United States
Environmental Protection
Agsncy
Robert S Kerr Environmental Research
Laboratory
Ada OK 74820
EPA-600/2 79-068
March 1979
Research and Development
Long Term
Recharge of
Trickling Filter
Effluent Into  Sand

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional  grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:

      1.  Environmental Health  Effects Research
      2.  Environmental Protection Technology
      3.  Ecological Research
      4   Environmental Monitoring
      5.  Socioeconomic Environmental  Studies
      6.  Scientific and Technical Assessment Reports (STAR)
      7.  Interagency Energy-Environment Research and Development
      8.  "Special" Reports
      9.  Miscellaneous Reports

This report has  been assigned  to the  ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required lor the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                            EPA-600/2-79-068
                                            March 1979
         LONG TERM RECHARGE OF TRICKLING
            FILTER EFFLUENT INTO SAND
                      by

              Donald B.  Aulenbach
Department of Chemical and Environmental Engineering
         Rensselaer Polytechnic Institute
              Troy, New York 12181
                Grant No. R-803452
                Project Officer

                Lowell E. Leach
           Wastewater Management Branch
  Robert S. Kerr Environmental Research Laboratory
              Ada, Oklahoma 74820
 ROBERT S. KERR ENVIRONMENTAL RESEARCH LABORATORY
        OFFICE OF RESEARCH AND DEVELOPMENT
       U.S.  ENVIRONMENTAL PROTECTION AGENCY
              ADA, OKLAHOMA 74820

-------
                                DISCLAIMER
      This report has been reviewed by the Robert. S. Kerr Environmental
Research Laboratory, U.S. Environmental Protection Agency, and approved for
publication.  Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or recom-
mendation for use.

-------
                                 FOREWORD

    The Environmental Protection Agency was established to coordinate administra-
 tion of the major Federal programs designed to protect the quality of our environ-
 ment.

    An important part of the agency's effort involves the search for information
 about environmental problems , management techniques , and new technologies
 through which optimum use of the nation's land and water resources can be
 assured and the threat pollution poses to the welfare of the American people can
 be minimized .

    EPA's Office of Research and Development conducts this search through a
 nationwide network of research facilities .

    As  one of these facilities,  the Robert S .  Kerr Environmental Research
 Laboratory is responsible for the management of programs to:  (a) investigate
 the nature,  transport, fate and management  of pollutants in groundwater; (b)
 develop and demonstrate methods for treating wastewaters with soil and other
 natural systems;  (c) develop and demonstrate pollution control technologies for
 irrigation return flows; (d) develop and demonstrate pollution control technologies
 to prevent,  control or  abate pollution from the petroleum refining and petrochemical
 industries;  and (f) develop and demonstrate technologies to manage pollution re-
 sulting  from combinations of industrial wastewaters or industrial/municipal waste-
 waters .

    This report contributes to the knowledge essential if the EPA  is to meet the
requirements of environmental laws that it establish and enforce pollution
 control  standards which are reasonable , cost effective  and provide adequate
protection for the American public .
                                                          6
                             William C . Galegar
                             Director
                             Robert S . Kerr Environmental Research Laboratory
                                    iii

-------
                                 PREFACE

    The area around Lake George was concerned with potential pollution of the
lake long before the passage of PL 92-500.  Pollution control for the most popu-
lated area around Lake George Village was accomplished by the installation of
a rapid infiltration  system for the equivalent of tertiary treatment of the secondary
treated domestic wastes. This system has been in operation continuously since
1939 and provided the facilities for a closer evaluation of the effectiveness of the
sand infiltration technique for final treatment of domestic wastewaters.

    By providing detailed information relating to the purification of wastewater
in the soil,  a decision r?n be made whether or not to install similar treatment sys-
tems at other locations.  The information provided should aid in the design of simi-
lar treatment systems.  Some of the major advantages of a land disposal system
include the savings in chemicals,  the lack of additional sludges to dispose of,
resistance to shock loading,  reduced cost where land is available at a reasonable
price, savings in energy where extreme pumping is not required,  and complete
treatment including removal of phosphorus and conversion of nitrogen compounds
to nitrates which may also be removed by appropriate treatment procedures.  Thus,
land application of wastewater must be considered as a viable means of domestic
wastewater treatment and disposal.
                                       IV

-------
                                  ABSTRACT
      The Lake George Village Sewage Treatment Plant has been providing the
equivalent of tertiary treatment of domestic wastewater by applying wastewater
which has been subjected to trickling filter purification and secondary sedimentation
to a natural delta sand deposit by the rapid infiltration method.  The sand system
has been shown to remove decomposable substances such as 5-day biochemical
oxygen demand (BOD-), chemical oxygen demand (COD),  alkylbenzensulfates,
coliforms and fecal coli, and ammonia and organic nitrogen.  Soluble inorganic
materials such as sodium,  chloride, potassium, and nitrate generally passed
through the bed unchanged.  Calcium and magnesium showed no significant
changes; whereas alkalinity increased due to carbonate and bicarbonate reactions
within the soil system.  Orthophosphates were  completely removed, and nitrates
were somewhat removed under reducing conditions produced in some of the deeper
sand beds.

      Operation has been continued successfully throughout the cold winter
months experienced in this area.  Allowing grass to grow in the infiltration beds
provided an increased infiltration rate at flooding depths exceeding 0.3 meters
(m) (1 foot,  ft);  whereas a decrease in infiltration rate was observed under
shallower depths of water.  The greatest removal of constituents occurred in
the top 10 m (30 ft) of most of the sand beds;  however,  continued quality improve-
ment was observed through further vertical flow and approximately 600 m (200
ft) of horizontal flow.  There  does not seem to be any loss of the capacity of the
complete system to remove the nutrients.

      Both vertical and horizontal flow studies were made. The average vertical
velocity in bed N-ll was 0.63 m/day (2 ft/day) .  The horizontal velocity varied
throughout the area but was in a general range of 3 to 12 m/day (1 to 4 ft/day) .

      Studies  at the Lake George Village Sewage Treatment Plant have shown
that a rapid infiltration system for purifying  secondary treated effluent can provide
the equivalent of tertiary treatment with nutrient removal for an extended period
of time exceeding at least 38 years.  This information should be useful in providing
data for the design of similar  systems serving other areas of small population.

      This report is submitted in fulfillment  of Grant #R803452 by Rensselaer
Polytechnic Institute (RPI)  under the partial  sponsorship of the U.S. Environmental
Protection Agency.  This report covers the period from November 1974 through
June 30, 1977, but also includes data collected by the New York State Department
of Environmental Conservation (NYSDEC) between June 1973 and October 1974,
and work was  completed as of November 17,  1978.
                                      v

-------
                                CONTENTS

Foreword	iii
Preface   	    iv
Abstract	    v
Figures	viii
Tables    	xiii
Acknowledgement	xiv

    1.   Introduction	    1
             Land Treatment of Wastewater   	    1
             Lake George and  the Lake George Village Sewage
             Treatment Plant	    3
    2.   Conclusions   	    8
    3.   Recommendations	   10
    4.   Materials and Methods	   H
    5.   Experimental Procedures	   22
             Sampling  Procedures	   22
             Analytical Procedures  	   23
                 Field Analyses   	   23
                 Laboratory Analyses	   24
    6.   Results  and Discussion	   28
             Changes in Quality Through the Soil	   28
                 Changes with Vertical Transport	   28
                 Changes with Horizontal Transport	   53
             Hydraulic Characteristics of the Sand System	104
                 Bed Dosing	104
                 Infiltration Studies	112
                 Percolation Rates   	119
                   Vertical Flow Through the Unsaturated Sand	121
             Soil Analyses and Well Tests	140

References   	,	143
                                    VI i

-------
Number                            FIGURES                              Page

   1      Diagrammatic Sketch of the Operation of a Rapid
            Infiltration System for Land Application of
            Water and Wastewater ....................   2

   2      Map Showing Location of Lake George  Village  .........   4

   3      Plan of the Lake George Village Sewage Treatment
            Plant  ...........................   5
   4      Photograph of the Treatment Plant
   5      Seasonal Flows Through  the Lake George Village
            Treatment Plant   ......................  12

   6      Map of the General Area Involved in This Study ........  13

   7      Sketch of  the Wells Used in the General  Study   ........  15

   8      Sketch Showing the Installation of the Suction
            Lysimeters .......................  17

   9      Plan of Bed N-ll    ......................  19

  10      Profile of  Bed N-ll   .....................  20

  11      Seasonal Variations in Temperature with Depth in
            Bed N-ll  .........................  29

  12      Seasonal Variations in pH with Depth in  Bed N-ll  ......  31

  13      Seasonal Variations in Total Dissolved Solids with
            Depth in Bed N-ll   ....................  3Z

  14      Seasonal Variations in Dissolved Oxygen with Depth
            in Bed N-ll  ........................  "
  15      Variation of Oxidation-Reduction Potential with
           Depth in Bed N-ll During Spring  1976
  16      Seasonal Variations  of Chloride with Depth in
           Bed N-ll    ........................  36

  17      Seasonal Variations  in Calcium with Depth in
           Bed N-ll    ........................  37

  18      Seasonal Variations  in Magnesium with Depth in
           Bed N-ll    ........................  39
                                      viii

-------
                               FIGURES (cont.)

Number                                                                 Page
  19      Seasonal Variations in Alkalinity  with  Depth in
            Bed N-ll	40

  20      Variations in the Forms of Nitrogen  Measured with
            Depth in Bed N-ll  During the  Summer of  1975	41

  21      Variations of Various  Forms of Nitrogen Measured
            with Depth in Bed N-ll During the  Fall of 1975	42

  22      Variation of the Various Forms of Nitrogen  Measured
            with Depth in Bed N-ll During Spring 1976   	43

  23      Variation in Ortho and Total Phosphate with Depth
            in Bed N-ll During Summer 1975    	45

  24      Variation in Ortho and Total Phosphate with Depth
            in Bed N-ll During Fall 1975	46

  25      Variation in Ortho and Total Phosphate with Depth
            in Bed N-ll  for Spring 1976   	47

  26      Seasonal Variations in Iron with  Depth in Bed N-ll	48

  27      Seasonal Variations in Sodium with Depth in Bed N-ll	50

  28      Seasonal Variations in Potassium with Depth in Bed N-ll ....   51

  29      Seasonal Variations in the Ratio  of Potassium to
            Sodium with  Depth  in Bed N-ll   	52

  30A     Seasonal Variations in Temperature	54

  30B     Seasonal Variations in Temperature	55

  30C     Seasonal Variations in Temperature	56

  31A     Seasonal Variations in Dissolved Oxygen	58

  31B     Seasonal Variations in Dissolved Oxygen	59

  31C     Seasonal Variations in Dissolved Oxygen	"0

  32A     Variation in Redox Potential During Spring  1976	61

  32B     Variation in Redox Potential During Spring  1976	62

  32C     Variation in Redox Potential During Spring  1976	63
                                        IX

-------
                                FIGURES (Cont.)

Number                                                                   Page

  33A     Seasonal Variations in pH	64

  33B     Seasonal Variations in pH	^5

  33C     Seasonal Variations in pH	*>6

  34A     Seasonal Variations in Nitrate	68

  34B     Seasonal Variations in Nitrate	"°

  34C     Seasonal Variations in Nitrate	70

  35A     Seasonal Variations in Ammonia Nitrogen	    1

  35B     Seasonal Variations in Ammonia Nitrogen	   '^

  35C     Seasonal Variations in Ammonia Nitrogen	   ''

  36A     Seasonal Variations in Total Kjeldahl Nitrogen	7^

  36B     Seasonal Variations in Total Kjeldahl Nitrogen	75

  36C     Seasonal Variations in Total Kjeldahl Nitrogen	'°

  37A     Seasonal Variations in Soluble Reactive Phosphate-
            Phosphorus 	   '9

  37B     Seasonal Variations in Soluble Reactive Phosphate-
            Phosphorus  	80

  37C     Seasonal Variations in Soluble Reactive Phosphate-
            Phosphorus  	

  38A     Seasonal Variations in Total Phosphorus	°^
                                                                           QO
  38B     Seasonal Variations in Total Phosphorus	OJ

  38C     Seasonal Variations in Total Phosphorus	°'

  39A     Seasonal Variations in Alkalinity	86

  39B     Seasonal Variations in Alkalinity	87

  39C     Seasonal Variations in Alkalinity	88

  4GA     Seasonal Variations in Chloride	8^

-------
                               FIGURES (Cont.)

Number                                                                 Page

  40B     Seasonal Variations in Chloride	90

  40C     Seasonal Variations in Chloride	91

  41A     Seasonal Variations in Dissolved Solids  	92

  41B     Seasonal Variations in Dissolved Solids	93

  41C     Seasonal Variations in Dissolved Solids	94

  42      Winter Operation of a Sand Bed (Bed S-4-
            February 7, 1975)   	113

  43      Infiltration Rates of Several Sand Beds During Normal
            Operation	115

  44      Changes in Infiltration Rates of Several Sand Beds
            Under Normal Dosing Operations  Including Normal
            Drying Periods	116

  45      Infiltration Rate in Bed S-5 Showing the Decline in
            Rate with Continued Normal Operation After
            Cleaning the Bed	117

  46      Infiltration Rate in Bed N-3 with Continuous Dosing
            Following Cleaning  of the Bed	118

  47      Comparison of Two Beds Under  Constant Inundation
            One with a Weed Cover and One  with the Weeds
            Removed	120

  48      Vertical Transport of Rhodamine WT in Bed  N-ll	122

  49      Plan Showing Location of New Wells Installed
            Specifically for the Tracer Studies    	127

  50      Occurrence of the  Tracers in Observation  Well  5	130

  51      Occurrence of the  Tracers in Observation  Well  15	131

  52      Occurrence of the  Tracers in Observation  Well  16	132

  53      Occurrence of the  Tracers in Observation  Well  17	133

  54      Occurrence of the  Tracers in Observation  Well  18	134

  55      Occurrence of the  Tracers in Observation  Well  20	135
                                       XI

-------
                              FIGURES (Cont.)

Number                                                               Page

  56      Occurrence of the Rhodamine WT  Tracer in Observation
           Well 9	136

  57      Occurrence of the Rhodamine WT  Tracer in Observation
           Well 6B	137
                                     Xll

-------
                                   TABLES

Number
                                                                      Pace
   1     Well Data	  16

   2     Fecal Coliforms	96

   3     Total Coliforms	98

   4     Enterococci	100

   5     Coliphage   	101

   6     Reduction in BOD in the Conventional Portion of the
          Lake George Village Sewage Treatment Plant	102

   7     BOD Reduction While Standing on a Slowly Infiltrating
          Sand Bed	103

   8     Biochemical Oxygen Demand (5-Day)	105

   9     Mean Monthly Flows	106

  10     Sand Bed Areas   	108

  11     Bed Dosing Frequency	109

  12     Depth of Applied Sewage	 H°

  13     Precipitation Effects   	HI

  14     Data  for Special Wells	128

  15     Velocity of Flow  in  Saturated Zone	138

  16     Sieve Analyses - Bed #13	140

  17     Sand Analysis    	141

  18     Sand Analysis -  Bed 11	142
                                     Xlll

-------
                            ACKNOWLEDGEMENT

    The author would particularly like to thank the students who conducted much
of the sampling and analytical work in the completion of this project. Those stu-
dents supported on this project include:   Louis Hajas, Stephen Beyer, Bradford
Middlesworth, Preston Chiaro, Robert Reach, and Jeffrey Bull.

    Special appreciation must be given to Harold Gordon, Superintendent of the
Lake George Village Sewage Treatment Plant,  and his crew who cooperated in
this study and who were at times inconvenienced by our studies.  Special thanks
must be given to the technical staff of the Rensselaer Fresh Water Institute at whose
laboratories numerous analytical measurements were made.

    The installation of the initial wells was conducted by Dr. T.  James Tofflemire
of the New York State Department of Environmental Conservation (NYSDEC) .  Some
initial quality measurements were also conducted by NYSDEC.

    A small portion of the results reported in this study was supported by a contract
from the US Army Corps of Engineers Cold Regions Research and Engineering Labora-
tories, Hanover, New Hampshire. RPI, itself, supported the major portion of the
studies conducted for the Corps of Engineers.
                                    xiv

-------
                                 SECTION 1

                               INTRODUCTION

LAND TREATMENT OF WASTEWATER

    Today's interest in land treatment of wastewater stems from man's increasing
concern for improving the quality of the environment.  Public Law 92-500 has influ-
enced the development of improved waste treatment methods within the framework
of increased efficiency,  greater energy conservation, and a general reduction of
operating costs. The potential water reuse factor achieved with land treatment has
received particular attention.  TheU. S. Environmental Protection Agency (EPA)
has supported research and promoted its public use.

    Concern for zero pollution discharge under Public Law 92-500 has rekindled
interest in land treatment of wastewater.  Land treatment is intimately linked with
ground water.  Analysis of the existing and expected ground water quality is vital
in the implementation of a land treatment system.  There is justifiable concern with
pollution of the ground water, capacity of the system (rising ground water tables
are very troublesome) ,  and the ultimate residence of the water.  The most exciting
parts of land treatment are the recycle aspects. Under ideal conditions both nutrient
recycle and water recycle are possible. Storage of water underground is desirable
due to minimal evaporative losses, cool temperatures, slow flow rates, and the
natural filtering capacity of the soil.

    Specifically, this report will concern itself with the rapid infiltration form of
land treatment which is one of the three basic forms of land treatment.  These forms
include:   spray irrigation (SI),  overland flow (OF), and rapid infiltration (RI) .
Rapid infiltration,  or infiltration-percolation as it is also referred to, (Figure 1)
involves:   (1) the flooding of a basin with between 0.5 ft. and 3.0 ft.  (0.15 m to  1 m)
of treated wastewater, (2) the infiltration of this through the soil  surface,  (3) perco-
lation through the soil matrix, and (4)  eventual entrance to the ground water. Reno-
vation of the wastewater occurs by physical, chemical,  and biological processes  (1).

    The Federal Water Pollution Control Act Amendments of 1972, PL 92-500,  has
brought land treatment into the spotlight.  There are four sections within which
land treatment systems are to be considered (2):  Section 208 - Areawide Waste
Treatment Management; Section  201 - Facilities Planning; Section 204 - Best Practi-
cable Treatment Technology; and Section 212 - Cost Effectiveness Analysis.  These
and other program memoranda to EPA regional administrators have resulted in the
growing interest in land treatment.

    The cost effectivensss  analysis has shown land treatment to be a very economical
treatment technique.  In many cases, a true cost effectiveness analysis is impossible

-------
                 RECHARGE  BASIN
Saturated
clogged  zone
                                            Ground  surface
                Unsaturated  transmission  zone
                   Ground  water mound
                                                 Water table
      Figure 1.  Diagrammatic Sketch of the Operation of a Rapid Infiltration
              System for Land Application of Water and Wastewater.

-------
to perform an 85-90 percent removal analysis.  Most land treatment systems approach
100 percent BOD removal. A recent study has shown that land treatment systems
are considerably less expensive, expecially in smaller treatment plants (3,4).  Of
the available land treatment techniques, rapid infiltration is the most economical
(1). Capital and operating costs are much lower than for OF and SI due to smaller
land requirements and  gravity distribution. Generally,  pretreatment costs are
higher than for other land treatment methods.

      The application of sewage to the land in one form or another dates back to
ancient times.  The EPA has identified soil treatment systems that were established
in the United States as  early as 1880 (5). In 1972, the EPA identified 571 municipal
wastewater facilities using surface application systems (1). The EPA has illustrated
the feasibility of the RI method by identifying several existing successful facilities.
These include  (1):  Hemet, California;  Phoenix, Arizona (Flushing Meadows);
Whittier Narrows,  California;  Santee, California; Vineland, New Jersey; Marysville,
California; Westby, Wisconsin; and Lake George, New York.  The effectiveness
of land treatment as an available method of advanced wastewater treatment has
been well documented.  More work must be  done on design and operational criteria
in order to design such systems for optimum efficiency.


LAKE GEORGE AND THE LAKE GEORGE VILLAGE SEWAGE TREATMENT PLANT

      Lake George is a beautiful,  clear lake noted for its tree-lined shores.  The
Lake is located in the southeast corner of the New York State Adirondack Park (Figure
2). The Lake George area is a tourist attraction in summer and in winter. Most
of its prosperity is due to the recreational value of the Lake.  Much credit must
be given to the Lake George Association for the maintenance of the Lake's purity.
The Lake George Association was organized in 1885  and, due to its efforts, the
Lake was given an "AA" classification (6).  This classsification prohibits sewage
discharges of any type into the Lake or any waters discharging  into the Lake (7) .
The Lake water is  used as a drinking water supply requiring only chlorination
prior to use as public drinking water (8).

      The Lake George Village sewage treatment plant (Figure 3) was constructed
in 1936.  The law concerning  discharges into the drainage basin was interpreted
to mean surface discharges. The law did not apply to subsurface discharges;
therefore, soil system  disposal (including septic tanks)  of effluent  was considered
within the law. The treatment plant was put into operation in 1939  and has run
continuously since that time.

      The treatment plant has had five additions to the initial six sand beds since
it was built. Bed N-7 was added in 1947, beds N-8 and 9 were added in 1950,
and beds N-10, 11, and 12 were added in 1956.  In 1965 the plant had a major expan-
sion with eight additional beds put into service (the South beds 1-6 and North beds
13 and 14).  In 1970, the final bed (S-7)  was put into use. The plant is located
(Figure 4) southwest of the lake.  The influent scheme is an intermittent pumped
system from two separate force mains located at the edge of the lake.

-------
                                                 Lake
                                                 Champlaln
                                                 II
Figure 2.  Map Showing Location of Lake George Village.

-------
          &
       ^
       •of
f    V
  ,s
      •§ ,,.«.
        Primary
        ScttliNi
        Tanks

      l*fl«*nt
      ChaMbcr
Figure  3.   Plan of the  Lake George Village Sewage Treatment Plant

-------
           FIGURE 4.  Photograph of the Treatment Plant

This is a view of the Lake George Village Sewage Treatment Plant
taken from the top of Prospect Mountain.  Looking southeast,
Interstate 87 (the Northway),  lies adjacent to the treatment plant
on its western boundary. West Brook is in the foreground hidden
by trees.  The road in the lower  right corner of the photo leads
to the West Brook upstream  staff  gauge.  Lake George is located
northeast of the treatment plant.
                               6

-------
      The plant was originally built in triplicate (9) .  Summer flows were approxi-
mately three times the winter flows .  Today the ratio is approximately two to one
(10) .  At the head of the plant are two nine-inch  (29.3 cm) Parshall flumes to re-
cord the flow of the Village and Town pumping stations .  Primary treatment con-
sists of one circular Imhoff tank and two mechanically  cleaned circular settling
digestion tanks  (clarigesters) .  Secondary treatment is accomplished by two high
rate rotating  arm trickling filters in  summer and one covered standard rate fixed
nozzle sprinkling filter in winter. Secondary sedimentation is accomplished by
two mechanically cleaned rectangular settling tanks and two circular settling tanks.
After  secondary sedimentation, the treated sewage is discharged without chlorination
to the sand infiltration beds .

      Gravity is used to convey the  treated sewage to  the 14 northern (lower)
beds. The sewage is pumped to the newer 7 southern (upper) beds. Normally,
the sewage is discharged to two beds at a time, one lower bed, and one upper bed.
Dosing is  changed at  approximately  8: 00am  and  at 4: 00pm; thus, the entire day's
flow is discharged to  a total of four beds. The effluent takes from  1/2 to 3 days
to seep into the ground, depending primarily  on  the size, age, and condition of
the bed. The newer south beds are  fairly uniform in size, but the areas of the
north beds vary significantly.  The  newer beds  have higher infiltration rates than
the older beds.  They all require periodic removal  of the surface mat which forms
on the sand and inhibits infiltration.  There is no set cleaning schedule; the clean-
ing of the beds is based on the observed condition of the bed, availability of the
bed for drying  and cleaning, and upon the time  spent by plant personnel on other
duties.  Cleaning consists of raking  and removing the  upper few cm of sand and
discing followed by releveling of the sand surface.  The mat which is removed
is either taken to a sanitary landfill  or it is  applied on  the forest floor in areas
surrounding  the treatment plant.  Sludges from  the settling tanks are dried on
sludge drying beds and are then disposed of in a sanitary landfill.

      Since this system has been in  operation for 38 years, it affords an ideal
opportunity to study the long term effects of treatment  using the rapid infiltration
technique. Many other areas may find such systems attractive. It is hoped that
this study of the Lake George  system may help resolve problems in the design of
future rapid infiltration systems .

-------
                                 SECTION 2

                               CONCLUSIONS

    The application of secondary treated effluent from a conventional domestic
sewage treatment plant onto natural delta sand beds using the rapid infiltration
technique with intermittent dosing has been shown to provide the equivalent of
tertiary treatment.  Such a system may be considered to be continuously effective
for long periods of time as exemplified by nearly 40 years of operation of the Lake
George Village Sewage Treatment Plant.

    A significant portion of the purification of the secondary  effluent applied to the
sand beds was accomplished in the upper unsaturated zone of the sand infiltration
beds during the vertical transport of the liquid. Biological oxygen demand (BOD) ,
chemical oxygen demand (COD), coliforms,  fecal coliforms, and streptococci were
effectively removed in the top  3 m  (10 ft) of sand.  Ammonia and organic nitrogen
were oxidized to nitrate in the top  3 m  (10 ft) of sand.  Data indicated a  further re-
duction of nitrate at greater depths, with almost total removal of nitrate  by the time
the wastewater passed through 18 m (60 ft) of sand.  There appeared to be a
positive correlation between DO, redox potential, and the oxidation of reduced
nitrogen compounds to nitrate, followed by the  reduction of nitrate to nitrogen gas.
This correlation is supported by measurable DO present at all depths within the
sand beds and within the aquifer below the sand beds to depths of 18 m  (60 ft) where
reduced nitrogen levels were observed.  There was further oxidation of reduced
nitrogen compounds in the horizontal transport of these compounds through the
soil.  Orthophosphate was generally reduced to concentration less than  1 mg/1
as P in  3 m (10 ft) of vertical transport, and concentrations seldom exceeded 0.2
mg/1 at the 7 m (23 ft) depth.  There was further removal of phosphate in the horizon-
tal transport  of phosphorus compounds in the saturated aquifer below the sand
beds.  Where the ground water reappears as seepage, approximately 600 m (2,000
ft) from the  sand infiltration beds, the soluble phosphate content averaged 14 Hg/1
which is approximately the same as the natural ground water in the area.

    Non-reactive soluble substances such as sodium, chloride, potassium,  and
total dissolved solids  showed no significant changes  during vertical or horizontal
transport through the sand.

    Low concentrations of viruses  (coliphage) were found in all the sampling wells
including the control wells. Highest values were found in wells 1 and 9, but there
was no particular pattern with respect to well location and concentration in any
of the other wells. Insufficient information was available to evaluate the significance
of these viruses or the degree of removal of them.

-------
    Copper and aluminum were found to be present in concentrations lower than
the detectable limits with normal atomic adsorption techniques.  The iron concentra-
tion in the ground water receiving sewage effluent was lower than in the natural
ground water in the surrounding area. Anionic non-soap surfactants were below
detectable limits at all locations .

    Although there was little significant change in pH with either vertical or
horizontal movement through the soil, there was an increase in alkalinity with
depth in the sand beds. This may be accredited to changes in the carbonate-
bicarbonate balance caused by production of CO- by biological activity within
the sand beds.

    Intermittent dosing of the sand beds is important in maintaining both efficiency
of treatment and speed of infiltration of the liquid onto the soil. Actual infiltration
rates ranged from a low of 0.07 m/day  (0.25 ft/day with a head of 0.15 m (0.05
ft) of sewage on the bed to a high of 0.7 m/day (2.25 ft/day) with a head of 0.46
m (1.5 ft) of  effluent on a  clean bed. Infiltration rates with a bed covered with
weeds  were lower than on a weed-free bed when the depth of liquid on the sand
bed was less than 0.3m (1 ft) and higher on a weed-free bed when the loading
depths exceeded this amount. The average  vertical velocity in bed N-ll was
.63 m/day  (2 ft/day) . The horizontal velocity  varied throughout the area but
was in the general range of 9 m/day (30 ft/day) .

-------
                                 SECTION 3

                            RECOMMENDATIONS

    For small installations treating primarily domestic sewage, a rapid infiltration
system applying secondary treated effluent to sand with intermittent dosing is
recommended where site conditions are favorable.  This system could provide an
efficient, inexpensive, and energy conserving method of tertiary treatment for the
removal of most common contaminants and nutrients.  Although each individual
situation should be monitored routinely, such a system can be expected to last for
many years and perhaps indefinitely.

    It has been shown that phosphorus may be completely removed in a sand
infiltration system,  even though mechanisms of removal are not understood.
Additional studies should be made to determine the methods of removal so
that this system could be applied in other similar situations.

    Complete nitrogen removal appears to be accomplished in the deeper (18 m)
sand beds. Additional studies should be made to determine if more complete re-
moval of nitrogen could be achieved in less vertical distance.  Methods  to vary the
DO and redox potential may be effective in achieving nitrogen removal.  In addition,
the removal of nitrogen by plant extraction, such as allowing the weeds to  grow on
the sand beds, should be studied.
                                     10

-------
                                   SECTION 4

                           MATERIALS AND METHODS

      The basic facilities used for this study were the existing Lake George Village
Sewage Treatment Plant (Fig. 3) and its variable inflows from Lake George Village
and the Town of Lake George.  The sewage influent to the treatment plant is primarily
of domestic nature received from both the Village and Town of Lake George which
are highly seasonal tourist areas. The flows have increased over a period of years
as shown in Fig. 5 with the summer  flows presently about double the winter flows.
The peak daily summer flows reach approximately 5 x 10  I/day (1.3 mgd). The
plant was put into operation in 1939  (9) to meet the requirements of the state regu-
lations (7) that there should be no discharge of sewage or sewage effluent into Lake
George or any tributary of Lake George.  It was considered that the application of
the treated effluent into sand beds would provide sufficient treatment that by the
time the water did reach Lake George it would have no effect upon the quality of the
Lake.

      The treatment plant has operated continuously since its completion; however,
no studies had been conducted to determine the quality of the effluent or  the final
destination of the  applied effluent in the sand beds until an initial attempt was made
by the author in 1968 (11) • These initial studies were begun even though sufficient
resources were not available for a detailed study at that time.

      In order to  determine the possible direction of flow of the treatment plant
effluent infiltrated into the ground through sand beds, additional studies were con-
ducted during 1972  (12) in which the resistivity of the soil including the ground water
was measured. The general direction of low resistivity (high conductivity) was ob-
served to follow a path nearly due north from  the treatment plant along Gage Road
toward West Brook. The study could not  conclude whether or not the zone of low
resistivity passed under West Brook  or terminated at this location.

      Based upon the above information,  a reconnaissance survey was conducted
of the area along the south bank of West Brook. Considerable seepage was observed
coming out of the base of the hill at the edge of the flood plain of West Brook. The
conductivity of the seepage was high, giving the first indication of the possibility
that the applied sewage effluent was reappearing from the ground and becoming
surface water which flowed into West Brook and ultimately into Lake George. With
this information available, the New York State Department of Environmental Con-
servation (NYSDEC) began installing observation and sampling wells in  several of
the sand beds and in the area between the sand beds and West Brook as the need
indicated to sample the water quality treatment efficiency of the sand beds.  The lo-
cations of all the wells  and other sampling points in the area are shown inFigure 6.
Wells 4,7, and 10 were designed to be control wells unaffected by the applied
                                      11

-------
    350-
    300
    250
    200

CO
(E
UJ

t   BO
    100
    50
                                                                            80
                                                                           -60
                                                                           -20
           1968    1969    1970    1971    1972    1973    I9T4    1975     1976
      Figure 5.  Seasonal Flows Through the Lake George Village Sewage Treatment
                 Plant.

-------
Figure 6.  Map of the General Area Involved in this Study Showing
          the Observation Wells and Other Sampling Points.
                            13

-------
 sewage effluent, and therefore indicative of the quality of the natural ground water
 in the area. An attempt was made to locate one additional control well, no. 13,
 along Prospect Mt. Road on the eastern side of Rt. 187 (the Northway); however,
 an obstruction was encountered before the ground water table was penetrated at
 this location. The elevations of the ground surface,  ground water level,  well screen
 elevations, and the bedrock elevations of each well are compared in Fig. 7 with
 actual data given in Table 1.

     Three different types of wells were utilized in this study.  The wells marked
 S in Table 1 were constructed of 3.15 cm  (1.25 in) steel pipe.  The first  section
 of pipe consisted of a steel tipped point and a 0.6 or 0.9 m (2 or 3 ft) well screen.
 Additional 1.3m (4 ft) sections of drive pipe were added as the point was driven
 into the sand.  A 113 kg (250 Ib) weight suspended from a tripod rig was used
 to drive the well point and pipe into the ground. This equipment was provided
 by NYSDEC. One disadvantage of steel pipe is that it rusts, thereby invalidating
 any values of iron determinations in the samples .  Thus, some  additional wells
 were installed using plastic pipe to eliminate sample interference from rust.  A
 commercial well  driller was contracted to install the plastic pipe wells.  Ten centi-
 meter (4 in) diameter holes were augered to bedrock and the 3.15 cm (1.25 in)
 plastic  screen  and pipe was then installed inside the auger tubing.  The  auger
 was then retrieved leaving the plastic well pipe in place.  The  holes were backfilled
 after the tools were retrieved.

     The third type of wells installed were constructed using the cable-tool precussion
 method of well drilling (13) . This involved raising  and dropping a heavy string
 of drilling tools into a bore hole. The reciprocating action of the tools mixed a
 slurry which was removed at regular intervals using a bailer.  As the soil was
 loosened and removed, a 15.25 cm (6 in)  steel casing was driven down the bore
 hole to  support the side-walls of the well. Within this  casing submersible pumps
 were installed  in each of the two wells constructed in sand bed  N-ll and  are labeled
 US and 11D (shallow and deep). The well screens in  these wells were 0.9m  (3
 ft) long and were located in the top 0.9m  (3 ft) of ground water for the shallow
 well and at the top of bedrock for the deep well. Appropriate wiring and piping
 were installed to provide continuous pumping of these  wells. During the installation
 of these large diameter pumping wells, samples of soil were taken for analysis
 at intervals of approximately 0.6m (2 ft) . These samples were sent to the U.S.
 Army Corps of Engineers Cold Regions Research and Engineering Laboratory,
 Hanover, New Hampshire, for chemical analysis and to NYSDEC for particle
 distribution and permeability analysis.

    In the  special case of bed N-ll, suction lysimeters were also installed for  sam -
 pling at various depths within the unsaturated portion  of the sand bed.  Holes were
 drilled by the cable-tool  percussion method as described above. The suction lysi-
 meters were lowered into the cased holes  attached to a  piece of PVC pipe  for support.
 The holes were backfilled with a slurry of 100 mesh silica sand placed around the
 porous ceramic cups of the lysimeters followed by a layer of sand that had been
 removed from the bore hole,  as shown in Fig. 8. A plug of bentonite clay was placed
 between each successive lysimeter to prevent the direct passage of any liquid down
 through possible voids surrounding the lysimeter.  The casing  was retrieved as
 backfilling progressed.  Several of the lysimeters sank when a  section of casing
 was removed, resulting in their exact depth not being accurately known.  Of the
13 lysimeters installed, only four were found  to operate satisfactorily.

                                      14

-------
151
^6

O
131'
12*
Ill-
Ill-
       2*  2 I  3* 31  JC  30
                                    S  CD  $1  7  M IK* (I  J  II  US 110 12A 12B 14
                                 WELL  NUMBER
  Figure 7.  Sketch of the Wells Used in  the General Study.   The  Upper Solid
             Line Indicates the Ground Surface  Level;  the  Dashed  Line  Indicates
             the Approximate Water Level  in the Well;  the  Next  Solid Line  at
             the Bottom of the Vertical Line Indicates the Depth  of the Well
             Point, the Lowest Solid Line Indicates the Depth of  the Bedrock.

-------
           TABLE 1.   WELL DATA
(Elevations in m.  Above Mean Sea Level)

Location
1
2A
2B
3A
3B
3C
3D
4
5
6A
6B
7
8AA
8A
8B
9
10
US
11D
12A
12B
14
Steel or
Plastic
P
P
S
P
P
P
S
S
P
P
S
S
P
P
P
S
S
S
S
P
P
S
Top of
Well
145.57
115.25
115.44
104.05
104.19
104.35
103.73
115.68
152.50
140.03
140.25
153.85
143.37
143.66
143.68
143.11
141.80
144.98
145.17
136.04
136.06
144.27
Ground
Surface
144.77
114.45
114.45
103.63
103.63
103.63
103.63
114.52
151.03
139.73
139.85
152.53
143.30
143.21
143.29
142.28
141.08
143.95
143.95
135.69
135.71
143.80
Approx.
Ground
Water
126.47
109.28
109.31
103.47
103.52
103.54
103.44
114.50
147.22
120.92
120.59
151.55'
123.48
122.03
122.02
123.10
134.66
124.06
123.81
117.54
117.04
125.81
Bottom of
Point
124.44
107.36
100.72
102.57
100.32
97.94
96.23
112.65
146.16
118.68
109.76
150.47
121.96
121.21
118.38
120.58
133.81
121.09
117.40
116.55
109.07
124.61
Bedrock
123.48
93.29
93.73
95.73
95.73
95.73
95.73
112.65
145.55
109.76
109.76
150.47
118.38
118.38
118.38
120.58
133.81
115.91
115.91
109.05
109.07
124.00
                 16

-------
                TUBING
                 CHAMBER
                 POROUS  CUP
                           SILICA   SAND



                           DELTA  SAND



                           BENTONITE




Figure 8. Sketch Showing the Installation of the Suction Lysimeters
                         17

-------
     The actual location of the wells and lysimeters in Bed N-ll is shown in
 Fig.  9, and the depths of the individual wells and lysimeters are shown in
 Fig.  10. Chemical analysis presented later indicates that the 4 operating lysimeters
 were located at depth intervals sufficient to obtain representative information
 on the change of quality with depth within the sand bed.
       •
     The two seepage areas which occur along the bottom  of the hill on the south
 bank of West Brook were probably originally one large seepage area that has
 been split by the construction of the road bed of Gage Road.  A small
 stream represents the total seepage which occurs above Gage Road; below Gage
 Road the seepage occurs over a wider area.  Flow of this large seepage area
 was channeled into a common ditch by connecting numerous areas of seepage
 into one main stream so that the seepage flow could be monitored before entering
 West Brook.

    In order to measure the flow of the two seepage streams, a small dam was
 constructed in each  stream, and a Stevens Type F water level recorder was
 installed. Cross-sectioned flow measurements were made a  number of times
 for each stream using a Gurley Pigmy meter.  These data were used to construct
 stage rating curves  for each stream which were then correlated  with the  stage
 of the Stevens water level recorder.  Thus , a continuous  record of flow was
 provided with flow-depth charts from the Stevens recorders.  These charts
 were changed on  a weekly basis.

    Tracer studies using both tritium and rhodamine dyes were conducted
 to follow both the vertical and horizontal movement of sewage plant effluent
 discharged onto the  sand beds. An initial study using rhodamine B dye was con-
 ducted without success.  Permission was then obtained from the NYSDEC Bureau
 of Radiation to conduct radio tracer studies.  In the tracer studies, both tritium
 (radioactive hydrogen-3) and rhodamine WT  (as 20% solution) were used as
 tracers.  Three 0.1 curie (Ci) portions of tritium were used, one  for each
 of three separate tracer studies. In order to differentiate between the three
 tritium tracer studies,  a tertiary tracer was added to the second and third
 studies.  In the second tritium tracer study, 108 kg  (240 Ib) of sodium chloride
 were dissolved in a 379 1 (100 gal) tank and introduced into North Sand Bed
 4 along with the tritium and rhodamine WT tracers.  The amount of sodium
 chloride added was  calculated to double the normal chloride content of the
 sewage.  In the third tritium tracer study, 22.7 kg (50 Ibs)  of KC1 were added
 as described in the second tracer study.  This amount of KC1  was designed
 to double the concentration of potassium normally found in the applied liquid.

    Special sampling equipment designed specifically for  this project included
 a bailer, a vacuum system, and a scooper.  The bailers were custom-made
 to fit the wells. They consisted of copper pipe approximately 1.52 to 1.83
 m (5-6 ft) long with a 1.91 cm (3/4 in) diameter.  A plastic  check valve was
 installed at the bottom of the pipe,  and holes were drilled at the top in order
 to attach a nylon cord.  The cord was approximately 24 m (80 ft) long ,thus
 allowing the bailer to reach the bottom of the deepest wells.  The  vacuum system
 consisted of a 1000 ml Erlenmeyer vacuum flask  connected to either a hand
vacuum pump or an upright vacuum-pressure pump.  The scooper was made
 of an  aluminum pole approximately 1.91 cm  (3/4  in)  in diameter,  1.52 to 1.83 m
 (5-6 ft) long, with an aluminum cup  at the bottom fastened with two screws.
 This scooper was used to collect samples of treatment plant influent.
                                     18

-------
SPLASH
                                          /DRIVE
             SAND  BED 11
LYSIMETERS-v
          O
                                     //
                SHALLOW PUMPING^
               WELL
            V
                                                          WELL POINTS
                                                          (blow-up)
DEEP PUMPING
WELL
                                                 OLD WELL
                                                 POINT-
                                                                   D
                                                                   G
scale
i—>—•-
                                                  20 ft
     Figure  9.   Plan of Bed N-ll Showing the Location  of the Driven Well Points,
                 the Lysimeters and  the Shallow and Deep  Pumping Wells.
                                         19

-------
               SARD  BED
          will     Points
(eel
11   PR 0 Fl LE
     Lysimelers
1975
Pumped walls
1
0
10"

20

30
40
50

80

70
80
90







\ B l







! I







l







I







i







(INK M R B


























unsaturated Zone



- -







Saturated



































,K 1














S 1














Zone




Bed Rock
B
















0

4

8

12

16

20

24
28
                                      meters
 Figure 10.  Profile of Bed N-ll Showing the Depth of the Driven Well Points,
            the Operational Lysimeters and the Shallow and Deep Pumped Wells.
                                   20

-------
    The water level in each well was measured using an electrical water-sensing
depth gauge.  This device was also used to find the elevation of the bottom of the
well points by lowering the  sensing probe into the well until it reached bottom.
Mean sea level elevations for tops of well pipes, ground surfaces, water levels in
wells,  and bottoms of wells were established.
                                       21

-------
                                 SECTION 5

                        EXPERIMENTAL PROCEDURES

SAMPLING PROCEDURES

    Intense efforts were made to secure representative water quality samples
of both ground waters and surface waters of the Lake George Treatment System.
Samples were collected from the inflow of raw sewage until it emerged from
underground and entered into West Brook.  Monitoring was conducted through
all seasons including two harsh New England winters when snows at the site
exceeded depths of 0.6 m (2 ft), and temperatures were well below -10°C for
several weeks . Monitoring included 13 well locations,  some of which were
well clusters from which  samples were collected from as many as four depths.
Other sampling points included two seepage areas (one upstream and one downstream
of Gage Road) , West Brook  (both upstream and downstream of the two seepage
areas), the raw influent to the plant,  and effluent discharged to various sand
beds.  The complete monitoring network included 28 separate water quality monitoring
locations.

    Ground water samples were obtained from both the observation  wells and
the pumping wells . In the observation wells, samples were  collected using either
a bailer or a vacuum pump depending upon the  depth to water in the well.  The
volume of sample  collected with the bailer at each dipping was approximately
500 ml.  The contents of the bailer were emptied into labelled plastic containers
which were first rinsed with the sample.  The dipping was repeated as often
as necessary to secure a sufficient sample volume.

    The vacuum pump was used for the shallower wells,  especially wells 2A and
B and 3A, B, C, and D. A plastic sampling tube  was lowered into the well, and a
vacuum was applied to the attached suction flask.  After approximately 200 ml of
sample was collected,  the flask was rinsed, and a sample was collected for analysis.

    Surface samples were collected at West Brook, the seepages, and the sewage
influent and effluent.  At West Brook and the seepages, the plastic sample containers
were first rinsed  with the sample and then simply filled by dipping the container
into the sample source. The sewage influent was collected using the scooper
described earlier.

    A pressure-vacuum pump was utilized for securing samples from the suction
lysimeters.  The lysimeters were designed with two tubes (Fig. 8); one for retriev-
ing the sample and the other to apply a pressure to force the sample from the porous
cup or to apply a  vacuum to enable the porous cup to become filled. To maintain a
vacuum, a C-clamp was attached to the exposed end of each tube. To secure a sample
the C-clamps were removed from both tubes, and pressure was  applied to the pressure-
vacuum tubing, forcing the sample through the collection tube.  After removing the
                                      22

-------
sample, the lysimeter was prepared for sample collection by placing the C-clamp on
the delivery tube and drawing a vacuum on the pressure-vacuum tube.  A C-clamp
was then placed on the vacuum tube before the vacuum pump was detached leaving
a negative pressure in the lysimeter.  Varying amounts of sample were secured
from the lysimeters depending on the time between sampling, the ground conditions,
and the water content of the soil system, which in turn was a function of bed dosing.
Volumes of samples varied from a few ml to as much as 600 ml.  Due to the scarcity
of samples, the sampling containers were not rinsed prior to filling .  Special pre-
cautions were taken to prevent contamination of sample  containers for the lysimeters.

      Wells IIS and 11D were generally sampled by means of submersible pumps.
The pump was operated for 5 to 10 minutes  prior to sample collection to assure
that a fresh sample was being collected from the aquifer. During the period from
July 25, 1975, until May 25, 1976, while the submersible pump on well US was
inoperable, samples were collected using a bailer.


      Plastic containers (1000 ml) were used to collect  samples.  Three containers
were filled for each sample.   Two containers of sample were preserved with 40 mg
HgCl-/! for nutrient analysis as  described in Standard  Methods  (14) .  Nutrient
analysis was conducted at the Fresh Water  Institute at Lake George.  Analyses were
performed for orthophosphate, total phosphorus, nitrate nitrogen, ammonia nitrogen,
and total Kj eldahl nitrogen.  The third sample was stored at 4 C and taken to the
Rensselaer Polytechnic Institute, North Hall Laboratory, for the determination of
alkalinity, chloride, biochemical oxygen demand (BOD), calcium, metals  and other
constitutents. All samples were prepared  and run as quickly as possible  after
arrival at the North Hall Laboratory.  Separate sterile containers were used to collect
bacteriological  samples.  These samples were collected simultaneously with all
other samples.

      Special difficulties were encountered when sampling during subfreezing
temperatures.  A temperature of -33°C (-27°F) was recorded during one sampling
trip. When temperatures became this low,  problems  arose with the sampling equip-
ment and the sample. Frequently the water samples froze as soon as  they came in
contact with the ambient air.  Ice clogged the sampling  tubes on the vacuum sampler
several times .  Another problem encountered was freezing of the observation wells.
It was learned later in the project that these wells had less tendency to freeze if the
cap was left off the top allowing  the water vapor  generated from the aquifer to escape,
thereby keeping the well ice  free above the frost line.  Heavy snowfall created a
problem of access to some of  the wells such as 2, 6,  and 10, resulting in sparse data
for these wells during winter monitoring.

ANALYTICAL PROCEDURES

Field Analyses

     Tests for unstable parameters, such as temperature and dissolved oxygen
 (DO), were performed in the field. Other field measurements include dissolved
 solids, redox potential, and  pH. Temperature and DO  were measured using a
 Yellow Springs Model 54 Oxygen Meter with a probe  on a 30 m (100 ft) cable.
 This portable instrument allowed ill situ measurements in all wells and surface
 samples. The  probe was maintained and operated according to the manufacturer's
 instructions.
                                      23

-------
     A TRI-R dissolved solids meter reading in parts per million was used to
 determine dissolved solids concentration. The sample cell was rinsed with sample
 prior to measurement.

     A Leeds and Northrup No. 7417 portable pH meter was used to determine
 pH and oxidation-reduction (redox) potential in the field.  A platinum electrode
 No. 117225 was required for redox measurements.  Due  to technical problems,
 this meter was only used near the end of the sampling schedule.

     Ground water elevations of the sample wells were determined  using a portable
 electric depth gauge (Soil Test, Inc. Model DR-760A) .  This instrument consists
 of an ammeter, which records the closed circuit of two electrodes  on a sounding
 cable lowered down the well bore when contact is made with the water surface.
 The cable is calibrated in 1.5 m (5 ft) increments, and a steel tape was used
 to obtain accurate measurements between the increments and the top of the  well
 pipe.

     The frequency of sampling was a function of the sample locations, the specific
 study being conducted at the time, and the season.  In general, samples were
 collected from the observation wells biweekly, with  half of the locations being
 sampled on  alternate weeks. Samples from the lysimeters  were usually collected
 weekly, but during some periods, samples were collected  three times per week so
 that enough sample volume could be accumulated for analysis.  Tracer study
 samples were  collected daily in observation wells surrounding sand beds where
 tracers were introduced.  In the tracer studies conducted in bed Nil,  the submersi-
 ble pumps were  operated continuously supplying samples to a Turner Model
 111 Fluorometer  equipped with a flow-thru sampling door and a continuous recorder.
 Stream flow measurements were also monitored continuously with Stevens Type
 F water level recorders.  Flows of the treatment plant influent were monitored
 by means  of the continuous flow recorders which are part of the treatment plant.

 Laboratory  Analyses

    The potentiometric method described in Standard Methods (14) was used
 to obtain accurate readings of alkalinity.  A Fisher Accumet Model 320 Expanded
 Scale Research pH Meter was used to determine the end points.

    The Technicon Auto Analyzer II was used to determine all forms of nitrogen
 and phosphorus  due to the large number  of samples  analyzed.  The component
 parts of the system included an Auto Analyzer II, sampler, pump,  manifold,
 S. C. Colorimeter, and Recorder. Technicon Methods Sheet No. 155-71W (Nov.
 1971) was followed for determinations of soluble reactive phosphorus  after filtering
 through a 0.45 um membrane filter. In analysis for total soluble phosphorus,
 the samples were first filtered and the persulfate digestion procedure described
 in Standard Methods  (14) was used to convert the total phosphorus to soluble
 reactive phosphorus  and then analyzed as above.

    Nitrate nitrogen and ammonia nitrogen concentrations were determined as
 outlined in Technicon Industrial Methods Sheet Nos. 158/71W/preliminary (Dec.
 1972) and 154-71W/tentative (Feb.  1973), respectively.  Kjeldahl nitrogen samples
 were first digested to convert the various organic forms  to ammonia as in Standard
Methods (14),  and then the procedure for ammonia nitrogen was followed.

                                     24

-------
    The membrane filter procedure outlined in Standard Methods (14) was used for
determining fecal coliform concentrations.  All sample bottles were sterilized in a Dis-
patch Oven, Style 287, Type 3-H hot air sterilizing oven at 170°C for a minimum of
two hours.  Graduated cylinders and medium containers were sterlized in a Wilmot
Castle Co.  #19465 autoclave for 15 minutes at 120°C.  Presterilized disposable
60x15 mm plastic petri dishes were used.  All pipets used were also the presteri-
lized disposable type.  A Gelman Manifold #4201 was used to hold three Gelman
Membrane Filter Funnels.  These funnels were also autoclaved at 120°C for 15 min-
utes for sterilization. Gelman presterilized 47 mm diameter 0.45 Jim gridded filters
with pads were used.  The medium was BBL-M-F  broth. All analyses were carried
out shortly after arrival at the North Hall Laboratory.  After sample preparation and
filtration, the petri dishes were placed in weighted plastic bags.  These plastic bags
were then immersed in a Blue M Magni-Whirl Constant Temperature Bath at 44.5 ±
0.2°C for 24 hours.  The colonies were counted using a Quebec Colony Counter
(Spencer Lens Company) .

    Total coliform analyses were conducted according to procedures outlined in
Standard Methods (14) which was basically the same as that for fecal coliforms. The
single step direct technic was used.  The medium was BBL-M-endo broth. The
plates were incubated at 35 +_ 0. 5 C in Eimer & Amend Type II incubator for 24 hours.
High humidity was maintained by placing an open beaker of distilled water in the
oven.  The colonies were counted with a lOx binocular microscope.

    Difco M enterococcus agar was used as a selective medium for recovering
fecal streptococci species. Two milliliters of sample were carefully transferred
using  sterile disposable pipets into 15x100 mm glass petri  dishes which had been
sterilized by autoclaving for 15 minutes  at 120°C. Next, 5 ml of agar were trans-
ferred into the dish, again using a sterile pipet.  The sample and agar were well
mixed and  allowed to harden.  The plates were then inverted and incubated in an
Eimer & Amend Type II incubator at 35 + 0.5°C for two days .  The streptococci
species appeared as pink-to-red circular colonies, and they were counted
using  a Quebec Colony Counter (Spencer Lens Company) .

     The double layer agar technique of Hershey et al (15) was used for  the deter^
mination of viral concentrations.  Five percent chloroform was added to  50 ml of
sample, and the mixture was stirred thoroughly for 30 minutes.  Air was then
bubbled through the mixture to remove the chloroform.  The samples were then
assayed according tn Hershey et al (15) .  Pure culture of E. coli B was supplied
by the Biology Department of RPI.  The top agar consisted of 10 g Difco Bacto-
Tryptone, 6 g Difco Bacto-Agar, 8 g NaCl, 1 g glucose, 1.5 ml of 1 N NaOH, and
 2  ml of 1 N CaCl? per liter.  The bottom agar contained the same ingredients,
but the Bacto-Agar was increased  to 10  g/1. All media and glassware were steri-
lized at 120°C for 15 minutes.  After preparation, the glass plates were inverted
and incubated at 35  + 0.5°C for approximately 8-10 hours. Plaques were then
 counted with  a Quebec Colony Counter (Spencer Lens Company) .

     Five day  BOD's were run as detailed in Standards Methods (14) . The only
 samples for which dilution was necessary were the treatment plant influent and ef-
 fluent.  The Azide Modification of  the Winkler Method (14) was used for determining
 dissolved oxygen concentrations.  The samples were incubated at 20 +_ 1.0  C in a
 Hythermco thermostatically  controlled incubator. In  addition  to the normal sampling
                                      25

-------
 sites,  a series of samples was taken of the supernatant liquid remaining on a
 sand bed before it was infiltrated into the ground.  These analyses were done
 to determine the BOD- reduction due to natural aeration while the effluent remained
 on the bed prior to inriltration.

     The chloride ion concentration was determined by the Argentometric method
 as described in Standard Methods  (14) and by the  specific ion electrode method.
 The latter was quicker and gave results consistent with the first method.  A Beckman
 Century S S pH meter was used with an Orion Analyzer double junction electrode,
 Model  90-02.  The inner chamber was filled with standard Orion 90-00-02 filling
 solution.  The outer chamber contained 4 M NHJSIO-.  Orion Research specific
 ion electrode model 94-17A was used as the chloride electrode.  Before  each set
 of analysis was run, the meter was calibrated to a  standard curve.

     Aluminum concentrations were determined with the use of a Perkin-Elmer
 403 Atomic Absorption Spectrophotometer equipped with a nitrous oxide burner
 while iron concentrations were determined  according to the tripyridine method,
 outlined in Standard Methods (14) . The tripyridine method is subject to interferences,
 as is the phenanthroline method; however,  there is no interference from phos-
 phates and heavy metals in concentrations which might be expected in the samples
 tested.

     The anionic surfactants (linear alkylate sulfonate, LAS) concentration was
 determined according to a recently developed, indirect two-phase titration method
 (16).  The method uses cetyldimethylbenzyl ammonium chloride (CDBAC) in
 excess to form a complex with the water-soluble anionic surfactant, and uses
 methyl orange (MO) as an indicator in the presence of chloroform.  The color
 of the CDBAC-MO complex in the chloroform phase  is yellow.  This water-chloroform
 two-phase mixture is then titrated  with sodium tetraphenylboron (STPB) reagent
 with intermittent shaking to ensure equilibrium between the chloroform and the
 aqueous phases.  The disappearance of the  yellow  color in the bottom chloroform
 layer indicates the endpoint of the  titration.  The actual concentration of anionic
 surfactant, such as LAS, is determined by the difference between the amount
 of the CDBAC  added to the sample and the amount of STPB needed for titration.

     Calcium, magnesium, copper,  lead, and some iron analyses were performed
 using a Beckman Atomic Adsorption System utilizing a laminar flow burner with
 an acetylene-air flame and using standard lamps with  detection DB-G Spectro-
 meter.  Sodium and potassium analyses were run using the  same instrument
 osram  lamps.

    Total carbon concentrations using 20fil  samples were analyzed with the use
 of the Beckman Infrared Analyzer according to the  combustion-infrared method
 described in Standard Methods (14) .

    Rhodamine WT was determined using a Turner Fluorometer Model 111 equipped
 with a  Model #110-851UV lamp, a #110-82 primary filter, and a #110-833 secondary
filter.  During continuous field monitoring the fluorometer was equipped with a flow-
thru sampling door and a continuous recorder (Gulton Industries Model 288) .  For
individual rhodamine measurements, the flow-thru door was replaced with a standard
cell holder door.
                                     26

-------
      Tritium samples were collected at the site and returned to North Hall for
analysis during initial tracer studies . In preparation for analysis, 7 ml of sample
were measured into 25 ml polyethylene counting vials equipped with screw caps.
After the samples were placed in the vials, 15 ml of Aquasol-2 was added  to each
vial. Aquasol-2 is a liquid scintillation cocktail prepared by the New England Nuclear
Company and is used as a solvent to keep isotopes in solution while counting.
A mixture of 15 ml of cocktail with 7 ml of sample was the optimum ratio recommended
for Packard Counting Instruments by company  representatives. After adding the
cocktail, the caps were replaced tightly  and the vials were gently shaken to mix
the contents.   A background sample was prepared using 7 ml of distilled water
and 15 ml of Aquasol-2. The samples were then placed in the refrigerator of the
scintillation counter described below and three hours were allowed to elapse prior
to  counting to  permit the decay of chemiluminescence and fluorescence. The samples
were then counted for 20 minutes each.  To obtain the net count per minute (cpm)
for each sample, the background gross count was substracted from each sample
gross count, and this net count divided by 20.

      The Packard Model 3003 Tri-Carb Scintillation Spectrophotrometer  used
to  count samples during initial radiotracer studies is a refrigerated unit with three
channels to allow comparison of counts.  The machine is a single sample manual
unit which required considerable time for sample handling and analysis.  This
instrument became inoperable on July 31, 1976.  All subsequent samples were
taken to the RPI Freshwater Institute Laboratory  at Lake George and counted on
a Beckman LS-133 Liquid Scintillation System. Although the Beckman machine
is  not a refrigerated unit, it has the advantage of a 100  sample automatic changer
plus a digital printout.  A tritium standard was prepared to determine the counting
efficiency of each liquid scintillation counter.  This standard indicated the counting
efficiency of the Packard Tri-Carb and Beckman  units was 39.4 and 56.4 percent,
respectively.

      The question arose at the beginning of the project as to whether the samples
should be distilled prior to tritium counting.  A vacuum distillation system was
assembled according to the technique devised by the Radiological Science Laboratory,
Division of Laboratories and Research, New York State Department of Health (17) .
This system was designed to separate tritium from other interferring radionuclides
and to remove chemical and/or physical quenching agents. Several samples were
counted with and without prior distillation.  Comparative results showed  very little
difference in counts between distilled and undistilled samples; therefore, the dis-
tillation procedure was abandoned.
                                      27

-------
                                 SECTION 6

                         RESULTS AND DISCUSSION

    The study of secondary sewage effluent applications onto sand beds construct-
ed in a natural glacial delta at the Lake George Village Treatment Plant included
evaluating both the renovation of wastewater moving through the natural delta
sand and the hydraulic characteristics of the sand system.  The changes in water
quality  are presented as those which occurred during vertical flow and those
during horizontal flow through the saturated aquifer.

    Massive amounts of data have been collected to determine the changes in
quality  of wastewater moving through the soil.  In order to condense the data,
seasonal averages  were made since the data reflect changes in the temperature
as well  as the flow regimes within the treatment system. Flows to the treatment
plants were significantly influenced by the seasonal tourist  influx in the Lake
George Area.  This major influx  occurred from about the middle of June through
the middle of September with the peak occurring during July and August. Thus,
the summer season represents the greatest sewage loads to the treatment plant.
For all monthly data reported herein,  the month begins with the 21st of the previous
month and continues  through the 20th day of the stated month.  The data are
reported in this way  in order that the monthly averages can be  converted to seasonal
averages without additional significant calculation.

CHANGE IN QUALITY THROUGH THE SOIL

Changes with Vertical Transport

    All  of the studies of the changes in quality of wastewater during vertical
transport  through  the unsaturated portion of the sand beds were conducted in
North sand bed 11. A series of well points was driven at 0.6m  (2 ft) spacings
to a maximum depth of 4.3 m (14  ft) as shown in Figure 9.  The well screens
on these wells were located in the unsaturated zone; therefore, samples could
only be  retrieved from the shallowest well in the series during  a flooding period.
Only 4 of the 13 lysimeters installed in the unsaturated zone were operable; therefore,
all samples retrieved through the unsaturated profile were collected from these
lysimeters. Wells IIS and 11D penetrated the saturated  aquifer  and were used
to monitor the quality of the wastewater after mixing with the native ground water.

    The average seasonal temperature fluctuations with depth for the year start-
ing in September 1975 through August 1976 are summarized in Figure 11.  The
temperature of the  sewage and applied effluent, as expected, followed the same
trends as the ambient air temperatures.  The temperatures near the bottom of
the unsaturated zone were somewhat more stable, and the temperatures  within the
saturated aquifer indicated only slight seasonal variations .  Thus, during the cooling
                                     28

-------
   Inf
    0--
   10-

Depth,
meters
    15
   20
    25-
   30
            T
                                                       Fall
                                                       Winter
                                                       Spring
                                                       Summer
                A   fb
                             15          20~
                         Temperature,  °C
          5           10


Figure 11.  Seasonal  Variations  in Temperature with Depth in Bed N-ll
                             29

-------
 seasons of fall and winter, the temperature increased with depth indicating heat
 stored in the soil profile warmed the infiltrating wastewater.  During the warming
 seasons temperature decreased with depth as the warmer infiltrating wastewater
 became cooled by soil which had been cooled by water which moved through
 the profile the previous season.  The greatest temperature fluctuation occurred
 at the 3 m (10 ft) depth which is interpreted  as effects of ambient temperature
 on the percolating effluent.  Temperature alone is not suggested to be a pollutant
 in this system;  however, the temperature has been shown to have  a dramatic
 effect on biological systems including the nitrification-denitrification processes.
 Thus, warmer temperature increasing biological metabolism should enhance
 purification of sewage (18, 19) .

     The seasonal variation of pH with depth  is  shown in Figure 12. The lowest
 average value  determined was 6.5 at the 11 m (36 ft) depth in the fall, and the
 highest was  7.4 during the summer in the upper part of the saturated zone.
 The difference in these values indicates there is not a significant change in pH
 with depth or season.  Although the spring and summer curves were similar
 below the 11 m  (36 ft)  depth, all seasons except the fall indicated a definite decrease
 in pH  in the  top 3m (9.8 ft) of the sand bed which could be attributed to the
 production of carbon dioxide produced by biological activity.  However, this
 explanation is  not well supported since there was an increase in pH at the 7 m
 (23 ft) depths in the spring and summer and  at the 11 m (23 ft) depth in winter.
 It should be noted that pH values tended to be higher in the saturated aquifer
 than in the unsaturated zone of the sand beds.

     There was a consistent slight increase in the dissolved solids  concentration
 with depth during all seasons in the unsaturated zone of the aquifer (Figure
 13) . The highest values were observed to occur during the summer and the lowest
 during the winter with essentially parallel gradients.  Values within the saturated
 aquifer were consistent throughout the year and were much lower  than the values
 in the unsaturated zone suggesting a considerable dilution of applied sewage
 effluent high in dissolved solids by the natural ground water which is normally
 low in dissolved solids.

     The dissolved oxygen (DO) content of the percolating effluent  is important
 to any biological process which may  be occurring in the sand beds.  Maintenance
 of aerobic conditions will enhance the oxidation of reduced nitrogen compounds
 to nitrate and generally stabilize other contaminants.  The DO concentration is
 a function of temperature with higher saturation values  occurring in colder water.
 This trend is illustrated by the generally lower values observed during summer
 and higher values during winter in the unconfined aquifer (Figure 14); however,
 the highest value observed was in the saturated zone during the summer.  As
 stated earlier,  the temperature in the saturated zone remained relatively constant
 throughout the year.  Thus,  the DO levels would  also be expected  to have  little
 variation.  There was  slightly more variation in DO than anticipated : however,
 slight  aeration  of the sample could have occurred due to the sampling technique
 used.  During the fall, winter and spring well IIS was monitored for DO by in-
 serting the DO  probe directly into the well.  During the summer when the  submers-
ible pump was out of service, the DO was measured in samples pumped to the
surface.  Data show that DO values measured in_ situ were  lower than values
measured from  the pumped samples,  indicating that all the DO values reported
                                     30

-------
                                    •    D
Depth,
meters
          - -o Fall
          - -• Winter
               Spring
               Summer
     30
                                                       7.3
     Figure 12,  Seasonal Variations in pH with  Depth  in Bed N-ll,
                                31

-------
Depth,
meters
                                                 Fall
                                                 Winter
                                                 Spring
                                                 Summer
     50


 Figure  13.
  100
  150       200       250
Dissolved  Solids, mg/l
Seasonal Variations in Total Dissolved Solids with  Depth in
Bed N-ll.
                                  32

-------
        Inf-r
     Depth,
     meters
              - - -o Fall
              - - •• Winter
                    Spring
                    Summer
         25
         30
                                                         i       I
           13579
                             Dissolved Oxygen, mg/l


Figure 14.  Seasonal Variations in  Dissolved Oxygen with Depth  in Bed N-ll.
                                     33

-------
 in the wells within the sand bed, with the exception of the fall, winter, and spring
 samples in well IIS, may be slightly high due to aeration during sampling. The
 DO concentration of the water standing on the surface of the sand beds was defi-
 nitely influenced by the  ambient temperatures.  The lowest DO values correlate
 with the highest temperatures in the summer; while the highest values correlate
 with the lowest temperature during winter.  In general,  the DO decreased at
 the 18 m (60 ft) depth  which may be correlated with the denitrification observed
 immediately above the saturated zone.

       The redox potential (Figure 15) was determined only during the spring
 of 1976 after the sampling instruments were received.  During treatment of sewage
 in the treatment plant, the redox potential was greatly increased. The redox
 potential in the sand profiles, as illustrated by data from bed  Nil, remained
 positive at all but the 18 m (60 ft) depth at which point it became slightly negative.
 This negative value correlates with the consistently low DO values at the 18 m
 (60 ft) depth described above.

       There was a  significant reduction in the chloride content of the wastewater
 in the conventional portion of the sewage  treatment plant, as shown in Figure
 16. Through the bed profile, the chloride concentration was essentially constant
 with depth ranging between extremes of 40 and 80 mg/1 in the unsaturated zone
 for the summer and spring, respectively. In the upper saturated zone of the
 aquifer seasonal variation ranged from about 40 to 60 mg/1, while values  near
 the bottom of the aquifer as shown on Figure 16 indicate low concentrations in
 the summer and higher concentrations during fall. However,  careful examination
 of data shows  that the high average concentration  during the fall was the result
 of a few exceptionally high values during November while other values observed
 during this season were  quite low, even lower than values at the  US well depth.
 These chloride concentration spikes  near  the bottom of the aquifer during fall
 cannot be readily explained.  The highest values of chloride in the influent raw\
 sewage occurred during winter. These high concentrations correlate directly
 with the time of salting of highways and subsequent infiltration into the sewage
 system.  The results indicated there was  no removal of chlorides in the sand
 during any  season and that chloride  could possibly have been used as a tracer
 of applied sewage effluent in a normal sand infiltration system. Unfortunately,
 the storage of road deicing salt on the ground in the area of the town garage
 (Figure 6) prohibited the use of chloride  as a tracer in the land application system
 under study.

      Even though there were no consistent trends of changes in calcium versus
 depth throughout the four seasons, as shown in Figure 17, some general interpreta-
 tions may be considered.  Summer seasonal averages versus depth indicated
 a general increase with depth in the unsaturated zone followed by a sharp reduction
in the saturated aquifer.   A possible interpretation of these data could be that
calcium is leached from the sand profile by the warm percolate during the summer
season then diluted by the lower concentrated ground water in the saturated
 zone.  During  winter calcium adsorption to the sand particles  is indicated by
the very general decrease in concentration of calcium versus depth. It was not
possible to correlate the calcium, pH, alkalinity, and orthophosphate concentrations
within the sand beds;  therefore, calcium should not be considered as a limiting
element in the operation of the infiltration  system.
                                      34

-------
          Inf
            i
           10

      Depth,
      meters
           15-
           20-
           25
           30
           •100
                    Spring
 -50           0          50

Oxidation • Reduction  Potential , mv
100
Figure 15.   Variation of Oxidation-Reduction Potential with  Depth in Bed N-ll
             During Spring 1976.
                                        35

-------
        Inf.
         0
          5i
         10


      Depth,
      meters


         15-
         20
         25-
         30
OD
                                   \
                                     X)
     - —o Fall
     	•• Winter
     	D Spring
         • Summer
                       40        _ 80          120          160
                                 Chloride, mg/l


Figure 16.  Seasonal Variations of Chloride with Depth in Bed N-ll
                                   36

-------
     Inf
      10-

  Depth,
  meters
     20-
      25-
     50
orm    •
                                 \
                             -o Fall
                             -• Winter
                                Spring
                                Summer
                       16             24
                           Calcium, ppm
                                     32
Figure 17.   Seasonal  Variations in Calcium with Depth in Bed N-ll
                               37

-------
       There was very little significant variation in the magnesium concentration
 with depth, as shown in Figure 18.  The lowest and highest values were approxi-
 mately 4 and 8 mg/1, respectively.  There was a slight trend toward increasing
 concentrations with depth for all seasons except winter.  The interpretation that
 magnesium in the sand may be leaching through during warmer seasons and
 stored during winter might be considered.  The natural ground water was low
 in magnesium and diluted the slightly higher concentrations infiltrated from above.

       Observation of Figure 19 indicates there was a consistent reduction in
 alkalinity of the sewage in transit through the treatment plant during all seasons.
 However, as wastewater infiltrated through the sand beds, there was a significant
 increase in alkalinity in the unsaturated zone.   As the percolate entered the natural
 ground water, alkalinity, as most other parameters,  was reduced by dilution,
 reaching concentrations lower than those of the influent.  The alkalinity measure-
 ments correlated very closely with pH and dissolved oxygen data.  As the alka-
 linity  increased, the pH and DO decreased indicating the presence of biological
 activity.  Microorganisms in the  sand profile utilized the DO in their biochemical
 mechanisms, thereby decreasing the DO.  During their respiration they  gave
 off carbon dioxide which reacted with the water, producing carbonates and hydrogen
 ions.  The increase in carbonate caused an increase in alkalinity and the increase
 in hydrogen ions decreased the pH.   Furthermore,  the alkalinity increases were
 greatest near the surface during the summer, indicating that increased tempera-
 ture encouraged biological activity.   The biological activity during the other
 seasons was less near the colder surface, but greater at the warmer 18 m (60
 ft)  depth.

       The interrelationship between organic, ammonia, and nitrate nitrogen is
 discussed by comparing the changes from one form to the other as the percolate
 moved through the soil profile  during the summer, fall, and spring seasons. In
 order  to compare these changes,  all  three forms measured (nitrate, ammonia,  and
 total Kjeldahl nitrogen) are shown in one figure for each season. Insufficient data
 were obtained during the winter to accurately support interpretations of changes
 for this season, therefore these data are not presented. The data for summer  and
 fall, Figures 20 and 21, respectively, showed similar trends.  In both seasons
 there was a decrease in ammonia and Kjeldahl nitrogen with a corresponding
 increase in the nitrate nitrogen at the 3m (10 ft) depth.  At slightly greater depths,
 there was a reduction in nitrate with a significant increase in the ammonia and
 Kjeldahl nitrogen, followed by reduction in both these parameters at the  18 m (60ft)
 depth. In both seasons, the nitrate content at the 18 m (60 ft) depth was less than
 1 mg/1; thus, there was apparently some loss of total nitrogen from the infiltrating
 water in the form of nitrogen which escaped to the atmosphere.   The only nitrogen
 form of significant concentration found in the pumped wells was  a relatively high
 nitrate concentration which occurred in well  IIS during the summer.  During
 the spring  the results (Figure  22) showed somewhat different trends.  There
 was a substantial increase in the  nitrate concentrations in the upper 3m  (10ft)
 of the sand beds with a subsequent gradual reduction with depth to less than
 1 mg/1 at the 18 m (60 ft)  depth.  At the 3m (10 ft) depth the ammonia and total
Kjeldahl nitrogen were reduced to less than 1 mg/1 followed by  a uniform concentra-
tion with depth down to the top of the saturated zone.  The results of the spring
 data indicated that there was an initial oxidation of the ammonia and total Kjeldahl
nitrogen to nitrate with a possible subsequent reduction of the nitrate directly
to nitrogen gas. It is also interesting to compare the nitrate content during the

                                      38

-------
   Inf.
     0-
Depth,
meters

    15-
   *
25-


30«



f - - -o Fall
\ ---•Winter
\ " •• Summer
[\

                         Magnesium, ppm

Figure 18.  Seasonal Variations in Magnesium with Depth in Bed N-ll,
                                 39

-------
        Inf
         OH
          5-
         icH

     Depth,
     meters
         15H
         20-
         25-
         30-
         O D
                    - --oFall
                    - - -• Winter
                    	O Spring
                          Summer
    100      150                250~
ALKALINITY,  mg/l  as  CaCOj
Figure 19.  Seasonal Variations in Alkalinity with Depth in Bed N-ll,
                                  40

-------
   Depth.
   meters
                                              Nitrate
                                              Ammonia
                                              TKN
                             8        12        16

                             Nitrogen,  mg/l
20
Figure 20.  Variations in the Forms of Nitrogen Measured with Depth in
            Bed N-ll During the Summer of 1975.
                                   41

-------
    Depth,
    meters
                                  	— • Nitrate
                                              • Ammonia
                                          —A TKN
                                    8           12
                              Nitrogen. mg/I
16
Figure 21.   Variation  of Various Forms of Nitrogen Measured with  Depth in
            Bed N-ll During the Fall 1975.
                                  42

-------
     Depth,
     meters
                                                Nitrate
                                                Ammonia
                                                TKN
                              8        12
                              Nitrogen, mg/l
16
20
Figure 22.   Variation of the  Various  Forms  of Nitrogen Measured with Depth
            in Bed N-ll  During Spring 1976.
                                   43

-------
 spring with the oxidation reduction potential measured during the same period
 as shown in Figure 15. The two curves are nearly identical in shape, with the
 possible exception of slightly higher values of redox potential in the saturated
 portion of the aquifer.  This correlation supports the theory that oxidation to
 nitrate occurred in the upper 3 m (10 ft) of the sand bed with subsequent reduc-
 tion of the nitrate  to nitrogen gas in the lower portions of the sand bed.  One
 question which these results do not resolve is the relatively high concentration
 (7 mg N/l) of nitrate in the seepage adjacent to West Brook. Since the total nitrogen
 at the 18 m (60 ft)  depth in bed N-ll was usually less than 7 mg/1, it is  obvious
 that the nitrate in  the seepage did not originate from bed N-ll. However, several
 of the sand beds, particularly the newer south beds, are much less than 20 m
 (65 ft) deep.  A bed with only 5 m (15 ft) of sand could contribute considerably
 more total nitrogen to the saturated ground water zone than deeper beds.  This
 may be the source of high nitrate water found at some locations in the upper
 portion of the aquifer and in the seepage adjacent to West Brook.

    The two forms of phosphorus studied were orthophosphate and total phos-
 phate. Total phosphate includes both ortho and the polyphosphate forms which
 generally are hydrolyzed to orthophosphate with time. Biological systems require
 phosphorus as a nutrient along with nitrogen.  The ortho and total phosphate
 phosphorus are compared for the  summer, fall, and spring in Figures 23,  24,
 and 25, respectively. Insufficient data were collected during the winter to accurately
 evaluate winter treatment; therefore,  the winter results are not presented.  In
 all cases the orthophosphate was reduced to less than 0.1 mg P/l by the time
 the effluent reached the 10 m (35 ft) depth.  The total phosphorus was also reduced
 significantly in the top 3 m (10 ft) in the summer and fall but was followed by
 an increase in concentration at the 8 m (25 ft) depth.  During the spring the con-
 centration of total phosphorus was very similar to that of orthophosphate.  Slight
 amounts of total phosphorus were observed in the shallow pumped well during
 the summer and fall, but during the spring in the deeper pumped well,  the levels
 were consistently  less than 0.1 mg/1. During the spring, a significant reduction
 of orthophosphate  occurred in the wastewater while in transit through the sewage
 treatment plant, but during  the fall, there was a slight increase in orthophosphate,
 possibly indicating a conversion of polyphosphate to orthophosphate in the treatment
 system.  In all instances, by the time the percolating wastewater reached the
 10 m  (35 ft)  depth, the orthophosphate was reduced to levels lower than could
 be achieved by conventional physical-chemical treatment methods for phosphate
 removal.

    There are many factors  that affect soluble iron concentration, including pH
 and biological activity. High concentrations  of iron in ground water indicate
 anaerobic conditions, since  in the presence of dissolved  oxygen,  iron will be
 precipitated and remain insoluble.  Therefore,  a high soluble iron content is
 indicative of low pH, high carbon dioxide concentrations, and high bacterial
 activity.   In addition, increasing  iron concentrations would be expected to increase
 alkalinity because the reaction converting iron  to a soluble reduced form also
 increases bicarbonate as shown in the following reaction:

                     Fe C03 + C02  + H20   <*•  Fe+2  +  2HCO '

Data collected for iron analysis versus depth were averaged for the various  seasons
 and presented in Figure 26. Low concentrations of iron were found in the raw

                                     44

-------
   Depth,
   meters
                                                  Orthophosphate
                                                  Total Phosphate
                              1.6        2.4       3.2
                                Phosphate . mg /I as P
4.0
Figure 23.  Variation in Ortho and Total Phosphate with Depth  in  Bed  N-ll
            During Summer  1975.
                                    45

-------
     Depth.
     meters
                                                  Orthophosphate
                                                  Total  Phosphate
                  0.4     0.8     1.2     1.6      2.0      2.4     2.8
                            Phosphate , mg/l as  P

Figure  24.   Variation  in Ortho and Total Phosphate with Depth  in  Bed N-ll
             During  Fall  1975.
                                       46

-------
    Depth,
    meters
         2QJ	/	
         25
         30-
                                         Orthophosphate

                                         Total Phosphate
                    0.8       1.6       2.4       3.2       4.0
                                Phosphate. mgX)  at P
Figure 25.  Variation in Ortho and Total Phosphate with  Depth  in  Bed N-ll
            for Spring 1976.
                                    47

-------
   50
                                                Ground
                                                 Water
                                                        b
                                              — -• Winttr
                                              	D Spring
                                                  • Summer
                           Iron ,  mg/l
Figure 26.  Seasonal Variations in Iron with  Depth  in  Bed  N-ll,
                              48

-------
sewage and applied effluent.  There was a general trend toward increasing concerr-
trations of iron with depth in the unsaturated zone, with the highest increase
observed during the summer and fall. During  the spring and summer the iron
concentrations were low in the saturated aquifer; however, during the fall and
winter,  high values were observed in well IIS. In evaluating the data, it must
be pointed out that during the fall, winter, and part of the spring the pump in
well US was inoperable and samples were collected using a bailer.  The well
was cased with steel pipe, and without the use of the submersible pump, it was
impossible to collect representative samples free of rust.  Thus,  the low values
observed in all the other wells and the low values measured most other times
in well US were probably more realistic values. There was a definite correlation
between the iron concentration, redox potential, dissolved oxygen,  and alkalinity.
There did not appear to be any correlation between iron concentration and phosphate
removal. It is possible that iron was solubilized in the upper portion of the sand
beds where  it immediately precipitated the phosphate, and then any further solubili-
zation of iron remained as an increase in iron content since there was  no phosphorus
remaining to precipitate the iron.

    Specific measurements for sodium and potassium were made  in order to pro-
vide some background data for use of these elements in the tracer studies.  Sodium
and potassium, being quite soluble, will follow the movement of water through
the soil. Sodium is found in some concentration in almost all natural waters,
but potassium is generally present only in waters containing human sewage effluents;
thus, by determining the potassium to sodium ratio, the presence of sewage effluent
can be detected.  This was particularly necessary in this study since  ground water
contamination from  road deicing salts is known to occur near Lake George.
Previous studies have shown that the road deicing salt used in this  area contains
no potassium (20).  Figure 27 shows all values of sodium to be within 10 to 18
mg/1 with no significant changes with depth or season.  Lower values were found
in the saturated portion of the aquifer indicating dilution of the sewage effluent
by the native ground water.

    The potassium  concentration of the applied effluent from the sewage treat-
ment plant also varied only slightly ( 4 to 10 mg/1) as shown in Figure 28.  There
appeared to be a very slight increase in concentration with depth; however,
dilution in the saturated aquifer was again illustrated with potassium  concentrations
less than 1 mg/1.  This evaluation confirms that the source of potassium is from
municipal wastewater, and potassium would be expected to be found in low concen-
trations in uncontaminated water.  The ratio of potassium to sodium is shown
in Figure 29. It may be seen that the ratio generally exceeded  0.4  in
areas where there was contamination from municipal wastewater and was in the
range of 0.1 in the relatively uncontaminated ground water  confirming the use of
the ratio as an indicator of the presence of human waste discharges.

     Measurements  were made for copper in all of the samples; however, all
of the concentrations were less than the lowest detectable limit on the Atomic
Absorption Spectrophometer System used in the study (0.05 mg/1). With concentra-
tions this low, it was impossible to evaluate any seasonal changes with depth.
These concentrations were less than the drinking water  standard (21).
                                       49

-------
      Inf
       0
      10

  Depth
  meters
      15-
      20
     25
     30
- —o Fall
— -• Winter
	D Spring
       Summer
       0         4         8        12        16        20
                            Sodium, ppm

Figure 27.  Seasonal Variations in Sodium with Depth in Bed N-ll,
                               50

-------
        0-
        10-

    Depth
    meters


        15-
        20
        25-
	o Fall
	-• Winter
        Spring
        Summer
          0         2         4         6         8        10
                              Potassium , ppm


Figure 28.  Seasonal Variations in Potassium with  Depth  in Bed  N-ll.
                                 51

-------
         Inf
          0
         10

      Depth,
      meters


         15-
         20-
         25
         so
           0.0
                   	o Fall
                   	• Winter
                            Spring
                            Summer
0.2          0.4           0.6
    Rat ie  Potassium: Sodium
0.8
Figure 29.  Seasonal  Variations in the Ratio of Potassium to Sodium with
            Depth  in  Bed N-ll.
                                    52

-------
Changes With  Horizontal Transport

      After the applied sewage flows vertically through the sand beds, it enters
the ground water aquifer below and then flows in a northerly direction toward
the seepage at West Brook approximately 600 m (2,000 ft)  away.  The studies
of change in quality with distance within the saturated aquifer were conducted
by two different researchers. The studies from November 1974 through September
1975 were conducted by Hajas (22)  and the study was continued by Chiaro (23)
from September 1975 through August 1976.  Hajas's results include all of the
data obtained prior to the initiation of this project, including studies by the NYSDEC.

      The data collected at each sampling location were condensed for interpretation
by averaging for each of the four seasons. The averaged seasonal data are displayed
in a series of bar graphs in Figures 30 through 41.  The order of seasons in
the bar graphs in Figure 30 through 39 is spring, summer,  fall,  and winter.
These graphs  are based on data from the spring of 1976, the summers  of 1975
and  1976, the fall of 1975, and winter of 1976.  Figures 40 and 41 show the seasonal
variations in chlorides and dissolved solids for all samples  collected through
September 1975. In these figures, the  seasons are listed  in the order of fall,
winter, spring, and  summer.

      The observation wells are identified according to the numbers as indicated
on Figure 6.  The letter following the well number indicates the relative depth
of the well point into the aquifer at locations where more than one point was installed.
These letters are AA, A, B, C,  and D indicating the order of shallowest to the
deepest, respectively.  Wells 11 are pumped wells and are labeled US and  11D
to indicate shallow and deep pumped wells, respectively. The West Brook surface
flow sampling sites upstream and  downstream of the seepage streams are designated
as stations WBUS and WBDS, respectively.  The stations monitoring the seepage
above Gage Road and below Gage Road are noted as  SA and SB while the treatment
plant influent  and effluent are designated as INF and EFF, respectively.  Well
elevations, water depth, and top of rock elevations are compared in Figure 7.

      Figures 30A, 30B, and 30C  show the seasonal variations in temperature
for each sampling point. As expected, the surface waters and the shallowest
wells (3A, 4,  5, 7, 10)  showed the largest fluctuations in temperature while
the deeper wells (2, 3) were relatively stable.  The sewage plant effluent being
applied to the infiltration beds was warmer in the summer (19.3°C) and colder
in the winter (6.0°C) than the raw sewage influent (18.6°C in summer and 6.9 C
in winter), reflecting the effect of the ambient air temperature on the sewage
as it passed through the treatment plant. The temperature of the effluent  in the
winter remained warm enough to melt the ice cover on the sand beds and thaw
the soil surface, allowing infiltration to continue through the  subfreezing winter
season.  Well  9 had temperature fluctuations about one season behind  those of
surface waters and other wells. The highest temperature in well 9 occurred
in the fall (15.26°C) and the lowest in  the spring (9.54°C) . One explanation
of this phenomenon was that the lateral flow from the sand beds to well 9  is several
months duration, causing delay in temperature fluctuations between the infiltration
beds and well 9. The cluster of wells at the well 6 location also exhibited similar
temperature lay characteristics, although not as dramatic.  Later, tracer studies
discredited  this explanation.


                                      53

-------
      II
      1C
§
a:
ui
a.


ui
      14
      12
      II
                   2«     2B      3A      3B      3C      3D




                                       SAMPLE   LOCATION
         Figure 30A.  Seasonal Variations in Temperature.  The Bars Represent,  in

                      Order, Spring,  Summer, Fall and Winter.

-------
o
•

-------
5
5
                        ill
                         IS
                         1C
                         14
                         12
                         10
ui

CTi
                              121
                       Figure 30C.
                  121     S(      SI    W B U S   W BOS    INF




                                      SAMPLE  LOCATION
                                                                                    F.F F
                Seasonal  Variations in Temperature.  The Bars  Represent, in

                Order, Spring,  Summer, Fall and Winter.

-------
The influence of the seepage stream temperature on West Brook was also apparent.
West Brook, downstream of the confluence of the seepage inflows, was from 0.5 C
to 1.6 C warmer in winter than upstream because of the warmer inflowing sewage
streams.

      Dissolved Oxygen (DO) concentration in water is a function of water tem-
perature;  therefore,  the seasonal DO changes can be directly correlated with
seasonal  temperature changes.   Figures 31A, 31B, and 31C illustrate the seasonal
DO changes at each water sampling station. In almost all cases as would be antic-
ipated, the DO was higher in the winter and lower in the summer. West Brook
and the seepages had the highest DO  concentrations of all stations measured.
DO levels  at WBUS were always  above 8.0 mg/1 and station SB always remained
above 5.9 mg/1 even  during the summer. These high concentrations might be
expected because these shallow, rapidly flowing streams tend to reoxygenate
quickly.  The DO of the treatment plant effluent was about 2 .5 mg/1 higher than
that of the influent due to the reduction of the BOD load and as a result of oxygena-
tion received when spraying the sewage over the trickling filters.  The control
wells (4,7, 10) had  relatively low DO concentrations, particularly in the summer
months when DO levels fell below 1 mg/1 in both wells 4 and 10. Wells 2A, 6A,
12A, and to some extent, 8AA had higher DO concentrations than respective deeper
wells in these individual well clusters indicating the upper portion of the saturated
zone is more oxygenated by oxygen rich percolate.  Wells 9, 6B, 2B, and 3D
had DO concentrations which were  generally less than 1 mg/1.  Wells 1 and 5
also exhibited depressed DO levels which were probably the result of BOD from
the effluent. Wells 3A and  4 had low DO concentrations which is attributed to
biological decomposition in the swampy area of these locations preventing reaeration
below the soil  surface. The DO data collected for wells 11 cannot be considered
reliable since samples were secured  by pumping, and data indicates air was
entrained during the sampling process.

      The redox potential of most wells was in the range of +50 to +150 mv as
shown in Figures 32A, 32B, and 32C.  These values indicate the presence of
oxidative conditions. The  summer redox value for well 11D was zero, indicating
low DO which further confirms the probable error in DO measurements for this
well and implies a similar error for well IIS as discussed above. The only other
low redox potentials were the summer values for the influent and effluent which
is normal since warm weather decreases the saturation concentration of oxygen
in water. The DO measured in the  influent  to the plant probably represented
oxygen that became  entrained as the raw sewage splashed into the Parshall flumes
leading to the primary settling basins.  The oxygen concentration was further
increased in passing through the trickling filters, and the amount of oxidizable
organic matter in the water was reduced as the sewage passed through the treatment
 system.  All these factors account  for the rise in redox potential of the sewage
water from the time  it entered the treatment plant until it was discharged onto
 the sand beds.

       Seasonal fluctuations of pH at each sampling point are shown in Figures
 33A, 33B, and 33C.  Two processes which are accepted as influencing pH are
 the production of organic acids  and  carbon dioxide by soil bacteria and, possibly,
 the conversion of ammonia to nitrates, both of which lower the pH.  The influent
 and effluent generally remained stable around pH 7.0.  In welk 1 and 5  the pH
                                      57

-------
                    14-
                    12

                    It
co
               CO
               CO
               2

                                         \s
                                 2*      21      3*      31      3C      3D


                                                    SAMPLE LOCATION
                        Figure 31A.  Seasonal Variations in Dissolved Oxygen.  The Bars Represent

                                     Spring, Summer, Fall and Winter, in that Order.

-------
      14
      12-
tu
g



o
UJ


o
V)
2?
      li-



            Ct      f I
SAI     SI     II       9       1*



             SAMPLE  LOCATION
                                                                IIS
                                              110      14
         Figure  31B.   Seasonal Variations in Dissolved Oxygen.   The Bars Represent

                       Spring,  Summer, Fall and Winter,  in  that  Order.

-------
jtf

X
o
o
CO
(O

o
     14-
      12
It
           ,2,     |2B      SI    SB
                                                 W80S     INF     IFF
                                      SAMPLE   LOCATION
    Figure  31C.   Seasonal Variations in Dissolved Oxygen.  The Bars  Represent

                 Spring,  Summer, Fall and Winter, in that Order.

-------
ZfO'
ISO
1 "•
< 51-
I—
UJ
5 I
a.
1 *
•101
•ISO
200
-250






























r-|

















































































































1
21     3D      38      3C
3D
                         SAMPLE    LOCATION




Figure 32A.  Variation  in  Redox Potential During Spring  1976,

-------
     ZOOi
-.    150
<
>-
z
o
a
x
o
a
ui
a:
     1QI
      SO
      so
     •100
     ISO
     200
            i I
                    (I
                            M»     8A
                                                             10
US
                                                                             11 D      14
                Figure  32B.   Variation in  Redox Potential  During Spring 1976.

                              (With some  Results for Summer).

-------
CTi
ui
                                  12*     121      S»       SI     NtUS   KIDS


                                                       SAMPLE   LOCATION
INF     t f F
                           Figure 32C.  Variation  in  Redox Potential During Spring 1976.
                                        (With  some Results for Summer).

-------
en
                      S.I
                      I.S
                      1.1
                      7.5
                      7.1
                      t.S
21     21      3«
                                                         31      3C
                                                          Lttititi
31
                      Figure  33A.   Seasonal Variations  in pH.  The Bars  Represent Spring,  Summer,
                                    Fall and Winter,  in  that Order.

-------
    IS
pi    7.J
     I
          il      fI      III
III     II      >
    Sinplt Itciliti
                                                      II     11$     111      14
     Figure 33B.  Seasonal  Variations in pH.  The  Bars  Represent Spring,  Summer,
                  Fall and  Winter, in that Order.

-------
Ch
                       Ml
                       1.5
                       1.1
                  ,1     7.5
-



—





-



—



-



-



-



                             121     121     S»
SI    NIDS   NIDS     INf     EFF
    Staple locatin
                     Figure 33C.  Seasonal Variations in pH.   The Bars Represent Spring,  Summer,
                                  Fall  and Winter, in that  Order.

-------
was somewhat below neutral and ranged from 6.5 to 6.7.  Most other wells exhibited
a rise in pH which was very pronounced in wells 6A, 6B, 2B, and 3D but not
as pronounced in wells IIS and D. This increase may have been the result of
the reduction of nitrate to ammonia (or nitrogen gas) under anaerobic conditions.
Increases in pH were most evident in the summer when DO concentrations were
low, and sewage flows were high. Wells 8,9, and 12 did not experience large
pH increases, indicating aerobic conditions were maintained at these wells.
Wells 3A, 3B , and  3C, the seepage streams, and West Brook all had similar pH
patterns indicating that the sewage flowed deep in the aquifer at well  3D and
that wells 3A, 3B,  and 3C were influenced by local runoff.  The sewage effluent
deeper in the ground water could be entering Lake George where the aquifer
emerges in the lake beneath the lake surface.  The pH of the natural ground water
was slightly less than 7.0, as measured in control wells 4,7, and 10.

    Log scales were used to construct bar graphs to illustrate seasonal averages
of concentrations of various nitrogen forms due to their wide variation in concentra-
tion.  The seasonal averages for nitrates are shown in Figures 34A, 34B,  and
34C, while ammonia data appear in Figures 35A, 35B,  and 35C, followed by  Figures
36A, 36B, and 36C which contain results of the total Kjeldahl nitrogen determinations.
These data are discussed together since they are all interrelated.

    Well 7, a control well which was theoretically uncontaminated, had relatively high
nitrate levels, especially in the fall when the nitrate level reached 2.20 mg/1.
The ammonia nitrogen concentrations in the well were less than 0.10 mg/1, but in
the fall total Kjeldahl nitrogen (TKN) concentration rose to 1.17 mg/1.
Well 7 is located in a small depression uphill from the treatment plant. It was
learned that plant personnel had been dumping the scrapings from the sand beds
in the area of this depression.  Apparently runoff from the dumped material
drained into the depression, infiltrated into the aquifer, and contaminated well
7 destroying its usefulness as a control well. Well 10, which was also designated
as a control well, had nitrate levels in  the range of 1 mg/1 in both the summer
and fall. Ammonia levels reached as high as 0.40 mg/1 in the fall, and TKN levels
were above 1.0 mg/1 every season, indicating well 10 may also have been a poor
selection for a control well.  The sewage may have flowed far enough east to
affect the ground water around well 10. Another possibility is that the natural
 ground water could have actually contained concentrations of nitrogen forms at
these levels; however, the data from well 4 tend to refute this possibility. Since
well 4 was  located on the north side of West Brook, it could not be contaminated
 by the  sewage effluent; therefore, it was more representative of the natural ground
 water.  Data from well 4 indicate the nitrate and ammonia concentrations were well
 below 0.10 mg/1, and the TKN concentrations did not exceed 1.0 mg/1.

    As  anticipated, the treatment plant had high TKN values ranging from 14.7
to 18.5 mg/1 and 6.3 to 9.7 mg/1 for the influent and effluent, respectively.
The ammonia concentrations were also  high but decreased during the treatment
process. These decreases occurred as the ammonia and TKN were converted
to nitrates  and were most evident in the summer season when an increase in nitrates
in the effluent occurred. Some of the reduction was a result of'partial removal
of organic forms in the treatment processes.
                                      67

-------
00
                      11.1
                 NO, «
                      1.10
                     O.I I
21     21      3«
                                                         31     3C
                                                           licit!*!
31
                       Figure 34A.  Seasonal  Variations in Nitrate.   The Bars Represent  Spring,
                                    Summer, Fall and Winter, in that  Order.

-------
CTi
                     11.11
                      1.1
                       l.ll
                       Ml
                                       n
                             II      II     M»
M     II     »
    Staple licititi
                                                                         ii     n:     in
                        Figure 34B.  Seasonal  Variations in Nitrate.  The  Bars  Represent Spring,
                                     Summer, Fall  and Winter, in that Order.

-------
       1.11
  NOj-ll
   • f/l
       1.11
       Ml

             12A     121     SX
SI    WIUS    NIDS     INF     F. f F
    Samplt Ltcatiti
Figure 34C.   Seasonal Variations in Nitrate.   The Bars Represent Spring,
              Summer, Fall  and Winter, in  that Order.

-------
    11.11
    1.M
N*}-N

•I/I
    111
               r
                  2*      21      3D      31      3C
                                     Smote location
                                                        3D
      Figure 35A.  Seasonal  Variations in Ammonia  Nitrogen.  The Bars  Represent
                    Spring, Summer,  Fall and Winter,  in that Order.

-------
   ll.OOi
    1.00
NN3 N
• (/I
    1.10
    0.11
           f *
                  (I     111     ID
                                                        10
11S     111
                                                                               M
                                           l«cati»n
       Figure 358.  Seasonal Variations  in Ammonia Nitrogen.   The Bars Represent
                    Spring, Summer,  Fall and Winter, in  that  Order.

-------
    lO.OOi
     100
 NH3 N
      1.10
     0.01
           12*     128     S»
SB    WBUS    W8DS     IMF     EFF
     Sample Location
Figure 35C.   Seasonal Variations  in Ammonia Nitrogen.   The Bars Represent
              Spring, Summer,  Fall and Winter, in  that  Order.

-------
    IM.t
    II.*
TKN
     1.1
     I.N
21     3»      31      3C
            Sample l»citi»i
                                                        31
                                                      Lb
         Figure 36A.   Seasonal Variations  in  Total  Kjeldhal Nitrogen.  The Bars
                      Represent Spring, Stammer,  Fall  and Winter, in that Order.

-------
                   1111
                    1M
01
TIM
• I/I
1.1
1.1




















-

















—























•• Cl IU III || J i|
SiBflt Ltcitin

11

S
im
in
                                                                                             14
                        Figure 36B.  Seasonal Variations in Total Kjeldhal Nitrogen.  The Bars
                                     Represent Spring, Summer, Fall and Winter, in that Order.

-------
   111.1
    ll.l
TKH
mt/\
     1.1
    0.1
         12ft
121
                         Sft
SI
                    nn
            n
                                      WHS    WHS
                                         Ltcititi
                                     INF
                             IFF
   Figure 36C.  Seasonal  Variations in Total Kjeldhal Nitrogen.  The Bars
               Represent Spring, Summer, Fall and Winter, in that Order.

-------
    During the vertical transport of wastewater applied to the sand beds, the
TKN and ammonia concentrations decreased only slightly but the nitrates generally
increased, particularly in the spring, fall, and winter as shown in wells 1 and
5. However, wells IIS and 11D  had nitrate concentrations generally below 1.0
mg/1, except for well US in the summer when it reached 4.0 mg/1.  Wells 8AA
and 9,  downgradient from the treatment plant, exhibited nitrate levels of over
5.0 mg/1 and TKN levels  of 1.0  mg/1.  The ammonia concentrations in well 9
were above 0.25 mg/1, but the concentrations in the well 8 series were not as
strongly influenced by the sewage percolate as in well 9.  These results indicate
that the well 8 cluster is  located on the fringe of the laterial sewage flow through
the saturated zone.  Nitrate observations further downgradient at the well 6 series
were about 4.0 mg/1 while the ammonia levels were higher than those at well
9 (6B was above 0.50 mg/1),  and TKN values were roughly the same in wells
6 and 9.  This suggests that denitrification of the percolating wastewater occurred
during its laterial movement between wells 9 and 6. In comparison of wells further
down gradient, the  well 12 series experienced high nitrate and low ammonia
concentrations with TKN  values slightly lower than those in the well 6 series.

    Wells 2 and 3 are the farthest from the sand beds and are located about 400
to 500 m  (1,300-1,650 ft)  downgradient. Well 2A had slightly higher nitrate
levels than 2B, indicating the higher DO concentrations allowed oxidation of
ammonia and organic forms.  The TKN levels in 2B were only slightly lower
than 2A, but the ammonia values in 2B were an order of magnitude higher than
those in  2A, 0.04 mg/1, and .40 mg/1, respectively.  These data may be correlated
with the  low DO concentrations  found in well 2B.  By the time the effluent reached
well 3, the nitrate concentration was almost  10 mg/1, the upper limit of the drinking
water standards (21) . The nitrate concentrations in the upper portion of the
aquifer,  as shown by well 3A, remained below 6.0 mg/1 which was probably
due to the influence of runoff and precipitation.  The ammonia levels in the lower
portion of the aquifer as  illustrated by data from well  3D reached 0.66 mg/1 in
the summer, confirming other findings which indicate that the sewage flow concentrates
in the lower regions of the aquifer. The lower seepage streams had minimum
nitrate levels of 7.50 and 7.70 mg/1 in the fall and  summer, respectively.  The
upper seepage stream had maximum nitrate levels of 3.27 and 2.84 mg/1 in the
fall and winter, respectively.  These data seem to  indicate that the seepage below
Gage Road was much more strongly influenced by the effluent than the seepage
above Gage Road.  The impact of the seepage streams flowing into West Brook
could be readily seen when comparing the upstream and downstream stations.
The nitrate concentrations in West Brook increased from a range of 0.11  to  0.17
mg/1 at the upstream station to  a range of  0.60 to 1.30 mg/1 at the downstream
station.

    General observations and interpretation of data have led to the conclusions
that the  nitrogen entering the treatment plant was primarily as organic
and ammonia nitrogen which was partially converted to nitrate as the effluent
passed through the treatment system, soil, and aquifer.  However, well
data indicated that  denitrification occurred in the zones of the soil and aquifer.
By the time the effluent emerged from the subsurface  to form the seepage
streams, the nitrogen had essentially been converted to the nitrate form.
                                      77

-------
 The high nitrate concentration of the seepage streams was diluted approximately
 an order of magnitude in West Brook, resulting in a yearly average nitrate level
 of about 0.73 mg/1 entering Lake George. Vollenweider recommended an upper
 limit of 0.3 mg/1 to control excessive algal growths (24) .

     Due to the wide range of values encountered at the diverse sampling locations
 in this study, a log scale  was used for bar graphs depicting seasonal changes
 in phosphorus compounds.  Figures 37A, 37B, and 37C illustrate seasonal changes
 in soluble reaction phosphorus (SRP) at all stations while Figures 38A,  38B,
 and 38C  illustrate seasonal changes in total phosphorus  (TP) .  The  treatment
 plant influent and effluent had high levels of both SRP and TP.  Well data presented
 indicate  the degree of phosphorus removal as wastewater percolates through the
 sand beds. Well 1 exhibited a slight drop in SRP levels from the effluent,  but
 SRP levels in well 5 actually increased to 0.70 mg/1 in winter and 3.80 mg/1 in
 summer.  Polyphosphates (TP) also showed an increase in well 5 but not so pro-
 nounced as SRP.  The high values in well 5 could be the result of a release of
 absorbed or precipitated phosphates in the shallow unsaturated soil profile due
 to pH changes or redox potential changes.  Wells 4,7, and 10 indicated  that the
 natural ground water had SRP concentrations within a range of 0.01 to 0.01 mg/1.
 The SRP levels at the well 6 series  (particularly at well 6B) dropped to less than
 0.02 mg/1.  The SRP levels increased to 0.05 rog/1 at the well 12 series and then
 decreased to less than 0.01 mg/1 at the well 2 series.  At the well 3 series, the
 two shallowest points had SRP concentrations comparable to those in control well 4.
 Well 3C,  however, had much higher levels of SRP than the shallower elevations,
 and well 3D at the bottom of the aquifer had very low levels. The seepage streams
 had SRP  concentrations approximately the same as native ground water,  causing
 no significant change in the SRP concentration of West Brook.

    In general,  those well locations which exhibited high values for ammonia
 concentrations had very low SRP concentrations.  The two conditions are apparently
 related, but conditions which favor the formation of ammonia (low DO and low
 redox potential) should inhibit precipitation of phosphates.  Instead, it appears
 that the orthophosphates were efficiently removed under those  conditions.  Total
 phosphorus concentrations were relatively high in wells 1 and  5 indicating that
 there was little TP removal in sand  beds N-4 or S-3. Well 9 showed no decrease
 in TP concentrations, but  the well 8 series exhibited depressed TP levels along
 with the well 6 and 12 series.  Although TP concentrations at well 2B were approxi-
 mately the same as those of native ground water, concentrations in the upper
 portion of the saturated aquifer at well 2A ranged from 1.80 mg/1 in the spring
 to 2.50 mg/1 in the winter. Wells 3A and 3B, also located in the upper portion of
 the aquifer, had TP levels slightly above those of native ground water, but the
 deeper wells, 3C and 3D,  had TP concentrations about the same as control well 4.
 Comparisons of TP measurements between the West Brook upstream and West Brook
 downstream indicated there was an insignificant contribution of TP to West Brook
 from the infiltration system.  Examination of data from the control wells indicated
well 7 appeared  to be definitely contaminated from surface dumping of sand bed
 scrapings while well 10 had TP concentrations only slightly above those  in control
well 4.
                                     78

-------
                      11.11
                      l.M
                      •.It
vo
                  SIP
                  • I/I
                      Ml
                     Mil
                     I.MIH
                                           21     3D      31      3C      30
                                                          tvLtcatid
                        Figure 37A.  Seasonal Variations in Soluble Reactive Phosphate-Phosphorus,
                                     The Bars Represent Spring, Summer,  Fall and Winter, in that
                                     Order.

-------
                    lO.OO-i
                     1.00
                     0.10
oo
o
SUP
•I/I
0.01
0.001
0.000













— l




























__








—




















. D a in MS 1 1 D
                            6 I
                                    f B      SI*
SI     SI      »
    Staple location
                                                                                                14
                      Figure 37B.   Seasonal Variations  in Soluble Reactive Phosphate-Phosphorus
                                    The Bars Represent Spring,  Summer, Fall and Winter,  in that
                                    Order.

-------
CD
                           1000
                            100
                            0.10
                       S RP
                            0.01
                           0.001
                          O.OOO
                                 12*     121     S I
SI    WIUS    MflDS    INF     IFF
    Sinplc location
                    Figure 37C.   Seasonal Variations  in Soluble Reactive  Phosphate-Phosphorus,
                                  The Bars Represent Spring, Summer,  Fall  and Winter, in  that
                                  Order.

-------
              3
              I
              o.
              8
              I
                  lO.OOi
                   1.00
                   0.10
                              r
00
              o
                   0.01
                  0.001
                                                         38
3C
3 D
                                                  SAMPLE LOCATION
                 Figure 38A.  Seasonal Variations in Total  Phosphorus.  The Bars Represent

                              Spring,  Summer, Fall and Winter,  in that Order.

-------
                   10.00!
                •§,
                V)
00
                a
                CO
                o
                X
                a.
               <

               o
                    0.10
                    0.01
                   0.001J
                          f A      SB     SAA      SA      81       J       10      US     110      14



                                                   SAMPLE   LOCATION




                       Figure 38B.  Seasonal  Variations in Total Phosphorus.   The Bars Represent

                                    Spring, Suiiimer, Fall  and Winter, in  that  Order.

-------
                     lO.OOi
                 -§   1.00'
                 <0


                 I

                 a.
                 CO
                 o
                     0.10"
CO
                 O
                      o.oi-
                     0.001J
                             12«     121      S*       $0     WHS    WHS     INF     EFF
                                                          SAMPLE  LOCATION



                    Figure 38C.   Seasonal Variations  in Total Phosphorus.   The  Bars Represent

                                  Spring, Summer,  Fall and Winter, in that Order.

-------
    Seasonal averages of alkalinity are presented in Figures 39A,  39B,  and 39C.
The native ground water had an alkalinity ranging from 40 mg/1 in well 10 to
60 mg/1 in well 4.  Data indicate the alkalinity of the effluent was somewhat seasonally
dependent, but to a much lesser degree than the influent.  Both had alkalinities
ranging from about 80 to 170 mg/1.  Well 1 had an alkalinity of about 100 mg/1
while well 5 had values as low as 61 mg/1 in the fall. Wells 5 and 7 experienced
very similar seasonal fluctuations.  Well 8A had an unexplained high alkalinity
in the summer while wells 8AA,  9B, and 9 had values around 100 mg/1.  The
well 11 series appeared to have alkalinities about the same concentration as  control
well 4.  Proceeding downgradient from well 6, a general increase in alkalinities
was evident with the highest values occurring in well 3 and station SB .  The
high alkalinity from the seepages increased the values in West Brook from about
25 mg/1 to about 40 mg/1.

    Chlorides (Cl)  are known to be present in sewage and are a good indication
of water course contamination.  The seasonal averages of chloride  concentrations
are plotted in Figures 40A, 40B, and 40C according to the previous conventional
listing of sampling sites. The treated  sewage chloride concentration fluctuated
from about 65 to 101 mg/1, with the highest concentration occurring in the winter
and the lowest in the fall. The concentration in wells 1 and 5 was  approximately one-
half that of sewage applied to the sand beds. High chloride levels were observed
in wells 2A, 2B, 3B, 3D, 6A, and 6B.  The high chloride concentrations in the
deeper wells at site 3 were attributed to infiltration of the stored road deicing
salt at the town garage, while wells 2 and 6 high concentrations are a result
of the infiltrating sewage moving through the aquifer.  Moderate chloride levels
were observed in wells 4, 8AA,  8A, 8B, 9,  US, and 11D which can also be attributed
to the infiltration of sewage in the sand beds. Low values were found in wells
7, 12A, and 12B. The lowest values were in control well 10 which  never exceeded
5 mg/1. The seepage above Gage Road had a moderate chloride concentration
indicating dilution by native ground water while the well 3 series  and station
SB had a high chloride level indicating the road deicing salt contamination had
not been diluted by the time it had reached these stations.  High chloride concentra-
tion from the seepage areas did reflect an increase in chloride in West Brook
as can be seen by comparing  data from stations WBUS and WBDS.

    The dissolved solids concentrations  of water samples from  all  the stations
were monitored regularly and used to indicate the degree of sewage contamination.
Figures 41A, 41B, and 41C show the average seasonal dissolved solids  concentrations
at all the sampling locations.  There seemed to be very little seasonal fluctuation
in the concentration of dissolved solids;  however, the location  of the sewage
plume can be fairly well defined by study of data from all the stations.  The highest
values were observed in the below Gage Road station and in the multiple depth
wells at the well 3 site. These high values can be attributed to the influence
of the aforementioned highway deicing salt.  An unexplained high  value was
also noted in well 6A during the summer. Average values of about 200  mg/1
were found in the treatment plant influent and effluent, in wells 1, 2A,  2B,  5, 6B,
8AA, 9, 12A, and 12B, and in the seepage above Gage Road. Lower values were
found in wells 4, 7, 8B,  10, US, and  11D.  There was a slight increase in the
dissolved solids in West Brook between the upstream and downstream stations.
                                       85

-------
                  300
oo
en
                  250
                   200
                   ISO
                   100
                    SO
                                         28      3A     3»      3C      3D



                                                   SAMPLE   LOCATION
                     Figure  39A.   Seasonal Variations in Alkalinity.   The Bars Represent Spring,

                                  Summer, Fall and Winter, in  that Order.

-------
03
                    JOOi
                    250
                    200
                    150-
                    100
                     SO
                           S A      SB     1AA      IX      SB      9       10      IIS      110      14

                                                     SAMPLE    LOCATION

                       Figure  39B.   Seasonal Variations  in Alkalinity.   The Bars Represent Spring,
                                    Summer, Fall  and Winter,  in that Order.

-------
CD
CO
    300i


    250


    200
••^
J
>-   ISO
z
_1
"*   100


     so
                                                              Tl
                              12X     12B     Sit      SB    NtUS    WBDS     INF     EFF
                                                      SAMPLE  LOCATION

                      Figure 39C.   Seasonal Variations in Alkalinity.   The Bars Represent Spring,
                                   Summer, Fall and Winter,  in that  Order.

-------
uu


80


60
40
20
0



F
W












•FALL
•WINTER
8P-SPRINO
•S
-

•SUMMER

FlW
WBI

SP
IS

s


—






80S














1
(






ft


10
0
r>
^





In Lh
Cl|
,3
^^\



















SB

l^J






-n










Ih















F
5100

80
60
40
20
n




•
-





1
sh
\

















































~|
o




































                 Sampling site
Figure 40A.  Seasonal Variations  in Chloride,
                    89

-------
uu
80
60
40
20
n


-




F



W


sr


s
o.
00
r»



<*»
o



o
.— -
**—







r^.
CN
&



^-f



—







•1
1
„.. — '




—^




















       3A
3B
3C
3D
r
-
K
w>









„







	 1
        6A      6B      7       8A      8AA
                  Sampling site
Figure 40B.  Seasonal Variations in Chloride.
                     90

-------
100


 80


 60


 4fl


 2
        8B
 100


  80


  60


  40
9
n-Tl
   10
11$     110
         12A      12B     14
                  sampling sue

 Figure 40C.   Seasonal Variations in Chloride,
                     91

-------
£      WBUS    WBOS     Sft      SB      INF
I
         EFF      1       21      2B      4
                    Sampling SUB
 Figure 41A.   Seasonal  Variations in Dissolved Solids,
                           92

-------
400r
300
200
10








F



W



IP



s
































































           3ft
  £ 4
                            n
                    6B      7       Oft
                      Sampling  sue
Oftft
Figure 41B.   Seasonal Variations  in Dissolved Solids
                         93

-------
   400


    300


    200


    100-
                           10
m     110
            12»     12B     M
                      sampling sue


Figure 41C.  Seasonal Variations in Dissolved Solids.
                          94

-------
    The biological parameters studied were not collected through all seasons,
thus the data are presented in tables instead of seasonal averages on bar graphs.
Table 2 contains daily results of fecal coliform analysis and average values for
the station. As can be seen in Table 2,  a 33 percent reduction in fecal coliform
concentrations was achieved through the primary and secondary treatment process.
The only wells which indicated definite signs of contamination were wells 1 and
5, both of which were located in sand beds.  Well 5 was only about 4 m (13 ft) below the
the surface of the sand bed and well  1 was almost 20 m (65 ft) below the sand bed surface.
The coliforms should have been filtered and adsorbed before they reached either
of these well points.  Possible explanations are that the bentonite seal around
the well tubing was ruptured allowing  the effluent to seep down the side of the
tubing. Another possibility is that the tubing may have become cracked during
installation and effluent may have seeped into the well through these cracks.
All other wells were essentially free of contamination.  Those few organisms
which did appear were probably the result of bailer contamination during  sampling
rather than  actual presence of coliforms in the wells.

    The surface waters contained significant numbers of fecal coliforms which
are apparently a result of contamination from animals. A number of reconnaissance
tours were made of West Brook. These tours extended upstream for several
miles through the forest and the only sources of contamination which were found
were some trails frequented by horses from a nearby  stable.  These trails were
located approximately one mile upstream of the seepage sites.  A small abandoned
reservoir once used to store water for  drinking supplies for the Village was
also observed in the same area. The number of coliform organisms present in
the surface waters seemed to be dependent on the amount of runoff occurring
at the time the samples were collected.  For example, samples collected on August
10 after a heavy rainfall in the area were high in coliforms. Heavy rains also
occurred on February 2 and 10 and on June 16 and 20.  Samples collected within
two days of these rainfall events also had  high coliform counts.  Therefore,  the
data suggest that non-point sources  were  responsible for at least a portion of
the coliform  organisms present in West Brook.

    Total coliforms showed the same basic trends as fecal coliforms.  The total
coliform data presented in Table 3 also indicate contamination of wells 1 and 5.
Surface waters contained low numbers (generally about 50 organisms/100 ml)
of coliforms. The treatment plant was effective in reducing total coliforms in
the influent by about 25 percent as shown by the effluent applied to the sand
beds.

     The presence of fecal streptococci species is an indication of fecal waste
contamination by warm-blooded animals.  Table 4 contains the data for the entero-
cocci determinations.  Enterococci are streptococci normally found in the intestinal
tract of warm blooded animals. In general, results were very similar to coliform
measurements; however, well 5 indicated no evidence of enterococci contamination.
The surface waters  exhibited low but measureable contamination similar to coliform
studies.  This contamination was probably also from non-point sources.

     The virus studies were limited to  measurement of only coliphage. These
viruses were collected at all stations on three separate occasions and summarized
 as shown in Table 5.  All wells, including the control wells, indicated the presence

-------
TABLE 2.  FECAL COLIFORMS (100 ml"1)

Date
1/6
1/16
1/28
2/4
2/13
2/20
2/25
3/5
3/10
3/31
4/6
4/15
4/22
4/29
5/20
5/27
6/16
6/24
7/17
8/10
8/18
Average
1
-
-
-
-
-
-
-
-
-
145
226
-
-
214
0
-
-
14
-
4
-
101
2A
-
0
-
-
0
0
-
0
-
0
-
0
-
-
-
-
0
0
-
0
0
0
2B
-
0
-
-
0
0
-
0
-
0
-
0
-
-
-
0
0
0
-
0
0
0
3A
-
-
-
1
-
0
-
-
0
-
-
0
-
-
-
0
-
-
-
-
0
0.2
3B
0
0
-
0
-
0
-
0
0
-
-
0
-
-
-
0
0
-
-
-
0
0
3C
0
0
-
1
-
0
-
-
0
-
-
0
-
-
-
0
0
-
-
-
0
0.1
3D
0
0
-
0
-
0
-
-
0
-
-
0
-
-
-
0
0
-
-
-
-
0
4
-
-
-
-
-
0
-
-
-
1
-
0
-
0
-
0
-
-
-
-
-
0.2
5
-
-
212
-
-
-
-
-
0
-
1
-
-
-
74
-
-
-
-
-
-
72
6A
-
-
-
-
-
-
-
-
-
-
-
-
-
0
0
-
-
1
-
0
0
0.2
6B
0
-
1
-
-
0
0
-
-
-
0
-
-
0
0
-
0
0
-
0
0
0.1
7 8AA
-
-
-
-
-
0
0
-
0
-
0
-
0
-
0
-
0
-
-
-
0
0 0
8A
-
-
-
-
-
-
-
-
-
-
-
-
0
-
0
-
0
-
-
-
—
0

            (continued)

-------
                                          TABLE 2.  (continued)

Date
1/6
1/16
1/28
2/4
2/13
2/20
2/25
3/5
3/10
3/31
4/6
4/15
4/22
4/29
5/20
5/27
6/16
6/24
7/17
8/10
8/18
Average
8B
0
-
0
-
0
-
0
-
-
-
0
-
-
-
-
-
-
-
-
-
-
0
9
0
-
2
-
-
-
-
-
-
0
0
-
-
0
-
-
0
0
-
0
0
0.2
10
-
-
-
-
0
-
0
-
-
0
-
-
0
-
0
-
-
-
-
-
-
0
us
-
0
-
0
0
-
0
-
-
-
0
-
0
0
-
-
0
-
0
0
0
0
11D
-
-
-
-
-
-
-
-
-
-
0
-
0
0
-
0
0
-
0
0
0
0
12A 12B
0
0
-
-
0
-
-
0
-
2
-
0
-
-
-
0 0
0
0 0
-
0 0
-
0 0.2
14 SA
0
0
-
0
0
1
-
1
-
-
-
0
-
2
-
-
4
0 9
-
0 4
0
0 1.8
SB
2
-
-
7
-
1
-
3
-
-
-
2
-
1
-
-
62
51
-
163
1
29
WBUS

-
-
25
-
47
-
0
-
-
-
0
-
2
-
-
28
58
-
12
0
19
WBDS

-
-
20
-
43
-
0
-
-
-
0
-
1
-
-
29
69
-
23
0
21
TNF

-
TNTC
-
TNTC
-
510*
-
470*
-
-
-
500*
-
-
-
300*
-
640*
-
780*
533*
F.FF

-
TNTC
-
TNTC
-
100*
-
350*
-
-
-
220*
-
600*
-
190*
-
620*
-
430*
359*
*  Expressed as 1000's per 100 ml.

-------
                                        TABLE 3.  TOTAL COLIFORMS  (100 ml"1)
00

Hatfi
T/6
1/16
1/28
2/4
2/13
2/20
2/25
3/5
3/10
3/31
4/6
4/15
4/22
4/29
5/20
5/27
6/16
6/24
7/17
8/10
8/18
Average
1
-
-
-
-
-
-
-
-
-
250
TNTC
-
-
200
0
-
-
0
-
0
-
125
2A
-
0
-
-
0
0
-
0
-
0
-
0
-
-
-
0
0
0
-
0
0
0
2B
-
0
-
-
0
1
-
0
-
0
-
0
-
-
-
0
0
0
-
0
0
0.1
3A
0
-
-
0
-
0
-
-
0
-
-
0
-
-
-
0
-
-
-
-
0
0
3B
0
0
-
0
-
0
-
0
0
-
-
0
-
-
-
0
0
-
-
-
0
0
3C
0 .
0
-
1
-
0
-
-
0
-
-
0
-
-
-
0
0
-
-
-
0
0.1
3D
0
0
-
0
-
0
-
-
0
-
-
0
-
-
-
0
0
-
-
-
0
0
4
-
-
-
-
-
0
-
-
-
0
-
0
-
0
-
0
-
-
-
-
0
0
5
-
-
385
-
-
-
-
-
0
-
2
-
-
-
0
-
-
-
-
-
-
97
6A
-
-
-
-
-
-
-
-
-
-
-
-
-
0
0
-
-
0
-
0
1
0.2
6B
0
-
0
-
-
0
0
-
-
-
0
-
-
0
0
-
0
0
-
0
0
0
7 8AA
-
-
-
-
-
0
0
-
0
-
0
-
0
-
0
-
0
-
-
-
0
0 0
8A
-
-
-
-
-
-
-
-
-
-
-
-
0
-
0
-
0
0
-
0
-
0

                                                 (continued)

-------
                                                TABLE 3.   (continued)
VO
V0

Date
1/6
1/16
1/28
2/4
2/13
2/20
2/25
3/5
3/10
3/31
4/6
4/15
4/22
4/29
5/20
5/27
6/16
6/24
7/17
8/10
8/18
Average
8B
0
-
0
-
0
-
0
-
-
-
0
-
-
-
-
-
-
-
-
-
-
0
9
0
-
0
-
-
-
-
-
-
0
0
-
-
0
0
-
1
0
-
0
0
0.1
10
-
-
-
-
0
-
0
-
-
0
-
-
0
-
0
-
-
-
-
-
-
0
us
-
0
-
1
0
-
0
-
-
-
0
-
0
0
-
-
0
-
0
-
0
0.1
11D
-
-
-
-
-
-
-
-
-
-
0
-
0
0
-
0
0
-
0
0
0
0
12A
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0
0
0
-
0
0
12B
0
0
-
-
0
-
-
0
-
0
-
0
-
-
0
0
0
0
0
-
0
0
14 SA
5
15
-
0
4
0
-
0
-
-
-
0
-
0
-
12
7
0 7
-
0 0
0
0 3.9
SB
4
-
-
11
-
7
-
0
-
-
-
1
-
0
-
0
30
8
-
169
2
21
WBUS
-
-
-
31
-
48
-
1
-
-
-
1
-
4
-
-
18
29
-
15
0
18
WBDS
-
-
-
30
-
42
-
0
-
-
-
0
-
4
-
17
21
41
-
29
0
18
INF
-
-
TNTC
-
TNTC
-
308*
-
540*
-
500*
-
400*
-
400*
-
400*
-
790*
-
900*
530*
EFF
-
-
TNTC
-
TNTC
-
230*
-
380*
-
170*
-
390*
-
-
-
190*
-
660*
-
760*
397*
      *  Expressed  as  1000's per 100 ml.

-------
TABLE 4. ENTEROCOCCI (ml )

Location
1
2A
2B
3A
3B
3C
3D
4
5
6A
6B
7
8AA
8A
8B
9
10
US
11D
12A
12B
14
SA
SB
WBUS
WBDS
INF
EFF
Blank
6/24
3
0
0
0
0
0
0
0
-
0
0
0
-
-
-
1
0
0
0
-
-
0
1
6
4
4
-
-
0
7/17
14
0
0
0
1
0
0
0
0
0
0
0
0
-
0
0
0
0
0
0
0
-
1
1
3
3
15,500
9,500
0
8/10
1
0
0
0
0
0
0
0
0
0
0
0
-
-
-
0
0
0
0
-
-
0
0
1
1
1
-
-
0
Average
5.8
0
0
0
0.17
0
0
0
0
0
0
0
0
0
-
0.17
0
0
0
0
0
0
0.5
2.3
2.7
2.7
15,500
9,500
0
100

-------
of coliphage.  The highest values were found in wells 1 and 9 but no pattern
seemed to be established for any of the wells.  The treatment plant influent and
effluent had relatively low concentrations along with the seepage streams and
West Brook. Eliassen et al (25) found that viruses were effectively removed
in soil systems. Robeck (26) found good removal in only 0.6m (2 ft) of sand.
Apparently, the adsorption capacity for coliphage (and for most virus as well)
has been exhausted in the Lake George treatment system;  however, the effects
of the virus on West Brook (and consequently on Lake George) may not be significant
when compared to background levels of the West Brook upstream station.  More
work needs to be  done in this area.  Specifically, the sources of contamination
in West Brook and the native ground water need to be identified.

TABLE 5
. COLIPHAGE (ml )

Location
1
2A
2B
3A
3B
3C
3D
4
5
6A
6B
7
8AA
8A
8B
9
10A
IIS
11D
12A
12B
14
SA
SB
WBUS
WBDS
INF
EFF
Blank
6/24
62
24
30
56
40
51
40
27
50
10
32
31
_
-
-
60
52
26
27
-
-
62
6
2
12
5
15
11
0
7/17
51
37
20
17
36
38
47
33
27
21
16
23
37
22
-
71
42
30
29
20
47
-
4
7
8
11
10
14
0
8/10
87
31
42
31
36
41
60
39
39
24
34
9
-
-
-
43
30
41
45
-
-
38
12
8
7
9
17
6
0
Average
67
31
31
35
37
43
49
33
39
18
27
21
37
22
-
58
41
32
34
20
47
50
7
6
9
8
14
10
0
                                      101

-------
     Three separate BOD, studies were conducted in this treatment process evalu-
 ation. The first study dealt with influent and effluent BOD,, of the conventional
 portion of the treatment plant, and the results are presented in Table 6.  The second
 study dealt with the BOD,- reduction in the effluent as it stood  on the sand beds
 prior to  complete infiltration, while the third study involved determining BOD,
 concentrations at all other stations monitored including wells, seeps, and streams.
             TABLE 6.   REDUCTION IN BOD IN TOE CONVENTIONAL PORTION
               OF THE  LAKE  GEORGE VILLAGE SEWAGE  TREATMENT PLANT
                            BOD5,  mg/1
   Date
              Influent
Effluent
   *
   **
Reduction
1/13 §
1/20/75 (Avg)
1/29/75
1/29/75
5/18/76
6/7/76
6/9/76
6/12/76
6/14/76
6/15/76
6/16/76
6/17/76
8/3/76
8/5/76
8/6/76
8/8/76
8/9/76
35.5

14.5*
19.9**
-
72.5
86.3
88.8
69.4
61.9
35.6
91.9
210.0
145.0
154.0
199.0
138.0
14.8

14.1
21.5
13.8
36.3
20.6
35.4
21.4
15.0
20.6
40.3
18.8
20.6
42.5
78.8
58

18
-
81
58
77
49
65
58
78
81
87
87
77
43

Sewer from Lake George Village
Sewer from Caldwell District (Town of Lake George)
   (1975  data ref.  [27],  1976  data ref.  [23])
    In January the BOD,, concentrations of the plant influent were very low and,
as a. result, BOD removed in the plant was also low.  The average for the January
13 and 20 samples  indicated a 54 percent reduction through the trickling filter
which was considered satisfactory for a period during which the temperature
of the wastewater was about 5 C. During June the influent BODr averaged 72
mg/1,  and the effluent averaged 23 mg/1 for a 68 percent reduction through the
treatment process with the trickling filter alone contributing to 32 percent reduction.
In August the influent BOD- rose to an average value of 169 mg/1 while the efflient
    r  was reduced to 40 mg/1 indicating a  76 percent reduction in the primary
                                      102

-------
and secondary processes.  Even though the percent reduction was high in the
conventional treatment process during the summer months, there was a large
increase in both the influent and effluent BOD,- and subsequent high BOD5 loading
of the sand beds.  The flows were also higher in August, thus making the total
bed loading much higher than other seasons.  The relatively low influent BODr
in January and June indicates that the sewage was being diluted by sewer line
infiltration while the higher flows in August masked this infiltration. If the infil-
tration could be eliminated, obviously a substantial hydraulic load would be
removed from the plant.

     In the second study evaluating BOD ^ reduction on the sand beds, bed
N-4 was chosen for  study because the effluent took longer to infiltrate from
this bed than others in the system. Most beds drained after only one or two
days; however, effluent remained on the surface of this bed for almost four days.
The effluent applied to the bed on May 18,  1976, had a BOD^ of about 22 mg/1
(Table 7). The sample collected from the bed on that day after several hours
of flooding had a BOD,, of about 13 mg/1. This reduction was probably due to
settling of particulate matter from the ponded wastewater onto the surface of
the sand bed.   By the following day,  the BOD,, of the wastewater had dropped
to 7 mg/1 which was a reduction of 59 percent. This reduction was theorized
to be due  to combination of additional sedimentation and biological removal.
By May 20, the BOD- had dropped to 5 mg/1 but increased slightly on the following
day.  This increase may have been due to algal growths in the ponded water
or due to wind mixing which would have disturbed the sediments in  the shallow
water depth.  On the last day before the effluent had disappeared into the subsurface,
the BOD5  had  again dropped to 5 mg/1.  These values indicate a substantial BOD-
reduction of the wastewater while  standing on the infiltration beds.  In this     3
test, a 76 percent reduction was  experienced during the four days it took for
the plant effluent to infiltrate into the sand.  As previously mentioned, the effluent
infiltrates into the subsurface much more rapidly in other beds, but most of
the BODr  reduction occurred within only one day of detention.
    TABLE 7. BOD5 REDUCTION WHILE  STANDING ON  A SLOWLY  INFILTRATING SAND BED
                      Date                      BOD5>

                  5/18/76 Applied Effluent         21.5
                  5/18/76 On Bed                   12.8
                  5/19/76 On Bed                    7.3
                  5/20/76 On Bed                    5.0
                  5/21/76 On Bed                    7.8
                  5/22/76 On Bed                    5.3
                                      103

-------
     The results in Table 8 show BOD^ values collected during the third phase
 of the study involving wells, seeps, and stream stations. These results indicate
 low, but measureable, BOD,- at every location including  West Brook and the con-
 trol wells. In most cases these low BOD-'s were satisfied by sufficient DO in
 the wells.  However,  wells 1, 2B, 3D, 5, 6B, 9,  and 14 generally did not contain
 enough DO to satisfy the BOD-. These wells also experienced relatively high
 ammonia concentrations which are indicative of anaerobic conditions.  The ammonia
 could have been produced by bacteria which used the oxygen  present in nitrates
 as a substitute for free DO.  Ammonia exerts an oxygen demand on the subsurface
 system and appears as BOD in that test.

 HYDRAULIC CHARACTERISTICS OF THE SAND SYSTEM

 Bed Dosing

    The flow of the treatment plant influent is gauged by two Parshall flumes.
 One of the flumes records the flow from the Caldwell Sewer District (Town of
 Lake George) and the other records the flow  from the Village of Lake George.
 These flows are combined, along with flow from the trickling filter recycle line,
 prior to primary sedimentation. Daily readings from the totalizer meter are  recorded
 by plant personnel at approximately 8: 00am.  Rather than use conventional monthly
 flows, seasonal months were set up since they better correspond to population
 variations.  These seasonal months start on the 21st of the previous month and
 run through the  20th of the stated month. The mean monthly flows are presented
 in Table 9 in units of liters per  second (1 mpd =  43.803 liters per second) .
 Total seasonal flows were previously shown in Figure 5.  The steady increase
 in peak summer flows during the study period was accompanied  by an even greater
 increase in minimum winter flows.  While August flows increased 45 percent
 from 1969 to 1976, January flows increased 129 percent during the same time
 span. The main reason for the increase in winter flows was the increase in winter
 tourism due to the rising popularity of skiing, snowmobiling,  and other winter
 sports,  plus a growth in attendance at the annual Lake George Winter Festival,
 normally held every weekend in February.  Based on earlier data, it would  appear
 that the maximum flow in the summer of 1976  should have exceeded 50 I/sec,
 but the lower value can be explained by  the relatively cold wet weather encountered
 that year which reduced tourism.  In contrast, the summer of 1975 was hot and
 dry, and the large tourist population accounted for extremely high flows during
 that summer.

    The treatment plant was originally designed  to treat  21.9 I/sec (0.5 mgd').
 but was subsequently expanded to a stated capacity of 76.6 I/sec (1.75 mgd ).
Currently peak daily flows are as high as 55  I/sec (1.25 mgd) in the summer.
A major problem in plant operation is caused by flow surges resulting from the
 intermittent sewage pumping to the plant.  The high rate pump from the Caldwell
Sewer District operates at 220 I/sec (5 mgd) . Even if flows from the Village
 and recycle line  were absent, this high pumping  rate would disrupt the system.
When sewage is pumped from both the Town and the  Village, the influent flow
to the primary settling tank is so large that it stirs up settled solids upsetting
both the trickling filters and the secondary settling tanks. This results in an
 effluent whose BOD is sometimes higher than the  simultaneous influent, which
leads to rapid sand bed clogging.
                                     104

-------
TABLE 8.   BIOCHEMICAL OXYGEN DEMAND (5-day), mg/1

Sample
1
1
2A
2A
2A
2B
2B
2B
3A
3B
3C
3D
4
5
5
6A
6A
6A
6B
6B
6B
7
9
9
9
10
r.s
11D
12A
12B
14
14
14
SA
SA
SA
SB
SB
SB
WBUS
WBUS
WBUS
WBDS
WBDS
WBDS
Date
6/24
8/3
6/24
7/17
8/3
6/24
7/17
8/3
6/24
6/24
6/24
6/24
6/24
6/24
8/3
6/24
7/17
8/3
6/24
7/17
8/3
8/3
6/24
7/17
8/3
6/24
7/17
7/17
7/17
7/17
6/24
7/17
8/3
6/24
7/17
8/3
6/24
7/17
8/3
6/24
7/17
8/3
6/24
7/17
8/3
BOD5
2.4
0.7
2.6
2.0
1.9
2.1
1.5
0.9
0.8
0.7
0.6
1.1
1.2
2.0
1.9
2.5
1.2
0.8
2.7
1.0
1.7
0.6
3.1
1.6
2.2
0.5
0.8
1.2
1.8
1.4
4.0
1.7
1.1
1.4
0.6
1.6
1.6
1.0
0.3
1.6
0.4
0.9
1.5
0.4
0.9
DO
2.5
0.7
5.7
6.0
6.0
1.3
1.7
1.7
1.6
0.9
0.8
0.9
1.2
1.1
1.1
6.1
4.6
4.6
0.9
1.0
1.0
2.0
1.1
0.5
0.5
3.4
6.8
7.1
5.0
2.0
1.0
1.1
1.1
8.4
7.4
7.4
7.0
6.0
6.0
10.4
8.1
8.1
8.5
7.6
7.6
  Note:  All dissolved oxygen (DO)  values are
         expressed as mg/1.
                        105

-------
TABLE 9.  MEAN MONTHLY FLOWS  (liters/second)

Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
1968
_
-
-
-
-
-
25.45
31.72
21.68
13.67
11.57
10.73
1969
9.46
8.59
11.96
16.43
21.86
21.60
27.99
36.23
25.50
14.94
13.14
13.19
1970
14.06
16.38
14.89
20.55
20.46
24.14
33.16
33.16
29.35
19.49
14.63
13.76
1971
14.81
15.20
16.21
22.74
23.74
26.37
36.89
44.16
37.54
23.61
16.82
15.86
1972
17.22
17.44
19.19
24.93
29.31
29.04
41.22
42.41
29.70
19.19
28.39
18.53
1973
17.17
19.89
21.03
24.18
24.97
29.66
39.25
43.68
33.95
22.74
18.05
18.88
1974
25.01
22.04
25.76
28.26
28.52
30.67
40.22
43.15
32.42
25.76
18.14
21.51
1975
21.86
22.47
24.88
30.62
28.96
34.13
42.49
49.42
35.75
32.86
25.76
23.44
1976
21.64
27.47
32.72
32.59
32.24
34.87
41.97
45.91
34.96
-
-
-

-------
    The effective infiltration areas of the individual sand beds were measured
by Beyer (10) in 1975, as shown in Table 10.  North beds 13 and 14 were treated
as one bed by plant personnel since they are always dosed simultaneously; hence
they are combined  in Table 10.  Apparently little attention was paid to constructing
the initial six and subsequent eight north sand beds of equal area; however,
the south beds were constructed much more uniformly.  The total area actually
in use for infiltration is somewhat less than previous  estimates (9) . Apparently
previous estimates included the separating dikes as well as the infiltration beds.

    The dosing records kept by plant personnel were used to determine the frequency
of bed use for each season. Table 11 shows the number of times each bed was
dosed per season for four years for which data were available.   The frequency
with which any one sand bed was dosed was determined by a daily evaluation
by plant personnel of the condition of the beds and the need for beds. Obviously,
those beds which drained and dried fastest were dosed more frequently.  Bed
selection was also  based on consideration of where the location of the distribution
chambers were. If, for example, both beds S-l and S-5 were available for dosing,
bed S-l would probably have been dosed because only one gate  at one distribution
chamber would have had to be opened. To  dose bed S-5, three  gates at three
distribution chambers have to be opened.  Since the normal procedure was to
dose one north and one south bed simultaneously, and since there are fewer
south beds, the south beds are dosed more frequently.  The south beds are also
newer and drain more rapidly  than the north beds.

    In calculating  the depth of effluent applied to each bed, it was assumed that
one half the daily flow occurred between the hours of 8: 00am and 4: 00pm. If
four beds were dosed daily with the dosing changed at 8: 00am and 4: 00pm, which
was the normal operating procedure, one fourth the daily flow was applied to
each bed.  If three beds were dosed between 8: 00am and 4: 00pm and two beds
were dosed from 4: 00pm to 8: 00am, the first three beds each received 1/6 the
daily flow (1/3 x 1/2) and the  second two beds each received 1/4 the daily flow
(1/2 x 1/2). Using similar reasoning, the daily flows to each bed were calculated
using the plant flow and bed dosing records.  These daily bed loadings Were
then totalled for each bed by season.  Finally, using the surface area of each
bed, the depth of effluent applied to each bed during each season was determined
as presented in Table 12 .

     Bed S-l received a significantly higher loading rate than any other bed,
indicating experienced plant personnel were utilizing its high infiltration rate.
Using similar reasoning, after review of the data, it might be theorized that
beds N-7, N-13-14, N-8, and N-5 had the lowest infiltration rates. However,
this reasoning is not entirely valid since some beds which were available for
dosing were not always needed and were only flooded when flows were high
enough to require their use.  During peak flow periods, it often became necessary
to dose beds which still contained effluent from previous dosings.  Also, the
time required for  cleaning each bed was highly variant which resulted in different
lengths of time each bed was out of service. Therefore, realizing the non-uniformity
of bed dosing operations, the  loadings presented in Table 12 should be considered
as only approximations for estimating the infiltration rates of each individual
bed.
                                      107

-------
TABLE 10.  SAND BED AREAS
Sand Bed


N-l
N-2
N-3
N-4
N-5
N-6
N-7
N-8
N-9
N-10
N-ll
N-12
N-13-14
S-l
S-2
S-3
S-4
S-5
S-6
S-7
TOTAL NORTH BEDS
TOTAL SOUTH BEDS
TOTAL
Area
2
m

908.0
986.0
853.0
1037.0
1185.0
955.0
1730.0
1062.0
964.0
809.0
1361.0
1154.0
1500.0
1118.0
1118.0
1135.0
1135.0
1193.0
1200.0
1234.0
14502.0
8134.0
22636.0

ft2

9,774
10,613
9,178
11,160
12,754
10,284
18,619
11,427
10,373
8,704
14,646
12,425
16,151
12,035
12,035
12,218
12,218
12,846
12,915
13,248
146,105
87,551
233,656
           108

-------
TABLE 11.  BED DOSING FREQUENCY

Bed
N-l
N-2
N-3
N-4
N-5
N-6
N-7
N-8
N-9
N-10
N-ll
N-12
N-13-14
S-l
S-2
S-3
S-4
S-5
S-6
S-7
S
1
1
2
4
0
5
0
2
7
5
4
11
1
6
7
2
6
8
11
6
1973
F W
14
15
11
13
12
15
6
7
19
13
17
29
13
37
12
30
24
25
38
20
13
15
6
8
8
16
9
7
24
18
13
33
11
38
26
23
19
22
23
28
S
10
9
16
12
7
23
10
10
25
16
11
37
17
45
19
28
25
28
26
26
S
9
12
15
18
13
21
10
10
28
23
25
42
17
51
21
38
25
39
42
28
1974
F W
13
12
11
14
10
18
6
7
18
15
19
25
15
28
25
17
31
29
32
22
10
14
7
12
10
10
9
6
29
16
12
33
12
39
0
31
27
30
38
15
S
9
11
11
12
9
16
8
9
22
12
22
30
13
42
0
28
30
26
35
25
1975
S F
9
7
11
11
11
17
8
11
13
12
22
29
13
27
19
26
26
24
28
19
20
13
13
13
12
19
10
9
15
10
12
14
12
25
26
15
24
22
21
21
W
15
12
13
15
8
9
8
9
15
14
13
19
11
31
20
28
12
22
23
21
1976
S S
12
18
10
13
9
25
9
9
14
12
15
15
10
29
20
26
28
9
26
22
7
16
18
15
14
22
9
10
18
16
14
16
10
32
29
25
25
29
26
14
Total
142
155
144
160
123
216
102
106
247
182
199
333
155
430
224
317
302
313
369
267

-------
TABLE 12.  DEPTH OF APPLIED SEWAGE (meters)
Bed
N-l
N-2
N-3
N-4
N-5
N-6
N-7
N-8
N-9
N-10
N-ll
N-12
N-13-14
S-l
S-2
S-3
S-4
S-5
S-6
S-7
1975
Fall
12.32
7.31
8.19
7.23
5.49
11.24
3.95
4.36
10.73
11.91
7.46
12.52
5.85
17.37
13.42
9.26
11.86
12.43
11.81
11.28
1976
Winter
9.47
6.82
9.20
8.22
3.59
5.21
2.63
4.54
10.77
13.02
8.49
15.44
4.08
21.66
10.73
18.79
5.28
11.42
12.30
10.43
1976
Spring
9.68
12.77
7.69
8.73
5.24
18.60
3.37
5.83
12.48
11.89
11.06
15.05
6.37
22.91
12.98
19.36
18.46
5.99
18.53
13.17
1976
Summer
7.98
12.03
17.51
12.76
8.66
18.64
4.06
6.96
16.93
16.93
12.04
16.26
5.09
27.58
21.54
22.16
18.07
24.05
20.03
8.27
Total
39.45
38.93
42.59
36.94
22.98
53.69
14.01
21.69
50.91
53.75
39.05
59.27
21.39
89,52
58.67
69.57
53.67
53.89
62.67
43.15
                    110

-------
    Total monthly precipitation,  calculated from data collected in a precipitation
gauge located at the treatment plant,  is presented in Table 13 along with the
monthly averages of depth of sewage plant effluent applied to each sand bed during
one complete year of study.  The last column of Table 13 shows the percent of
precipitation effecting the total hydraulic loading.  The yearly average contribution
of precipitation to the bed loading was 2.59 percent.  This average contribution
was insignificant to the total volume infiltrated; however, in August 1976, a heavy
rainfall occurred during a period of peak flow. This incident not only added an
additional hydraulic load to the plant, but also disrupted cleaning operations.
Although precipitation may not have a significant direct effect on bed loadings, it
raay indirectly affect the plant hydraulics and operation.


 	TABLE 13.  PRECIPITATION EFFECTS	

             Depth  of AppliedPrecipitationPercent - Precip.
 Month       Sewage (meters)             (cm)                of Total
9/75
10/75
11/75
12/75
1/76
2/76
3/76
4/76
5/76
6/76
7/76
8/76
4.330
3.833
3.206
2.761
2.631
3.348
3.626
3.969
3.876
4.163
5.012
5.646
12.70
13.16
7.57
4.52
9.22
8.76
10.49
6.93
20.27
5.97
10.92
11.43
2.85
3.33
2.32
1.60
3.38
2.56
2.81
1.71
4.98
1.42
2.13
1.98
     During periods of high flows, the primary and secondary treatment processes
 operate very inefficiently, resulting in a high carryover of solids to the sand
 beds, which decreases the infiltration rate.  Although the beds should be cleaned
 frequently, they cannot be taken out of service for long periods during high
 flow periods.  Bed cleaning is normally slowed during summer months due to
 additional time required for weed removal in the beds. Severe conditions occurred
 during August 1975 and 1976 requiring sewage effluent to be pumped to adjacent
 forest lands for periods exceeding two weeks.

     Winter operation is of primary interest in this study since the site is located
 in the cold upper New York State climate. Within the primary treatment system,
 the only modifications necessary during sub-zero winter temperatures and
 heavy  snow are covering the Parshall flumes and the 2-compartment settling
 tanks with boards.  Secondary treatment is modified during winter by diverting
 the flow from the rotary high rate trickling filters to the standard rate fixed nozzle
                                       111

-------
sprinkling filter which is covered with boards.  Temperatures are maintained
above freezing at all times within the covered trickling filter. Heavy snow cover
provides insulation retaining heat from the wastewater.

    Primary interest is the operational kinetics of the sand beds during the winter.
The treatment system is operated continuously through the harsh winters without
effluent storage.  Normal dosing continues using two beds per cycle. Reed et
al (28) have indicated that an infiltration-percolation system could not work in
the harsh climate of the northeast.  He suggested that freezing of the moist soil
surface would create an impervious layer halting percolation of effluent.  Figure
42 shows this did not occur in the Lake George System.  The first photo shows
the initial dosing of an ice and snow covered bed. The ice and snow near the
splash pad melted very quickly during flooding. From this point the warm effluent
flowed under the ice simultaneously melting the  ice above it and the ground below
it.  The second photo indicates continuous melting as the depth of effluent in
the bed increased.  Within a few hours all that remained of the ice in the beds
was a few floating pieces that eventually  melted.  Infiltration proceeded, as in
other seasons,  with the ambient temperature cooling the effluent standing on
the bed. By the time the ponded  effluent had cooled and begun freezing again,
the majority of the effluent had entered the subsurface. An ice layer formed in
this manner may be beneficial, serving as an insulating layer for the soil surface.

Infiltration Studies

    The principal advantage  of the infiltration-percolation method of land treatment
is that the application rate is  greater per unit area than with other land  treatment
systems.  Studies were made to establish the approximate infiltration rates and
to determine the dynamics of  infiltration  existing at the Lake George Village Sewage
Treatment Plant.

    The first infiltration rate studies were attempted using a staff gauge mounted
in a sand bed from  which periodic water  level readings were recorded.  This
method failed to give an accurate  view of the process, since insufficient readings
were taken. Results from the recording  gauge in the seepage above Gage Road
Station indicated a virtually constant flow in the seepage stream. Therefore,
this Stevens Type F recording gauge was removed from the seepage  above Gage
Road Station and installed in  a sand bed. A second gauge was installed in another
bed under identical conditions. Stilling  wells were not necessary when using
the Type F gauge in the sand beds since  the beds were recessed below ground
level and wave action due to wind was minimal.

    The sand beds  used to study  infiltration rates were selected on the basis
of availability.  Work schedules of the treatment plant personnel with respect
to cleaning and dosing were often the determining factors in choosing a bed for
study during a specific time, particularly since  normal operational procedures
of the facility included a  drying period between  dosings.

    One of the major problems in  measuring accurate infiltration rates was the
inability to determine the initial rate during dosing. Reed et al.  (28) stated
that the greatest infiltration rate occurred at the time of initial soil wetting .
                                     112

-------
             Dosing an ice and snow covered bed
Melting of ice and snow as the bed fills with treated sewage
 Melting of ice and snow almost complete within a few hours
         FIGURE 42.  Winter Operation of a Sand Bed
                 Bed S-4 - February 7,  1975.
                             113

-------
 Similar results were observed at Lake George.  For example,  bed N-9 dosed on
 August 20, 1975, to a depth of 0.15 m (0.5 ft) recorded an extremely high infiltration
 rate of 0.9 m/day (2.94 ft/day) .  In all probability this infiltration rate was even
 higher at the beginning of bed dosing.  However, infiltration  measurements in this
 study were begun on each bed at the time flooding stopped. Thus,  these measure-
 ments reflect a more accurate evaluation of the actual uniform  infiltration rate
 during continuous operation.

     A general comparison of infiltration rates of several sand  beds  at corresponding
 heads of effluent is  shown in Figure 43. These rates represent normal operation
 which is considered to be the operational period following a week of flooding
 cycles after the beds had been cleaned and scraped. The infiltration rate appeared
 to be a function of head after the initial soil wetting.  This function tended to
 be invalid at lower heads, especially in the older sand beds.  The oldest north
 beds generally had  slower infiltration rates than the newer north beds; however,
 the highest rates in all beds were generally found in the south beds.

     A study was made to  determine if the infiltration rate declined with bed use
 and to what extent it was  renewed with a drying period between dosings.  The
 beds examined were N-5 and N-ll.  Bed N-ll had drying periods of approximately
 24 hours between draining and dosing while  bed N-5 had a 12  hour drying period.
 Longer drying periods would have been desirable but the summer months at
 Lake George do  not  allow  the plant operators this luxury.  Even with these short
 recuperative periods, there  was little reduction in infiltration rates as shown
 in Figure 44.

     Figure 45 illustrates the effect of bed cleaning in increasing the infiltration  rate .
 This Figure also shows a  decline in the infiltration rate after successive dosings.
 The initial rates recorded exceeded 0.6 m/day (2 ft/day), and higher rates would
 probably have been measured if recording equipment would have been operating
 properly during initial bed filling. It was anticipated that infiltration in bed S-5
 would have continued to decrease during the  next few dosings until it approached
 the curve of bed S-7, which is shown to illustrate normal conditions without
 recent bed cleaning.

    These studies were followed by an evaluation of the infiltration rate when
 beds are continuously flooded after cleaning.  As can be observed in Figure
 46, the response of the bed was similar to that of a bed operated with drying
 periods between dosings. There was a general reduction in infiltration rate
 with time after each  successive dosing when the bed was refilled while still  inundated
 from the previous dosing.  The rate quickly decreased to a rate approaching
 that of a normally operated bed.  The curve of bed N-5 was included to compare
 this study with normal operation where beds  are rested between dosings.

    The treatment plant was plagued during the summer of 1975 with weed growths
in the beds.  The major problem caused by the weeds was that they interferred
with raking the beds to remove the organic mat on the bed surface.  This occurrence
presented an opportunity  to evaluate the effect of weed growth in the beds on
the infiltration rate.  The  infiltration rate was found to be very low when measured
                                     114

-------
I
£
   o*
oc
    Figure 43.   Infiltration  Rates of Several Sand Beds During Normal Operation.

-------
~   1.0
I
                 I	1	till
                       05
  IO

HEAD (FT)
15
2.0
  Figure 44.   Changes in Infiltration Rates of Several Sand Beds  Under Normal
               Dosin& Operations Including Normal Drying Periods.

-------
 Ul
 I-
 tr
ir
 2.4
 22
2.0
 1.8
 1.6-
 1.4
 1.2
 1.0
08
0.6
O.4
0.2
                                      8-20
                                                        8-23
                            S-
                                                     8-25.
                 S-5
                              _--*—
                                             „_—*-
                                        K-—""          un
             NOTE BED S-7 CURVE
                   ADDED FOR
                   COMPARISON TO
                   UNCLEAN BED
          0.1
  1.0
HEAD (FT)
                                                         L5
    Figure 45.   Infiltration Rate in Bed S-5  Showing  the Decline in Rate With
                 Continued Normal Operation after  Cleaning the Bed.

-------
    1.0
I
111
I-
(C
    0.5
          8-26
                                                NOTE: BED N-5 CURVE ADDED
                                                     FOR COMPARISON TO
                                                     UNCLEAN  BED
                       05
   1.0
HEAD (FT)
2.0
     Figure 46.   Infiltration Rate in Bed N-3 with Continuous Dosing Following
                 Cleaning of the Bed.

-------
NJ
O
               1   »
               ife
               UJ
                   0.5
               <
               cr
                                            BED S-3 HAD A COVER  OF VEGETATION
                                            BED S-2 HAD A COVER  OF NORMAL  BARE SURFACE
                   NOTE:
                    HEAD  QUANTITIES
                    FOR  S-3 ARE AN
                    ESTIMATE
                                      0.5
     1.0

HE AD (FT)
1.5
                      Figure 47.   Comparison of Two Beds Under Constant Inundation One with a
                                  Weed Cover and One with the Weeds Removed.

-------
Vertical Flow Through the Unsaturated Sand —

    After any applied effluent infiltrates into a sand, it first percolates downward
through the unsaturated sand in essentially a vertical direction until it reaches
a saturated zone.  As wastewater reaches the saturated zone, it becomes dispersed
by diffusion and moves along with the native water in the saturated zone generally
at the same velocity and direction as the native water.  The first studies to determine
the vertical percolation rate were conducted by Glavin and Romero-Rojas in 1968
(29) .  Using rhodamine B dye, they established a vertical percolation rate of
approximately 8 m/day (27 ft/day) in the top 1.5m (5 ft) of sand beds N-ll and
N-13.  This value appears extremely  high and after comparing the results with
more recent work, it seems likely that the applied effluent in which the dye was
placed was moving down the outside of a poorly sealed well pipe to the well screen,
giving erroneous results.

    All the more recent tracer studies were performed in bed N-ll utilizing well
points, lysimeters , and sampling wells equipped with submersible pumps
as shown in Figures 9 and 10.  The first attempted dye studies for this project
began March 26, 1975, using rhodamine B dye.  The dye was added  at 10: 45am
after 2 hours and  45 minutes of bed dosing. Mixing of the dye was accomplished by
stirring the ponded water by wading and by the natural wind action. No usable
data were collected due to equipment failure.

    The second attempt to monitor dye movement through the sand beds was
performed during the month  of June 1975 using rhodamine WT which is less subject
to adsorption onto sand particles than rhodamine B.  Bed N-ll was dosed with
sewage effluent from 10: 10 am June 2 through 8: 00 am, June 3.  The dye was
added at 4: 00 pm on June 2,  and  all the lysimeters were evacuated that day.
Figure 48 is the diagrammatic result of this  experiment. The effluent moved very
rapidly into the soil column, and dye was observed in well point 11-A by 9: 00
am on June 3 in very strong  concentrations.  In the afternoon of June 3, the 3
m (10 ft) deep lysimeter was found to contain dye that registered as a maximum
reading on the fluorometer.  Sampling continued each day for the first week and
then each Monday, Wednesday, and Friday until completion. Two days were
found to be the minimum time required for the lysimeters to fill.  Well 11-S was
pumped continuously during the  study and, each night, a flow through flourometer
was attached to the well 11-S pump discharge line for continuous monitoring.
The peak dye concentration appeared at the 7.6 m (25 ft) depth on June 11.
The first dye appeared at the 10.6 m (35 ft) depth 22 days after the study began.
The first appearance of dye in the ground water from well 11-S was  observed
on June 27 25 days after its addition to the flooded bed.  The dye had traveled
through approximately 3 m (10 ft) of ground water to enter the well screen in
well 11-S.  Pump  drawdown  may have had some effect in pulling the  dye through
the ground water, but the effect  was  localized due to the low pumping rate.

    The peak concentration of dye reached the 18 m (60 ft) lysimeter 29 days
after the study began.  The curve in Figure 48 connecting the 3 m, 7.6m, and
18 m lysimeters is essentially a straight line indicating uniform vertical flow
through the unsaturated zone. The peak dye concentration reached well 11-S
in the saturated zone during the night of July 3-4, 31 days after the study began.
                                     121

-------
. SURFACE
u
10
2O
30

40

50
DEPTH- 60
70

80
90
\ TIME OF TRAVEL
*Xv\
" ^^Ni
• x"\.
^N. X
^\.
^\£PEAK CONCENTRATION
^\CURVE
WATER TABLE ^x
: \
AQUIFER /a *
^ FIRST APPEARANCE •
I BEDROCK 'N ^°mD WATER -
0
3
6
9

12

15
18 DEPTH-
METERS

24
27
0 10 20 30 40
TIME- DAYS
Figure 48.    Vertical Transport  of Rhodanine WT  in  Bed



                                 "*  «"
                                   122

-------
 The average velocity based on the movement of the peak concentration from the
surface to well 11-S was 0.7 m/day (2.32 ft/day) .  This is about 0.15 m/day
(0.5 ft/day) less than the velocity calculated based on the first appearance of
dye in the ground water.  The velocity through the unsaturated zone between
the surface and the 18 m (60 ft) depth was 0.55 m/day (1.79 ft/day) .  No dye
was ever recovered from well 11-D near the bottom of the aquifer.  This well
was pumped only for a few minutes during each sample collection period.

    The dye study was designed to determine:  1) the vertical flow rate of sewage
effluent through the unsaturated zone of sand below the infiltration beds; 2)
the depth of penetration of sewage effluent in the saturated zone and,  3) the plume
configuration and movement within the aquifer. It was hoped that the plume
could be followed downgradient, and its changes characterized by sampling the
network of test wells between the infiltration beds  and seeps entering West Brook.
Had the latter portion of this  study been successful, it would have  confirmed
that the seepage areas were the result of ground water recharge from the upgradient
sewage infiltration basins. Unfortunately,  either the  dye was  diluted and/or
absorbed upon entering the saturated zone,  or  the observation wells were not
located in the plume from bed N-11 because the dye could not be traced in the
ground water.

Horizontal Flow in the Saturated Zone—

    The previously discussed rhodamine WT tracer studies were unsuccessful
in measuring the horizontal movement of sewage;  therefore, separate tracer
studies were conducted to determine the horizontal transport of infiltrated sewage
within the saturated portion of the aquifer below the sand beds. It was concluded
after careful consideration that tritium would be the most useful tracer to measure
movement within the saturated zone since it essentially is the same as water.
It can be detected at very low concentrations so its addition would  not affect
either ground water or surface water supplies into which it ultimately might
flow.  Initially it was planned that a  1 curie (Ci) dose of tritium would be used
for each of 3 separate tests; however, permission  could not be obtained from
the New York Department of Environmental Conservation, Bureau of Radiation,
for doses this large.  Permission was given for use of 0.1 Ci doses for the three
separate studies.  An estimate was made of the anticipated concentration of the
tritium in the ground water after dilution with one day of dosing of sewage effluent
onto a sand bed.  The counting rates of samples collected from the dosed bed
and samples of that water upon arrival at the seepage areas were required to
be significantly higher than background tritium in the native ground  water if
valid flow rate determinations were to be made from the studies. These estimated
calculations are as follows:

    Flooding of bed N-4 on June 30, 1976 =0.24 million gallons = 0.89 megaliters

    Amount of tritium introduced in  the dosing = 0.1 Curie

    0.1 Ci/0.89 megaliters = 0.11 Ci/megaliter

     (0.11 Ci/megaliter) (1 megaliter/106liters) (106  \i Ci/Ci)  =

       0.11 n Ci/1 in the applied effluent where n Ci  = micro Ci =10   Ci.

                                      123

-------
       The combined flow of the seepages at West Brook above and below Gage Road
 was approximately 3.785 megaliters/day (1 mgd) . Thus, the tritium concentration
 in the combined seepage should have been:

     (0.11 n Ci/1)  (O.B9 megaliters)/S.785 megaliters = 0.03 n Ci/1
     (0.03 [l Ci/1)  (10pCi/n Ci)  (2.2 dpm/pCi)  (1 liter/10 ml) =
       66. 0 dpmj£ml in the combined seepage where:
       uCi =10  curies      _12
       pCi = pico curies = 10   curies
       dpm = disintergrations per minute

    A 7 ml sample of water from the seepage area should produce 462 dpm. Based
on a counting efficiency of 39.4 percent for the Tri-Carb counting system used,
a count of 182 counts per minute (cpm) above background could be expected.
Counts of this magnitude are considered adequate for detection.  The Beckman
liquid scintillation counting system with a counting efficiency of 56.4 percent
should yield counts of about 260 cpm above background.  The flow in West Brook
averages 37.85 megaliters/da (10 mgd) providing an additional 1: 10 dilution
of the seepages . Thus, the count above background for West Brook should be
approximately 18.2 and 26.0 cpm for the Tri-Carb and Beckman counters,  respectively
These values are considered high enough to provide statistically significant
results.  In all of the above calculations, it was assumed that complete mixing
of the introduced tritium in the applied effluent,  in the ground water, in the seepages,
and in West Brook  was obtained. If the tritium was not well mixed in the applied
effluent and infiltrated into the ground water as a relatively small plug  flow,
the "hot front end" could give counts significantly higher than calculated, but
detection with wide spread wells would be much more difficult.

    Rhodamine WT dye was also considered as a potential tracer.  This tracer
has the advantages of being detectable in low levels by use of a fluorometer and
being visually observable at higher concentrations.  At sufficient concentrations
the presence of the dye could be observed in the field as  samples are collected
from the observation wells.  The significant disadvantage of using rhodamine
WT as a tracer is that it is readily absorbed on soil particles, particularly on
clay.  Since there  is very little clay in the Lake George sand filtration system,
rhodamine WT was used in all three tracer studies as a secondary tracer to provide
a potential visual observation of the presence of the applied effluent at the numerous
sampling stations.

    Normally sodium chloride provides a satisfactory tracer for monitoring the
flow of ground water.  Unfortunately, due to contamination b/ highway deicing
salt stored  at the town highway garage located near the sewage treatment system,
sodium chloride could not be utilized as a tracer with the accuracy desired.
However, due to the lack of a better tracer, 108 kg (240 Ib) of NaCl was added
to the second tritium tracer study to provide a tertiary tracer to differentiate
between the second study and any other tritium tracer study. The amount added
was calculated to double the amount of chloride normally present in the applied
sewage, and it was also significantly higher than the chlorides in the ground
water near the deicing  salt storage area.
                                      124

-------
    Potassium, in the form of potassium chloride (KC1) ,  has also been used as a
secondary tracer and can be measured in low concentrations.  In the third tritium tracer
study, 22.7 kg of KC1 (50 Ibs) was added to differentiate this study from
previous studies.  The amount added was calculated to double the concentration of
potassium normally found in the applied sewage effluent.

    In each of the separate tracer studies, samples were secured on a daily basis
from all of the downgradient observation wells,  the seepage, and West Brook begin-
ning from the time of the tracer injection. During the third tracer study, the
sampling frequency was decreased to a biweekly schedule after the first two
months due to financial constraints. This schedule was considered satisfactory
since, as the duration of the study increased, the tracer would become more
dispersed, and it would require more than one day to pass an observation well.
This fact was confirmed by earlier results from  the third tracer study.  Prior
to introducing the tritium, background counts were taken on all well points between
the point of application and West Brook.  Each well point was bailed dry and was
allowed to refill for each sample collection assuring fresh aquifer samples.

    In the first tracer study, rhodamine WT was injected into well IIS on May
10, 1976, one day prior to injecting tritium into the same well.   The tracer was
injected directly into the upper portion of the aquifer, immediately below bed
N-ll.  To ensure that all the tritiated water reached the  ground water without adhering
to the walls of the well casing, well US was continuously flushed by pumping
into it from well 11D for approximately three weeks.  After the flushing period,
well US was sampled and found to contain no tritium.  Monitoring continued
daily in all surrounding wells, at the seep areas,  and in West Brook through
August without recovery of any dye or tritium .  There are a number of possible
reasons why this first tracer study was unsuccessful. One possible reason could
have been that the tritium traveled as a plug and was not detected because the
sampling frequency was not short enough.  If the tritium did not move into the
aquifer as a plug, it could have been small enough to have passed by all monitoring
wells without being  detected.  Another obvious  explanation could be that not
enough tritium was used to be detected after dilution in the ground water and
infiltrating sewage effluent.  The  most reasonable explanation is that the tritium
moved through the aquifer as a narrow tear-like plume and did not pass any of
the sample wells and was missed in the seeps either because it became diluted
in transit to this station or passed through between sampling periods. To assure
that this situation could not occur in subsequent studies, tritium and dye were
mixed with sewage effluent in infiltration basins then allowed to percolate into
the aquifer covering a much larger area, thereby, greatly increasing the chance
of detection at downgradient stations ,

    The second tracer study began on June 30, 1976,  in bed N-4 (Figure 6) .
The study was conducted in the manner described above to evaluate both the
vertical flow and the direction and time of horizontal flow.  Flooding of bed N-4
began about 8: 00 am and at 11: 00 am,  while the bed was being flooded, 108
kg (240 Ibs) of NaCl and 18 kg (40 Ibs) of 20 percent rhodamine WT dye solution
were mixed in  a 379 liter (100 gal) tank and applied to the bed influent on  the
splash apron.  To ensure uniform distribution of the tritium, the tritiated water
 (0.1 Ci of tritium) was mixed with city tap water in a 95 liter (25 gal) tank from
                                       125

-------
 which it was discharged onto the bed N-4 splash apron beginning at 2: 00 pm.
 Immediately after introduction of tritium in bed N-4,  the wells located between
 bed N-4 and West Brook were sampled  daily and analyzed in the North Hall Labora-
 tory for the three tracers.  On July 31, 1976, the Tri-Carb liquid scintillation
 counter became inoperative, and all subsequent samples were counted on the
 Beckman system in the laboratory located at Lake George. In order to assure
 retrieval of the tracers, an additional well (well 14)  was installed by July 3,
 1976, approximately 15  m (50 ft) from the northeast corner of sand bed N-4.
 This well was downgradient of bed N-4 and located in the top meter (3 ft) of
 the aquifer where there would be a  high probability of intercepting the tracers.
 Sand bed N-4 was monitored until the tritium no longer appeared in the sewage
 ponded on the bed.  After approximately  one week of sampling, during which
 time the bed was flooded twice, no tritium could be detected.  The initial infiltration
 period for the tracer spiked flooded bed was approximately four days while the
 second dosing took approximately three days to infiltrate.  Sampling was continued
 on a weekly basis until August 1976 without detecting any tracer at any sampling
 station including well 1  which was located in bed N-4 at a depth of about 2 m
 (6 ft) into the aquifer.  Sampling ceased  in August, and preparation for a third
 and final tracer study was begun.

    In view of the failure to detect any  tracers in the first two tritium tracer
 studies, it was important that the third and final tracer study be successful.
 In order not to lose the tracer by means of dilution, it was decided the south
 sand beds should be used to conduct the  test since there is less distance to the
 ground water and less ground water available for dilution. Sand bed S-3 was
 chosen for the tracer study since this bed had only 5.5m (18 ft) of sand above
 the bed rock, and the top of ground water was only about 4.2m (14 ft) from the
 sand bed surface (aquifer thickness of about 1.3m (4 ft). Furthermore, obser-
 vation well 5 was located in this  sand bed and penetrated the aquifer. A ring
 of additional observation wells was installed adjacent to the south beds on the
 downgradient side (Figure 49).  These new wells were required since all other
 observation wells were downgradient some distance from  bed S-3, and ground
 water monitoring close to the dosed  bed was imperative. During well construction,
 the intent was to locate the well screens alternatively at the top  and bottom of
 the aquifer.  It may be seen from the summary of the  data in Table 14 that since
 the aquifer was less than 2 m (7  ft)  thick, it made little difference whether the
 0.7m (7 ft)  long well point was located at the top or bottom of the aquifer.  A
number of attempts  to install observation  well 21  along the north edge of bed
S-4 was conducted and rock was encountered each time; therefore, its installation
was abandoned due to lack of time to locate a suitable site.  Wells 15 through
 19 were installed during August  1976, and well 20  was completed the second
week of September.  Sampling was begun on these  wells on August 29, 1976,
to determine the background levels for  the three  tracers which would  be introduced
into the ground water during the final tracer test.

    The tracers for the final study  were  introduced  into bed S-4 on November 1,
 1976, four hours after start of dosing.  This bed had been recently cleaned
in preparation for the test and had a relatively high infiltration  rate as noted
by the  fact that it took four hours to maintain surface ponding.  A factor  influ-
encing the four hour filling period was  the low flows  coming into the treat-
ment plant.   The pump which supplied sewage effluent to the south


                                     126

-------
         Sludge
          Beds
    Primary
   Settling
   Tanks

Influent
Chamber
  South
  Beds
                                                               Final Settling
                                                                  Tanks
 Figure 49.  Plan Showing Location of New Wells  Installed Specifically for
             the Tracer Studies.
                                   127

-------
sand beds was operating intermittently due to the low plant inflow.  The cycle
during the flooding period was 8 minutes of pumping and 8 minutes of resting,
which obviously lengthened the normal time of filling.  At 12: 40 pm 55 kg  (25
Ibs) of 20 percent rhodamine WT was added during one 8 minute dosing cycle.
The alternating pumping cycle and the wind appeared  to uniformly mix the dye
in the flooded bed. Starting at 1: 00 pm,  110 kg (50 Ibs) of KCL was added in
two equal doses over two dosing cycles.  Each dose of KCL was first mixed in
a 190 liter (50 gal) plastic drum prior to application at the splash apron in the
sand bed. Finally, the 0.1 Ci of tritium was diluted in a 60 liter (16 gal) container
and added to the applied effluent at the splash apron beginning at 1: 50 pm and
continuing through the 8 minute dosing cycle.  Flooding of the bed continued
for an additional hour and one half before sampling, which allowed  time for the
tracers to become well mixed in the bed.  Samples were then taken from each
corner of the bed to evaluate how well mixing  had been accomplished in the flooded
bed.  Dosing was then continued until 8: 00 am on November 2 after  which the
bed was allowed to dry.  By November 3 when the sampling team arrived, the
applied sewage effluent and tracers had  entered the subsurface, and the bed
was dry.  Subsequent bed floodings were initiated at 4: 00 pm on November 4,
8: 00 am November  6, and 4: 00 pm November 8.  Each  of these dosings was for
a single period consisting of approximately one fourth  of the day's sewage plant
inflow.  The bed was essentially dry when the next observation was made on
November 11.

                    TABLE 14.  DATA FOR SPECIAL WELLS
                  (Elevations in m. Above Mean Sea Level)

Location
15
16
17
18
19
20
21
Steel or
Plastic
S
S
S
S
S
S
S
Top of
Well
148.3
147.9
148.0
148.4
148.4
148.4
—
Ground
Surface
147.8
147.3
147.3
147.9
148.0
147.9
—
Approx.
Ground
Water
140
139.6
138.3
137.3
137.3
137.3
—
Bottom of
Point
139.9
137.9
137.3
135.6
136.6
135.8
—
Bedrock
_
137.9
-
135.6
-
135.8
—

    Sampling indicated the tracers were being retrieved in the observation wells.
Both the dye and tritium appeared in well 5 on November 3,  2 days after dosing.
As stated earlier, well 5 was located in the aquifer directly below sand bed S-
3. The dye and tritium was observed in well 15 on November 8, in wells 16,
17, and 20 on November 9, and well 18 on November 14.  After well 19 had been
installed, apparently the ground water elevation dropped below the well screen,
and samples could not be retrieved. The dye was observed  to move farther down-
gradient as indicated by its first appearance in wells 9 and 6B on December 4
and 27, respectively.
                                    128

-------
    The patterns of occurrence of the tracers in the observation wells are shown
in Figures 50-57.  In all of the figures, day O corresponds with the time of appli-
cation of the tracers into the flooded sand bed S-3 on November 1, 1976. In well
5 (Figure 50) , which was  located in the center of bed S-3, the tracers appeared
on the second day  after application with the tritium arriving a few hours prior
to the dye.  These tracers peaked on the 3rd and 4th day and then dropped
sharply to background by the 8th day. In well 15 (Figure 51) the first appearance
of the tracer was on the 7th day. A second lower peak occurred in this well
on the 10th day possibly indicating pulsing due to the intermittent bed loading.
By the 12th day the tracer concentrations returned to their background levels.
In well 16 (Figure 52) the  first appearance of the tracers was on the 8th day,
with the highest concentrations of both the tritium and dye occurring on the 9th
day. The tracers indicated a second peak on the 12th day,  again indicating pulsing
within the aquifer. By the 13th and 14th days, the tracers had returned to back-
ground levels.  The highest concentration of tritium observed in this study from
any sampling station occurred in well 17 as noted in Figure 53. Well 17 is located
approximately in the middle  of the ring of  new observation wells installed for
the final tracer study.  The first indication of tritium was observed on the 8th
day with the highest concentration of both tracers occurring on the 9th day.
This peak was followed by a second peak on the 12th day suggesting pulsing
characteristics described  earlier.  By the  14th day the tracers in well 17 had
returned to background.  Well 18 was located behind the rock which was encountered
during attempts  to install well 21. Apparently, the flow had to pass around this
rock to reach well 18 since the first appearance of the tracers above background
levels was not observed until the 13th day as shown on Figure 54.  The peaks
were much lower and broader than other wells indicating more dilution due to
longer contact time in the  aquifer. The peaks occurred on the 18th and 19th days
followed by a slow return  of tracer concentration to background levels.  Well
20 was located west of well 18, apparently on the opposite side of the rock previously
mentioned. Tracers first  appeared in well 20 on the  8th day and peaked about
the llth day, indicating the  flow may have passed well 20 prior to reaching well
18.  The tracer concentrations slowly returned to background as in well 18.
The rhodamine WT was detected in wells 9 and 6B beginning on the 33rd and
56th days,  respectively, as  shown in Figures 56 and 57. The tritium was not
recovered in either of these  wells or any other sampling locations although sampling
was  continued through June 1977.  Since rhodamine WT  identified as that added
to bed S-3 was positively monitored in wells 9 and 6B and no tritium could be
detected above background in these wells, it must be deducted that the tritium
concentration was diluted  to background level before reaching these wells.

    Calculations were made  of the time of flow to the various observation wells
using data  collected in the third tracer study as shown in Table 15.  Accurate
measurements were made  of the distance from the north  edge of sand bed S-3
to the individual observation wells.  The first appearance of the tracer was utilized
as the basis for determining the velocity of flow within the ground water.  The
reason for using the first  appearance of the tracer for horizontal velocity calculations
was  that the tracer was infiltrated over a wide surface area (the bottom of sand
bed S-3) resulting in considerable dispersion by the time it reached the saturated
                                     129

-------
              300
              HJ
             CPM
                                                   DYE
                                                    PPM
u>
o
              200
              100
                         TRITIUM
                     	RHODAMINE  WT
                                                                                            0
                 0
10
 15
DAY
20
                                    25
                                                                                          30
                          Figure 50.   Occurrence of the Tracers in Observation Well  5.

-------
300
CPM
                                                                            DYE
                                                                             PPM
200
 100
                                                      • TRITIUM
                                                    	RHODAMINE WT
                           10
 15
DAY
20
25
30
          Figure 51.   Occurrence of the Tracers  in Observation Well  15.

-------
300
CPM
                                       DYE
                                        PPM
200
100
                                                       •TRITIUM
                                                    	RHODAMINE WT
  0
                           10
 15
DAY
20
                                                               25
                                                                          30
            Figure  52.   Occurrence of the Tracers  in Observation Well 16.

-------
                400 -
                CPM
                300
                200
u>
00
                100
                   0
                                        DYE
                                        PPM
                                                                         TRITIUM
              	RHODAMINE WT
                                            10
 15

DAY
20
25
30
                         Figure 53.   Occurrence of the Tracers in Observation Well 17.

-------
300 -
CPM
                                                                             DYE
                                                                              PPM
200
               -TRITIUM
 100
	RHODAMINE  WT
                                          *
                            10
                           15
                          DAY
20
                                                                25
                                                              30
              Figure 54.    Occurrence  of the  Tracers  in Observation Well  18.

-------
                300
                HJ
                CPM
                                       DYE
                                       PPM
                200
Ul
                 100
                                                                        TRITIUM
                                                                    	RHODAMINE  WT
                                                                                             0
                                           10
 15
DAY
20
25
                                                                                          30
                        Figure 55.   Occurrence of the Tracers in Observation Well 20.

-------
   .06
   .05

DYE
 PPM
   .04
   .03
   .02
   .01
                                         I
                         	RHODAMINE  WT
              10
20
30
 40
DAY
50
60
70
       Figure 56.    Occurrence of the Rhodamine WT Tracer in  Observation Well 9.

-------
  .06
  .05

DYE
 PPM
  .04
   .03
   .02
    .01
I
	 i
10
r~
i
i
i
	 RHODAMINE WT J
i
1
1
1
1
1 1 1 1 / 1
20 30 40 50 60
DAY
I
70
         Figure 57.    Occurrence  of the  Rhodamine WT Tracer in Observation Well 6B,

-------
zone. The tracer which entered the ground water at the north edge of bed S-3
would reach the observation wells first, and the peak of the tracer would arrive
at the observation Dwells at approximately the time it took for the flow from the
south edge of the bed to reach the observation wells. Therefore, basing the
flow velocity on the distance from the north edge of bed S-3 to the observation
well and the first observation of the tracer in that well should provide the most
valid determination of ground water velociti^   F,,~+u         ,
all horizontal veloci.y Lasuremen.s was coTsid.^dt faTS    '
appearance in the saturated ground water which was the tim» n, + *v,  *      *•  *
appeared in well 5 directly beneath bed S-3.  Thus  the t£e fSft^J^Scal
flow from the surface of the sand bed to the ground water is not considered in
horizontal flow calculations .  It must be emphasized that all the  results are considered
to have an accuracy of ±1 day, including  the time to appear in weu 5
              TABLE 15.  VELOCITY OF FLOW IN SATURATED ZONE
From
Bed S-3
Bed S-3
Bed S-3
Bed S-3
Bed S-3
Bed S-3
Bed S-3
Well 15
Bed S-3
Well 9
To
Well 15
Well 16
Well 17
Well 18
Well 19
Well 20
Well 9
Well 9
Well 6B
Well 6B
Distance, m
50.60
62.64
69.72
75.90
76.50
72.34
249.63
199.03
321.11
71,48
Time for First Tracer
Appearance, days
5
6
6
11
-
6
31
26
54
23
Rate
m/day
10.1
10.4
11.6
6.9
_
12.1
8.0
7.7
5.9
3.1
ft/ day
33.2
34.2
38.1
22.6

39.6
26.4
25.1
19.5
10.2

    As seen in Table 15, the velocity in the area north of the south sand beds
ranged between 10 and 12 m/day (33-40 ft/day) .  The slower transit time of
7 m/day  (23 ft day) to well 18 may be explained by the presence of the large
rock area near well 18 that was discussed earlier. Since the liquid would have
to flow around this area, the actual distance of flow was probably greater than
the 75,9  m (250 ft) measured. The dye was observed in well 9,  31 days after its
observation in well 5.

    Based upon 31 days of transit time for the dye to move from well 5 to well
9, the velocity was calculated to be 8 m/day (26 ft/day) which was slightly less
than the  velocity immediately up gradient.  Calculating the velocity on the 26
days transit from well 15 to well 9  gave only a slightly lower rate of flow in this
area. The dye occurred at well 6B, 54 days after observation of the dye in well
5.  Calculations indicate the ground water velocity between these two observation
points averaged about 6 m/day (20 ft/day) .  By calculating the velocity  of flow
from well 9 to well 6B for the 23 day transit period between these points, an appreci-
ably lower velocity of 3 m/day (10 ft/day) was indicated.  This lower velocity
                                     138

-------
 could be attributed to either an actually slower flow rate in this area, to lateral
 dispersion, or to lateral channelization of the tracer which would effectively
 increase the distance traveled causing the velocity to appear to be slower.

     The next sampling well downgradient in the suspected flow path of the sewage
 effluent was well 12.  Unfortunately, both sample wells at site 12 were inoperable
 during the period of this study.  None of the three tracers added in the third
 study were ever found in wells 2 and 3, in the seepage above Gage Road, in
 the seepage below Gage Road, or in West Brook downstream of the seepage areas
 through the conclusion of the study on June 30, 1977. It is unfortunate that
 the flow could not be traced beyond well 6; however, the results do agree with
 Bever's calculations (19) which  indicated that a velocity of approximately 11
  /d    (36 ft/day) would be necessary to prevent excessive ground water mounding
 m  tvf  area of the sand beds.  One minor discrepancy between the data presented
 ^  th^s study and Beyer's is that in the area between wells 6 and 9, Beyer indicated
 ^horizontal velocity similar to that of the area farther upgradient; whereas,
 this study indicates an apparent horizontal velocity of approximately 1/3 that
 of Beyer's.

     Comparison of vertical velocity measurements of sand beds S-3 and N-ll
  pvealedsome very interesting data.  The vertical velocity in bed S-3,  based
    the appearance of the tracer in well 5, indicated a velocity of 2.5 m/day (8
 ft/dav) while the  vertical velocity measured earlier in bed N-ll was 0.7 m/day
 (?  32 ft/day) .   The difference in the two values indicates significant differences
 •  sand hydraulic characteristics.  These characteristics can be attributed to
 "^tural geologic phenomena that occurred during  the formation of the glacial
 dlta in which the sand beds are located. During formation of deltas, soil particles
      rmally distributed with the very coarse material being  deposited near
 th 6 ice melt area while the progressively finer particles are deposited farther
 v16 cracjient. Thus, particle size distribution greatly effects both the vertical
 downg    ntaj rates Of water movement at different locations in the delta.

    B  sed on a conservative velocity of 10 m/day  in the  saturated zone, it should
 t-    ftir/m  approximately 60 days for the tracers to travel the 600 m (2,000 ft)
  W6 t Brook   As can be seen in Table 15, it actually took 54 days for the  tracers
  t    1 from S-3 to well 6B which is approximately half the distance between
 th    uth sand beds and West Brook.   Since sampling was continued for 8 months
 the so    jmateiy 240 days, it was felt that ample time had elapsed for the tracers
 or*PP^"reached West Brook during the study. Since tracers were never observed
 K   a d well 6, it may be concluded that either an  insufficient amount had been
 dd°d or that the tracers  had been adsorbed in the soil system prior to reaching
     ,.   noints farther downgradient.  It is possible that the rhodamine WT dye
 "^adsorbed by the soil or diluted to background levels; however, it is normally
WaS 'dered that tritium is not adsorbed and,  therefore, insufficient tritium may
htve been added to be detected the entire distance required after dilution in
th  ground water. Since  the rhodamine WT was detected at a greater distance
from the application point in this study than tritium, it became a better tracer
than the small amount of tritium used.
                                      139

-------
SOIL ANALYSIS AND WELL TESTS

      Numerous samples of the sand were collected for analysis but few extensive
analyses were performed due to time limitations .  The first set of sand analyses
was conducted in 1968 by Glavin and Romero-Rojas (29) as shown in Table 16.
These analyses were conducted on sand collected from various depths in bed
N-13, and the results indicate the filter media is primarily a fine uniform sand.

 	TABLE  16.   SIEVE ANALYSES -  BED #15	

 Sample  Location            Effective Size         Uniformity Coefficient

 Test boring hole  #3            0.19 m                        2.6
    Approx.  30-33'  depth
 Test boring hole  #3            0.135 mm                      3.4
    Approx.  10' depth
 Center  of  bed                 0.25 mm                       3.6
    13" below surface*


*Sample  collected and  analyzed by New York State Department of Health

      Additional analyses were performed by the New York State Department
of Environmental Conservation (NYSDEC) on samples taken from the well bores
of a number of wells during their construction. It may be seen in Table 17 that
soil taken during construction of well 1, located in bed N-4, consisted primarily
of fine sand. Well 5, located in bed S-3, was also quite high in fine sand.  Well
6 had a nearly equal distribution of fine and coarse sand while well 3 in the seepage
area had a mixture  containing mostly fine sand with relatively large quantities
of gravel and coarse sand. Well 2 located near the seepage area above Gage
Road consisted primarily of coarse sand. Comparison of the soils at these five
locations illustrates that the soil is not very homogenous or that the sampling
technique did not accurately characterize the soil at each location. Permeability
measurements conducted on the samples from wells 1,2,  and 6 showed considerable
variation. Well 2 had a permeability of 19.5 m/day (64 ft/day) while wells 1
and 6 had 8.5 m/day (28 ft/day) and 4.6 m/day (1.51 ft/day), respectively.

      One other set of soil  samples was  collected on January 9, 1975.  These
samples were forwarded to the US Army Corps of Engineers Cold Regions Research
and Engineering Laboratory, Hanover, New Hampshire.  The laboratory then
sent the samples to  the United States Testing Co. , Inc. , for extensive analysis.
These samples were secured at 0.6 m (2 ft) intervals during the installation of
the two pumping wells located in bed N-ll.  Analyses conducted for organic
carbon, total phosphorus, organic phosphorus, exchangeable phosphorus, soluble
phosphorus, organic nitrogen,  and the cation exchange capacity were all high
in  the top 3 m (10 ft) of the sand bed, but they gradually decreased with soil
depth in  relatively low values. Analyses made to determine the level of free iron
oxides, pH,  soluble salts, and the exchangeable cations (calcium, magnesium,
potassium, and sodium) indicated little variation with depth.
                                      140

-------
    Permeabilities measured within the aquifer ranged from 0.4 m/day (1.37 ft/day)
to 19.5 m/day (64 ft/day) .  Based on a mean slope of 0.032 m/m from well 11 to
well 2 and using the maximum permeability, the estimated horizontal travel time
to well 2 would be  2.67 years.  This velocity obviously does not agree with that
described earlier in these ground water tracer studies.
                          TABLE  17.   SAND ANALYSIS
  In Aquifer
Well 1
Well 2
Well 3
Well 5
Well 6
Effective Size, mm
Mean Size, mm
Uniformity Coeff.
"S" Shape Curves
Gravel %
Fine Sand %
Coarse Sand %
Permeability
m/day
ft/ day
0.08
0.15
2.13
?
1
94
6

8.5
28
.37
1.40
4.32
Yes
16
13
71

19.5
64
0.15
0.27
2.53
No
20
62
18

-
-
0.12
0.30
2.83
Yes
8
73
19

-
-
0.15
0.38
2.93
Yes
1
58
41

4.6
15.1

    An attempt was made to determine the permeability of the aquifer below bed
N-ll using the pumped wells.  Well 11-D was used for the tests conducted during
August 11, 13, and 18, 1975.  The data were evaluated using three methods:
time-drawdown, time-recovery, and residual drawdown (13).  Of the three methods,
the residual drawdown method is preferred for transmissibility determinations,
particularly when an observation well is not available, and the pumped well
is used for drawdown measurements (13) .  These tests resulted in permeability
values on the lower end of the range of those previously reported by the NYSDEC.
All these tests were taken under conditions of partial penetration of the aquifer
and should only be considered approximations.

    Permeabilities were measured in  the unsaturated soil using the borehole
test in well points 11A and 11F at depths of 0.6 m (2 ft)  and 3.6m  (12 ft) , respectively.
The results of these tests  showed that the permeabilities in these respective zones
were 2.9 m/day (9.5 ft/day) and 2.8  m/day (9.15 ft/day) . The range of permeabi-
lities calculated by this method again fell within the range of those calculated
by the NYSDEC.  A final check on the permeabilities was performed  using a sample
from a 26.2 m (86 ft) depth from bed N-ll.  These calculations using Hazen's
Formula ( 13) , indicated a permeability of 19 .4 m/day (63 .8 ft/day) . The corres-
ponding permeability estimated by the NYSDEC was  2.8 m/day (9.18 ft/day)
as shown in Table 18.  In  general, the velocities  calculated from the permeability
studies were much lower than those determined in the tracer studies.
                                     141

-------
                   TABLE 18.  SAND ANALYSIS - BED 11
Depth, m                     20.7            23.7              26.2

Effective Size, mm           0.07            0.09              0.15
Mean Size, mm                0.32            0.48              0.70
Uniformity Coeff.            5.57            6.78              6.0
"S" Shaped Curve             Yes             Yes               Yes
Gravel %                     4               8                14
Fine Sand %                 45              38                26
Coarse Sand %               36              46                55
Permeability
   m/day                     0.4             1.3               28
   ft/day                    1.37            4.27              gj]_g
                                  142

-------
                                  REFERENCES
  1.    Pound, C. E.  and R. W. Crites.  Wastewater Treatment and Reuse by Land
       Application-Vol.  II.  EPA-660/2-73-006b, U.S. Environmental Protection
       Agency, Roberts. Kerr Environmental Research Laboratory, Ada, Oklahoma,
       August 1973.  241 pp.

 2.    Evaluation of Land Application Systems. EPA-430/9-75-001, U. S .
       Environmental Protection Agency, Washington, D.  C. ,  March 1975.  182 pp.

 3.    Wastewater Reclamation at Whittier Narrows .  Publication No. 33,
       California State Water Quality Board, Sacramento,  California, 1966.  99pp.

 4.    Laverty, F. B., R. Stone, and L.  A. Meyerson. Reclaiming Hyperion
       Effluent.  J. San. Eng. Div.,  ASCE  87(SA6):  1-40, 1961.

 5.    Sullivan, R. H. and M. M. Cohn.  Survey of Facilities Using Land
       Application of Wastewater. EPA-430/9-73-006, U.  S. Environmental
       Protection Agency, Washington, B.C., July  1973.  377 pp.

 6.    New York Classifications and Standards Governing the Quality and
       Purity of Waters .  Part 702. 1975.

 7.   The Environmental Conservation Law. Chapter 664.  Sec. 17, Title 17,
      Art. 1709, Albany, New York,  1977.

 8.   Official Code Rules and Regulations of the State of New York.  Part 830,
      Title 6, Item 430,  Water Index  No.  C-101-T367,  Class AA.

 9.   Vrooman, M. ,  Jr. Complete Sewage Disposal for a Small Community.
      Water Works and Sewerage 87(3): 130, 1940.

10.   Beyer, S.  M.  Flow Dynamics in the Infiltration-Percolation Method of
      Land Treatment of Waste Water. M.S. Thesis, Rensselaer Polytechnic
      Institute, Troy, New York, 1976.   156pp.

11.   Aulenbach, D. B., T. P. Glavin, and J. A. Romero-Rojas. Protracted
      Recharge of Treated Sewage into Sand:  Part I - Quality Changes in
      Vertical Transport Through Sand.  Ground Water 12(3): 161-169, 1974.
                                    143

-------
12    Fink  W. B.  and D. B.  Aulenbach. Protracted Recharge of Treated
      Sewage into Sand:  Part II  -  Tracing the Flow  of Contaminated Ground
      Water with a Resistivity Survey.  Ground Water 12(4): 219-223, 1974.

13    Ground Water and Wells. Edward E. Johnson, Inc. , St. Paul, Minnesota,
      1966.  440pp.

14    APHA, AWWA , and WPCF . Standard Methods for  the Examination of Water
      and Wastewater. American Public Health Assoc. ,  13th ed.,  1971, and
      14th ed. , 1975.

15.   Hershey, A. D., G. Kalmonson, and J. Bronfenbrenner.  Quantitative
      Relationships  in the Phage-Antiphage Reaction:   Unity and Homogeniety
      of the Reactants.  J. Immunol. 46:281-99, 1943.

16.   Wang, L. K.  and  P.  J. Panzardi. Determination of Anionic Surfactants
      with  Azure A and Quaternary Ammonium Salt.  Analytical Chemistry 47(7):
      1472-1477, 1975.

17.   Determination of Tritium as HTO (Distillation Method) .  Radiological
      Sciences Laboratory, Div. of Laboratories and  Research,  New York
      State Department of Health, February 1973 (Revised May  1974).

18.   Harris, R.  The Effect of Depth on Water Quality in a Sand Bed.  M.E.
      Project, Rensselaer Polytechnic Institute, Troy,  New York, December 1975.
      68 pp.
 19.
Reach, R.  G. Tertiary Treatment as a Function of Depth in an Infiltration-
Percolation System.  M. E. Project, Rensselaer Polytechnic Institute, Troy,
New York,  May 1977.  201pp.
 20.    Lipka, G. S.  Chloride and Potassium Studies at Lake George, New York.
       M. E. Project,  Rensselaer Polytechnic Institute,  Troy, New York, May 1975.
       97pp.

 21.    U.S. Public Health Service Drinking Water Standards, U. S. Public
       Health Service  Publ.  No. 956, Washington,  D. C., 1962.

 22    Hajas, L. Purification of Land Applied Sewage Within the Ground Water .
       M.S. Thesis, Rensselaer Polytechnic Institute, Troy, New York, December
       1975. 242pp.

 23 .    Chiaro, P. S.  Hydraulic and Water Quality  Aspects of a Municipal Land
       Application Wastewater Treatment System.  M.E. Project, Rensselaer
       Polytechnic Institute, Troy, New York, December 1976.  169 pp.

 24.   Vollenweider, R. A.  Scientific Fundamentals of the Eutrophication of
       Lakes and Flowing Waters, with Particular Reference to Nitrogen and
       Phosphorus  as  Factors in  Eutrophication.  Organization for Economic
       Cooperation and Development,  Paris, France, 1971.   159pp.
                                      144

-------
 5.    Eliassen, R. et al.  Studies on the Movement of Viruses with Ground-
      water.  Water Quality Research Laboratory, Stanford University, Palo
      Alto, California, 1967.  137 pp.

 6.    Robeck, G. G.  Microbial Problems in Groundwater,  Ground Water 7(3):
      33-35, May-June 1969.

 7.    Aulenbach,  L.  M.  Analysis of the Lake George Sewage Treatment Plant.
      Wheaton College, Norton, Massachusetts,  January 1975.  19pp. (Unpublished)

 8.    Wastewater Management by Disposal on the Land. S. C. Reed,  Coordinator,
      Special Report 171, Corps of Engineers, U. S. Army, CRREL, Hanover,
     New Hampshire, May 1972. 183 pp.

:9.   Glavin,  T. P. and J. A. Romero-Rojas.   Nutrient Removal Through
     Natural  Sand Beds. M. E. Thesis, Bio-Environmental Engineering Div.,
     Rensselaer Polytechnic Institute, Troy,  New York, August 1968.  106 pp.
                                 145

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 1. REPORT NO.
  EPA-600/2-79-068
                3. RECIPIENT'S ACCESSION-NO.
 4. TITLE AND SUBTITLE

  LONG TERM RECHARGE  OF  TRICKLING FILTER EFFLUENT
  INTO SAND
                5 REPORT DATE
                  March 1979
               6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)
  Donald B. Aulenbach
               8. PERFORMING ORGANIZATION REPORT NO
  Rensselaer Polytechnic  Institute
  Troy, New York  12181
                                                               IOGRAM ELEMENT NO
                                                                1BC822
                1. CONTRACT/GRANT NO.

                 R803452
  Robert  S.  Kerr Environmental Research Lab.
  Office  of  Research and Development
  U.S.  Environmental Protection Agency
  Ada,  Oklahoma  74820	
   - Ada, OK
              "-0 oE/3OVERED
                  EPA/600/15
  The rapid infiltration  of trickling filter effluent onto natural  delta sand beds at
  the Lake George Village Sewage Treatment Plant has been shown  to  produce the
  equivalent of tertiary  treatment to the domestic wastewater  since 1939 with no
  indication of exhaustion of the purification capacity.  Most of the  purification
  took place in the  top 10 m of the sand.  BOD, COD, alkylbenzenesulfonates,  total
  and fecal coliforms, and phosphates were essentially completely removed in  the
  sand system.  Ammonia and organic nitrogen were converted  to nitrates,  some of
  which were removed under reducing conditions.  Vertical velocities in'the sand we
  measured between 0.6 m/day and 2.5 m/day.  Horizontal velocities  varied between  ^
  3 m/day and 12 m/day.   Allowing weeds to grow on the sand  beds  increased the infil
  tration rate when  the depth of liquid on the bed exceeded  0.3  m,  but decreased the"
  infiltration rate  at shallower depths.  A rapid sand infiltration system is
  recommended as a suitable means of providing tertiary treatment to domestic
  wastewater.
 7.
                  DESCRIPTORS
                                KEY WORDS AND DOCUMENT ANALYSIS
  Groundwater
  Purification
  Quality control
  Sewage  treatment
 3. DISTRIBUTION STATEMENT


   RELEASE  TO  PUBLIC
EPA Form 2220-1 (9-73)
  ^IDENTIFIERS/OPEN ENDED TERMS
     Land application
     High rate infiltration
     Tertiary  treatment
     Sewage tracing in
        groundwater
  19. SECURITY CLASS (This Report)
      UNCLASSIFIED
  20. SECURITY
      ICURITY CLASS (Th
      UNCLASSIFIED
This page)
146
                                                                        c. poSATl Reid/Group
                 68D
                 48B,  E,  G
                                                                        22. PRICE

-------