.0 18.0 77-1 2.99 • 75 14 119 15.2 81 .3 70.2 83.3 2.98 .72 19 219 7.1 79.5 55.6 81 .3 3.02 .87 21 19 11 .0 80.3 65.2 82.2 2.99 .79 20 120 97.1 97.1 3.01 .67 28 220 97.2 97.2 3.00 .80 23 20 97.1 97.1 3.00 .73 25 112 93.7 93.7 3.03 .75 16 212 95.8 95.8 3.01 .87 16 12 94.8 94.8 3.02 .81 16 71 < \ cr in a. en »— - -*-> => c <: 1 z a> Q) vD 2: — ' — Z 1- XI o_ a; CO U. a. h- O >- 4-1 CC c ^ CC CD z 2: 2: L, => O. CD co O. < H- -4-1 < c O PMNIN PTNR Percent Perce Q. O X Z LU c 00 la vO v£> cm 4 la r-» vO — 4 O 00 4 lT\ en Ln4" tA lt\ n — — — CM CM r- — — CM CM 4 4 4 4 LA _ _ _ ^_ ^_ iaoo r^-4 .3- n \o — r-. o 4 cm r~- cm o 4 cm r~»oo O OS O o — — o a\ o o o o o\ ronroN ro roN MN co ro n ro rotN -3" OAOO O CM 00 — -3" LA O r^vfl no M 44444 cr\o — 4 — 1 4vDNLn\0 00 4 — CM la 4 oa4 oa oa — LA CM 4 4 VO CM 4 O 00 MrorOM — CM CM CM CM CM — CM oa4 O0 CTv oa \£> LA 4 OA CT\ v£> LA O vD vD O LA4 OOOOO 000 — ooooo PA oa OA oa OA 4 MD O 4 (J\ 1 1 1 1 1 1 4 OAOO O O 1 1 1 1 1 1 CM CM 4 CM CM OA vO CT\ O CO CM LA CA LA4 CT> CT\ CT\ OOOOO oaoaoaoaoa 44444 cmcmcmcmcm cmcmcmcmcm — CM OA4 — CM OA4 LA O