&EPA
             United States
             Environmental Protection
             Agency
            Robert S. Kerr Environmental Research EPA-600/2-79-174
            Laboratory           August 1979
            Ada OK 74820
             Research and Development
Treatment of
Secondary Effluent by
Infiltration-Percolation

-------
                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 for 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-600/2-79-174
                                              August 1979
             TREATMENT OF SECONDARY
      EFFLUENT BY INFILTRATION-PERCOLATION
                       by

                   D.G.  Smith
                  K.D.  Linstedt
                  E.R.   Bennett

           City of Boulder,  Colorado
                      and
            University of Colorado
           Boulder, Colorado  80309
              Grant No.   R803931
                 Project Officer

                  Lowell 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 endorse-
ment or recommendation for use.
                                    11

-------
                               FOREWORD


     The Environmental  Protection Agency was established to coordinate
administration of the major Federal  programs designed to protect the
quality of our environment.

     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 tech-
nologies for irrigation return flows; (d) develop and demonstrate pollution
control technologies for animal  production wastes; (e) develop and demon-
strate technologies to prevent, control or abate pollution from the petro-
leum refining and petrochemical  industries; and (f) develop and demonstrate
technologies to manage pollution resulting from combinations of industrial
wastewaters or industrial /municipal  wastewaters.

     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.
                    William C. Galegar       \J
                    Director
                    Robert S. Kerr Environmental Research Laboratory
                                    iii

-------
                                  ABSTRACT


      The objective of this study was to evaluate the performance of an
Infiltration-percolation system for Improving the quality of secondary
effluent from a municipal wastewater treatment system.  This was done by
constructing three Infiltration-percolation basins and monitoring their
Influent and effluent quality over a two year period.

      The facility consisted of three Infiltration-percolation basins of
sizes ranging between 0.24 hectares (ha) [0.6 acres(ac)] and 0.36 ha (0.9 ac).
Unchlorlnated secondary wastewater effluent was applied twice a week to each
basin at loading rates which varied between 12.2 meters/year (m/yr) [40 feet/
year (ft/yr)] and 48.8  m/yr (160 ft/yr).  The wastewater percolated through
loamy and clay sands covering alluvial  sand and gravel.  The percolate was
collected 2.4-3.0 m (8-10 ft) below the surface by underdralns for discharge.

      Analyses of the basin Influent and effluent collected from the under-
dralns Indicated that the systems were  generally effective 1n reducing the
wastewater concentrations of COD, conform organisms, and ammonium nitrogen.
Phosphorus leakage occurred to some extent 1n each of the basins, with the
most heavily loaded basins yielding the highest phosphorus concentrations 1n
the discharge water.  The nitrate concentration of the water Increased
significantly because of nitrification  of the ammonium nitrogen retained
within the soil matrix.  The concentrations of hardness, alkalinity, and
chlorides also showed significant Increases 1n the percolate water.

      This report was submitted 1n fulfillment of Grant No. R803931 by the
City of Boulder, Colorado, and the University of Colorado, under the sponsor-
ship of the U. S. Environmental Protection Agency.  The report covers the
period August 1, 1975, to June 30, 1978, and the work was completed as of
October 15, 1978.
                                      1v

-------
                               CONTENTS

Foreword	ill
Abstract	iv
Figures	vi
Tables	ix
Acknowledgements 	    x
   1.  Introduction  	  1
   2.  Conclusions 	  2
   3.  Recommendations 	  3
   4.  Experimental System and Procedures  	  4
            Site Description and Physical Facility 	  4
            Basin Modification 	  7
            Basin Loading	7
   5.  Sampling and Analysis	13
            Sampling	13
            Analytical Methods	13
   6.  Hydraulic Characteristics of the Site	21
            Ground Water Profile	21
            Basin Hydraulics	21
            Infiltrometer Studies	32
            Factors Affecting Infiltration Rates	41
   7.  Treatment Performance of the Infiltration-Percolation
       System	51
            General Considerations	51
            Phosphorus Behavior	51
            Refractory Organics	57
            Nitrogen	57
            Dissolved Salt Species	65
            Coliform Organisms	70
            Heavy Metals	76
   8.  Column Studies	83
            Column Operation	83
            Column Performance	84
                 Chemical Oxygen Demand	84
                 Phosphorus	86
                 Nitrogen	86
                 Dissolved Solids	89
                 Coliforms	98
                 Heavy Metals	99

References	101

-------
                                  FIGURES
Number
   1   Schematic of Boulder wastewater treatment plant 	    5
   2   Infiltration-percolation system layout  	    6
   3   Basin underdrains  	    8
   4   Typical  section through basins   	    9
   5   Well  point locations	   10
   6   Modified basin configuration   	   11
   7   Analytical  procedure for heavy  metals  	   19
   8   Ground water contours and flow  directions (May,  1975)  	   22
   9   Ground water elevations surrounding  basins  (August,  1976)  ...   23
  10   Ground water mounding pattern within basins  	   24
  11   Typical  discharge  profile for Basin  1  	   27
  12   Typical  discharge  profile for Basin  2  prior  to
         modification	   28
  13   Typical  discharge  profile for Basin  3  prior  to
         modification  	   29
  14   Normalized  discharge profiles prior  to modification  	   30
  15   Peak  discharge flows prior to basin  modification   	   31
  16   Infiltrometer test locations  	   33
  17   Infiltrometer test results  for Basin 1  	   36
  18   Infiltrometer test data on  Basin 2 prior to  modification   ...   37
  19    Infiltrometer test data on  Basin 3 prior to  modification   ...   39
  20   Pattern  of  infiltration rate decline in  Basin 1	43
                                   vi

-------
                            FIGURES (continued)
Number                                                                 Page
  21   Pattern of infiltration rate decline in Basin 2	44
  22   Pattern of infiltration rate decline in Basin 3	   45
  23   Infiltration rate as a function of suspended solids
         loading in Basin 1	   46
  24   Infiltration rate as a function of suspended solids
         loading in Basin 2	   47
  25   Infiltration rate as a function of suspended solids
         loading in Basin 3	48
  26   Infiltration rate as a function of wastewater temperature ...   50
  27   Weekly precipitation and minimum temperature of renovated
         water	   53
  28   Effluent phosphorus as a function of time	54
  29   Effluent phosphorus variation within loading cycles 	   56
  30   Effluent COD as a function of time	58
  31   Effluent ammonium nitrogen as a function of time   	   60
  32   Effluent nitrate nitrogen as a function of time 	   61
  33   Cumulative nitrogen applied and discharged from Basin 1  ....   62
  34   Suspended solids variation in applied wastewater   	   64
  35   Nitrate variation in effluent within loading cycles 	   66
  36   Effluent ammonium variations within loading cycles  	   67
  37   Effluent hardness as a function of time	68
  38   Change in hardness concentration through the basins 	   69
  39   Change in calcium concentration through the basins  	   71
  40   Change in magnesium concentration through the basins  	   72
  41   Change in alkalinity through the basins 	   73
  42   Change in chloride concentration through the basins 	   74
                                    vii

-------
                            FIGURES (continued)
Number                                                                 Page
  43   Wastewater cadmium concentrations 	   77
  44   Wastewater copper concentrations  	   78
  45   Wastewater chromium concentrations  	   79
  46   Wastewater nickel  concentrations  	   80
  47   Wastewater lead concentrations  	   81
  48   Wastewater zinc concentrations  	   82
  49   COD concentration  variation  with column depth and time  ....   85
  50   Wastewater phosphorus  concentration as a function of
         soil  depth	87
  51   Nitrate concentration  as  a function of column depth
         and sampling time	90
  52   Column underdrain  flow as a  function of time after
         loading	91
  53   Mass flow of nitrogen  from column as a function of
         time after loading	92
  54   Wastewater calcium concentration as a function of
         column depth  	   93
  55   Wastewater magnesium concentration as a function of
         column depth  	   94
  56   Wastewater alkalinity  as  a function of column depth 	   95
  57   Wastewater hardness as a  function of column depth 	   96
  58   Wastewater chloride concentration as a function of
         column depth  	   97
  59   Wastewater heavy metal concentration variations
         with depth	100
                                   viii

-------
                                    TABLES
Number                                                                      Page
  1   Loading and Sampling Schedule (October, 1976 to January, 1977)  ...    14
  2   Loading and Sampling Schedule (February, 1977 to June, 1978)  ....    16
  3   Ground Water Elevations  	    25
  4   Base Flows	    26
  5   Summary of Infiltrometer Test Results  	    35
  6   Summary of Infiltration Rates  	    40
  7   Seasonal Averages for Wastewater Constituents in
        Boulder Secondary Effluent 	    52
  8   Coliform Removal Data  	    75
  9   Selected Nitrogen Removal Data 	    88
  10   Coliform Removal with Depth  	    99

-------
                              ACKNOWLEDGEMENTS

     The authors wish to thank the many people who contributed to the
completion of this research effort.  The support of the City of Boulder waste-
water treatment plant staff is gratefully acknowledged.  Mr. Arthur Dike,
Mr. Charles Tregay, and Mr. Alton Ragsdale were particularly helpful in the
operation and maintenance of the test facility.  Similarly, Mr. Chris Rudkin
and Mr. Steve Miller provided much information and assistance in the develop-
ment of sampling procedures at the site.

     Sampling, analysis, and operational coordination were provided by the
following individuals during their graduate training at the University of
Colorado:  Mr. Bob White, Mr.  Paul  Hamilton, Mr. Dan St. John, Mr. Bill
DeOreo, Mr.  Bill Earley, Mr. Joe Tamburini, and Mr. Paul King.

     The authors wish to express appreciation to the Robert S. Kerr Environ-
mental  Research Laboratory for its  support of this project, and especially to
Mr. Lowell  Leach, and Mr. Richard Thomas who served as Project Officers
during portions of the study.   The  assistance of Mr. Roger Dean, Region  VIII,
EPA, in providing coordination is also acknowledged.

-------
                                  SECTION 1

                                INTRODUCTION

     In 1972 the United States Congress passed Public Law 92-500,  a  bill
which committed this country to the task of upgrading polluted waters  and
preventing degradation of clean waters.  Three levels of treatment were
identified as the primary goals of this act.  By 1977, secondary treatment
or its equivalent would be required; by 1983, the best available treatment
would be required; and by 1985, the goal was to eliminate the discharge of
pollutants to our waterways.  In order to comply with the strict standards
set forth in this law, communities and industries alike were faced with pro-
viding some form of highly efficient wastewater treatment.  This treatment
would be directed at removing undesirable wastewater constituents  such as
nitrogen, phosphorus, suspended solids, refractory organics, and microorgan-
isms.  Their removal could be accomplished through application of either
in-plant physical, chemical, and biological processes, or by land  treatment.

     In addition to establishing strict discharge requirements, P.L. 92-500
also specified that land treatment be considered as an alternative to  other
advanced treatment processes.  With the implementation of this requirement,
it has become apparent that land treatment combines the various physical-
chemical, and biological processes into a single process which is  very
effective in many wastewater treatment situations.  Land treatment is  espe-
cially attractive for smaller communities where the land is readily  available
and the cost effectiveness of physical-chemical processes is somewhat
questionable.

     Three major methods of land application have been demonstrated  to pro-
vide effective wastewater treatment.  They include:  irrigation, overland
flow, and infiltration-percolation.  Successful application of one of  these
land treatment methods depends on several interrelated factors such  as the
soil type, geology, topography, ground water characteristics, and climate.
As a result of these considerations, land treatment processes are very site
specific.  Selection of the best land use system for a particular location
may depend on any one, or all of these considerations.

     The study discussed in this report involved the operation of three small
infiltration-percolation (i-p) basins in Boulder, Colorado.  These infiltra-
tion-percolation basins were installed as a demonstration project in the
spring of 1976 to treat a portion of the secondary effluent discharged by  the
City of Boulder wastewater treatment facility.  The treatment efficiency of
the infiltration-percolation system was monitored throughout the period of
operation by regular sampling of the applied secondary effluent and  the
renovated water.  The performance of this system was the subject of  this
report.

-------
                                  SECTION 2

                                 CONCLUSIONS

     The infiltration-percolation system demonstrated excellent capability
for polishing secondary effluent in the context of the Boulder, Colorado
wastewater treatment situation.  Under proper conditions of hydraulic loading
and loading cycle control, the system was capable of providing very high
levels of removal for virtually all wastewater constituents of major pollu-
tion significance.

     Infiltration-percolation systems were shown to be capable of operation
throughout the year in either flooded basins or ridge and furrow configura-
tions where appropriate surface maintenance was provided.  The necessary
surface maintenance consisted of scarification to control plant development
in the bare basins so that ice which formed in the cold winter months would
not be anchored, but could float free with successive wastewater applications.

     The infiltration rate in the ridge and furrow system appeared to be less
affected by wastewater suspended solids deposition than the flat basins which
were completely flooded.  This suggests that it would be possible to operate
ridge and furrow systems longer between extended drying periods than flooded
basins loaded at comparable rates.

     When a basin was loaded at a sufficiently high hydraulic rate to stress
the infiltration capabilities of the soil system, one of the indications of
the stressed condition was a deterioration of the quality of the effluent
from the basin.  This quality deterioration was evidenced by increased con-
centrations of both phosphorus and ammonium nitrogen, as well as a reduction
in the amount of nitrate nitrogen in the basin discharge.

     The basins demonstrated good capability for making major reductions in
the concentration of fecal coliforms in the unchlorinated secondary effluent.
Nevertheless, significant concentrations of these bacteria were still
detected after passage of the wastewater through 2.4-3.0 meters (8-10 feet)
of the composite soil mixture found in the Boulder treatment location.

     Significant removal of heavy metals was observed in the infiltration-
percolation soil system.  Most of this removal occurred in the tight clay
loam layer at the top surface of the soil profile.

-------
                                  SECTION  3

                               RECOMMENDATIONS

     From the observations developed in conjunction  with  this  investigation,
the following recommendations seem appropriate:

         The capability of the infiltration-percolation basins for
     treating a wide range of wastewater pollutants  suggest that these
     systems would be appropriate for treating  primary as well as
     secondary effluent.  It is recommended that a comparative study
     be performed which would assess the acceptable  hydraulic  loading
     rates and product water characteristics with these two different
     levels of pre-application treatment.

         From the aesthetic and operational perspectives, it would
     appear that there might be advantages to the use of a ridge and
     furrowed application system rather than a  flat basin having no
     vegetation.  A study should be undertaken  to identify the loading
     schedule which would provide the best combination of hydraulic
     loading and treatment performance in this  configuration.   The  aim
     of this effort should be to minimize the total  land area  require-
     ments for providing this alternative application method.

         In the study discussed in this report, little total nitrogen
     removal was observed.  It was shown in the operation of beds 1
     and 2 that if the loading rate was maintained at an appropriate
     level, the effluent ammonium nitrogen could be controlled at a
     concentration of less than 1 mi 111gram/liter (mg/1).  However, with the
     loading sequence practiced 1n this Investigation, essentially  all  of the
     nitrogen was discharged from these beds as nitrate.   Other Investi-
     gators have demonstrated that substantial  denitriflcatlon can  be
     achieved In rapid infiltration systems if the loading sequence Is
     appropriately managed.   Modification of the loading sequence  should
     be attempted at the Boulder facility for the purpose of maximizing
     the total nitrogen removal.

         Significant phosphorus leakage was observed  in  the product water
     from  all three of  the basins.  It is  suspected that this leakage
     occurred in part because of  the short hydraulic  detention from the
     basin surface to the underdrains.  It would be of interest  to
     determine  the impact of  varying the  underdrain placement to increase
     the contact time between the wastewater and the  soil  profile.   This
     could be done by selectively closing  off some of the  existing drains
     in the  Boulder  infiltration-percolation system.

-------
                                  SECTION  4

                      EXPERIMENTAL SYSTEM  AND PROCEDURES

 SITE  DESCRIPTION AND  PHYSICAL  FACILITY

      The wastewater treated  in the demonstration project facility was drawn
 from  the effluent of  the City  of  Boulder, 75th Street Wastewater Treatment
 Plant.  This  treatment facility processed approximately 0.53 cubic meter/
 second  (nr/sec) [12 million  gallons per day (MGD)] of wastewater by means of
 a  standard  rate trickling filter,  as 1s shown schematically in Figure 1.

      The site of the  infiltration-percolation system consisted of approxi-
 mately  1.0  hectares (ha) (2.5  ac)  of land located adjacent to the City of Boulder
 75th  Street Wastewater Treatment  Plant, and about 152 meters (500 feet)
 south of Boulder Creek.  The land, which was formerly rangeland, had a slight
 eastward slope ranging between 0-1%.  The soil was classified as the Niwot
 series  soil which has been described by the U. S. Soil Conservation Service
 as a  layer  varying from sandy clay loam to light clay loam superimposed
 over  sand and gravel  (1).  These  soils have been described to have moderate
 permeability  and a high seasonal water table ranging from 0.15 meters to
 0.46  meters (6 to 8 1nches)(l).  Underlying this soil at a depth of approxi-
 mately  3.7  meters (12 feet)  was the very  impermeable Pierre Shale formation
 (2).  The depth to ground water on the site was found to vary from 0.9 to
 1.5 meters  (3 to 5 feet) In  May,  1975(2).

      The City of Boulder Wastewater Utility Department designed and
 constructed the pilot plant  facility.  Construction began 1n December 1975
 and was completed in April  1976.   Three Infiltration-percolation basins were
 constructed,  with each basin separated by a berm approximately 0.76 meters
 (2.5  feet)  high.  All  three  basins were surrounded by an impermeable clay-
 core  dike which was also 0.76 meters (2.5 feet) high.  The 1.83 meter (6 feet)
 wide  clay-core dike extended from  the ground surface to the impermeable
 bedrock, and  served to completely  enclose the system, and minimize the
 Interaction between the beds and the surrounding ground water.   The South
 basin, referred to as Basin  1, was 0.35 ha(0.87 ac); the Middle basin, Basin
 2, was 0.24 ha (0.60 ac); and the  North basin, Basin 3 was 0.26 ha (0.65 ac).
 Figure 2 shows a plan view of the  three basins and their respective sizes.

     After the basin area was completely sealed, an underdrain system was
 installed to  lower the existing water table and to collect the applied water
during the operation of the system.  The underdrain system for each basin
consisted of  two 0.18 meter  (7 Inch) perforated PVC drain pipes located

-------
            Headworks
           Flow
        Diversion
           Box
                                 Primary      Trickling      Secondary
                                 Clarifiers   Filters        Clarifiers  Chlorination
   City
Collection
  System
   \
    \
     \
      \
      \
   Grit to
Land Disposal
    Site
                                     
-------
                               Clay Dike
Distribution
Box
In-Line Flow
Meter
  Secondary
  Clarifier
                                            BASIN  1 - X'ov
                                                     0.87 ac
No Scale
                                                                                          ump House
                                                          Clay Dike

                           Figure 2.  Infiltration-Percolation System Layout

-------
2.43 to 3.05 meters (8 to 10 feet) under the basin surface.  The collected
water flowed by gravity to a manhole in each basin, and then to a central
manhole where the monitoring and sampling took place.  From this manhole the
water flowed to a wet well and was pumped to its discharge point in Boulder
Creek.  A plan view showing the underdrain system is shown in Figure 3, and a
cross section of the basins is shown in Figure 4.

     Well  points were installed in each basin to measure the rise and fall of
the water table during the loading cycles.  These well points were developed
by augering holes through the beds, and installing 0.04 meter (1.5 Inches) PVC
pipes in the holes.  These pipes extended from 0.91 meters (3 feet) above the
ground surface to the bedrock.  The locations and elevations of these well
points have been indicated in Figure 5.  The depth of the water table was
measured by lowering a float down the pipe until the water surface was inter-
cepted, and then recording the depth.


BASIN MODIFICATION

     Following six months of treatment operation on the beds which have been
described, a decision was made to modify Beds 2 and 3 in an attempt to
improve their hydraulic performance.  As such, in January, 1977 the top tight
loamy layer of soil was removed from both of these beds.  This necessitated
the removal of the top 0.46 meters (18 inches) of soil from Bed 3, and the
top 0.61 meters (24 inches) of soil from Bed 2.  Both beds were subsequently
graded to a flat surface with a slight,eastward slope, similar to that in
their initial condition.

     While Basin 3 remained in the modified form throughout the rest of the
study, Bed 2 was further altered in March, 1977 by the construction of a
ridge and furrow system.  The furrows were 0.46 meters (18 inches) deep and
spaced approximately 1.98 meters (6.5 feet)on center.  The furrows averaged
1.02 meters (40 inches) in width, and the ridges averaged  0.91 meters  (36
inches) wide.  There were 9 furrows constructed in Bed 2,  each approximately
85 meters (280 feet) long.  A furrow at each end was  constructed to facili-
tate loading.  A plan view of the three basins, and  the modified areas, is
shown in Figure 6.  It should be noted that Bed  1 was not  altered  in any way,
after the initial construction.
BASIN LOADING

     Loading of the basins was accomplished by  pumping  the effluent  from a
secondary clarifier with a centrifugal  pump which was powered by a portable
gasoline engine.  The  unchlorinated wastewater  from  this  point was pumped
through approximately  297 meters  (975  feet) of  0.36  meter (14 inch)  PVC pipe
to a distribution box  which  directed the water  to the desired basin.

     Each basin was loaded twice  a week at lh day intervals. From October,
1976 to January,  1977, Bed 1 (surface  of  basin) was  loaded at approximately
8:00am  on Mondays and  7:00pm on  Thursdays; Bed  2, 8:00am  on  Tuesdays and  7:00pm

-------
00
               5115.36
                                                                                        Discharge to
                                                                                        Boulder Creek
                                             0.40% slope             ~~ "~
                                                                                         I          No Scale
                                                                                          5101.49
                  0.40% slope
                                         5104.81

                 0.40%  slope             """
     178 mm(7 in)perforated  PVC underdrain
J_	0.40%  slope	
                                                                                           Gnd
                                     Figure 3.  Basin underdrains.
                                                (Elevations in feet.)

-------
                      Standpipe Well
                                                     Clay Dike
Loam to
Clay Loam
 Sand and
 Gravel
                         .Underdrafn -~)
'.   • N  •    .unoerarain —n
   Pierre
   Shale
                                                   Manhole
.X V
                        Figure 4.  Typical section through basins.

-------
        5117.66
£     N.E. Corner
    Distribution
       Box
                              5115.88
   5114.78  /*


5116.43
                                                                                         No Scale

                                                                                  1114.36    |
            N. Rim M.H.

                   ^5114.02
                  B.M.  Well C-19X
                  (CH2M-Hill,      »
                        1975)
                                                                                        Pump House
                                                                                        5114.37
                                                                                    B.M. Well C-20~

                  Figure 5.  Well point locations (.)  and elevations  (feet).        CH2"-H1". 1975)

-------
                            Clay Dike
Distribution Box
    In-Line
    Flow Meter
 Secondary
 Clarifier
                                                                                             No Scale
                                                                                     A	Pump House
                                   Figure 6.   Modified basin configuration.

-------
on Fridays; Bed 3, 8:00am on Wednesdays and 7:00pm on Saturdays.  The beds
were loaded on alternating days to facilitate the ease of operation, sampling,
and analysis.  In February 1977, the loading cycle was shifted one day later
to begin on Tuesday morning instead of Monday morning due to a change in
operator schedules.  After six weeks of loading, all basins were dried and
then scarified.  The required length of drying time varied from one week in
the summer to two or three weeks In the winter, depending on the weather
conditions during the drying period.
                                     12

-------
                                  SECTION  5

                            SAMPLING AND ANALYSIS
SAMPLING
     Samples of both the influent and effluent of the  three  basins were  taken
at regular intervals.  The influent samples  were taken at the  inlet  pipe of
each bed during loading.  Following renovation treatment, the  basin  effluent
was sampled at the center manhole, with samples collected independently  for
each basin at specified times after loading.  Tables  1 and 2 summarize the
loading and sampling schedule.   Only the first loading of each week  was
sampled and monitored.   The temperature, dissolved oxygen level,  nitrogen,
hydraulic flow, and infiltration rate measurements were monitored with time
after the loadings.  All other parameters were measured only once for each
loading cycle on the sample collected 24 hours after  loading.

     The water samples  were preserved immediately after collection.  Samples
collected for testing of nitrites and nitrates were preserved  with 40 mgA
HgCl2, and stored at 4°C.  Samples to be analyzed for organic  nitrogen and
ammonia were acidified  with concentrated sulfuric acid to a  pH of less than
2, and refrigerated at  4°C.  These procedures were in accordance  with those
outlined in Methods for Chemical Analysis of Water and Wastes  (3).   Samples
collected for all of the tests were refrigerated at 4°C, with  the exception
of those used in determining temperature and dissolved oxygen.


ANALYTICAL METHODS

     All analyses, except those noted, were performed at the Sanitary
Engineering Laboratory at the University of Colorado.   The samples were  sub-
jected to the following series of analyses for purposes of this investiga-
tion:  total solids, suspended solids, phosphorus, COD, temperature, coli-
forms, and the nitrogen series.  In addition, selected samples were  analyzed
for Cd, Cu, Cr, Ni, Pb, and Zn during November-December, 1977.

     Temperature measurements were taken immediately  following the sampling
at the pilot plant site.  A mercury-filled centigrade thermometer, calibrated
to 1°C, was used for these measurements.

     Total kjeldahl nitrogen (TKN) was determined by  acid digestion  of 250
mill niters (ml) samples and distillation Into boric  add.  The TKN  was  titrated
with standard sulfuric add to the pH of the blank carried through the same
procedures.  The difference between the TKN and ammonium nitrogen was the organic


                                      13

-------
                  TABLE 1.  LOADING AND SAMPLING SCHEDULE

                       (OCTOBER 1976 TO JANUARY 1977)
   Day
Time
Basin 1
Basin 2
Basin 3
Monday     8:00 a.m.  Load Basin
                      Measure Flow
                      Sample Effluent

           9:00 a.m.  Measure Flow
                      Sample Influent
                      Sample Effluent

          11:00 a.m.  Measure Flow
                      Sample Effluent
           2:00 p.m.  Measure Flow
                      Sample Effluent

           8:00 p.m.  Measure Flow
                      Sample Effluent
Tuesday    8:00 a.m.
           9:00 a.m.
         Measure Flow
         Sample for
         Complete Analysis
          11:00 a.m.


           2:00 p.m.


           8:00 p.m.  Measure Flow
                      Sample Effluent

Wednesday  8:00 a.m.  Measure Flow
                      Sample Effluent


           9:00 a.m.
          11:00 a.m.


           2:00 p.m.
                Load Basin
                Measure Flow
                Sample Effluent

                Measure Flow
                Sample Influent
                Sample Effluent
                Measure Flow
                Sample Effluent
                Measure Flow
                Sample Effluent
                Measure Flow
                Sample Effluent

                Measure Flow
                Sample for
                Complete Analysis
                                               Load Basin
                                               Measure Flow
                                               Sample Effluent

                                               Measure Flow
                                               Sample Influent
                                               Sample Effluent

                                               Measure Flow
                                               Sample Effluent
                                               Measure Flow
                                               Sample Effluent
                                                              (continued)
                                     14

-------
                            TABLE 1.  (continued)
   Day
Time
Basin 1
Basin 2
Basin 3
           8:00 p.m.  Measure Flow
                      Sample Effluent

Thursday   8:00 a.m.



           7:00 p.m.  Load Basin
           8:00 p.m.


Friday     8:00 a.m.


           7:00 p.m.

           8:00 p.m.


Saturday   7:00 p.m.
                            Measure Flow
                            Sample Effluent

                            Measure Flow
                            Sample Effluent
                            Measure Flow
                            Sample Effluent
                            Load Basin
                                   Measure Flow
                                   Sample Effluent

                                   Measure Flow
                                   Sample for
                                   Complete Analysis
                                   Measure Flow
                                   Sample Effluent

                                   Measure Flow
                                   Sample Effluent
                                               Measure  Flow
                                               Sample Effluent

                                               Load  Basin
                                      15

-------
                  TABLE 2.  LOADING AND SAMPLING SCHEDULE

                       (FEBRUARY 1977 TO JUNE 1978)
   Day
Time
Basin 1
Basin 2
Basin 3
Tuesday    8:00 a.m.



           9:00 a.m.



           12 Noon


           4:00 p.m.


           8:00 p.m.


Wednesday  8:00 a.m.



           9:00 a.m.



           12 Noon


           4:00 p.m.


           8:00 p.m.


Thursday   8:00 a.m.



           9:00 a.m.



           12 Noon


           4:00 p.m.
         Load Basin
         Measure Flow
         Sample Effluent

         Measure Flow
         Sample Influent
         Sample Effluent

         Measure Flow
         Sample Effluent

         Measure Flow
         Sample Effluent
         Measure Flow
         Sample Effluent

         Measure Flow
         Sample for
         Complete Analysis
         Measure Flow
         Sample Effluent

         Measure Flow
         Sample Effluent
                Load Basin
                Measure Flow
                Sample Effluent

                Measure Flow
                Sample Influent
                Sample Effluent

                Measure Flow
                Sample Effluent

                Measure Flow
                Sample Effluent

                Measure Flow
                Sample Effluent

                Measure Flow
                Sample for
                Complete Analysis
                Load Basin
                Measure Flow
                Sample Effluent

                Measure Flow
                Sample Influent
                Sample Effluent

                Measure Flow
                Sample Effluent
                Measure Flow
                Sample Effluent
                                                              (continued)
                                     16

-------
                           TABLE 2.  (continued)
   Day
Time
Basin 1
Basin 2
Basin 3
           8:00 p.m.  Measure Flow
                      Sample Effluent

Friday     8:00 a.m.
           8:00 p.m.  Load Basin
Saturday   8:00 a.m.

           8:00 p.m.
Sunday     8:00 p.m.
                            Measure Flow
                            Sample Effluent

                            Measure Flow
                            Sample Effluent

                            Measure Flow
                            Sample Effluent
                            Load Basin
                                   Measure Flow
                                   Sample Effluent

                                   Measure Flow
                                   Sample for
                                   Complete Analysis
                                   Measure Flow
                                   Sample Effluent

                                   Measure Flow
                                   Sample Effluent
                                   Measure Flow
                                   Sample Effluent

                                   Load Basin
                                      17

-------
nitrogen.  The organic nitrogen procedure has been described in detail in
Section 135 of Standard Methods (4).

     Ammonium determinations were made with an Orion Research ammonium
electrode Model 95-10.  The instructions provided with the electrode were
followed, with a new calibration curve prepared for each group of samples.

     The nitrite determinations were initially made by following the method
outlined in Section 420 of the 14th Edition of Standard Methods (4).  How-
ever, after January, 1977 nitrite analyses were modified to facilitate the
use of a Technicon Auto Analyzer II.  The automated procedure was an adapta-
tion of the diazotization method outlined in Standard Methods (4).

     Similarly, nitrates were initially determined by the Brucine method
outlined in Standard Methods (4).  A Bausch and Lomb Spectronic 70 spectro-
photometer was used for transmittance readings with new standards run for
each set of nitrate samples.  After January, 1977 an automated procedure was
adapted using the Technicon Auto Analyzer II.  Nitrates were determined by
the automated copper-cadmium reduction method described in Section 605 of
Standard Methods (4).

     Total phosphorus determinations were made by first acidifying the
samples, and then digesting them by the Persulfate Digestion Method outlined
in Section 223C-III of Standard Methods (4).  Color development was accom-
plished by the Stannous Chloride method outlined in Section 223E of Standard
Methods (4).  The transmittance was measured with a Bausch and Lomb Spec-
tronic 70 spectrophotometer at a wave length of 690 nanometers (nm).

     Total and fecal coliforms were reported as coliforms per 100 ml, and
were determined by use of the membrane filter technique described in Section
408 of Standard Methods (4).

     Both the suspended solids and total dissolved solids were run using some
variations to the process as described in Standard Methods (4).  Suspended
solids were determined by the modified process described by Harada and his
co-workers (5).  Total solids were determined by the following procedure:
first, evaporating dishes were dried at 103°C for one hour, then dessicated
for one-half hour, and weighed for tare weight.  One hundred milliliters of
sample were then evaporated in the dishes on a hot plate.  Following this,
they were dessicated for one-half hour and weighed.

     The chemical oxygen demand (COD) analyses were determined by the method
outlined 1n Section 220 of Standard Methods (4).  The soluble COD was
determined by following the Indicated procedure after filtration of the sample
through a 450 nm mllUpore filter.

     During the loading cycle of November-December 1977, a special set of
heavy metal samples was taken from the secondary effluent entering the beds,
and the bed product water 24 hours after loading.  These samples were stored
in new polyethylene bottles which had been acid washed three times in 1:1
HN03, 1:1 HC1, and distilled water.  The samples were subsequently subjected
to the sample treatment Indicated in Figure 7.

                                     18

-------
    Secondary Eff.
    [Unchlorinated)
       I.P.Basin|-
             Unacidified Sample
                     from
               Basin Influent
                50 ml
50 ml
          i
       Filtered
     Within  2  Hours
         0.45  y
                    Reagents
                 1 ml,  1:1  HN03
                1 ml,  1:1 U3PQk
                 2 ml,  30%  H202
                   Digestion
                80°C for 45 min
        Dissolved
         Metals
              Unacidified Sample
                     from
                Basin Effluent
                                         50 ml
50 ml
Dist. Water
   Blank
        Total
       Metals
                  Zn by flame
                      AA
              Cd, Cu, Cr, Ni, Pb
                by flameless AA
Figure 7.  Analytical procedure for heavy metals.
                            19

-------
     As indicated in this figure, the concentrations of cadmium, copper,
chromium, nickel, and lead were determined by flameless atomic absorption.
Zinc was determined by flame atomic absorption since it was impossible to use
the flameless method for measurement of zinc in these samples because of the
very high sensitivity of the instrument and the excessive amounts of back-
ground zinc.  The use of flameless atomic absorption allowed concentration  of
the samples to be avoided, as well as the vigorous digestion necessary with
flame atomic absorption.
                                     20

-------
                                 SECTION 6

                    HYDRAULIC CHARACTERISTICS OF THE SITE
GROUND WATER PROFILE

     Prior to basin construction, the ground water characteristics of the
proposed location were measured and recorded.   From these measurements,  the
ground water table contour of Figure 8 was developed to define the conditions
in May, 1975.  From this figure it can be seen that the general  pattern  of
ground water flow was away from Boulder Creek and towards the north and  west
sides of Basin 3.  The flow was also directed towards the west end, and  the
west and south sides of Basins 2 and 1, respectively.  For comparison,
Figure 9 shows the ground water levels at selected sites surrounding the
basins on August 10, 1976, which was after several months of basin operation.
The water levels at that time were actually somewhat.1ower than those prior
to initiation of basin operation, reflecting the change in ground water  table
with the time of the year.

     Within the basins, the ground water elevations were lowered by the
underdrain svstem to aoproximately 2.4 meters (8 feet) from the surface
throughout the area of the infiltration-percolation system.  These ground
water elevations were monitored on a time profile basis during a typical week
of loading, with the elevations reported in Table 3.  Mounding occurred  to a
small degree, but serious ground water mounding was prevented by the under-
drain system.  The maximum mounding condition is indicated in Figure 10,
which shows a cross-section of the area from Boulder Creek to the
infiltration-percolation system.

     With the steep gradient of ground water elevations across the
peripheral clay dike, it was expected that some flow would occur.  As a
result, prior to application of any wastewater to the beds, the base ground
water flow discharged from each basin was measured.  These flows are indi-
cated in Table 4, and show that the two outside basins had significant flow
across the dike, while the middle basin received essentially no ground water.
Under the minimum loading condition on Bed 3, the ground water constituted
about 2Q% of the average underdrain flow following basin loading.  However,
with the higher loading following modification of the basins, the ground
water contribution to the underdrain flow was reduced to less than 5% of the
total underdrain flow in all the basins.


BASIN HYDRAULICS

     Following flooding of each basin, the wastewater percolated downward

                                     21

-------
PO
PO
                      Figure  8.   Ground water contours and flow directions (feet) (May, 1975)

-------
ro
to
                                            Well C-25
                                            Water - 5108.13
                                            Ground - 5112.58
                                                               Well  C-19
      - 5106.27
Ground - 5111.73
                                                                Well  C-20
                                                                Water - 5106.43
                                                                Ground.- 5112.54
0)
£
1/1
                     Figure 9.  Ground water  elevations  (feet)  surrounding basins
                               (August,  1976)

-------
ro
ELEV. 5115 ft.
(15.59.m)


         Clay


ELEV. 5110 ft.
(1557.4m)
                Ground
                Water
                Table-.
      ELEV. 5105 ft.
      (1555.9m)•
                 n-
Access Road
to Treatment
PI ant-7     BASIN  1
                                                BASIN 2
                                                       BASIN 3
                              Bed Surface
                                                                                        Boulder Creek
                                        Bottom of Furrow
                                                                       Ground Water.Table
                               •Ground Water
                             • •  Table
                                                                           Bedrock
                         Figure 10.   Ground water mounding pattern within  basins.

-------
                                                                      TABLE 3.   GROUND WATER ELEVATIONS (feet)
Bed and
Time After
Loading
Bed 1 - IP 0 hrs
19 8
1012
1024
1932
1948
1072
Bed 2 - 20 0 hrs
29 8
2924
2032
2948
Bed 3 - 39 0 hrs
30 8
3024
3972
Hell Point Location
1-1 1-2 1-3 2-1 2-2 2-3 2-4 3-1 3-2 3-3 3-4 3-5 3-7 3-8 A B C
Well Point Elevation
5116.27 5116.44 5116.16 5111.54 5115.02 5115.34 5115.14 5114.94 5115.30 5116.06 5116.43 5116.06 5115.78 5115.88 5114.78 5114.02 5114.37
5104.60 5105.69 5104.99 5103.54 5104.19 5103.52 5104.31 5103.19 5103.97 5104.39 5107.43 5104.14 5106.70 5108.21 5107.78 5106.77 5107,95
5104.52 510S.86 5104.99 5103.29 M04.10 5103.44 5104.14 5103.44 5104.02 5104.64 5107.43 5104.06 5105.89 5108.13 5107.78 5106.85 5107.70
5104.69 5106.44 5105.24 5103.37 5103.94 5103.44 5103.97 5103.44 ---------
5104.69 5106.44 5105.24 5103.37 5103.94 5103.44 5103.97 5103.44 5103.52 5104.56 5107.26 5104.14 5105.48 5107.88 5108.03 5107.10 b!08.20
5104.52 5106.44 5105.24 5105.96 5104.85 5104.35 5104.72 5103.44 5103.86 5104.39 5107.43 5104.06 5105.89 5108.38 5108.03 5107.02 5108.37
5104.77 5106.44 5105.24 5104.46 5104.27 5103.77 5104.22 5103.44 5103.97 5104.56 5107.51 5104.23 5105.73 5108.21 5107.86 5107.10 5108.04
b!04.81 5106.44 5105.33 5104.04 5104.35 5103.69 5104.31 5103.86 5103.94 510b.39 5107.76 5104.48 5106.31 5108.33 5108.03 5107.10 5108.20
5104.69 5106.44 5105.24 5103.37 5103.94 5103.44 5103.97 5103.44 5103.52 5104.56 5107.26 5104.14 5105.48 5107.88 5108.03 5107.10 5108.20
5104.52 5106.44 5105.24 5105.96 5104.35 5104.35 5104.72 5103.44 5103.86 5104.39 5107.43 5104.06 5105.89 5108.38 5108.03 5107.02 5108.37
5104.77 5106.44 5105.24 5104.46 5104.27 5103.77 5104.22 5103.44 5103.97 5104.56 5107.51 5104.23 5105.73 5108.21 5108.86 5107.10 510S.O-!
5104.77 5106.44 5105.33 5104.29 5104.02 5103.52 5104.22 5103.61 5104.19 5105.31 5107.43 5104.39 5106.39 5108.30 5107.70 5105.94 5108.37
5104.81 5106.44 5105.33 5104.04 5104.35 5103.69 5194.31 5103.86 5103.94 5105.39 5107.76 5104.48 5106.31 5108.33 5108.03 5107.10 5108.20
5104.77 5106.44 5105.24 5104.46 5104.27 5103.77 5104.22 5103.44 5103.97 5104.56 5107.51 5104.23 5105.73 5108.21 5107.86 5107.10 5108.04
5104.77 5106.44 5105.33 5104.29 5104.02 5103.52 5104.22 5103.61 5104.19 5105.31 5107.43 5104.39 5106.39 5108.30 5107.70 5106.94 5108.37
5104.81 5106.44 5105.33 5104.04 5104.35 5103.69 5104.31 5103.86 5103.94 5105.39 5107.76 5104.48 5106.31 6108.39 5108.03 5107.10 5108.20 |
5104.60 5105.61 5105.16 5105.04 5104.19 5103.52 5104.31 5103.87 5103.77 5104.73 5107.68 5104.27 5105.73 5108.21 5107.78 5107.02 5107.87 |
ro
                           ft x 0.305 > m

-------
                            TABLE  4.  BASE FLOWS


                                      Base Flow

                        Basin     (cfs)    (m3/sec)


                          1       0.013   3.68 x 10-1*

                          200

                          3       0.025   7.08 x lO'4*
through the soil, was collected by underdrains, and was pumped to the surface
and discharged into Boulder Creek.  With each wastewater application, the
discharge flow of Basin 1 exhibited a rapid increase to a peak within 12
hours after loading.  This was followed by a gradual decline in flow as is
shown by a representative discharge hydrograph in Figure 11.  Prior to basin
modification, a representative hydrograph for Basin 2 exhibited a rise and
fall similar to that shown in Figure 12.  While the peak occurred at about
the same time after loading, the curve was much broader, indicating a lower
percolation rate, or slower water mass flow through the soil.  Similarly,
Basin 3 had comparable hydraulic characteristics, but with a dampened peak
occurring at about 15 hours as shown in Figure 13.  To facilitate comparison
of the profiles for each bed, the flows from Basins 2 and 3 were normalized to
that of Basin 1 by multiplying their flows by the fraction of loading time
for Basin 1 divided by the loading time for each of the other basins.  This
was done to correct the hydrographs for the different amounts of water
applied to each basin.  With this normalization, the hydraulic dampening of
Basins 2 and 3 was more apparent, as can be seen in Figure 14.  The lower
infiltration rates which are suggested by these curves for Basins 2 and 3
were also indicated by ponding of the water in these basins for several days
after loading.  It can be speculated that this extended ponding time further
lowered the infiltration rates because of increased algae growth, and the
accompanying fouling of the surface with suspended solids.

     In addition to the differences in infiltration rates which have been
noted for the three basins, the infiltration rates of all the basins gradu-
ally declined during the first few months of operation.  This trend is
indicated in Figure 15, which is a plot of the peak flow of the underdrain
discharge during each of the first several loading cycles.  The reduced peak
flow rates were characteristic of similar reductions which occurred in the
acceptable application rates during the first months of operation.  The ini-
tial loading rates in May, 1976 were equivalent to 48.5 m/yr (159 ft/yr) on
Bed 1, 36.0 m/yr (118 ft/yr) on Bed 2, and 53.6 m/yr (176 ft/yr)  on Bed 3.
After two weeks of loading at these rates, ponding conditions developed on
all three beds.  As a result, all of the beds were allowed to dry for one
week and then scarified.  Subsequently, the loading rates were reduced to
27.4 m/yr (90 ft/yr), 12.2 m/yr (40 ft/yr) and 15.2 m/yr (50 ft/yr) for
beds 1, 2, and 3, respectively.  Bed 1 functioned acceptably at this load-
ing rate, and the rate was increased to 30.5 m/yr (100 ft/yr) in October,

                                     26

-------
ro
                                                                                                    -  25
                                                                                                    -  20  _
                                                                                                           ro
                                                                                                           O
                                                                                                            X
                                                                                                            o
                                                                                                       15   21
                                                                                                       U   (/)
                                                                                                           CO
                                                                                                            E
                                                                                                    -  10  ^
                                                                                                    -   5
                                                   36           48          60
                                            Time after Loading (hours)
                          Figure 11.  Typical  discharge hydrograph for Basin  1.

-------
0.4
 .3
q-
£> .2
o
         o
               12
24          36          48
     Time after Loading (hours)
                                                                   60
72
                                                                                            10.0
                                                                                             7.5
                                                                                             5.0
                                                                                            2.5
                                                                                                  m
                                                                                                  O
                                                                                                  X
                                                                                                  
-------
ro
IQ
        0.4
          .3
«*-
o
                                                                                                     10.0
                                                                                                      7.5
                                                                                                      5.0
                                                                                                           O

                                                                                                           0)
                                                                                                      2.5
                                                                           I
                                                                                  I
                        12
24
                                                                    60
72
                                36           48


                       Time after Loading (hours)



Figure 13.  Typical discharge  hydrograph  for Basin 3 prior to modification,
84

-------
in
to
(13
CO
-a

N


03


O
                                                                    0 Basin  1

                                                                     Basin  2

                                                                    0 Basin  3
                   12
                                24          36           48

                                      Time after  Loading (hours)
60
72
                                                                                               -   20
                                                                                               -   15
                                                                                               -   10
                                                                                               -    5
84
                                                                                                      CO

                                                                                                      O
                                                                                                       O
                                                                                                       d>
                                                                                                       c
                                                                                                       •r-
                                                                                                       VI

                                                                                                       CO
                                                                                                       •a
                                                                                                       a)
                                                                                                       N
                                                                                                       o
                Figure  14.   Normalized discharge  hydrographs prior to modification.

-------
to
M-
O
    1.0


     .9


     .8


     .7


     .6
U-    .5
(O
CD
D_
.4


.3


.2


.1


 0
                                                                  I       I        I

                                                               o Basin 1
                                                               a Basin 2

                                                               a Basin 3

                                                              =—  Drying Period
                             I
                                                          J_
                                                               I
           1      8       15     23     29

           I             August

Figure 15.  Peak discharge flows prior to basin modification.
              18     25

                July
                                                             5       12     19

                                                                   September
                                                                        25
                                                                        20
                                                                           ro
                                                                           O
                                                                                          15
                                                                                             tfl

                                                                                            CO
                                                                                          10 o
                                                                                             fO
                                                                                             cu
                                                                                            D_
                                                                                       26

-------
  1976 without causing any hydraulic problems.   However,  Beds  2  and  3  continued
  to demonstrate hydraulic difficulties,  and by December,  1976 their loading
  rates had been effectively reduced to 3.96 m/yr  (13  ft/yr) and 5.79  m/yr
  (19 ft/yr),  respectively.


  INFILTROMETER STUDIES

       It was  theorized that  the surface  layer  of  soil was restricting the
  wastewater infiltration  and, thus, was  responsible for the poor hydraulic
  performance  of Basins 2  and 3.  The Niwot  Series soil of the area  has been
  characterized by the Soil Conservation  Service as consisting of 0-0.30 meter
  (0-12  inch)  depth loams  to clay loams,  and  0.30-1.52 meter (12-60  inch)
  depth  coarse  sand (1).   If the surface  soils  of  Beds 2 and 3 were  indeed
  clay  loams,  it was suspected that  the theory  would be correct.   As a result
  infiltrometer tests were performed on all  three beds to determine-    (1) the*
  existing  infiltration rates on the surface,   (2)  the type and depth of the
  overlying soil layer, and  (3) the infiltration rate of the underlying soil
  and its soil   classification.  A double ring infiltrometer was used for per-
  formance of these tests.  The inner ring was 0.20 meters (8 inches) in
  diameter and  the outer ring was 0.38 meters (15 inches)  in diameter,  with
  the length of both at 0.36 meters  (14 inches).  The rings were  concentric,
  and were connected by metal  plates welded between the two rings.

      Test sites were  located within each of the basins which  had no
  significant surface disturbance, and which had soil  textures  representative
 of the area.   At these sites,  the  infiltrometer was  pressed into the  soil  to
 a depth of approximately 0.15  meters  (6  inches).   Installation  was  performed
 with care to  minimize the soil  disturbance around the cylinder.  The  area
 between the inner and outer  ring provided a buffer  pond which served  to  mini-
 mize the  radial  flow  of  water  away from  the inner ring.   This area  was  filled
 first with water and  kept at a  constant  level  throughout  the  testing.  The
 water used in the  testing was  unchlorinated secondary effluent  from the  75th
 Street Trickling Filter  Plant;  the same  wastewater which  was  applied  to  the
 basins.  After the outer ring was  filled,  the  inner ring  was  filled to a
 0.15 meter (6  inch) depth with  care taken  to minimize the disturbance of the
 soil  surface.   The water level  in  the inner ring  was  measured at the  start of
 testing and at time intervals ranging from  1 minute to 30 minutes,  depending
 upon the infiltration rate.  Sites  with  high infiltration rates were  tested
 several times, with the  last test  used to determine the average  infiltration
 rate.  The average infiltration rate was determined by dividing  the total
 drop in the water level by the corresponding elapsed  time.  These infiltra-
 tion rates were reported  in centimeters  per  hour.   The above procedure was
 patterned after similar procedures  described by Haise and Johnson (6,7).

     Infiltration tests were performed at each of the locations indicated by
a dot in Figure 16. The lettered dots represent sites tested prior to surface
modification of the beds, and the numbered dots represent sites tested after
some soil  was   removed  from the surface of Beds 2 and 3.   Surface infiltra-
tion tests were performed in Bed 1  at the locations identified by points A,
D, and F in Figure 16.  These tests yielded infiltration  rates ranging from
0.58-7.13 cm/hr (0.23-2.81 in/hr),  with an average of 4.42 cm/hr (1.74 in/hr).

                                     32

-------
GO
                                                                                      No Scale

                                                                                      -   I


                                                                                   Pump  House
                                                                                  LEGEND
                                                                                      Infiltration Test
                                                                                      Location

                                                                                      Soil Log Location
                            Figure 16.   Infiltrometer test locations.

-------
 Details  of  the  rates  at each  site  have been  included  in Table 5.  The soil
 within this basin was a silty loam which extended from the surface to
 0.38-0.66 meters  (15-26 inches) from  the surface.  The underlying soil was
 well  rounded gravel mixed with sand and silt-sand loam.   Infiltration tests
 at  a  0.38 meter (15 inch) depth at location  C and a 0.66  meter  (26 inch)
 depth at location E yielded rates  of  110 cm/hr  (43.5  in/hr) and 88 cm/hr
 (34.5 in/hr), respectively.   However, the presence of gravel at these depths
 prevented the desired 0.15 meter (6 inch) penetration of  the double ring
 infiltrometer,  which  may have contributed to the very high observed rates.
 A soil log  2.44 meters (8 feet) in depth revealed the same gravel-sand soil
 mixture  from a  0.51 meter (20 inch) depth to the invert of the underdrain
 system at approximately 2.44  meters (8 feet).   It can be  seen in Figure 17
 that  a significant increase in the measured  infiltration  rate was noted once
 the surface soil layer of silty loam was penetrated.  Because of the rela-
 tively homogeneous nature of  the soil below  this point, it was expected that
 the infiltration rate would become relatively constant at depths of greater
 than  0.61 meters (2 feet).  Based on the favorable infiltrometer results,
 and the  successful operation  at a  loading rate  of 30.5 m/yr (100 ft/yr), it
 was decided to  continue loading Bed 1 at 30.5 m/yr (100 ft/yr) with no
 physical alteration of the bed surface.

      Infiltration tests on the surface of Bed 2 at points A, D, and E
 yielded  infiltration  rates ranging from 0.10-1.41 cm/hr (0.04-0.16 in/hr),
 with  an  average rate  of only  0.23 cm/hr (0.09 in/hr).  The soil on the sur-
 face  of  this bed appeared to  be a silty loam with some heavy clay.  At
 depths of 0.46  meters (18 inches) and 0.76 meters (30 inches), the measured
 infiltration rates increased  to 0.89 cm/hr (0.35 in/hr) and 0.99 cm/hr
 (0.39 in/hr), respectively.   These measurements were made at points C and B.
 The soil texture at each location was similar to the silty-clay loam of the
 surface  soil, but also contained some sand.   At point H and a depth of 1.07
 meters (42  inches), the soil was a sandy loam,  and an infiltration rate of
 22.6  cm/hr  (8.89 in/hr) was observed.  A soil log on Bed 2 showed that
 gravel and  a sandy loam existed from approximately 0.91 meters (3 feet) to
 the invert  of the underdrain  system at about 2.44 meters  (8 feet).  A plot of
 the infiltration rate as a function of depth is shown in  Figure 18.  From
 this  figure  it  was determined that the top 0.91-1.07 meters (3-3.5 feet) of
 overburden were unsuitable for an infiltration-percolation system and should
 be  removed.

     As  has been indicated, the method which was ultimately utilized to
 physically  alter Bed 2 consisted of stripping the top 0.61 meters (2 feet)
 of soil  and then constructing a ridge and furrow system, as was outlined in
 Section 4.   Following bed modification, six infiltration tests were per-
 formed on the bottom of the 0.46 meter (18 inch) deep furrows, at the
 locations numbered 1 through 6 in Figure 16.  The results of these tests
yielded  infiltration rates which varied from 3.2 cm/hr (1.25 1n/hr) to
 27.9 cm/hr  (11.0 in/hr), with an average of 12.1 cm/hr (4.76 in/hr).  The
 predominant soil type found in the bottom of the furrows was a sandy loam.
 Loading on Bed  2 following the surface modification was limited to the
 amount of wastewater which filled the furrows without overflowing onto the
 ridges.   These  loadings were applied on a bi-weekly basis.  This practice


                                     34

-------
TABLE 5.  SUMMARY OF INFILTROMETER TEST RESULTS

Test
Bed No. Location
1
Soil Log 1:
0-0.51 m - silty-loam
0.51-2.42 m - sand and
gravel
2
Soil Log 2:
0-0.91 m - silty-loam
w/ heavy clay
0.91-2.43 m - gravel
and sandy loam








3

Soil Log 3:
0-0.46 m - silty-sandy
loam w/clay
0.46-2.44 m - sandy loam







A
D
F
B
C
E
A
D
E
F
G
C
B
I
H
1
2
3
4
5
6
A
F
c
D
G
E
2
3
4
5
6
1
B
Depth from
Original Surface
(meters)
Surface
Surface
Surface
0.051
0.381
0.660
Surface
Surface
Surface
0.051
0.051
0.457
0.762
0.864
1.067
1.067
1.067
1.067
1.067
1.067
1.067
Surface
Surface
0.254
0.254
0.457
0.330
0.457
0.457
0.457
0.457
0.457
0.559
0.610
Infiltration
Rate
(cm/hr)
5.5
0.6
7.1
0
110.5
87.6
0.20
0.10
0.41
0.13
0.30
0.89
0.99
-
22.6
27.9
41.9
9.1
3.5
16.7
3.2
2.0
0.53
0.97
4.2
52.0
18.1
4.7
3.0
1.3
3.5
15.2
11.4
38.1
                        35

-------
                              Soil  Depth  (m)
                0.2       0,4                 0.8       1.0
O)
O
                                                                  120
                                                                  100
                                                                   80
                                                                      0)
                                                                   60 §
                                                                   40
                                                                   20
                                    2              3
                            Soil Depth  (ft)
           Figure 17.   Infiltrometer test  results  for Basin 1.
                                  36

-------
  25
  20
£  lOh
              0.2
    Soil Depth (m)



0.4                0.8
1.0
                   ±
                                                                 60
                                                                 50
                                                                 40
                                                                     (O
                                                                     ce.
            30  o

                +»

                £
                •p

                •r—
                q-

            20  ~
                                                                 10
    "01234


                           Soil Depth  (ft)



 Figure 18.  Infiltrometer test data on  Basin 2 prior to modification,
                                  37

-------
 resulted in an effective loading rate of 41.2 m/yr (135 ft/yr) over the
 furrow bottom area, or 12.8 m/yr (42 ft/yr) over the entire area of Bed 2.
 This proved to be a successful operating level.

      Surface infiltration tests performed on Bed 3 at locations A and F
 resulted in infiltration rates of 2.0 cm/hr (0.78 in/hr) and 0.53 cm/hr
 (0.21 in/hr), respectively.   Location A consisted of silty-sandy loam, while
 location F consisted of a clay-silt loam.  At a depth of 0.25 meters
 (10 inches), an average infiltration rate of 2.59 cm/hr (1.02 in/hr) was
 observed, and the soil  was predominantly a silty-sandy loam.   These tests
 were made at points C and D.   Infiltration tests performed at sites E and G
 were at a depth of 0.46 meters (18 inches), and yielded an average rate of
 35.0 cm/hr (13.8 in/hr).   At  the 0.61 meter (24 inch) depth,  an infiltration
 rate of 38.1 cm/hr (15.0 in/hr) was measured at point B, with the soil  con-
 sisting primarily of a  sandy  loam containing some clay.   Figure 19 summarizes
 these results graphically as  a plot of the variation  of infiltration rate
 with soil  depth.

      In order to  improve  the  hydraulic performance of Basin  3  it was  decided
 that the top 0.46 meters  (1.5 feet) of tight soil  should be  removed   The
 stripping  operation was performed without developing  a ridge  and furrow
 system as  was constructed in  Bed 2.   Following  the stripping, surface infil-
 tration tests were performed  at each of five representative  locations,
 points  2-6,  within Bed  3.  The infiltration rates  measured at these sites
 yielded an average infiltration rate of 5.6 cm/hr  (2.19  in/hr).   The soil at
 this depth was  predominantly  a sandy loam with  varying amounts  of clay also
 present.   The loading rate of 48.8  m/yr (160 ft/yr) which  was  applied  fol-
 lowing  bed modification was determined  by an empirical method  used for esti-
 mating  loading  rates  for  infiltration-percolation  systems  (Personal  communi-
 cation,  R.E.  Thomas).   The loading  rate was established  at ten  percent  of
 the  theoretical loading for one year at the average infiltration  rate  of
 5.6  cm/hr  (2.19 in/hr)  as shown below:

                Loading  rate =  0-1(2.19  in/hr) (8760 hr/.vr).

                Loading  rate =  160 ft/yr   =48.8 m/yr

 When  loading was  resumed, Bed  3 was  erroneously loaded at  a rate  of 57.3 m/yr
 (188  ft/hr)  for 3 weeks because of a  failure  to account  for the reduced  sur-
 face area developed as  a result of incorporating a side  slope into  the
 excavation.  At this loading rate, the bed  failed  to  drain between  succes-
 sive loading cycles.  When the  loading was  reduced to 48.8 m/yr (160 ft/yr)
 on the actual area, the basin  operated In a successful manner.

     Following the physical alteration of Beds 2 and  3,  infiltration rates
 of the system were recorded during the operation of all  three beds.  This
was done by measuring the change in water level  with  time after inundation
of each of the beds.  The average infiltration rates were determined by
 dividing the decline in water  level by the elapsed time.  A summary of these
 infiltration rates is shown in Table 6 for the three beds.  The average
                                     38

-------
    25
    20
-  15
 o
•^
•P
S  10
             Soil Depth (m)
0.2      0.4                 0.8
                                                      1.0
                                                                  60
                                                                  50
                                                                  40
                                                   30
                                                                  20
                                                                  10
                                                                      §
                                                                      O£
                                                                      4J
      01234
                               Soil  Depth (ft)
    Figure  19.   Infiltrometer test data on Basin 3 prior to modification.
                                    39

-------
TABLE 6.  SUMMARY OF INFILTRATION RATES
Loading
Date
M3
Loaded
Infiltration
Rate (cm/hr)
Bed 1
2-8-77
2-15-77
2-22-77
3-1-77
3-8-77
3-15-77
4-12-77
4-19-77
5-10-77
5-17-77
5-31-77
6-7-77
6-14-77
6-21-77
6-28-77
7-5-77
Average
1024
1256
1158
1166
1193
1240
817
1146
1195
957
986
1079
1212
1257
1239
1114
1128
0.99
0.84
0.84
0.64
0.48
0.48
2.5
0.38
2.9
0.71
2.8
2.0
1.7
1.6
1.4
1.3
1.3
Loading
Date
M3
Loaded
Infiltration
Rate (cm/hr)
Bed 2
4-13-77
4-20-77
4-27-77
5-4-77
5-18-77
6-1-77
6-8-77
6-15-77
6-22-77
6-29-77
7-6-77

Average




255
273
-
263
261
238
320
248
250
525
253

289




9.5
2.2
2.4
2.5
4.8
5.1
4.8
5.4
5.4
3.3
5.1

4.6




Loading
Date
M3
Loaded
Infiltration
Rate (cm/hr)
Bed 3
2-10-77
2-17-77
2-24-77
4-14-77
4-21-77
5-12-77
5-19-77
6-2-77
6-9-77
6-23-77
6-30-77
7-7-77

Average



1262
997
1330
1220
984
1191
1080
1186
1160
946
1169
1147

1139



3.2
0.58
0.58
0.61
0.56
2.3
1.1
2.3
1.9
1.6
1.4
1.2

1.4




-------
infiltration rates are also shown for Beds 1, 2, and 3 at 1.4 cm/hr
(0.53 in/hr), 4.6 cm/hr (1.81 in/hr), and 1.45 cm/hr (0.57 in/hr),
respectively.

     Comparison of the recorded infiltration rates with the average rates
obtained in the infiltrometer tests shows a consistent relationship between
the two.  The average infiltration rate for the surface of Bed 1 was found
to be 4.4 cm/hr (1.74 in/hr) in the infiltrometer tests.  The corresponding
recorded infiltration rate for Bed 1 was 1.4 cm/hr (0.53 in/hr), or 30% of
the infiltrometer test value.  Similarly, the recorded infiltration rate for
Bed 3 was 26% of the infiltrometer test value, and the recorded infiltration
rate for Bed 2 was 38% of the infiltrometer test value.

     The higher infiltration rates obtained from the infiltrometer tests may
have been associated with any of several factors.  However, the major factor
which contributed to the high rates was believed to be lateral flow from the
infiltrometer rings.  The combination of lateral and vertical flow through
the soil medium served to increase the rate at which the soil would accept
the applied water.  The U.S. Salinity Laboratory Staff has reported that the
impact of lateral flow increases as the infiltration area decreases (7).  The
infiltration area used in the Boulder infiltration-percolation project was
very small compared to the total area of the basins.  Although lateral flow
was minimized by the outer buffer pond in the double ring infiltrometer, it
was still felt to be primarily responsible for the higher rates observed in
the infiltrometer studies.
FACTORS AFFECTING INFILTRATION RATES

     The infiltration rates in an infiltration-percolation system are
apparently affected by operation and management practices, and wastewater
characteristics, 1n addition to site conditions at a specific location, as
reflected by the data collected during this study.

     Operation and management practices of an infiltration-percolation
system consist mainly of the loading schedule, the basin surface management,
and the method of wastewater application.  Varying the loading schedule,
i.e., the inundation and drying period of an infiltration-percolation system,
has been shown to affect the infiltration rates of that system.  The extent
of this impact is difficult to determine, but it has been stated in the
literature that maximum infiltration rates are achieved by shorter inunda-
tions and longer drying times (8,9,10,11).

     The longer drying times allow for aeration of the soil, and enhance the
decomposition and dessication of organic material deposited during the inun-
dation period.  The microbial population will also decrease as substrate is
utilized, thereby increasing the pore space available for infiltration (11).
In the Boulder study, the loading schedule was not intentionally varied, but
observations made during bed operation showed the approximate drying times
for Beds 2, 1, and 3 to be 3 days, 2 to 2.5 days, and 1 to 2 days, respec-
tively.  A plot of the recorded infiltration rate as a function of time is


                                    41

-------
  shown  for  Basins  1, 2, and 3  in Figures 20, 21, and 22, respectively.  From
  these  figures  it  can be seen  that the infiltration rate following drying and
  scarification  was  the highest.  In Basins 1 and 3, the high initial infiltra-
  tion rate  was  followed by a significant rate decrease as the loading cycle
  progressed.  The  high initial infiltration rate was attributed to the soil
  aeration and scarification procedure which served to remove any organic
  matter that may have formed during the preceding loading cycle.  The highest
  infiltration rate  for Basins 2 and 3 followed a long drying period which
  included the period of bed modification.  These high rates, observed on
  February 10, 1977 for Basin 3, and April 13, 1977 for Basin 2, resulted primarily
  from the bed modifications which created a new soil-water interface for
  infiltration.  Since the literature suggests that soil clogging is a surface
  phenomena, it follows that stripping of the beds removed the clogged surface
  and created a "fresh" surface (12, 13).

      The effect of basin surface management on the infiltration rate was
  investigated to only a limited degree at the Boulder site.   Basin surface
 management practices may include the growth of vegetation on the surface,
 placement of a layer of sand  or gravel,  or operation with a bare soil  sur-
 face.   Surface management may also encompass the type and frequency of scari-
 fication during the drying periods.   All  beds  in the Boulder system had bare
 surfaces, so comparison  with  vegetated or gravel surfaces was  not possible.
 The scarification  practiced in this  study undoubtedly contributed to the
 restoration of high infiltration rates observed at the beginning of each
 loading cycle.   However,  since the same  type and frequency  of  scarification
 was used throughout the  operation  of the basins, comparison with other methods
 was not possible.

     A  gradual  decline  in  infiltration rates was exhibited  as  the loading
 cycle  progressed on Beds  1  and 3.  Many  factors likely contributed to  this
 decline in  infiltration  rates.  Other investigators  have  suggested that the
 major  factors  contributing  to  such declines  are the  accumulation of suspended
 solids,  and the microbial  activity which  is  effective  1n  degrading the
 applied organics.   However, Bed  2  did not  exhibit  these declining trends,
 even though it  was subjected  to  application  with the  same strength waste-
 water.   The differences  in  operation  between basins which may  have caused  the
 infiltration  rate  in Basin  2  to  remain constant were:   (1)   the water  was
 applied  to  Beds 1  and 3 by  total surface flooding, and  (2)  Bed  2  was loaded
 at  a lower  loading  rate than Beds  1 and 3.   Although Bed 2  was  not scarified
 after the drying period, while Beds 1 and 3 were, this fact would only
 affect the  initial   infiltration rate and would not be responsible for main-
 taining infiltration rates during the loading cycle.  The fact that Bed 2
 was a ridge and furrow system  may explain why infiltration rates were main-
 tained during the loading cycle.  It has been suggested by McGaughey that a
 ridge and furrow system would  not be affected by suspended solids to the
 same degree as a flat surface  (8).   Since the solids that settle will only
 alter the surface at the bottom of the furrow,  the sides of the furrow
 remain relatively free of clogging solids and,  therefore,  allow the Initial
 Infiltration rate to be maintained.  Figures 23, 24, and 25 show the Infil-
 tration rate plotted as a function of accumulated suspended  sol Ids for each
of the three basins.  From these figures it can be seen that the accumulated
                                      42

-------
            2.0
co
         (U
         >o
        CtL
         O
            1.5
1.0
            0.5
              0
                                                                                                         i.
                                                                                                         .c

                                                                                                         E



                                                                                                         (U
                                                                                                         C
                                                                                                         O
                                                                                                        4J
                                                                                                        
-------
    Drying and

Bed Modification
                                                                                         u
                                                                              to
                                                                              cc.
                                                                              c
                                                                              o
                                                                              (O
                                                                              S-
                                                                              M-
                                                                              c
                                                                         2.0
                                                                 July
12 19 26 2 9 16 23 30 6 13 20 27 6 13 29  27  10  17  24 1 8  15  22 29 5 12 19  26 3 10 17

 December    January     February    March    April   May        June

   1976        1977                                               1977

                                    Time  (weeks)


          Figure 21.  Pattern  of  infiltration  rate decline in Basin 2.

-------
cn
                                                                                                           o
                                                                                                           0)
                                                                                                           4->
                                                                                                           (0
                                                                                                           CC.


                                                                                                           O
                                                                                                           £
               0
                 12 19 26 2 9 16 23  30  6  13  20 27
                 December    January   February
                   1976        1977
6 13 20 27  3  10 17 24  1  8 15 22 29 5 12 19 26  3 10
  March     April     May         June        July
                                 1977
                            Figure 22.  Pattern  of infiltration  rate decline in Basin 3.

-------
    2.0
    1.5
Q)
rtJ
o:
to
t-
    1.0
   0.5
     0
       0
                  100
                 _L
1000
               Accumulated Suspended Solids (mg/m2)


              200       300        400        500

                             	1	
                                        T
                                                      T
                                LOADING CYCLE OF 5-31-77  to  7-5-77
                                                                         600
                      _L
                                                                      JL
                                                                            700
                                                                                             5.0
                                                                                 01
                                                                                 4->
                                                                                 
-------
    2.0
    1.5

                                                                                                O
1000
                    2000      3000      4000        5000       6000       7000      8000

                        Accumulated Suspended Solids  (mg/ft2)

Figure 24.  Infiltration rate as a function of suspended solids loading  in  Basin  2.

-------
00
                         100
            2.0





        IT
        JC


        3  1,5
         o>
         +J
         
-------
suspended solids did not appear to have an effect on the infiltration rates
in Basin 2, but appeared to affect the Infiltration  rates of  Basins  1  and  3.
Based on these data, it was concluded that the existence of a ridge  and fur-
row system in Basin 2 was partially responsible for  maintaining  the  relatively
high initial infiltration rates.

     The lower loading rate applied to Bed 2 was undoubtedly responsible,  in
some part, for the relatively constant infiltration  rates.  This lower load-
ing rate permitted longer drying times between the loading times on  Bed 2,
when compared to Beds 1 and 3.  As stated previously, Bed 2 was  dry  for
approximately 3 days while Beds 1 and 3 were dry from 2 to 2.5 days, and 1 to
2 days, respectively.  These longer drying times allowed for a more  complete
aeration of the soil in Basin 2 and enhanced the decomposition and dessication
of the organic material deposited during the flooding.  Thus, more pore vol-
ume became available for the transport of water, resulting in the mainten-
ance of initial infiltration rates throughout the cycle.

     The influent temperature also appeared to have some effect  on the
infiltration rate.  Figure 26 indicates the average infiltration rate over a
complete loading cycle as a function of the average temperature  of the
influent over the same period.  The high Infiltration rates observed in Basins
2 and 3 immediately following their modification were not included in these
averages because these high rates resulted mainly from the bed modification,
as was previously discussed.  The trends illustrated in Figure 26 were attri-
buted to the lower viscosity of the water at higher temperatures, and to the
fact that the quality of the wastewater influent to the basins was generally
better during the warmer weather.

     While temperature did seem to affect the infiltrative capacity of the
soil, the cold temperatures encountered did not necessitate discontinuing the
operation of the Infiltration-percolation system.  Even though there was some
freezing of the beds during February and March, the ice did not appear to
Interfere with bed performance.  The ice layer apparently served to insulate
the underlying water and collapsed as the water level declined.   Subsequent
loadings melted the broken ice, after which a new ice layer formed and the
cycle was repeated.  Prolonged periods of sub-freezing temperatures could
freeze the beds solid, which would cause severe operational problems.  How-
ever, this did not occur in the Boulder situation.
                                      49

-------
JL.U
s_
^.
c

&a
(0 Q£ U • 0
OJ C
> o

1 o
1— 1
1 1 1 1
Basin 1
_

^^°
r^ 	
- ^^"^
^^^^ -
^^^
O



1 1 1 1
0 5 10 15 20

"t
2. of

u
o>
(O
i.og
•r-
+J
OS
t-
•p
c
C
l— r
Ave. Temperature - C°
2.0

*£
c
7 1-5
0) •»->
CO «
 o

5 i.o
c
c 1
1— 1
o

i 1 I 1
Basin 2 y°
/
/
/
J
f
J
J
J
f
O _

t
, +


S_
JC
"e
4.0^
•p
IO
a:

c
o
•I—
3.0 «
£
* c
1* ""
i i i i T
0 5 10 15 20
Ave. Temperature - C°
~ 1.0
.c
c

a» as
as ce.
OJ c 0. b
3^ O


2.0^
S.
c
o
1.05
to
fv
-p
c
                            10          15          20
                       Ave.  Temperature  -  C°
Figure 26.  Infiltration rate as  a function of wastewater temperature.
                               5Q

-------
                                 SECTION  7

                          TREATMENT PERFORMANCE OF
                     THE INFILTRATION-PERCOLATION  SYSTEM


GENERAL CONSIDERATIONS

     In assessing the treatment performance of the infiltration-percolation
system, the primary focus was directed toward monitoring the fate of the fol-
lowing major wastewater constituents:   phosphorus, organics, nitrogen, hard-
ness, alkalinity, and coliform organisms.  In addition, the behavior of
selected heavy metals was assessed in  a specific short term study.  The aim
of the study was to evaluate the treatment efficiency and its variation with
seasonal and climatic changes, as well as long term equilibration effects.

     Table 7 summarizes the average seasonal constituent concentrations in
the secondary effluent applied to the beds.  Seasonal values were derived
by averaging the constituent concentrations of all the secondary effluent
samples taken during a particular season.  For example, the average fall COD
concentration was the average of the COD concentrations in all of the secon-
dary effluent sampled between September 22 and December 22.  Some of the
seasonal trends observed in the renovated water quality, e.g., COD, were a
result of the seasonal variation of the pollutant concentrations in the
applied secondary effluent.  Other seasonal trends observed, such as the
phosphate level in the renovated water, seemed to be less affected by the
average seasonal pollutant concentration in the applied secondary effluent.

     Figure 27 summarizes the annual precipitation pattern as well as the
minimum weekly temperature in the renovated water from Basin 1 for the first
year of study.  These data have been presented for reference, since the mini-
mum renovated water  temperature reflected  an environmental condition within
the soil system that may  have affected the removal efficiencies  for certain
constituents.  The weekly precipitation  represents a moisture effect that
may have been effective  in causing some  of the observed seasonal variations.


PHOSPHORUS BEHAVIOR

     Data compiled during the course  of  this  study showed  the phosphorus
concentrations  in the basin  effluents  to vary over quite a broad range.
This variation  is demonstrated  in  the  total  phosphorus  concentration  profiles
of Figure 28.

      In all  three of the basins,  there appeared to be  a significant seasonal


                                      51

-------
        TABLE  7.   SEASONAL AVERAGES FOR WASTEWATER
                  CONSTITUENTS  IN  BOULDER SECONDARY EFFLUENT
Constituent
Total Nitrogen1
Ammonia
Organic Nitrogen
Nitrate
Total Phosphate2
COD
Alkalinity3
Hardness3
Cal ci urn
Magnesium
Chlorides
Suspended Solids
Total Solids
Spring
13.8
7.3
4.7
1.8
4.8
58.0
128
132
34.6
11.0
27.4
14.6
311
Summer
13.2
6.0
4.7
2.5
5.1
58.1
118
154
35.6
14.8
23.2
11.7
333
Fall
14.6
5.3
6.6
2.7
7.2
72.1
125
119
-
-
32.1
21.6
286
Winter
24.5
11.7
12.2
0.6
7.9
118.3
164
112
32.2
7.4
27.3
25.4
302
Annual
16.5
7.6
7.0
1.9
6.2
76.6
134
129
34.1
11.1
27.5
18.3
308
     nitrogen concentrations in mg/£ as N

2mg/£ as P

3mg/«, as CaCO
                                 52

-------
CO
          C  2.0-
          _o
          '£
           o
           0)
           2
           2.
                July August  Sept.  Oct.  Nov.    Dec.

                                1976
                                                a.
                                                •r-
                                                O
Jan.   Feb. March  April   May  June   July

                     1977
                    Figure 27.  Weekly precipitation  and  minimum temperature of renovated water.

-------
in
                           I    I    I   I    I   I    I   I    1   I   I    I   !    I   I    I   I
                                        Basin
                                        Modification
              J   A   S  0   N   D

                      1976
 I   I    I  TTTTTT'l
JF   N   A  M   J   JA   S   ON   D
  I   I   I   I   1
J  F   M   A  M   J
                  1977
       1978
                                Figure 28.  Effluent phosphorus  as  a  function of time.

-------
pattern to the leakage of phosphorus through the basins.  The concentration of
total phosphorus appearing 1n the renovated water increased markedly during
the winter loading cycles when the wastewater temperatures  were the lowest,
and the influent phosphorus concentration was the highest because of reduced
infiltration in the municipal collection system and lower phosphorus uptake
in the trickling filter system.

     It has been demonstrated that phosphates introduced to the soil system
are initially adsorbed in the mineral or organic fraction of the soil
(14, 15).  The adsorbed phosphate is subsequently precipitated or mineral-
ized as inorganic compounds into the soil matrix by a slower reaction.  This
precipitation reaction apparently serves to release the adsorption sites
within the soil for additional phosphorus removal.  The adsorption process
will limit the rate at which phosphorus can be removed by the soil until the
soil adsorption sites become saturated.  Further phosphorus removal will be
controlled by the slower rate of precipitation of relatively insoluble com-
pounds (16).  The increase in the phosphorus levels applied to the beds
during the fall and winter apparently saturated the phosphorus adsorption
capacity within the soil system.  At this point, the slower precipitation
reaction effectively controlled the rate and limited the effectiveness of
the phosphorus removal process.  This effect was shown most dramatically in
the winter of 1977-78 after Basins 2 and 3 had been modified.  During modi-
fication, Basin 3 had most of the fine grained soil removed from  the  upper
portion of the soil profile.  Following modification, this basin was  loaded
at the highest equivalent hydraulic loading rate of 48.9 m/yr  (160  ft/yr).
With this combination of basin modification and loading characteristics, the
phosphorus removal efficiency was reduced in Basin 3 to about  40% during
January of 1978.  Conversely, Basin 2, which was loaded at the lowest
rate, demonstrated a much higher and more consistent phosphorus  removal
capability.  These data suggest that by  increasing the detention  time in the
soil system, the phosphorus  removal efficiency might be controlled  at a very
high level.

     Analysis of the phosphorus profile  for Basin  1 indicates  that  the phos-
phorus leakage peaked during  both winters of operation.  However,  the
effluent concentration during  the second year of  operation was about twice
that observed during the first winter.   These data suggest that  the system
had not  fully equilibrated  in  the first  year of operation.  This  finding is
consistent with the  findings  reported  from  the Lake George, N.Y.  system by
Aulenbach  (17,  18).  Apparently,  the  regeneration of adsorption  sites by
mineralization  is not complete,  and a  continual  degradation of system per-
formance can be expected in phosphorus removal.

      In  addition  to  the  seasonal  trends  which have been discussed,  signifi-
cant variation was also  noted in  the  phosphorus  leakage during different
loadings within a cycle.   In most of the loading  cycles, the concentration
of total phosphate in  the  renovated water increased from the first loading
of the cycle  to the  last.   Selected examples showing this  trend  have been
presented  in  Figure  29.   In keeping'with this trend, the total phosphate
concentration  in  the renovated water from the  first loading of a loading
cycle was  usually less  than the phosphate concentration  in the renovated
water from the last  loading of the  previous loading  cycle.  These

                                     55

-------
en
               2.0
            s-
            o
            CL

            I/)
               1.0
                        I    I
                 ol
                  8/30 9/6 9/13 9/20  10/11 10/18 10/25 11/1 11/8 11/29 12/6 12/13 12/20 12/27 1/3


                           1976                    1976                      1976              1977



                                                     (A) Basin 1
      10/14 10/19 10/26 11/2      8/25  9/1  9/8  9/15   6/2 6/9 6/16  6/23  6/30 7/7  8/4  8/11 8/18 8/25


               1976                           1976                      1977                       1977


             (B)  Basin  2                                           (C)  Basin 3
                        Figure 29.   Effluent phosphorus  variation  within  loading cycles.

-------
observations can be explained in terms of the slow precipitation of phosphate
in the soil matrix.  During the resting period between loading cycles,
rejuvenation of the soil  phosphate adsorption capacity apparently occurred.
The soil was able to remove the initial application of phosphate most
efficiently during the first loading of the new cycle.  The subsequent  appli-
cations of effluent during the cycle showed that the capacity of the soil  to
remove phosphate was reduced as the cycle continued.  By the last loading  of
the loading cycle, the soil adsorption capacity for phosphates was in its
most exhausted state, and the greatest leaching of phosphates through the
soil system occurred at that time.


REFRACTORY ORGANICS

     Refractory organics were monitored during the study period by sampling
the secondary effluent during application to each bed, and the renovated
water from the basin 24 hours after application commenced.  Figure 30 pre-
sents a summary of the average COD level for each loading cycle in the basin
effluents.  As can be seen from this figure, the infiltration-percolation
system was generally effective in reducing the product COD to a residual
level of 10-20 mg/x,.  These levels represented a reduction of the influent
COD by 70-80%.  The only significant deviation from this treatment pattern
occurred in the December-February period of 1976-77.  During that time the
influent had unusually high levels of COD which were attributed to highly
concentrated industrial waste discharges into the Boulder Treatment Plant.

     The successful operation of the basins through the winter demonstrated
the fact that cold conditions did not  reduce the capability of the soil for
removing organics  from the applied wastewater effluents.  This performance
supports the findings of other authors regarding the  temperature  insensi-
tivity for organic removal in high rate systems.   In  one study of this
phenomenon, Thomas, et al. loaded six silica sand lysimeters with  secondary
and septic tank effluent to evaluate the effect of  temperature on COD
removals (19).  Two lysimeters were incubated at 28°C, two at 18°C to 30°C,
and two at 18°C to 35°C.   All six lysimeters produced comparable  COD removal
efficiencies.  Schwartz, et al. also reported that  cold climate did not
restrict the COD  removal capability of a high-rate  land treatment system.(20).

     The relative positions of  the three curves in  Figure 30 suggests that
the organic content of the product water may bear  some relationship to the
hydraulic  loading Of  the basin.   Following bed modification, Basin 3 was
loaded  at  the highest rate, and the product water  from this basin was
typically  the highest in  COD.   Consistent with this observation,  Basin 1, the
next most  heavily loaded,  had a lower  effluent COD, and Basin 2 had both the
lowest  hydraulic  loading  and the  lowest product COD level.


NITROGEN

     Nitrogen was the most difficult  constituent  to evaluate with respect  to
the treatment  performance  of the  infiltration-percolation  basins.  The  reason


                                      57

-------
en
00
      o
      8
                                                              1    I—I—I
                                                              o       Basin 1
                                    Basin
                                 Modification
               JASONDJFMAMJJASOND
                      1976
1977
J   F   M   A  M   J

          1978
                              Figure 30.  Effluent  COD as  a  function of time.

-------
for this is that the nitrogen cycle contains several  pathways for transfor-
mation of the organic and ammonium nitrogen applied to a soil system in
secondary effluent.  The three major sequences involve the following
reactions:  (1)  The ammonium ions may be held in the soil  by adsorption  or
fixation processes; (2) The ammonium ions retained within the soil  by adsorp-
tion or fixation may be nitrified to nitrate and leached through the soil
system with the renovated water; and (3)  The nitrate, once formed,  can be
denitrified in the presence of a carbon source and anoxic conditions.  The
The adsorption and nitrification processes were clearly the most important
during the operation of the Infiltration-percolation  basins in this study.
However, the relative Importance of the fixation process, which allows
Indefinite nitrogen storage, and the den1tr1fication  process, was difficult
to evaluate.

     Nitrate and ammonium ions were the principal nitrogen forms discharged
in the basin effluents, with the nitrate ion accounting  for about 98% of  the
total nitrogen discharged from Basin  1.  Figures  31  and  32 show the  average
ammonium  ion and nitrate concentrations, respectively,  for each of the load-
ing cycles during  this study.  The average nitrate and  ammonium ion  levels
for a given weekly loading  period were calculated by a  flow weighting  pro-
cess.  This involved development of a hydraulic  discharge  curve by plotting
the hydraulic  discharge from the basin as a  function of sampling time.  The
total basin discharge was determined  from this figure by integrating under
the hydraulic  discharge curve.  The total pounds of  nitrogen  discharged by a
basin during  the week were  determined by a  similar method.   A nitrogen mass
flow curve was generated by plotting  the product of  the ammonium or  nitrate
nitrogen  concentration and  the  flow rate as  a  function  of sample time.   Fol-
lowing  normalization of units,  the  area  under  this curve yielded the total
pounds  of the  ammonium or nitrate nitrogen  discharged by the basin during
the loading period.  The average  ammonium  or nitrate concentration was found
by dividing the  pounds of ammonium  or nitrate nitrogen  discharged  by the
total  hydraulic  discharge.

      Through  an  analysis of the input and output nitrogen quantities, a
nitrogen  balance was made  for Basin 1 which covered  the first full year
of operation.   Figure  33  shows the  cumulative nitrogen balance through
Basin  1 during this period.  The upper line is the cumulative total  nitrogen
applied,  and  the lower line is the cumulative nitrogen discharged  from the
basin.   The difference between the lines at any given date shows the total
nitrogen  storage or loss  in the basin from July 12,  1976 through the loading
cycle beginning on that date.

      From these curves it is apparent that the basin tended to store nitro-
 gen during the cold winter months of the first year and release it as the
water temperature wanned in the spring.  This figure suggests that the
 nitrification process was inhibited during the coldest period and ammonium
nitrogen  was  stored by fixation or  exchange 1n the basin.   With the  onset of
the warming  trend, nitrification  of the ammonium stored 1n the basin seemed to
be more complete,  and  nitrogen was  released from the basin as the  mobile
nitrate Ion.   This observation Is  reinforced by the  increased NH*  leakage
                                      59

-------
       0?


       U>
tTt
O
          5  -
          4  -
          J  •*••*
          2  -
          0
                                                                                 I
                                            Basin Modification
               J    A  S    0   N  D

                      1976
JFMAMJJASONDJFMAMJ
                    1977
1978
                       Figure 31.  Effluent ammonium nitrogen  as  a function of time.

-------
CT>
         I co
         O
          01
                    I—I—I—I—I—I
                                                   Basin Modification
         "-  10  —
         i • i  AW
                  JASONDJFMAMJJASOND
J   F  M   A   M   J
                         1976                              1977                             1978

                           Figure 32.  Effluent nitrate nitrogen as a function of time.

-------
           2000
ro
           1000  -
         
                                                                                                - 800
                    o Nitrogen  Applied


                    • Nitrogen  Discharged
                        Figure 33.  Cumulative nitrogen applied  and  discharged from Basin  1.

-------
from Basins 1 and 2 during the colder winter months.   A similar observation
and explanation has been offered by Bouwer,  et al.  (21).

          "This seasonal trend is probably caused by  greater  drying  of
     the soil between flooding periods in the summer  than  in  the winter,
     which allowed more oxygen to diffuse to greater  depths  in the soil
     in the warmer months.  Thus, more adsorbed NHj could  be  nitrified
     in the summer, causing more NH$ to be adsorbed during flooding  and,
     consequently, a decrease in the NHiJ level in the renovated water.
     This continued until the fall when poorer drying limited the amount
     and depth of penetration of oxygen.  The seasonal trend  could also
     be the result of the higher amount of SS in the  effluent in fall  and
     winter.  These solids began to accumulate on the soil during winter
     and early spring.  The resulting sludge layer could have acted as
     an oxygen sink during drying, thus reducing the  amount of oxygen
     entering the soil."

     This comment was made in relation to the Flushing Meadows system, but
also appears applicable to the infiltration-percolation system operated by
the City of Boulder.  A plot of the suspended solids  in the secondary efflu-
ent during the study period, Figure 34, did show that an increase in
suspended solids occurred in the winter, as was the case at Flushing Meadows.

     Seasonal nitrogen data from the Lake George system demonstrated the same
general variations in bed performance which have been noted in this study.
Ammonium concentrations taken from two sample wells were highest  in the win-
ter and lowest in the summer.  Conversely, the nitrate concentrations were
generally higher in the summer and fall.  A similar ammonium  leakage  trend
occurred during the operation of the Brookings infiltration system  (22).   The
ammonium ion concentrations in the product water were lowest  in  the summer
and highest  in the winter.  This trend likely reflects both the  variation of
the ammonium concentration in the applied wastewater, as well  as  the  seasonal
variation in the ability of the basins to convert ammonium to  nitrates.

     The nitrogen behavior in the basins during the second year of operation
showed a somewhat less predictable pattern than that of the first year.  The
level of nitrification seemed to be higher and somewhat more consistent
throughout the year, although there was a slight tendency toward peaking of
effluent nitrates in the spring, and increased leakage of ammonium ions dur-
ing the coldest winter months.  However, in Basin 3,  which was the most
heavily loaded during this period, there seemed to be some very
significant  departures  from the expected pattern.  The ammonium level in the
effluent was significantly higher than in the other two basins, and increased
rather dramatically during the last 7-8 months of operation.   This was
accompanied  by a nitrate  level that was substantially below that of the dis-
charge from  Basins  1 and  2.  These data  Indicate that the basin was unable to
effect nitrification of an increasing  amount  of the applied nitrogen.  This
pattern seems  to suggest  that the basin was being stressed to  the point that
the hydraulic, ammonium,  and organic loads exceeded the capacity of the sys-
tem for transferring adequate oxygen to  permit complete oxidation of  the
applied wastewater  ammonium.  Clearly,  this loading would have to be  reduced
to provide  acceptable  ammonium removals  on a  long  term basis.

                                      63

-------
     30
           T—i—I—i—r
T—i—i—i—i—r
                                                    i—i—i—r
     20
0)
4J
«/»

-------
     In addition to the observed seasonal  trends,  nitrate  and  ammonium
concentrations in the renovated water followed a general pattern  within  the
individual  loading cycles.   The average nitrate concentrations tended  to
decline with successive loadings within a  cycle.  Figure 35 shows data selec-
ted from loading cycles that clearly illustrate this trend. The  maximum
nitrification occurred when oxygen penetrated deepest into the basins  during
maximum drying.  For this reason, the maximum nitrate concentrations  in  the
renovated water occurred during the first  loading following the extended dry-
ing periods.  Subsequent loadings during the cycle tended  to limit the basin
aeration and reduced the degree of nitrification that could occur.  As a
result, the basins produced lower nitrate  levels as the loading cycle
continued.

     A reverse trend was observed with respect to the ammonium ion concentra-
tion, particularly during the winter loading cycles.  This is  shown for  two
cycles in Figure 36.  As shown in this figure, the ammonium ion level  in the
product water increased with each successive inundation within a  loading
cycle.  This trend was apparently caused by a reduction in the available
ammonium adsorption capacity within the soil.  Rejuvenation of the soil  ammo-
nium adsorption capacity generally occurred during the drying  period, when
the ammonium ions held on adsorption sites were nitrified.  This  nitrifica-
tion freed the adsorption sites in the soil for readsorption of ammonium ions
during subsequent loadings of the basin.  When ample drying of the basin did
not occur between inundations, rejuvenation of the soil ammonium adsorption
capacity did not occur.  In these cases, greater amounts of ammonium leached
through the soil profile with each succeeding loading period, resulting  in
an increase in ammonium concentration in the renovated water with successive
loadings of the loading cycle.  The most obvious examples of this trend
occurred during the winter when an abnormally high ammonium concentration
was present in the applied effluent.  During this time, the ammonium adsorp-
tion capacity was the most heavily taxed because of high ammonium loading and
low rates of nitrification.  Apparently, the ammonium adsorption capacity of
the soil was not fully taxed during the other seasons, and the indicated
trend did not occur except in the winter.


DISSOLVED SALT SPECIES

     Water  percolating through the infiltration-percolation soil  system
carried with  it soluble salts.  The dissolved inorganic constituents moni-
tored in this  study were hardness, calcium, magnesium, alkalinity, and
chloride.   An  effort was made to  correlate the  concentrations of the various
salt constituents leached from the soil with operating conditions and the
seasonal variations of environmental conditions.

     The average hardness concentrations of the renovated water, and  the
changes observed in hardness as the water percolated through the  basins are
shown in Figures 37 and 38, respectively.  The  most pronounced characteris-
tic of both these curves was the  peaking pattern observed  during  the  period
February-May,  1977.  This peaking was apparently related to the  addition of
alum  in the secondary  treatment system  for suspended solids control during


                                     65

-------
   10
   i
       15
       10
            I  I   I  I
9/   9
                 13  /20    15*
                                 i  i   i
              1976
                     1977
       10
   (0

   .oi
   £
            i  i   i  r
            I  i   i
         i  i^»^
                                    i  i   i  i
     1976
                                       1977
                                                          Basin 1
                                                          Basin 2
                                                          Basin 3
            1976                    1977
Figure 35.  Nitrate variation in effluent within  loading  cycles.
                               66

-------
          2.0
      I    r
      E
          1.5
          1.0
          0.5
                                 Basin 1
                	I    I     I     I     I
                2/8  2/15 2/22 3/1  3/8  3/15
                            1977
          2.0
           1.5
      re
      o?
           1.0
           0.5
                  T
                 2/10
2/17
1977
2/24
Figure 36'.  Effluent ammonium variations with loading cycles.
                          67

-------
           350
00
       vt
       

       "O
                                                                   -  Basin 2



                                                                   -  Basin 3
                0—av-fr-..
           150 -
            100
                JASON   DJFMAMJJASOND
                       1976
1977
J  F  M   A   M  J

      1978
                            Figure 37.  Effluent hardness as a  function of time.

-------
            250
VO
        en
        to
        ca
       •o
                        I   I    I    I    I   I    I    I   1   I    I    I   I    I    I   I    |    I   I   I    I    I


                                                                         Basin  1
                                                                  ----  Basin   2
                                                                O——  Basin   3
                                        J  F   M   A  M
        °T   150  -
        en


        1
            100  -
             50  _
J  A
ONDJFMAMJ

                   1978
                        Figure 38.  Change in hardness concentration through the basins.

-------
this period.  With the reduction in pH which accompanies alum addition, it
was not surprising to observe increased leaching of hardness ions.  The
increased uptake of calcium and magnesium from the soil profile can also be
seen by analyzing the changes in concentration of these two constituents from
the influent to the effluent, as is shown in Figures 39 and 40.

     Accompanying the release of the cations causing hardness, the alkalinity
of the water collected in the underdrain system also increased markedly dur-
ing the period February-May, 1977.  This is shown very clearly in the curves
of alkalinity change presented in Figure 41.  The behavior of both the hard-
ness and alkalinity constituents during the period of chemical addition
apparently showed the impact of solution pH variations on the inorganic qual-
ity of the percolate water.

     Apart from the unusual data developed during the spring of 1977, there
seemed to be a general decline in the level of leaching of hardness and alka-
linity from the soil profile.  In fact, during the last two loading cycles of
Basins 2 and 3, there was little change in the concentration of hardness and
alkalinity from the influent to the effluent.  This would seem to indicate
that an equilibrium condition was being approached between the applied waste-
water and the soil profile.  The reasons for the deviation from this pattern
in Basin 1 were not readily apparent.

     The chloride concentration of the renovated water was monitored
to assess the level of interaction between the ground water and the
infiltration-percolation system.  As can be seen in Figure 42, the chloride
concentration in the water percolating through the basin seemed to show little
change over the period of this study.   The only significant exception to this
was during the first six months of operation of Basin 3.  During this
period before the bed modification, the infiltration rate was very low, and
the rate of application was continually reduced to the point that the
equivalent loading in December, 1976 was only about 4.6 m/yr (15 ft/yr).
With this low wastewater loading, dilution of the wastewater chloride con-
centration by the limited flow of ground water coming through the clay dike
apparently made a detectable impact on the underdrain chloride concentration.
When the basin surface was modified and the loading rate increased, this
effect was not apparent.


COLI FORM ORGANISMS

     During the early stages of the study, grab samples were collected
periodically to monitor the performance of the infiltration-percolation sys-
tem in removing bacterial organisms.  The data which summarize these analy-
ses are presented in Table 8.  The influent values in this table represent
averages for each of the three basins, while the effluent levels are specific
for each  basin.   As indicated In these data, the basins were very effective in
making a substantial reduction in the  coliform concentrations.  The effici-
encies shown compare well with those which have been reported for similar
high rate systems.  Nevertheless, the effluent from each basin contained a
significant concentration of residual  fecal organisms.  The difference in
the effluent fecal coliform levels between the Basin 1 effluent and those in

                                     70

-------
    60
u>
CO
'£  50

  40

e
5  30

    10
I   I    I    I    I
I    i    I    I   i   1
                                                                                            i    i
I    I    I    I   I
        J   A   S    0  N   D
                                  I    I   I   I   I    I    I   I    I    i    I
                     JFMAMJJASOND
                                            J   F  M   A    M   J
                 1976                               1977                               1978



                  Figure 39.   Change in calcium concentration  through the basins.

-------
           30
ro
-P


en


8
.c
4-9
           20
0)
u
c
o

0  10
        CT>
        0)


        c
        
-------
          200
CO
       lO
      o
      o
      (O
      o
      (0
      .01
      0}
      CO


      0)
      1
c
•r-


-------
    +15
    +10
     +5
T   I   I    I—I—I—I—I—I—I—I—I—I—i—\—i—i—|—i—i—i—r


                           Basin Modification
0?



OJ

O
                        «v

                                                               Basin  1


                                                            — Basin  2


                                                       n— .— Basin  3
1 1 1 1 1
J A S 0 N D

0
i I
F M A
1
1
M J
I
J A
I I
1 	 1 	 1 	 1
SON

D

J
I i
	 1 	 1 	
F M A

1 	 1 	 '
M J
                 1976                              1977                              1978


                 Figure 42.  Change  in  chloride concentration  through  the basins.

-------
                TABLE 8.   COLIFORM  REMOVAL DATA
Parameter                    Total  Coliforms     Fecal  Coliforms
                              (per 100 ml)        (per 100 ml)
Composite Influent Avg.       2.175 x 106            1 x 105
Effluent Avg.
   Basin 1                       5700                 3800
   Basin 2                       4025                  900
   Basin 3                       4250                 1025
% Removal                    Total Coliforms     Fecal Coliforms
   Basin 1                         99.7                 96.2
   Basin 2                         99.8                 99.!
   Basin 3                         99.8                 99.0
                                75

-------
Basins 2 and  3 was  likely related to the greater soil permeability and
higher hydraulic  loading rate in this basin during the period of collection
of these data.
HEAVY METALS

     In a separate short term study phase, the removal of several important
heavy metal wastewater constituents was assessed.  The details of the
research approach for this portion of the investigation have been presented
previously in Section 5.

     The results of this heavy metal investigation have been summarized in
Figures 43 through 48.  Each figure presents the concentrations of the dis-
solved, suspended, and total metals in the influent to the basins, and the
total metal concentration in the effluent from the basins.  The concentrations
are averages of six determinations from each bed over a six week loading
cycle.  The data from all three of the basins were plotted on a single figure
for each metal.  The figures show the part per million concentrations as bar
graphs with the means, standard deviations, and ranges tabulated below.  The
percent removal of each metal during infiltration-percolation was calculated
and presented next to the mean value of the metal concentration in the
effluent. The concentration of the metals are given in parts per billion (p.p.b.).

     These data show that all of the metals studied were removed to some
extent from the wastewater during the infiltration-percolation process.  The
levels of removal ranged from high efficiencies of greater than 80% for lead
and zinc, to low levels of less than 50% for nickel and chromium.

     As seen in the previous figures, the metals which occurred in the Boul-
der wastewater were rather evenly distributed between the dissolved and sus-
pended fractions.  In fact, in several cases the measured mean concentrations
of metals in the dissolved fraction exceeded those in the suspended fraction.
This was not expected since it has been reported in the literature that heavy
metals are usually strongly associated, with the suspended fraction.  However,
in most cases the results which have been reported were from studies with
either sludges or raw wastewater having much higher suspended solids
concentrations.

     The metals associated with the suspended material undoubtedly accounted
for much of the observed removal.  Since most of the suspended solids were
removed from the wastewater during infiltration-percolation, it appears that
a large portion of the metal  removal observed in this study was a result of
filtration of the solids.  The remaining metal  removal was likely due to a
combination of exchange and precipitation reactions.
                                     76

-------
      20
      15
   tL  10
   o
   o
   o
       0
                   Diss
             Sus
           Influent
            Tot.
                                     Tot. (A=68%)
                                   Effluent
Mean
1.26
1.21
2.47
0.78
Stnd.Dev.
Range
1.09
4.50
2.91
11.00
2.65
11.00
1.56
6.00
                         Figure 43.   Wastewater cadmium concentrations.

-------
               80 r-
oo
         Mean
                            Diss.
19.75
  Sus.
Influent
 24.27
                        Tot.
54.02
                                      Tot. (A=65%)
Effluent

  18.75
         Stnd.Dev,
13.83
  9.67
16.05
   8.72
         Range
50.00
 37.00
52.00
  26.00
                                  Figure 44.  Wastewater copper concentrations.

-------
        80
        60
    Q.
    a.   40
    §
    o
       20
Mean
                   Diss.
16.93
  Sus.
Influent
  9.00
                         Tot.
25.93
  Tot.  (A-47%)
Effluent
   14.60
Stnd.Dev.
Range
14.11
49.00
7.38
22.00
14.10
41.00
9.97
38.00
                           Figure  45.   Wastewater  chromium concentrations,

-------
          .a
          Q.
          Q.

          O
          O
          O
00
o
       Mean
                           Diss.
 8.59
  Sus.
Influent
  5.27
                        Tot.
13.86
                                     Tot.   (A=36%)
                                   Effluent
                                                                                           8.94
       Stnd.Dev.
 5.58
  4.36
 8.31
4.55
       Range
17.00       14.00       31.00
     Figure 46.  Wastewater nickel concentrations.
                                                   12.00

-------
oo
         Mean
2.
10.15
  Tot.  (A=83%)
Effluent
  1.75
Stnd.Dev.
Range
2.55
7.00
3.41
11.00
3.75
13.00
2.54
9.00
                               Figure 47.  Wastewater lead concentrations.

-------
              400
               300
           i  200
           o
          o
00
ro
               100
         Mean
                            Diss.
49.17
  Sus.
Influent
 11.11
                         Tot.
60.28
  Tot.   (A=87%)
Effluent
   7.94
Stnd.Dev.
Range
41.06
150.00
11.67
25.00
35.76
120.00
15.92
40.00
                                    Figure  48.   Wastewater  zinc  concentrations.

-------
                                 SECTION  8

                               COLUMN  STUDIES
COLUMN OPERATION
     In order to assess the patterns  of constituent removal  with  soil  depth,
a small column was constructed to simulate the soil  profile  of Basin 1.
The laboratory column consisted of a  3.7 meter (12 foot)  section  of 0.20
meter (8 inch) PVC pipe.  It was capped on one end-, and a 0.15 meter (6 inch)
layer of gravel was placed in the bottom to serve as an underdrain system.
A 2.1 meter (7 foot)layer of sand was placed directly above  the  gravel, and
a 0.20 meter (8 inch) silt-loam layer was added on top of the sand.  Samp-
ling ports were placed at depths of 0.20 meters (8 inches),  0.56 meters
(1 foot 10 inches), 1.2 meters (3 feet 10 inches), 1.8 meters (5 feet
10 inches) from the top, and at the bottom of the column.

     The soil used for the laboratory column study was taken from throughout
the depth of Basin 1.  The top silt-loam layer was cohesive  and  facilitated
collection of an undisturbed sample,  while the sand was too  loose to permit
collection of undisturbed samples with depth.  As a result,  composite sam-
ples were collected at several basin  depths and Introduced into  the column  1n
the same order to simulate the gradation of soils found in the field.

     The sample collection apparatus at each port consisted  of a 65 mm
funnel attached to a section of rubber tubing at the discharge end.  To pre-
vent clogging of the funnel, a small  piece of fine screen was placed within
the mouth of each funnel.

     Since the amount of sample that could be collected at each sampling was
quite small, it was necessary to make a series of column runs to adequately
evaluate the soil column performance.  This series consisted of several
column loading cycles for assessing the organic removal behavior of the soil
column.  These were followed by two column runs for evaluating the removal
of phosphorus in the soil.  The following two loadings provided the data for
evaluating the flow, nitrate, nitrite, and ammonium profiles.  The next
loading was tested for calcium, hardness, alkalinity, and chlorides, and a
run for col 1 form and phosphorus analyses completed the series.

      In each series of tests, 0.30-0.33 meters (12-13 inches) of secondary
effluent was applied to the top of the column.  During the wastewater appli-
cation, a portable gravel layer was set on top of the soil while the waste-
water was applied slowly through small holes  in a plastic bag to prevent
disturbance of the top  loam layer.


                                     33

-------
 COLUMN  PERFORMANCE

      The  removal of COD, phosphorus, nitrogen, dissolved solids, and
 coliforms by  the infiltration-percolation soil column was evaluated under a
 number  of different loading cycles.  The results of these evaluations have
 been  summarized graphically in this section.

 Chemical Oxygen Demand

      The first column run was made to assess the removal of COD by the pilot
 soil  column.  Samples for testing the levels of COD removal were collected
 every 6-12 hours from each of the sample ports.  A graphical representation
 of  the  results of this run has been presented in Figure 49.  This figure
 indicates the COD concentration in the percolate water as a function of depth
 and time after loading.

      In analyzing the data on this figure, it is apparent that one of the
 most  significant characteristics of the figure was the large increase in the
 COD concentration during passage of the water through the top 0.23 meters
 (9  inches) of soil.  In fact, the COD of water removed from the 0.23 meter
 (9  inch) sampling port never dropped below the soluble COD of the applied
wastewater.   Since the COD associated with suspended solids is typically
 removed in the top few inches of soil by filtering and straining, it was
expected that the COD in the column would decrease very rapidly to a level
below the soluble COD concentration of the influent (23).  However, in this
 run there was no net removal of COD until the water had percolated through
approximately 0.46 meters (18 inches) of soil.  An increase of COD within the
 top few centimeters (cm) of soil  has not been reported by previous  investigators.
However, it has been noted that the suspended solids are rapidly removed in
 the top few centimeters of soil by filtering and straining (23).  As oxygen
enters the soil during a subsequent drying period, the retained organics are
 usually decomposed (24).  Since the oxygen diffusion rates are typically
high  at shallow depths, the trapped COD would be expected to decompose
quickly (23).  However, in this study the moisture content of the top soil
 layer was high during the entire loading schedule.  This high moisture con-
tent may have slowed the rates of oxygen transport and thus affected the
rates of organic decomposition (23, 25).  This factor may have contributed
to incomplete oxidation of the organics, resulting in some solubilization.
Subsequent loadings of the secondary effluent could have flushed this solu-
ble COD out of the top few centimeters of soil.  In time, as the soluble organics
were flushed from the soil, the COD concentration in the collected water
would decrease.  This pattern was observed to occur as can be seen by the
curves developed at the later sampling times.

     The COD increase in the upper levels of the column had no effect on the
effluent quality.   The high initial COD levels in the samples from the top
sampling port were quickly reduced by the underlying sand layer in the
column.   The COD concentration decreased with depth to a constant COD con-
centration of approximately 20 mgA at the bottom of the column.  The efflu-
ent level  of 20 mg/«,  of COD represented 50-75% removal of the applied COD.
                                    84

-------
00
                     2  -
                     4  -
                     6  -
                     8
                                                 Soil Influent, 39.9 mg/£
                                                               Influent, 72.3 mg/a
                                                                  Soil Surface
                                                               Time of Sampling
                                                   60        80
                                                   COD (mgA)
                       Figure 49.  COD Concentration variation with column depth and time.

-------
The results are consistent with those of previous studies and the actual field
operations in this study (19, 20, 26).

Phosphorus

     The behavior of phosphorus in the soil column was evaluated through
three separate loadings of secondary effluent.  Two of the phosphorus runs,
Nos. 6 and 7, were made early in the loading history of the column, and the
third was made after many column loading cycles.  This was done to gain some
insight regarding the phosphorus reduction capacity with repeated loading of
a soil matrix.  A graphical summarization of the phosphorus depth relation-
ship has been presented in Figure 50.

     The three curves plotted in Figure 50 represent the phosphorus concen-
tration of the soil solution as a function of the soil depth for each of
three separate loadings.  In all three of these curves the phosphorus con-
centration of the applied wastewater declined sharply in the top silt loam
layer.  This highly efficient removal suggests that the silt loam provided
numerous sites for phosphorus adsorption.  The rate of phosphorus removal
with increased depth declined significantly when the water reached the sand
layer.  This decline in the removal rate could be due to many factors,
although the most probable cause is related to the characteristics of the
sand media.  Because of their relatively large particle size, sandy soils
provide relatively few adsorption sites for phosphorus removal (27).  As a
result, the effectiveness of sand for phosphorus removal is not as great as
that of finer textured soils.

     While the results of the column study seemed reasonable, the effluents
collected in the field operation had phosphorus concentrations which were
significantly higher than those determined in the laboratory column study.
Since the soil used in the column was taken directly from Basin 1, it was
thought that changes in the soil matrix must have occurred.  The soil sam-
ples taken for the column study were obtained in the middle of January, 1977,
while the column construction was not completed until the end of June, 1977.
This resting period may have provided time for nearly all of the adsorbed
phosphorus to become mineralized (28).  The result of this phosphorus pre-
cipitation was the freeing of adsorption sites for the subsequent fixation
of phosphorus within the column.

Nitrogen

     In a separate loading cycle, secondary effluent was applied to the soil
matrix for analysis of the behavior of the nitrogen species.   The data
obtained from this run consisted of time profiles at each sampling port of
the ammonium, nitrite, and nitrate concentrations, and the flow quantities.

     The ammonium level  of the applied wastewater was very low in this phase
of the study, containing only 5.44 mg/a .   As can be seen in Table 9, the
ammonium concentration decreased rapidly to less than 1.0 mg/i as the water
moved through the silt loam layer.  The concentration decreased further when
the water entered the sand region.  Sampling Port No. 3 yielded ammonium
                                     86

-------
                                  Soil Surface
CO
2.0
7.0       8.0
                                         3.0      4.0       5.0        6.0
                                            Total Phosphorus (mg/Jt)
                 Figure 50.  Wastewater phosphorus concentration as a  function of  soil depth.

-------
TABLE 9.  SELECTED NITROGEN REMOVAL  DATA  (mg/Jl)
Sample Port
Influent
Port 1



Port 3



Port 5



Collection
Time
Hrs. Min.

3
5
8
19
3
5
8
19
3
5
9
19

00
30
55
40
00
30
55
45
00
30
00
50
Nitrite
.48
.46
.70
.14
.02
.02
.01
.01
.02
.05
.01
.01
.01
Nitrate
1.63
16.59
14.00
8.75
3.08
9.95
12.67
13.61
14.48
29.34
34.02
30.42
22.18
Ammon i a
5.44
.92
.77
-
.22
.23
.09
.08
.07
_
.08
.06
.07

-------
levels in the percolate water of less than 0.10 mg/Ji.   This  indicates  that
most of the ammonium ion was adsorbed or fixed in the  upper  one-half of the
column depth, with little change in the ammonium levels noted below that
point.

     As seen in Table 9, the nitrite level of the influent wastewater  was
0.48 mgA as N.  At the first sampling level  this concentration had
increased slightly to levels ranging from 0.32-0.71 mg/st, as  N.  Below  the
first sampling port the nitrite concentration dropped  to less than 0.05 mg/x,
and remained at a very low level throughout the rest of the  soil  matrix.
The increased levels of nitrite at the top sampling port would point to the
initiation of nitrification in the applied wastewater, or incomplete nitri-
fication during the previous drying period (25).  As shown by the extremely
low nitrite concentrations at Port No. 2, a majority of the  nitrite pro-
duced in the silt loam layer was quickly converted into the  nitrate form.

     The nitrate form of nitrogen was present at significantly higher
concentrations than the other nitrogen forms throughout the column depth.
The time profiles of the nitrate concentrations in the water from the various
ports are shown in Figure 51.  As can be seen from this figure, the concen-
trations of nitrate in the column increased as the water moved through the
soil.  In addition, the shape of the nitrogen discharge curves from Port
No. 1 and the effluent, displayed the characteristic first flush nitrate
peaking which has been reported from several other studies of cyclically
loaded soil systems (10, 29).

      Flow measurement on the column effluent provided  the data for the curve
of discharge as a function of time in Figure 52.  These data were  used in
conjunction with the effluent nitrogen data to  provide  the nitrogen mass dis-
charge pattern shown in Figure  53.  The  pattern  shown  in these figures com-
pared favorably with that of the actual  field  experience.

Dissolved Solids

      In  the  portion of  this column study dealing with  the behavior of dis-
solved solids  in soil systems,  analyses  were made  for  calcium, magnesium,
total hardness, alkalinity, and chlorides.  These  determinations were made on
the total water volume  collected  from each port during one  loading of secon-
dary  effluent.  The data describing  the  fate of these  constituents have  been
summarized  graphically  in  Figures  54  through  58.   Each figure presents  the
variation  in concentration  of  the  dissolved material with soil depth.

      The curves of  alkalinity,  calcium,  magnesium,  and total  hardness dis-
played similar variations  of concentration with depth.  The  concentration of
each  of  these constituents  increased substantially in  the top portions of
the soil.   The rate of concentration increased, then  gradually  declined with-
 in  the 0.61-1.2 meter (2-4 foot)  level.   At  depths greater than 1.2 meters
 (4  feet) the concentrations of these materials remained essentially constant.

      A slight difference in the concentration behavior of these four  consti-
 tuents was noted  in the top silt loam layer.   The concentrations of hardness
 and calcium increased significantly as  the wastewater moved through the fine

                                      89

-------
                                                                      Nitrate Concentrations
VO
o
                 30
             .«*  20
0)
-IJ

-------
     400
     300
rO
     200
     100
       0
         0        4

          Figure  52.
8
                                      I         1         I          1
                                           Flow Rate of Column Effluent
20
28
               12       16
                   Time (hrs)
Column underdrain flow as a function of time after loading.

-------
10
ro
                0)
• 11


 10



  9


  8


  7
                
-------
V£>
OJ
                                                     Soil  Surface

                                                             silt-loam
                                                  40       50

                                                 Ca+2 (mg/*)

                 Figure  54.   Wastewater calcium concentration as a function of column depth.
100

-------
VO
                    8
                      0
                               Soil Surface
                               T         I     v
                               Silt-loam      \
                               	*  x
                               Sand
_L
                                                   I
I
10
         25
 I
30
                                                                                            50
                                                                                           100
                                                       Q.


                                                  150  Q
                                                                                           200
35
                                15        20


                                Mg+2 (rng/A)


Figure 55.  Wastewater magnesium concentration as a function of column depth.

-------
vo
Ul
           8
                     100
                                                                   Soil Surface
150                200

 Alkalinity (mg/i  as CaC03)
                                                                               250
                        Figure  56.  Wastewater  alkalinity  as  a  function of column depth.
                                                                                                       0
300

-------
                         Soil  Surface
                         Silt-loam
                                                                                               0
                                                                                              50
                              Sand
to
               o
               CO
                   8
                                                           I
                                                                                              100
150
                                                                                             200
                                                                                                  ex
                                                                                                  01
                                      100                200                 300
                                         Total Hardness (mg/«, as CaCo3)
                          Figure 57.  Wastewater hardness as a function of column depth.

-------
        Soil Surface
        Silt-Loam
             Sand
o
to
   8
              10
30
              15        20       25


                Chlorides  (mgA)

Figure 58.   Wastewater chloride concentration  as

            a function of  column  depth.
                                                                50
                                                               100
                                                               150
                                                               200
                o
                00
35

-------
grain soil.  However, the alkalinity and magnesium showed little change in
concentration during percolation through the silt loam layer.

     Testing for chlorides in the soil column effluents showed variations
similar to those of the other dissolved constituents, although the changes
in concentration were not as large.  The chloride concentration increased
rapidly in the top 0.61 meters (2 feet) of soil and remained essentially
constant throughout the remaining column depth.

     The observed variations in concentration of the dissolved materials
within the soil matrix can be attributed to a number of reactions (30).  The
dissolved material in a soil solution may interact with the soil matrix by
ion exchange, precipitation or dissolution with the solid phase, ingestion
by microorganisms, incorporation into the soil organic matter, and reactions
with the soil air.  The large increases observed in the concentration of
alkalinity, calcium, magnesium, and total hardness indicate that a dissolu-
tion process was likely occurring as a result of the contact between the
soil and the applied wastewater.  Chemical principles suggest that mineral
dissolution would be expected when a soil solution contains dissolved salts
at a concentration level which is below equilibrium with respect to any
solid phase or mineral present.  Thus, the composition of the renovated
water ultimately would be controlled by the solubility of the various miner-
als within the soil matrix.  However, for some precipitation and dissolution
reactions, the rates are extremely slow and are controlled by kinetic and
thermodynamic constraints (30).  The materials tested in this study have
been classified as being very reactive within the soil matrix (30).  This
would account for the very rapid change of concentration within the top
centimeters of soil.

     All of the materials tested did display the same characteristics in the
lower regions of the column.  In each case, the dissolved solids showed
little change in concentration once the 1.22 meter (4 foot) column depth was
reached.  Thus, the reaction rate of the dissolution reaction had slowed and
the soil solution was near equilibrium.  This slowing of the reaction rate
would suggest that the soils below the 1.22 meter (4 foot)'level were not
significantly involved in the dissolution reaction.

Coliforms

     The concentrations of both fecal and total coliforms were determined
during the study.  A summary of these results has been presented in Table 10.
As can be seen in this table, the column removed 95% of the total and fecal
coliforrris applied within the first 0.23 meters (9 inches) of the silt loam
layer.  As the wastewater moved further through the soil matrix, the coli-
form count was reduced somewhat more, but at a slower rate.  These results
were predicted since bacterial  organisms are removed in soil by straining,
die-off, sedimentation, entrapment, and adsorption (23, 24,31).  The finer
grained silt loam layer would be expected to provide more locations for
these processes to occur.  The bottom layer was much coarser sand, and pro-
vided lower removals by the indicated processes.  The results of this column
study compare favorably with other data on the removal of bacteria in soils
(32, 33).

                                     98

-------
                  TABLE 10.  COLIFORM REMOVAL WITH DEPTH

Port 1
Port 2
Port 3
Port 4
Port 5
% Removal
Total
95
99.6
—
99.75
99. 5
% Removal
Fecal
95
99.8
—
99.98
99.98
Heavy Metals

     Figure 59 provides a summary of the behavior of heavy metals 1n the col
umn as a function of column depth.  It 1s clear from this figure that most
of the heavy metal removal occurred 1n the top 25 cm (10 in) of the soil
profile.  It is not apparent from these data whether the  decreased removal
below 25 cm (10 1n) was a mass action effect, which would be observed 1n
a homogenous soil column, or whether the surface removal was simply due to
the presence of the sllty-loam soil 1n that region.  However, 1t 1s likely
that both mass action effects and soil type effects combined to yield the
observed results, with soil type providing the predominant influence.
                                     99

-------
  0
 50
100
150
200
250
                20
             9  Cadmium
             w  Copper
             O  Nickel
             o  Zinc
             a  Chrome
             •  Lead
     40          60
Cone, (p.p.b.)
80
     Figure 59.  Wastewater Keavy metal concentration
                 variations with depth.
                          100

-------
                                REFERENCES

 1.  U.S. Department of Agriculture, Soil Conservation Service, "Soil Survey
    of Boulder County Area, Colorado."  1975.  86 pp.

 2.  CH2M-Hill, Inc. and City of Boulder, Colorado.  "Wastewater Facilities
    Plan for Boulder, Colorado."  1975.  256 pp.

 3.  U.S. Environmental Protection Agency. "Methods for Chemical Analysis of
    Water  and Wastes."  EPA-625-/6-74-003, National Environmental Research
    Center, Cincinnati, Ohio, 1974.  298 pp.

 4.  Standard Methods for the Examination of Water and Wastewater.  Four-
    teenth Edition, Academic Press* New York.  APHA, AWWA, WPCF.  1976.
    1193 pp.

 5.  Harada, H.M.,  G.H. Reid, E.R. Bennett, and K.D. Linstedt.  "A Modified
    Filtration Method for the Analysis of Wastewater Suspended Solids,"
    Journal of the Water Pollution Control Federation. 45:1853-1858.   1973.

 6.  Haise, H.R., W.W. Donnan, J.T. Phelan, L.F. Lawhon, and  D.G. Shockley.
    "The Use of Cylinder Infiltrometers to Determine the  Intake Character-
    istics of Irrigated Soils."  U.S.D.A. Agricultural Research Service
    No. 41-7.  1956.  10 pp.

 7.  Johnson, A.I.  "A Field Method for Measurement of Infiltration." Paper
    1544-F, U.S. Geological Survey Water Supply, U.S. Government Printing
    Office.  1963. 25 pp.

 8.  McGaughey, P.M. and J.H. Winneberger.  "Studies of the Failure of  Sep-
    tic Tank Percolation Systems."  Journal of the Water  Pollution Control
    Federation,  36:593-606.   1964.

 9.  Laverty, F.B., R. Stone,  and F.A.  Myerson. "Reclaiming Hyperion
    Effluent."   J. Sanitary Engineering  Division, American Society of  Civil
    Engineers, SA6:1-40.19617

10.  Bouwer, H.,  R.C.  Rice, and E.D. Escarcega.   "Renovating  Secondary  Sew-
    age by Groundwater Recharge with  Infiltration Basins." EPA-16060-DRV-
    03/72, U.S.  Environmental Protection Agency,  Washington, D.C.   1972.
     102 pp.

11.  Bouwer, H.   "Land Treatment of Liquid Waste:  The Hydrologic  System."
     IN: Proceedings of the Joint Conference  on  Recycling  Municipal  Sludges
    and Effluents  on  Land, Champaign,  Illinois.1973.   pp.  103-112.


                                    101

-------
                           REFERENCES (continued)

12.  Rice, R.C.  "Soil  Clogging During Infiltration with Secondary Effluent."
     Journal of the Water Pollution Control  Federation,  46:708-716.   1974.

13.  McGaughey, P.M. and R.B.  Krone.  "Soil  Mantle as a  Wastewater Treatment
     System."  SERL Report No.  67-11, University of California,  Berkeley.
     1967.  200 pp.

14.  Hsu, P.M.  "Adsorption of Phosphate by Aluminum and Iron in Soils."
     Soil Sci. Soc. Amer. Proc.» 28:474-478.   1964.

15.  Harter, R.D.  "Phosphorus Adsorption Sites in Soils."   Soil Sci.  Soc.
     Amer. Proc.. 33:630-632.   1969.

16.  Barrow, N.J. and T.C. Shaw.  "The Slow Reactions Between Soil and
     Anions: 2 Effect of Time  and Temperature on the Decrease in Phosphate
     Concentration in the Soil  Solution."  Soil Sci.. 119:167-177.  1975.

17.  Aulenbach, D.B., N.L. Clesceri, T.J. Tofflemire, S. Beyers, and
     L. Hajas.  "Water Renovation Using Deep Natural Sand Beds." J.  American
     Water Works  Association.  67:510-515.  1975.

18.  Aulenbach, D.B., J.J. Ferris, N.L. Clescert,  and T.J.  Tofflemire.
     "Protracted  Recharge of Treated Sewage into Sand Part III:  Nutrient
     Transport Through the Sand."  Ground Water. 12:161-169.   1974.

19.  Thomas, R.E. and T.W. Bendixen.  "Degradation of Wastewater Organics  in
     Soil."  J. Water Pollution Control Federation, 41:808-813.   1969.

20.  Schwartz, W.A. and T.W. Bendixen.  "Soil System for Liquid  Waste Treat-
     ment and Disposal:  Environmental Factors."  J. Water Pollution Control
     Federation.  42:624-630.  1970.

21.  Bouwer, H.,  J.C. Lance, and M.S. Riggs.   "High-Rate Land Treatment II:
     Water Quality and Economic Aspects of the Flushing  Meadows  Project."
     J. Water Pollution Control Federation, 46:844-859.   1974.

22.  Dornbush, J.N.  "Infiltration Land Treatment  of Stabilization Pond
     Effluent Technical Progress Report #1 and Addendum."  Unpublished.
     1976.  44 pp.

23.  Pound, Charles E.  and R.  Crites.  Wastewater  Treatment and  Reuse by
     Land Application—Volume  II.  EPA-660/2-73-006b, U.S.  Environmental
     Protection Agency, Ada, Oklahoma, 1973.   249  pp.

24.  Powell, G.M.  Land Treatment of Municipal Wastewater Effluents Design
     Factors—II.  U.S. Environmental Protection Agency  Technology Transfer,
     January 1976.  72 pp.
                                     102

-------
                           REFERENCES  (continued)

25.  Lance, J.C., F.D.  Whisler, and H.  Bouwer.   "Oxygen  Utilization  in Soils
     Flooded with Sewage Water."  Journal  of Environmental  Quality,
     2:345-350.   1973.

26.  Robeck, G.G., T.W. Bendixen, W.A.  Schwartz, and R.L. Woodward.   "Factors
     Influencing the Design and Operation  of Soil  Systems for Waste  Treat-
     ment."  J.  Water Pollution Control Federation,  36:971-983.   1964.

27.  Hinesly, T.D.  "Land Treatment Process for Wastewater  Renovation."
     Public Works, 105(2):62-66.  1974.

28.  Ellis, B.6.  "The Soil as a Chemical  Filter."  "Conference on Recycling
     Treated Municipal  Wastewater through Forest and Cropland."  EPA-660/2-
     74-003, U.S. Environmental Protection Agency, Ada, Oklahoma, March,
     1974.  pp.  47-72.

29.  Lance, J.C. and F.D. Whisler.  "Nitrogen  Balance in Soil Columns
     Intermittently Flooded with Secondary Sewage Effluent."  Journal of
     Environmental Quality, 1:180-186.  1972.

30.  Lindsay, W.L.  "Inorganic Reactions of Sewage Wastes with Soils."
     IN: Conference on Recycling Municipal Sludges and Effluents on Land,
     Champaign,  Illinois.  1973.  pp. 91-96.

31.  Gerba, C.P., C. Wall is, and J.L.  Melnick.  "Fate of Wastewater Bacteria
     and Viruses  in Soil."  Journal Irr. and Drain. Div. Proc. Amer. Soc.
     Civil  Eng.,  101:157-174.   1975.

32.  Aulenbach,  D.B.,  T.P. Glavin,  and J.A.R.  Rojas.  "Protracted Recharge
     of Treated  Sewage into Sand—Part I Quality  Changes in  Vertical Trans-
     port  through Sand."   Ground Water, 12:161-169.   1974.

33.  Page,  H.G.  and C.H. Wayman.   "Removal  of  ABS and Other  Sewage  Compon-
     ents  by  Infiltration  through  Soils."   Ground Water, 4:10-17.   1966.
                                      103

-------
                                    TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 1. REPORT NO.

  EPA-600/2-79-m
                              2.
 4. TITLE ANDSUBTITLE
 TREATMENT OF SECONDARY EFFLUENT BY  INFILTRATION-
 PERCOLATION
                                                            3. RECIPIENT'S ACCESSION NO.
             5. REPORT DATE
                August 1979 Issuing  date
             6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)
 D.  G. Smith, K.  D.  Llnstedt, and E.  R.  Bennett
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
                                                            1O. PROGRAM ELEMENT NO.
  City of Boulder, Colorado
         and
  University of Colorado
  Boulder, Colorado   80309
               1BC822
              11. CONTRACT/GRANT NO.
               R803931
 12. SPONSORING AGENCY NAME AND ADDRESS
                                                            13. TYPE OF REPORT AND PERIOD COVERED
  Robert S. Kerr  Environmental Research  Lab - Ada, OK
  Office of Research  and Development
  U.  S.  Environmental  Protection Agency
  Ada, Oklahoma   74820
               Final - 8/75 -  9/78
              14. SPONSORING AGENCY CODE
               EPA/600/15
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
         Treatment  performance of three rapid infiltration  basins receiving  secondary
    treated municipal  wastewater is  evaluated at the City of Boulder, Colorado,  Sewage
    Treatment Plant.   The prime objectives of the project are evaluation of  the  pre-
    treatment provided,  the hydraulic  load to the soil,  duration of wetting  and  drying
    cycles, and climatic influences  on  system performance.   Two of the basins  have
    unscarified beds while one has a bed surface constructed in a ridge and  furrow
    arrangement.  Year-round operation  allowed evaluation of applied effluent  and
    infiltrated water  collected from underdrains during  extreme climatic conditions.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS  C.  COS AT I Field/Group
    Land use
    Ground water
    Purification
    Quality control
    Sewage treatment
    Nutrient removal
 Land Application
 High rate infiltration
 Tertiary  treatment
 Ridge and furrow basins
 Underdrain system
 Sewage  effluents
68D
48B, E, G
 8. DISTRIBUTION STATEMENT

    RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
 UNCLASSIFIED
                                                                          21. NO. OF PAGES
 114
                                              20. SECURITY CLASS (Thispage)

                                                UNCLASSIFIED
                                                                          22. PRICE
EPA Form 2220-1 (9-73)
                                             104
                                                                     * u.s.GovEmMcmntiinim;OFFICE: 1979 -657-060/5424

-------