WATER POLLUTION CONTROL RESEARCH SERIES
13030 ELY 04/71-8
REC-R2-TI-8
DWR NO.
        BIO-ENGINEERING ASPECTS OF AGRICULTURAL. DRAINAGE

             SAN JOAQUIN VALLEY, CALIFORNIA
  DENITRIFICATION BY ANAEROBIC FILTERS
                  AND PONDS
ENVIRONMENTAL PROTECTION AGENCY0RESEARCH AND MONITORING


-------
    BIO-ENGINEERING ASPECTS OF AGRICULTURAL DRAINAGE
             SAN JQAQ.UIH VALLEY. CALIFORNIA
The Bio-Engineering Aspects of Agricultural Drainage
reports describe  the results of a unique interagency study
of the occurrence of nitrogen and nitrogen removal treat-
ment of subsurface agricultural wastewaters of the San
Joaquin Valley, California.

The three principal agencies involved in the study are
the Water Quality Office of the Environmental Protection
Agency, the United States Bureau of Reclamation, and the
California Department of Water Resources.

Inquiries pertaining to the Bio-Engineering Aspects of
Agricultural Drainage reports should be directed to the
author agency, but may be directed to any one of the three
principal agencies.
                       THE REPORTS

It is planned that a series of twelve reports will be
issued describing the results of the interagency study.

There will be a summary report covering all phases of
the study.

A group of four reports will be prepared on the phase of
the study related to predictions of subsurface agricul-
tural wastewater quality — one report by each of the
three agencies, and a summary of the three reports.

Another group of four reports will be prepared on the
treatment methods studied and on the biostimulatory
testing of the treatment plant effluent.  There will be
three basic reports and a summary of the three reports.
This report, "DENITRIFICATION BY ANAEROBIC FILTERS AND PONDS",
is one of the three basic reports of this group.

The other three planned reports will cover (1) techniques
to reduce nitrogen during transport or storage (2), possi-
bilities for reducing nitrogen on the farm, and (3) desali-
nation of subsurface agricultural wastewaters.

-------
          BIO-ENGINEERING ASPECTS OF AGRICULTURAL DRAINAGE
                    SAN JOAQUIN VALLEY, CALIFORNIA
                            DENITRIFICAT10N
                                   BY
                    ANAEROBIC FILTERS AND PONDS
                          Study conducted by
                      Robert S. Kerr Water Research Center
                    Treatment and Control Research Program
                             Ada,  Oklahoma
The agricultural drainage study was coordinated by:

Robert J. Pafford, Jr., Regional Director, Region 2
UNITED STATES BUREAU OF RECLAMATION
2800 Cottage Way, Sacramento, California 95825

Paul DeFalco, Jr., Regional Director, Pacific Southwest Region
WATER QUALITY OFFICE, ENVIRONMENTAL PROTECTION AGENCY
760 Market Street, San Francisco, California 94102

John R. Teerink, Deputy Director
CALIFORNIA DEPARTMENT OF WATER RESOURCES
1416 Ninth Street, Sacramento, California 95814
                               PROGRAM * 13030 ELY
                                     April 1971

   For sale by llio Superintendent of Documents, U.S. Government Printing Ofliec, Washington, IXC. 20-102 - Price 75 conts

-------
           REVIEW NOTICE
This report has been reviewed by
U. S. Bureau of Reclamation, and the
California Department of Water Resources,
and has been approved for publication.
Approval does not signify that the
contents necessarily reflect the views
and policies of the U. S. Bureau of
Reclamation or the California Department
of Water Resources.

The mention of trade names or commercial
products does not constitute endorsement
or recommendation for use by either of the
reviewing agencies or the Water Quality
Office of the Environmental Protection
Agency.
               ii

-------
                               ABSTRACT
The removal of nitrogen from tile drainage by means of bacterial
reduction was investigated at  the Interagency Agricultural Wastewater
Treatment Center near Firebaugh, California.  The major nitrogen form
in tile drainage is nitrate  (approximately 98 percent) .  The process
required that an organic  carbon source be added  to the waste to accom-
plish reduction of the nitrogen.  The bacterial  process was used in two
configurations, anaerobic filters and anaerobic  deep ponds.  It was found
that with the addition of approximately 65 mg/1  of methanol 20 mg/1
nitrate-nitrogen could be reduced to 2 mg/1 or less of total nitrogen
within one hour of treatment by filter denitrification at water temper-
atures as low as 14°C.  The same removal was achieved at 12°C in a filter
operating at a detention  time  of two hours.  A covered deep pond required
an actual detention time  of eight days at water  temperatures of approxi-
mately 22°C and a theoretical  detention time of  15 days at temperatures
of approximately 16°C to  accomplish the same removal.  An uncovered pond
was not able to achieve the same results at theoretical detention times
as long as 20 days.  The  projected costs for both processes are approxi-
mately 90 dollars per million  gallons.
This report was submitted in fulfillment of Project No. 13030 BW under
the sponsorship of  the Water Quality Office of the Environmental Protection
Agency .
Key Words:  Agricultural Wastes, Denitrification,  Irrigation Return  Flows,
Nitrate Removal, Anaerobic Treatment
                                  iii

-------
                              BACKGROUND
This report is one of a series which presents the findings of intensive
interagency investigations of practical means to control the nitrate
concentration in subsurface agricultural wastewater prior to its discharge
into other water.  The primary participants in the program are the Water
Quality Office of the Environmental Protection Agency, the United States
Bureau of Reclamation, and the California Department of Water Resources,
but several other agencies also are cooperating in the program.  These
three agencies initiated the program because they are responsible for
providing a system for disposing of subsurface agricultural wastewater
from the San Joaquin Valley of California and protecting water quality in
California's water bodies.  Other agencies cooperated in the program by
providing particular knowledge pertaining to specific parts of the
overall task.

Ultimately,  the need to provide subsurface drainage for large areas of
agricultural land in the western and southern San Joaquin Valley has been
recognized for some time.  In 1954, the Bureau of Reclamation included a
drain in its feasibility report of the San Luis Unit.  In 1957, the
California Department of Water Resources initiated an investigation to
assess the extent of salinity and high ground water problems and to
develop plans for drainage and export facilities.  The Burns-Porter Act,
in 1960, authorized San Joaquin Valley drainage facilities as a part of
the State Water Facilities.

The authorizing legislation for the San Luis Unit of the Bureau of
Reclamation's Central Valley Project, Public Law 86-488, passed in June
I960, included drainage facilities to serve project lands.  This Act
required that the Secretary of Interior either provide for constructing
the San Luis Drain to the Delta or receive satisfactory assurance that
the State of California would provide a master drain for the San Joaquin
Valley that would adequately serve the San Luis Unit.

Investigations by the Bureau of Reclamation and the Department of Water
Resources revealed that serious drainage problems already exist and that
areas requiring subsurface drainage would probably exceed 1,000,000 acres
by the year 2020.  Disposal of the drainage into the Sacramento-San Joaquin
Delta near Antioch, California, was found to be the least costly alternative
plan.

Preliminary data indicated the drainage water would be relatively high in
nitrogen.  The then Federal Water Quality Administration conducted a study
to determine the effect of discharging such drainage water on the quality
of water in the San Francisco Bay and Delta.   Upon completion of this study
in 1967, the Administration's report concluded that the nitrogen content
of untreated drainage waters could have significant adverse effects upon
the fish and recreation values of the receiving waters.  The report
recommended a three-year research program to establish the economic
feasibility of nitrate-nitrogen removal.

-------
As a consequence, the three agencies formed the Interagency Agricultural
Wastewater Study Group and developed a three-year cooperative research
program which assigned specific areas of responsibility to each of the
agencies.  The scope of the investigation included an inventory of
nitrogen conditions in the potential drainage areas, possible control of
nitrates at the source, prediction of drainage quality, changes in nitrogen
in transit, and methods of nitrogen removal from drain waters including
biological-chemical processes and desalination.
                                 VI

-------
                                CONTENTS


SECTION                                                                PAGE

           Abstract	iii

           Background 	     v

           Table of Contents	vii

           Figures	vlii

           Tables 	     X

   I       Summary and Conclusions	     1

              Summary of Anaerobic Filter Results 	     1
              Summary of Anaerobic Deep Pond Denitrification  ....     2

  II       Introduction 	     3

              Water Quality Criteria	     3
              Process Background	     3
              Waste Treatment by Dissimilatory Nitrogen Reduction  .  .     7

 III       Experimental Procedures  	    10

              Water Quality	    10
              Apparatus	    10
              Methanol Additions  	    15
              Process Evaluation Procedures  	    15

  IV       Results and Discussion	    19

              Filter Denitrification	    19
              Pond Denitrification	    43
              Consumptive Ratio for Field Processes  	    52
              Regrowth Studies  	    52
              Botulism Studies	    52
              Process Cost Estimates	    53

   V       Acknowledgments	    63

  VI       References	    64

 VII       Publications	    67
                                   vii

-------
                                   FIGURES


                                                                        PAGE

 1     The Nitrogen Cycle As Presented By Sawyer  (2)                       4

 2     Anaerobic Processes For Denitrification                            8

 3     Pictorial Of Pilot Scale Filter                                    13

 4     Example Of Hydraulic Tracer Response Curve                         17

 5     Influent Pressure Versus Time For Activitated  Carbon and
       One Inch Diameter Aggregate Medias                                 24

 6     Total Nitrogen Removal And Influent  Pressure Versus  Time
       For Sand Filled 18-Inch Diameter Filter With One  Hour
       Detention                                                          25

 7     Predicted Seasonal Variation Of Nitrogen Concentration And
       Temperature                                                        28

 8     Effluent Nitrogen Versus Water Temperature                         30

 9     Predicted Detention Times For Projected Seasonal  Variations
       Of Nitrogen And Projected Minimum Water Temperatures               31

10     Total Effluent Ammonia Plus Organic  Nitrogen Versus  Time            34

11     Results Of Hydraulic Tracer Studies  On Anaerobic  Filters
       Containing One Inch Diameter Aggregate                             35

12     Results Of Hydraulic Tracer Studies  On Anaerobic  Filters
       Containing One Inch Diameter Aggregate After 12-14 Months
       Of Continuous Operation                                            36

13     Results Of Hydraulic Tracer Studies  Performed  On  Pilot
       Scale Filter                                                       41

14     Results Of Pressure Profiles For Pilot Scale Filter                 42

15     Hydraulic Tracer Results For The Uncovered Deep Pond               45

16     Predicted Detention Time For Treatment Of  Agricultural Return
       Waters By Covered Pond Denitrification                             49

17     Hydraulic Tracer Results For The Covered Deep  Pond                  50
                                    viii

-------
                            FIGURES - (Continued)

                                                                        PAGE

18     Predicted Seasonal Variation Of  Tile Drainage Flow and
       Nitrogen Concentrations From San Joaquin Valley,  California         54

19     Schematic Diagram - Filter Denitrification                          55

20     Schematic Diagram - Pond Denitrification                           59
                                     IX

-------
                                 TABLES
NO.                                                                   PAGE

 1     Characteristics Of Tile Drainage Used At The Interagency
       Agricultural Wastewater Treatment Center And Average
       Mineral Concentrations Of Irrigation Waters                      11

 2     Laboratory Analysis                                              18

 3     Summary Of Media And Operational Criteria                        21

 4     Nitrogen Removal Efficiencies Of Experimental Media              22

 5     Influent Pressures And Nitrogen Removal For Experimental
       Media                                                            23

 6     Comparison Of Profile Nitrate Plus Nitrite Nitrogen
       Concentrations At Various Temperatures For Short And Long
       Term Operation                                                   29

 7     Comparison Of Effluent Nitrogen Concentrations At Various
       Temperatures For Short And Long Term Operation                   33

 8     Total And Volatile Solids Effluent Data For Filters Filled
       With One-Inch Diameter Aggregate                                 34

 9     Effluent Nitrogen Concentrations At Various Temperatures For
       An Algae Supplemented Filter Influent                            38

10     Operations And Effluent Nitrogen Summary For Pilot Scale
       Filter                                                           40

11     Operation And Nitrogen Concentration Summary For The
       Uncovered Deep Pond                                              44

12     Data Summary Of The Total And Volatile Suspended Solids  For
       The Uncovered Anaerobic Deep Pond                                46

13     Operating And Effluent Nitrogen Concentration Summary For
       The Covered Deep Pond                                            47

14     Data Summary Of Suspended And Volatile Suspended Solids  In
       The Covered Anaerobic Deep Pond                                  51

15     Filter Denitrification Design Criteria                           56

16     Capital Costs For Filter Denitrification Design
       Capacity - 228 MGD                                               57

-------
                              TABLES - (Continued)
NO.                                                                  PAGE

17    Treatment Costs for Filter Denitrification Design
      Capacity - 228 MGD                                              58

18    Pond Denitrification Design Criteria                            59

19    Capital Costs for Deep Pond Denitrification Design
      Capacity - 228 MGD                                              60

20    Treatment Costs for Pond Denitrification                        61
                                   XI

-------
                              SECTION I

                        SUMMARY AND CONCLUSIONS
The field feasibility studies of anaerobic denitrification of agricultural
tile drainage were completed in December 1969.  The experiments were
designed to investigate the possibility of removing nitrogen by means of
anaerobic filters and anaerobic covered and uncovered deep ponds.  It
was concluded from the presented experimental work that removal of
nitrogen from agricultural tile drainage is technically feasible by means
of bacterial denitrification in anaerobic filters and covered anaerobic
ponds.  Furthermore, according to preliminary cost estimates for the two
processes, the cost of treatment for agricultural tile drainage will be
92 dollars per million gallons for anaerobic filters and 88 dollars per
million gallons for covered anaerobic ponds.  Summaries of secondary
conclusions and experimental findings for each of the above processes
follow.
                  Summary of Anaerobic Filter Results

In evaluation studies of various media for use in anaerobic filters, it
was determined that the most feasible medium in terms of performance,
cost, and operational control was rounded aggregate having a size range
of 0.75 to 1.5 inches in diameter.  Early performance of an anaerobic
filter containing one inch diameter aggregate and operated at a 1-hour
hydraulic detention time within a water temperature range of 14 to 20°C
showed that an influent nitrate-nitrogen concentration of 20 mg/1 could
be reduced to 2 mg/1 or less.  A filter containing the same size medium
and receiving the same influent nitrogen concentration but operated on a
2-hour detention time within a temperature range of 12 to 22°C produced an
effluent having a total nitrogen concentration of 2 mg/1 or less.  Pre-
liminary investigations indicate that hydraulic detention times as long as
7.5 hours may be necessary to produce an effluent containing a 2 mg/1 or
less total nitrogen concentration at water temperatures lower than 14°C
and influent nitrate-nitrogen concentrations of up to 35 mg/1.  Long term
experimentation with anaerobic filters demonstrated that operational
problems will eventually occur due to excessive bacterial growth within the
filters.  The excess growth led to an increase in total Kjeldahl nitrogen in
the effluent discharged from a filter.  Furthermore, the bacterial mass
results in a change in hydraulic patterns within the filter.  The change
had an adverse effect on the nitrogen removal capacity of the filters.
Future experimentation is necessary to determine methods of controlling
and removing excess bacterial growth from the filters.  Studies with a
pilot scale filter have not as yet led to the nitrogen removal results
comparable to those obtained with the smaller filters.  The operation of
the pilot scale filter was adversely affected by operational start-up
problems and detention time changes imposed before a sufficient bacterial
mass had accumulated.
                                   -1-

-------
Results of studies on the use of algae-laden water as an influent to
an anaerobic filter indicate that the algae do not interfere with
nitrate-nitrite removal.  The accumulation of algal cells within the
filter and the decomposition thereof, along with suspended algae passing
through the filter eventually resulted in the production of excessively
high effluent concentrations of total ammonia plus organic nitrogen
(range 2 to 5 mg/l-N).  Based on studies by the California Department of
Fish and Game, it is not expected that full-scale anaerobic filters would
offer any major threat to waterfowl by furnishing a possible habitat for
Type C Botulism.
             Summary of Anaerobic Deep Pond Denitrification

Experimental data obtained with the uncovered pond show that at no time
did the pond meet the 2 mg/1 total nitrogen effluent criterion.  Operational
problems which would occur in such a pond would include the prevention
and/or elimination of algal blooms that may occur; oxygen production by
the algae would hinder the establishment of the anaerobic conditions;
and in large ponds, wind mixing would bring about an input of dissolved
oxygen that would be much more thoroughly distributed than, was the case
in the pond used for the studies reported herein.  Although an outbreak of
botulism in uncovered ponds could possibly occur; outbreaks can be prevented
by following suitable operational practices.

The covered pond was capable of reducing an influent nitrate-nitrogen
concentration of 20 mg/1 to 2 mg/1 or less of total nitrogen at water
temperatures as low as 14°C at a 15-day theoretical hydraulic detention
time.  With the temperatures at 20° to 22°C, it could produce an effluent
having a total nitrogen concentration of 2 mg/1 or less at an actual
detention time of 8.2 days.  When the actual detention time was shortened
to 5 days, the effluent requirement of 2 mg/1 total nitrogen could not be
met.  Instead the average effluent concentration remained at approximately
4 mg/1 total nitrogen.  In the experiments which have been completed,
successful operation at temperatures below 14°C has not been achieved.
However, continuing experimentation is expected to show that the 2 mg/1
total nitrogen criterion can be met during the colder months of the year.
Operational problems with a covered pond would be limited to those per-
taining to the hydraulic regime; however, if properly designed these should
be minimized.  Due to the covering of the pond, no operational problem would
be encountered with respect to algal blooms or wind mixing.  Additionally,
the possibility of Type C Botulism is eliminated by the use of a covered
pond, since the water surface would not available to waterfowl.
                                     -2-

-------
                              SECTION II

                             INTRODUCTION
This report presents the results of the experimental studies on bacterial
denitrification of agricultural tile drainage performed at the Interagency
Agricultural Wastewater Treatment Center (IAWTC) near Firebaugh, California.
Specifically, the report deals with the feasibility portion of a continuing
study on the removal of nitrate-nitrogen from subsurface tile drainage
waters occurring in the San Joaquin Valley by the use of anaerobic filters
and/or anaerobic deep ponds.  The studies were initiated in June 1967 and
will be continued through December 1970.  The last year is being devoted
to making operational refinements, the results of which will be published
in a future report.

The objectives of the first phase of the studies were to determine the
feasibility of the processes under study to remove nitrogen from tile
drainage under field conditions, and to develop preliminary information
on the treatment costs.  To satisfy these primary objectives, the pilot-
scale studies discussed in this report were designed to determine:  (1) the-
relationship between the maximum practical nitrogen removal efficiency and
the minimum hydraulic detention time for anaerobic covered ponds, uncovered
ponds, and filters; (2) the actual organic carbon requirement; (3) the
significance of different media on anaerobic filter performance; and (4)
the effect of prolonged operation on treatment efficiency and process
operation and maintenance costs.
                        Water Quality Criteria

In a report by the Federal Water Quality Administration, it was recommended
that the total nitrogen content in the treated tile drainage not exceed
2 mg/1 (1).  This recommendation was made the effluent criterion that
determined whether or not any treatment process under consideration was
technically acceptable.  In this report the term "total nitrogen" refers
to the sum of the nitrogen concentrations of the following compounds;
nitrate, nitrite, ammonia, and organic nitrogen.
                          Process Background

The relationships existing between the various forms of nitrogen in the
biosphere are best illustrated by the diagram of the nitrogen cycle
developed by Sawyer (2) and reproduced in this report in Figure 1.  The
basic biological phenomena involved in the nitrogen cycle are:  fixation
of atmospheric nitrogen, assimilation of inorganic nitrogen into cellular
material, decomposition of cellular organic nitrogen to inorganic nitrogen,
and reduction of oxidized inorganic nitrogen forms by bacteria to gaseous
nitrogen.  The processes described herein entail only the latter part of
the above cycle.  Bacterial nitrogen reduction is divided into two classes:
                                    — 3—

-------
assimilatory and dissimilatory nitrogen reduction.  In assimilatory
nitrogen reduction, nitrate and/or nitrite-nitrogen is reduced to ammonia
prior to incorporation into cellular material.  In dissimilatory nitrogen
reduction, the oxidized anions serve as exogeneous hydrogen acceptors
for the oxidation of organic carbon (3).  The principle end product of
dissimilatory nitrogen reduction is gaseous nitrogen.  The process
described in this report is primarily dissimilatory in nature.
Dissimilatory Nitrogen Reduction

Dissimilatory nitrogen reduction, commonly spoken of as denitrification,
can be brought about by a large number of commonly occurring facultative
bacteria (4).  It occurs most frequently under anaerobic, or nearly anaer-
obic, conditions.  The exact degree of the inhibition of denitrification
by the presence of oxygen is not known (5) (6).   However, it is a well
accepted fact that the inhibition is a non-competitive type due to a
large difference in the reaction rates for oxidized nitrogen utilization
and oxygen utilization (7).

Some bacteria can only reduce nitrate to nitrite, others can only reduce
nitrite to molecular nitrogen, while a third group can bring about the
               FIGURE I - THE NITROGEN CYCLE AS PRESENTED BY SAWYER (2)
                                    -4-

-------
reduction of nitrate and/or nitrite to molecular nitrogen  (8).  However,
the enzymatic reactions involved for all of the different  organisms are
basically the same.  Taniguchi, et al, proposed a biochemical pathway for
nitrate reduction, which with the exception of the terminal enzyme, is
the same as the pathway involved when oxygen is used as the terminal
hydrogen acceptor  (9).  In denitrification, nitrate reductase is the
terminal enzyme while cytochrome oxidase is used when oxygen is the
terminal acceptor.  Heredia and Medina (10) discovered an  alternate
pathway involving Vitamin K~ while studying Escherichia coli.  This
pathway is operative under both aerobic and anaerobic conditions, and is
most likely the one associated with assimilatory nitrate reduction.  Others
have determined that the proposed pathways apparently are  reasonably repre-
sentative of the general pathway of nitrogen reduction (3).

Assuming a heterogeneous population of bacteria, an adequate supply of
organic carbon, and relatively anaerobic conditions, any or all of the
following half reactions may take place:

     1/2 N03~ + H+ + e~ = 1/2 HO + 1/2 NO ~	(1)

     1/3 N02~ + H+ + e~ = 1/3 H20 + 1/6 N2 + 1/3 OH~	(2)

     1/5 N03~ + H+ + e~ = 2/5 H20 + 1/10 N2 + 1/5 OH~	(3)

By combining these half reactions with the half reaction for the oxidation
of a biologically degradable organic carbon source, a set  of balanced
stoichiometric equations for dissimilatory nitrogen reduction can be
derived.  For example, the following equations are derived if the organic
carbon source being oxidized is methanol (CHoOH):

     N03~ + 1/3 CH3OH = N02~ + 1/3 C02 + 2/3 H20	(4)

     NO ~ 4- 1/2 CH OH = 1/2 N  + 1/2 CO  + 1/2 H 0 + OH~	(5)
       Z          J          &-         £-        £.
     NO ~ + 5/6 CH3OH = 1/2 N2 + 5/6 C02 + 7/6 HO + OH~	(6)

Equations 4, 5 and 6 constitute the theoretical foundation upon which
waste treatment by dissimilatory nitrogen reduction is  based.   They are
used to determine the amount of organic carbon required for the reduction
of a given quantity or concentration of nitrogen in a waste.  However, if
the quantity of organic carbon determined .by these equations were used,
denitrification would not be complete.  Additional carbon must be added to
supply that needed for cell growth.  There are little data in the literature
to indicate what this additional requirement might be.   But it is known
that it varies from organism to organism, with the nature  of the organic
carbon source being used and the environment in which the denitrification
is taking place (11)(12).
                                     — 5—

-------
Consumptive Ratio

The consumptive ratio is defined as one obtained by dividing the actual
quantity of organic carbon source required to denitrify and deoxygenate a
waste plus the carbon required for cell growth by the stoichiometric
requirement for denitrification and deoxygenation of the waste (13).  It
is used to quantify the carbon source needed to complete the denitrification
process.  A consumptive ratio value of one would indicate that no chemical
was required for cell synthesis.  A ratio greater than one is to be expected.
The higher the ratio the greater is the chemical requirement for biological
growth.

The actual consumptive ratio for a particular set of conditions must be
determined experimentally.  Once a consumptive ratio value is determined,
equations can be developed for predicting the amount of the bacterial
cells that will be synthesized.  Using a typical empirical formulation
of C5H702N for bacterial cells  (14), the following equations were deve-
loped by McCarty, et al, for the denitrification process when methanol is
used as the carbon source with a consumptive ratio equal to 1.3, as deter-
mined in their study (11).

Overall nitrate removal:

     N03~ + 1.08 CH3OH + H+ = 0.065 CjH^N + 0.47 N2 + 0.76 C02 +

     2.44 H20	(7)


Overall nitrite removal:

     N02 + 0.67 CH3OH + H+ = 0.04 C HyOjN + 0.48 N£ + 0.47 (X>2 +

     1.7 H20	(8)


Overall deoxygenation:

     02 + 0.93 CH3OH + 0.056 N03~ + 0.056 H+ = 0.056 C5H?02N +

     0.65 C02 + 1.69 H20	(9)


The consumptive ratio of 1.3 was used for Equations 7, 8, and 9 even though
the consumptive ratio for deoxygenation alone probably is different from
that for denitrification alone  (11).  However, as long as the ratio of
nitrogen to oxygen in the influent waste stream is fairly large, the
results of using such predictive equations should be sufficiently accurate
for most cases.  It should also be noted that even though dissimilatory
nitrogen reduction is the principle mode of'nitrogen removal, some assimi-
latory nitrogen reduction takes place.  For example, if the influent waste
stream contains 20 mg/1 of nitrate-nitrogen and 8 mg/1 of dissolved oxygen,
                                     -6-

-------
according to Equations 7, 8, and 9, 1.5 mg/1 of the influent nitrate-nitrogen
would be assimilated into the organic nitrogen of the 12.1 mg/1 of cellular
biomass produced.  These relationships are more clearly illustrated by con-
verting the molar equivalents in Equations 7, 8, and 9 to milligrams per
liter as follows:

          10 mg/1 of NOo-N plus 24.7 mg/1 of methanol will produce
               5.25 mg/1 of cells which contain 0.65 mg/1 of Organic-N

          10 mg/1 of N02~N plus 15.3 mg/1 of methanol will produce
               3.22 mg/1 of cells which contain 0.4 mg/1 of Organic-N

          10 mg/1 of dissolved oxygen plus 9.3 mg/1 of methanol plus
               0.35 mg/1 of N03-N will produce
               2.0 mg/1 which contain 0.25 mg/1 of cells of Organic-N

Because of its relation with process efficiency and treatment costs, the
determination of the consumptive ratio for denitrification of agricultural
wastewater was of importance in the field work.
           Waste Treatment by Dissimilatory Nitrogen Reduction

As with most biological waste treatment techniques, there are several
different process designs that may be used for denitrification.  In
Figure 2, three denitrification process designs are diagramed:  anaerobic
activated sludge, anaerobic ponds, and anaerobic filters.  In an activated
sludge plant, the waste is aerated until it is nitrified.  It is then mixed
with a small proportion of untreated plant influent which serves as the
organic carbon source and is then treated under anaerobic conditions to
allow denitrification to take place (15).  Because of the short detention
times in such a plant, it is necessary to separate the bacteria cells
produced from the effluent of the denitrification unit, and recycle them to
the influent of the unit.  Because of the low cell production that takes
place during denitrification, cell separation and recycling can be relat-
ively difficult to accomplish.  In addition, due to variations in the
quality of municipal and industrial wastes, the anaerobic activated sludge
method of denitrification has been found difficult to control (16).

In the anaerobic pond denitrification process the bacteria are "free
floating" in the waste.  In order to accomplish successful treatment, it
is necessary that the reproduction rate of the bacteria be equal to or
greater than the hydraulic detention time of the treatment vessel.  If this
criterion is not met the bacteria population will be washed out of the system.
Studies conducted with the use of completely and partially mixed simulated
anaerobic ponds, in which methanol was used as the organic carbon source,
show that the required hydraulic detention times are on the order of days,
rather than of hours as for anaerobic activated sludge process (12)(17)(18).
The long detention times requires substantial land area for treatment
                                    -7-

-------
CHEMICAL
INFLUENT

1 .
i

^
a

'•o
RECYCLE

_J
              ANAEROBIC ACTIVATED  SLUDGE
CHEMICAL
                                            EFFLUENT
                RECYCLE
                    ANAEROBIC  POND
                                     EFFLUENT
              CHEMICAL
              INFLUENT
                               1 1,
                    ANAEROBIC FILTER
       FIGURE 2 - ANAEROBIC PROCESSES FOR DENITRIFICATION

-------
plants using the anaerobic pond system.  When land availability is not
a problem, the anaerobic pond can be an attractive method.

The third anaerobic process for denitrification utilizes a reactor vessel
which is filled with an inert biological support medium.  The support
medium (or packing material) provides a means of retaining the bacteria
necessary for the attainment of efficient denitrification within the
treatment vessel.  By retaining the bacterial culture on an inert medium,
the solids retention time of the process can be prolonged beyond the
hydraulic detention time without solids separation and recycle.  This type
of unit is referred to as an anaerobic filter.  The necessary anaerobic
conditions are maintained by operating the unit in such a way as to keep
the media bed completely submerged with the waste being treated.  In 1959,
Finsen and Sampson (19) reported on an experimental study in which they
passed a nitrified waste upward through a column filled with 1-inch dia-
meter glass marbles using cane sugar molasses as the organic carbon source.
The unit was operated with a hydraulic detention time of 2.5 hours, but
most of the denitrification took place within 37 minutes after entry into
the unit.  A few other studies of denitrification in anaerobic filters
have been conducted in which glucose or methanol was used as the organic
carbon source (20)(21)(22).  Findings made in these investigations were
similar to those by Finsen and Sampson.  They all indicate that anaerobic
filter denitrification is a process characterized by high removal efficiency
and a short required hydraulic detention time comparable to that for
anaerobic activated sludge denitrification, and yet not beset with the
operating difficulties characteristic of the latter process.
                                     -9-

-------
                              SECTION III

                        EXPERIMENTAL PROCEDURES
To determine the technical feasibility of denitrifying agricultural tile
drainage under field conditions, it was essential to test the processes
under the nearest to actual conditions that were attainable at the time.
The following is a description of the water, apparatus and procedures
used in the experimental work at the Interagency Agricultural Wastewater
Treatment Center (IAWTC) for this purpose.
                             Water Quality

The water used for the experimental work was taken from a tile drainage
system which serviced a 400-acre field.  The land was used primarily for
the cultivation of rice during the summer and for barley during the winter.
The water requirement for the two crops differs greatly, thereby causing
the quantity of irrigation water applied to have a large seasonal variance.
The variance in turn led to a seasonal change in the dissolved minerals
of the drainage from this field.  In Table 1 are listed the mineral con-
centration ranges of the tile drainage used at the IAWTC, and also the
average mineral concentrations of the irrigation water.  The higher
dissolved mineral concentrations are present during the winter when tile
drainage return flows decrease.

The nitrogen in the tile drainage was essentially in the form of nitrate.
Nitrite-nitrogen was not detected at concentrations greater than 0.1 mg/1;
while organic nitrogen generally was constant at a concentration of approx-
imately 0.4 mg/1, and ammonia-nitrogen concentration was usually undetectable.
It should be noted that the nitrogen concentration in this particular tile
system varied to such an extent that it was necessary during high flow
seasons to supplement the tile drainage with sodium nitrate so that the
desired nitrate concentrations be maintained.  During the low flow season,
it was sometimes necessary to supplement the quantity of tile drainage
flow in order to have a sufficient supply for the experiments.  The
supplemental water was taken from the Delta-Mendota Canal which supplies
irrigation water for the central San Joaquin Valley.  Water from the canal
has a total dissolved solids of less than 500 mg/1 and a negligible nitro-
gen concentration.  Therefore, when canal water was used, it was necessary
to increase the nitrate concentration.
                               Apparatus

To fulfill the objectives of the study, various pieces of experimental
equipment had to be fabricated.  The majority of the apparatus used in
filter experimentation was made of commonly available materials and was
constructed by the Center's personnel.  Other apparatus such as large-scale
ponds were constructed by contract.  The following is a summary of the
field apparatus used at the Interagency Agricultural Wastewater Treatment
Center.

                                   -10-

-------
                                TABLE 1

        CHARACTERISTICS OF TILE DRAINAGE USED AT THE INTERAGENCY
                AGRICULTURAL WASTEWATER TREATMENT CENTER
                                  and
          AVERAGE MINERAL CONCENTRATIONS OP IRRIGATION WATERS
      CONSTITUENT
RANGE OF MINERAL
CONCENTRATIONS IN
  TILE DRAINAGE
      mg/1
AVERAGE MINERAL
CONCENTRATION OF
IRRIGATION WATER
       mg/1
Bicarbonate
Boron
Calcium
Chloride
Magnesium
Nitrogen
Phosphate
Potassium
Sodium
Sulfate
Pesticides (CHC)
Total Dissolved Solids
5 Day BOD
COD
Dissolved Oxygen
280-330
4-15
160-390
310-640
70-230
5-25
0.13-0.33
4-11
620-2050
1500-3900
0.001
2500-7600
1-3
10-20
7-9
90
0.3
20
60
10
1
0.5
3
50
65

300



Anaerobic Filters

Anaerobic filters were constructed in three sizes.  Initially, small-scale
units were used in studying the process feasibility under field conditions.
Once shown that the process did work, larger-scale units were constructed
to continue the feasibility work and to conduct operational studies.
Having successfully completed this phase, a pilot-scale unit was erected
to study the effect of larger-scale operations on process efficiency.

Small-Scale Units.  Units used for the initial feasibility studies of
filter denitrification were constructed of a 4-inch diameter polyvinyl
chloride (PVC) pipe.  These filters contained 6.5 feet of media and were
plumbed to provide an upward flow of water.  Larger-scale experimental
units were constructed of 18- and 36-inch diameter reinforced concrete
pipe containing six feet of media.  In the 18-inch diameter filters the
influent distribution system consisted of a perforated PVC pipe extending
across the filter bottom.  The 36-inch diameter filters had a media support
screen located approximately four inches above a flow distribution system
which consisted of perforated PVC pipe placed in a cross-configuration.  The
flow to these units was controlled by rotameters and the methanol was injected
by positive displacement chemical feed pumps immediately prior to entrance
                                   -11-

-------
to the filter.  To permit extractions of samples from within the filter,
profile sample ports were located at the depth quarter points of the
18-inch diameter filters, and at one-foot intervals of the 36-inch
diameter filters.  The sampling devices consisted of lengths of perforated
PVC pipe placed diametrically across the filter diameter to make it pos-
sible to take representative samples.

Media selected for the experimental filters were obtained from local
commerical sources or from natural deposits and were graded with two
sets of screens.  The smaller media were graded to pass a 3/8-inch sieve
and be retained on a 1/4-inch sieve.  The larger media were graded to pass
a 1-1/4-inch sieve and be retained on a 7/8-inch sieve.  The media used
were activated carbon, aggregate, coal, volcanic cinders, sand, and a
commercially produced artificial media.

Pilot-Scale Filter.  The successful operation and performance of the small
scale experimental units justified the construction of a pilot-scale
filter similar in hydraulic and structural design to a projected full-scale
unit.  Such a pilot filter was constructed in the spring of 1969.  A
pictorial diagram of the unit is shown in Figure 3.  The unit is of wood
construction having dimensions of 10-feet by 10-feet square and with a
water depth of 7 feet.  The medium depth was set at 6 feet, and was
supported by a false bottom similar to that used in sand filtration units.
The medium surface support was constructed to form an 8.5-inch deep
reservoir at the filter bottom.  The medium selected for this filter was
rounded aggregate graded to pass a 1-1/2-inch sieve and be retained on a
3/4-inch sieve.  This selection was based on the success of the smaller
filters which contained 1-inch diameter aggregate.  The filters influent
system was a series of five perforated pipes 9 feet in length installed
at equal intervals across the plenun chamber.  During most of the time
in which the filter was operated, only the center influent pipe was used
for distribution of the water.

Flow through the unit was measured by a 60° "V" notch effluent weir.
Methanol addition was accomplished by its injection into the influent
line just prior to entering the filter.  An extensive series of sample
taps were installed which allowed a total of 75 samples to be taken from
within the media bed.  As shown in the pictorial diagram in Figure 3 five
sampling levels were provided, and at each level 3 sample manifolds were
installed.  Each manifold had 5 sampling ports connected to PVC pipe of
either 1, 3, 5, 7, or 9 feet in length, thus forming a sampling pattern
which transected the medium bed.  The sample ports made it possible to
monitor nitrogen and pressure profiles.  Also installed within the filter
were 4 temperature probes, one near each wall approximately at midmedium
depth.  The probes were used to detect any large temperature differential
which might exist between sections of the filter and which could alter
the hydraulic regime within the filter.
                                   -12-

-------
Temperature Controlled Filter.  As experimentation progressed,  it became
apparent that an in-depth study was needed of the temperature effect on
filter denitrification.  For this purpose a filter was  placed in a tem-
perature controlled environment.  A 4-inch diameter filter constructed  of
PVC pipe was installed within an incubator capable of maintaining high
and low temperatures.  The total medium depth of the filter was 6 feet
divided into 3 sections of 2-foot lengths.  Each section was fitted with
profile sample taps similar to the larger filters.  The medium selected
for this filter was .375-inch diameter glass beads.  Glass beads were
selected to provide a known surface area for the bacteria population.
Operation of the unit involved placing a 24-hour tile drainage supply in
the incubator prior to introducing it into the filter in order that it
could attain the proper temperature.  Thus, two water reservoirs were
contained within the incubator at one time, one being brought to the
                                                        OVERFLOW  WEIR
                              NFLUENT
                               ISTRIBUTION
                                SYSTEM
    WATER VOLUME 310 FT.3
    MEDIA VOLUME 600 FT.3
               FIGURE 3- PICTORIAL  OF PILOT SCALE  FILTER
                                   -13-

-------
required temperature while the other was being pumped to the filter.
Methanol was premixed with the water entering the filter, precaution being
taken to eliminate the development of bacteria within the holding tank
which would lead to denitrification of the irrigation return water before
it entered the filter.  The temperature environment for this filter was
varied from 7.5°C to 25°C.
Anaerobic Ponds

Deep pond anaerobic denitrification was studied in two phases.  In the
first phase, feasibility studies were done in small-scale simulated ponds.
One of the objectives of this phase was to confirm under field conditions
the laboratory work performed by McCarty (17).  In the second phase the
results of the initial studies were used in designing and conducting
large-scale operations.

Simulated Deep Ponds.  A series of six small scale deep ponds, 3 feet in
diameter and varying in depth from 6 to 11 feet were constructed of rein-
forced concrete pipe.  Three of the ponds were equipped with covers to
eliminate wind mixing and algal growth, while the remaining three were
left as open ponds.  Flow through the ponds was regulated by positive
displacement pumps, and methanol was added to a common intake line for
all simulated ponds.

Large-Scale Ponds.  Two large-scale earthen ponds were constructed.  The
larger pond had a water surface dimension of 50 feet by 200 feet and
a capacity of approximately 750,000 gallons at a depth of 14 feet.  It
was covered with 2-feet x 8-feet x 3-inch styrofoam planks.  This floating
cover served to eliminate light, wind mixing, and provided an insulating
layer which reduced temperature differentials within the pond.  The cover
also functioned as a means of collecting in-pond samples without the need
for construction of elaborate scaffolding.  Flow through the pond was
monitored by means of a 60° "V" notch weir located at its effluent.

A second pond was  constructed adjacent to the above described pond.  This
pond had a water surface dimension of 50 feet x 50 feet and a capacity
of 220,000 gallons at a depth of 14 feet.  The pond was not covered in
order that the effects of wind mix, light, and differential temperatures
could be monitored.  Flow through the pond was monitored by a recording
flow meter.  Methanol injection to both ponds was accomplished by positive
displacement chemical feed pumps, injection being made into the influent
pipes of the ponds.

To maintain bacterial biomass within the ponds, flow from the ponds bottoms
between the influent and effluent was recirculated and mixed with the
respective influents.  The recirculation rates for the ponds were about 50
and 35 gpm for the covered and uncovered ponds respectively and were constant
during most of the study period.
                                    -14-

-------
                          Methanol Additions

In the initial studies, the concentration of methanol added to the
filters was determined according to the following equation which was
developed by McCarty (11) ;
     C  = [1.90 Nft + 1.18 N, + 0.67 Dn] C  .......... (10)
      m          u         1         u   r

        C  = required methanol concentration, mg/1

        Nn = initial nitrate- nitrogen concentration, mg/1
        N  = initial .nitrite-nitrogen concentration, mg/1

        DO = initial dissolved oxygen concentration, mg/1

        C  = consumptive ratio for the particular nitrified waste.

Based upon the work of McCarty, et al, a consumptive ratio of 1.3 was
assumed to be applicable to agricultural was tewaters (11).  Making an
allowance for the dissolved oxygen content of the waste being treated
(7 to 9 mg/1), an estimate based on Equation 10 indicated 55.5 to 57.2
mg/1 of methanol would be required to remove 20 mg/1 nitrate-nitrogen.
In practice, however, due to normal daily variations of approximately
3 mg/1 in the influent nitrate concentrations of the tile drainage, 65 mg/1
of methanol were added as a. safety- factor for the expected average of 20 mg/1
nitrate-nitrogen.  A similar increase in methanol addition over that required
according to Equation 10 was made for the higher nitrate concentrations used
in filter experiments.
                     Process Evaluation Procedures

Evaluation of the aforementioned apparatus required monitoring of both
physical and chemical parameters.  A description of the procedures .employed
in maintaining operational control and making process evaluations follows.

Operating and Sampling Procedure

To insure that all experimental factors under consideration were controlled
and maintained at the required levels, periodic maintenance and inspection
checks of all apparatus were performed at 4-hour intervals for a minimum
of 16 hours per weekday and 12 hours per day on weekends.   This coverage
was found to be sufficient for control of standard operational parameters
such as flow rates, methanol addition, etc., and to prevent or correct
equipment breakdown.

Hydraulic tracer analyses were performed as needed.  Tracer studies were
analyzed by means of the volume apportionment technique (23)(24).  The
advantage of the volume apportionment technique is that a definite tracer
response curve is obtained, which allows simultaneous identification of such
hydraulic characteristics as short circuiting, stagnant zones, plug flow,
                                  -15-

-------
and complete mixing.  A step increase or decrease of a  tracer element in
the influent of the reactor is used for this method.  Differential equations
have been developed which identify the relationships which can occur.  The
majority of the reactor vessels used at the treatment center had hydraulic
response curves which indicated that the vessel volume was partially com-
pletely mixed in series with plug flow and contained some stagnant zones.
In such a system a step decrease in a tracer concentration from C  to C
would give the following equation:
                     C        I/a Iqt/v -  (1-a-b)]
                      /CQ = e-

where

        C  =  Effluent tracer concentration at time = t

        C  =  Effluent tracer concentration when time = o
        /-i
         /CQ =Fraction of tracer element remaining at time = t

        V  =  Total reactor volume

        e  =  Natural logarithm base

        Q  =  Flow rate

        t  =  Time after the change in tracer concentration

        c]_t/v  =  The number of detention times since the change in tracer
                 concentration.  (When cj_t/v = 1;  a = 1;  b = o then
                 C/CQ = e-1 = .368)

        a  =  Fraction of total volume which is completely mixed

        b  =  Fraction of total volume which is stagnant

        1-a-b  =  Fraction of total volume which is plug flow


A typical curve for this equation when plotted on semilog paper is shown
in Figure 4.  Points A and B on the abscissa are defined as:

                     A = 1-a-b

                     B = 1-b

In practice the tracer elements used were  the chloride ion for the filters
and Rhodamine B dye for the deep ponds.  A tracer response curve such
as Figure 4 is plotted using the values of the vessel's effluent  tracer
concentration.  The numerical values of A and B are then known and the
hydraulic patterns for the vessel can then be classified.  For more
detailed information the reader is referred to the above noted references.
                                  -16-

-------
Grab samples for chemical analysis  were  collected during the morning and
were analyzed within a few hours.   Studies  to determine diurnal variations
of chemical and physical factors were  also  conducted as deemed necessary.
Water temperatures were monitored daily  with maximum-minimum thermometers
and periodically with 8-day- recording  thermographs.  The influent pressure
required for an anaerobic filter to maintain a  constant hydraulic detention
time also was monitored daily.   The rotameters  used for flow control were
calibrated volumetrically.
Analytical Procedures

Laboratory analyses were conducted according  to the schedule and method
given in Table 2.   The majority of the  techniques were used as described
in Standard Methods (25).   Until September  11, 1968, effluent methanol was
originally determined by a chromatropic acid  method (26) (27)(28).   After
that date methanol determinations were  made with the use of gas chromat-
ograph equipped with a carbowax column  and  a  flame ionization detector.
Average influent methanol was calculated over a 24-hour period by
measuring amounts injected by the chemical  feed pumps.
               o
              o
              o
               I
              LJ
              o:
              z
              u
              S
              u
              _l
              LU
              a:
              LJ
              o
              <
              o:
              U.
              o
g
o
tr
u.
                 0.368--
                      0      A   0.5    B    1.0          1.5
                      THEORETICAL DETENTION TIMES-qt/v
                    FIGURE  4-EXAMPLE OF HYDRAULIC
                         TRACER RESPONSE CURVE
                                  -17-

-------
                                                   TABLE 2

                                               LABORATORY ANALYSIS
             CONSTITUENT
                                        FREQUENCY
                                  FILTERS        PONDS
                                        METHOD OF ANALYSIS
I
i-1
CD
I
Nitrate-Nitrogen

Nitrite-Nitrogen

Total Kjeldahl Nitrogen
Ammonia Nitrogen
Organic Nitrogen
Orthophosphate
PH
Alkalinity
Dissolved Oxygen
Suspended Solids
Vol. Suspended Solids
Methanol
3-6/week

3-6/week

  I/week
  I/week
  I/week
  I/week
  I/week
  I/week
Daily
  I/week
  I/week
3-4/week
3/week

3/week

I/week
I/week
I/week
I/week
I/week
I/week
Daily
I/week
I/week
3-4/week
      Electrical Conductivity
      Total Dissolved Solids
      Algal Cell Counts & Identification
      Chloride
                                    I/week       I/week
                                          As Needed
                                          As Needed
                                          As Needed
Brucine Method and/or Specific Ion
        Electrode
Diazotization  a - napthylamine
        Method
Kjeldahl Method
Distillation Method
Kjeldahl Method
Stannous Chloride Modification
Glass Electrode
pH Titration
Winkler Method
0.45 y Glass Paper, 103°C
0.45 y Glass Paper, 600°C
Gas Chromatograph Carbowax Column
        Flame loniz. Detector
Wheats tone Bridge
Gravimetric Method
Sedgewick-Rafter Cell
Silver Nitrate Titration

-------
                               SECTION IV

                         RESULTS AND DISCUSSION
The results of the anaerobic denitrification experimental work performed
at the Intera.gency Agricultural Wastewater Treatment Center are presented
below.  The data obtained  for filter denitrification are presented first,
followed by a description  of the pond denitrification results.  Included
are sections on the results of botulism research conducted by the California
Department of Fish and Game, regrowth studies and projected costs for the
denitrification processes.
                         Filter Denitrification

Field evaluation of filter denitrification was initiated in October 1967.
The 4-inch diameter PVC filters and media previously described were used
for this phase of the study.  It was found that all filters removed a
minimum of 80 percent of the incoming nitrogen for at least 30 consecutive
days at detention times ranging from 1 to 6 hours.  Although results showed
that the process did work, the smaller filters were abandoned because their
functioning was greatly affected by ambient air temperatures.  Larger scale
filters were constructed for the subsequent study on start-up procedures,
temperature effects, nitrogen loading, and long-term operation effects on
denitrification.
Start-Up Procedures

To establish the necessary bacterial populations in the early studies the
filters were started on a program of 6- to 8-hour theoretical hydraulic
detention times based on void volumes.  No bacterial inocula were used.
Once adequate denitrification was occurring, the hydraulic detention times
were shortened to the desired values.  Since the initial experiments took
place during the warmer months when water temperatures were as high as
24°C, this procedure proved satisfactory.  In fact, some filters placed
in operation during the warmer seasons at hydraulic detention times as
short as two hours were reducing nitrate-nitrogen from 20 mg/1 to less than
2.0 mg/1 of total nitrogen within four days.

Placing a new filter into operation as a functioning unit when water
temperatures were below 15°C required a more complicated approach.  It
was necessary to use a mass inoculation of denitrifying bacteria, and to
allow the filters to remain stagnant until nitrogen levels are reduced
to the desired 2 mg/1 of total nitrogen.  As the denitrification rate
increased, flow-through operations were commenced at hydraulic detention
times sufficient to continue the desired rate of denitrification.  These
experiments demonstrated that a 24-hour start-up detention time may be
necessary to achieve adequate nitrate-reduction at water temperatures from
12-15°C and a start-up detention time as long as 72 hours may be required
at water temperatures below 12°C.  These detention times may need to be
maintained for several detention periods before a sufficient bacterial


                                   -19-

-------
population has been established  to permit reduction of  the detention
period.  Reduction of detention  times after start-up during cold weather
became a matter of operating at  the minimum detention time (5-7 hours),
which effects  the required reduction of nitrogen.
Media Evaluation

Media evaluation for anaerobic filter denitrification was studied from
July 1968  to January 1969.  An influent containing 20 mg/1 of nitrate-
nitrogen was used  in making the evaluation.  The  types, size, hydraulic
loadings,  and  filter sizes for all media used in  the experiments are
summarized in  Table 3.
Aggregate media  offers  a relatively smooth, nonporous, nonsorptive surface,
and  in  addition  is very durable.  Coal has a limited sorptive capacity
(29), and its  irregular surface provides more surface area per unit volume
than does rounded aggregate.  Volcanic cinders were selected because, like
coal, they  are highly porous and roughly textured with a large surface area
per  unit volume, although,  the adsorptive capacity of cinders is less than
that of coal.  The activated carbon selected for the study was of the
largest size commercially available (0.16-inch diameter).  It was used as
an extreme  in  that part of  the investigation concerned with the significance
of adsorptive  quality.  Sand was included in the study because of its
fineness of gradation and durability.  It also provided a good extreme in
the  study of the significance of medium size on anaerobic filter operation.
A filter was also constructed for experimental use with SURFPAC ±.' , a
commercially available  plastic medium designed- for use in trickling filters.
This medium has  a fluted design and a void ratio of approximately nine tenths,

Effect  of Medium Characteristics on Nitrogen Removal. In evaluating the
effect  of medium characteristics on filter performance, a 2-hour detention
time was chosen  for  all media except the SURFPAC which was operated at a
16-hour detention time.  The 2-hour detention was selected because the
early feasibility results indicated that this was an intermediate detention
time at which  all media could be fairly evaluated.  Furthermore, the
detention time required a flow rate through all the experimental filters,
which was easily- attained and controlled.  The SURFPAC required a longer
detention time because  of the large void volume of the medium.  Excepting
for  the larger aggregate and SURFPAC, nitrogen removal efficiencies attained
•in the  evaluation were  essentially equal regardless of the medium.  The data
are  summarized in Table 4.

It was  apparent  that texture of the medium surface and sorptive quality
did  not appreciably  affect  removal efficiencies.  There was also no apparent
difference  between the  removal efficiencies of filters containing media of
the  same type, but of different sizes up to 1 inch in diameter.  For these
_!/ Registered  trademark of  the Dow Chemical Company.


                                    -20-

-------
                                                                 TABLE 3


                                                                 SUMMARY

                                                                   OF
                                                     MEDIA AND OPERATIONAL CRITERIA
I
N3

MEDIUM



Activated
Carbon



Rounded
Aggregate



Angular
Bituminous
Coal
Volcanic
Cinders
DOW
SURFPAC _!/
MEAN
DIAMETER
(Inches)

0.07

0.16

0.38


1.0

2.0

0.31
1.0

0.31
0.62
1.0

-
FILTER
HYDRAULIC DETENTION
TIME RANGE
(Hours)
2-9

0.5 - 6

2-6


0.5 - 7

2-4

2
2

2
2-8
2

8-16
FILTER
HYDRAULIC LOAD
RANGE
(Gal/Ft2/Min.)
0.13 - 0.03

0.50 - 0.04

0.13 - 0.04


0.48 - 0.03

0.62 - 0.31

0.12
0.15

0.16
0.18 - 0.05
0.21

0.08 - 0.04
FILTER
DIAMETER
(Inches)

4

4
18
4
18
4
18
20
18
36
18
18

18
18
18

20 (square)
SIZE
DEPTH
(Feet)

6.5

6.5
6.0
6.5
6.0
6.5
6.0
.6.0
6.0
6.0
6.0
6.0

6.0
6.0
6.0

6.0
                          _!/ Registered trademark of the Dow Chemical Company

-------
                                TABLE 4

                      NITROGEN REMOVAL EFFICIENCIES
                         OF EXPERIMENTAL MEDIA
                                              TOTAL NITROGEN REMOVAL
                                                     PERCENT

0.016- Inch Activated Carbon
Washed Sand
0.31-Inch Angular Coal
0.31-Inch Volcanic Cinders
0.38-Inch Rounded Aggregate
0.62-Inch Volcanic Cinders
1.0-Inch Angular Bituminous Coal
1.0-Inch Volcanic Cinders
1.0- Inch Rounded Aggregate
2.0-Inch Rounded Aggregate
DOW SURFPAC
Min.
89
84
80
85
82
87
81
89
89
45
80
Max.
99
97
98
98
97
97
98
97
98
92
90
Average
96
93
93
94
94
91
93
96
94
72
86
Note:  Nitrogen removal based on 20 mg/1 of nitrate-nitrogen in the
       influent for a period of 80 days at idenitcal climatic and
       operational conditions.
reasons and considering the relative merits of each medium, the 0.31-inch
coal and the 0.31-inch volcanic cinder-filled filters were abandoned near
the middle of October 1968.

Effects of Medium Characteristics on Long-Term Operation.  The second phase
of the work was concerned with media evaluation.  In this phase emphasis
was placed on the effect of long-term operation on unit efficiency and on
the minimization of hydraulic detention times used.  The filters that
contained the 0.31-inch coal and volcanic cinders were emptied and refilled
with washed sand and 1.0-inch aggregate, respectively.  These units were
operated at a hydraulic detention time of 0.5 hours.  In addition, two
18-inch diameter filters were filled with washed sand and 1.0-inch aggregate
and were operated with 1-hour hydraulic detention times.  Operation of the
remaining filters was not changed except the detention time of the SURFPAC
filter was reduced to 8 hours.  The factors given major attention in this
phase of the medium evaluation were influent pressure required to maintain
a constant flow rate through the filters, changes in influent pressure with
time, and nitrogen removal variations with time.

The influent pressures required to maintain a constant hydraulic detention
time and the corresponding nitrogen removal efficiencies are summarized in
Table 5.  The data show a definite contrast in required influent pressure
and nitrogen removal between the majority of media less than 1 inch in
diameter and those 1 inch or larger in diameter.  Generally, the trend was
an increase in required pressure with decrease in the particle size of the
medium.  Also, the average nitrogen removal efficiency with the very fine
media was lower than that with the coarser media.  A good example of these

                                   -22-

-------
                            TABLE 5




INFLUENT PRESSURES AND NITROGEN REMOVAL FOR EXPERIMENTAL MEDIA




Data Obtained With The Use Of 18- and 36- Inch Diameter Filters
MEDIUM
0.16 Inch Activated Carbon
Washed Sand
Washed Sand
ts? Washed Sand
us
1
0.38 Inch Rounded Aggregate
0.62 Inch Volcanic Cinders
1-Inch Angular Bituminous Coal
1-Inch Volcanic Cinders
1-Inch Rounded Aggregate
1-Inch Rounded Aggregate
1-Inch Rounded Aggregate
2.0 Rounded Aggregate
2.0 Rounded Aggregate
SURFPAC
SURFPAC
_!/ Pressure data based on
"11 Nitrogen removal based
HYDRAULIC
DETENTION
TIME
HOURS
2
0
1
2


2
2
2
2
0
1
2
2
4
8
16
.0
.5
.0
.0


.0
.0
.0
.0
.5
.0
.0
.0
.0
.0
.0
equal periods
on 20 mg/1 of
LENGTH OF
OPERATION
DAYS
193
51
93
183


182
138
132
130
92
106
183
168
282
402
67
REQUIRED INFLUENT
PRESSURE, P.S.I.G.JL/
MIN. MAX. AVERAGE
7.0
4.1
5.5
4.9


9.6
8.3
4.0
3.1
6.6
3.2
3.5
3.6
3.6
Equal
Equal
15
6
32
12


21
12
7
6
6
5
7
6
5
to
to
.4
.2
.0
.1


.5
.0
.8
.8
.6
.5
.5
.8
.5
11
5
10
8


15
10
6
5
4
4
5
4
4
.7
.2
.6
.1


.4
.6
.0
.1
.7
.2
.5
.4
.0
Static Head
Static Head
of operation. Pressure includes
nitrate-nitrogen
influent.


2.6

psi

TOTAL
NITROGEN REMOVAL
PERCENT 2.J
.MIN. MAX. AVERAGE
86
0
12
23


91
60
81
85
53
80
84
45
50
30
80
water head,

98
95
98
96


97
97
97
98
91
97
97
92
94
94
90


96
19
58
82


95
89
93
95
75
92
93
75
80
70
86



-------
characteristics "was the pressure  and  efficiency performances obtained
with the sand and the  1-inch diameter rounded aggregate.

For the first 130 days of operation  the filter filled with 1-inch diameter
aggregate and operated at a 1-hour hydraulic detention time had an average
effluent total nitrogen concentration of 1.4 mg/1 and an influent pressure
for this 130-day period which  did not exceed 5.5 psig.  The highest influent
pressure required by this filter  for  its entire operational period of 470
days was only 11.5 psig.  After 250  days of  operation the pressure appeared
to become cyclic, which was attributed to a  buildup of bacteria within the
voids and its dislodgement by  the pressure increases.  The historical
pressure pattern of this filter is shown in  Figure 5.

The performance of the 1-inch  aggregate filter may be compared with that
of the sand-filled filter operated at the same hydraulic detention time.
Influent pressure and  total nitrogen  removal data for the filter are
plotted in Figure 6.   Examination of  Figure  6 shows that by the 52nd day
of operation, the required influent  pressure exceeded 30 psig.  Because
of operating difficulties which developed due to this required pressure,
a surge of water had to be forced through the filter on the 55th day of
operation to break up  the bacterial mass.  Prior to breaking up the
bacterial mass, the effluent total nitrogen  concentration was less than
2 mg/1.  However, following the bacterial dislodgement the effluent total
nitrogen concentration increased  to  as high  as 13 mg/1.   This efficiency
head loss relationship was observed  again approximately one month later
       60-
    u)
    o:
    3
       45-
    LJ
    o:
    a.
    2
    LU
       30-
       15-
                             ACTIVATED CARBON
                                        300
        100
                150
                                                         400
             200      250      300      350
                DAYS OF CONTINUOUS OPERATION
FIGURE 5 - INFLUENT PRESSURE VERSUS TIME  FOR ACTIVATED CARBON
         AND ONE INCH DIAMETER AGGREGATE MEDIAS
                                                                 450
                                                                         500
                                    -24-

-------
when the influent pressure and nitrogen removal rate of the unit had
increased to approximately 22 psig and 95 percent,  respectively.   Directly
thereafter, and without surging water through the column,  the pressure
dropped to less than 10 psig and the nitrogen removal efficiency to less
than 20 percent.  This drop was attributed to short circuit paths,  which
were forced through the sand medium by the higher pressure which again
caused dislodgement of the bacteria.  The average effluent nitrogen con-
centration for the period of time covered by Figure 6 was  8.8 mg/1.   This
unit was taken out of operation in January 1969, after 128 days  of  con-
tinuous operation.

The sand-filled filter operated at a 2-hour detention time also  had a
high average influent pressure and low average nitrogen removal  for its
period of operation.  The low efficiency was assumed to be due to poor
hydraulics within the unit.  This assumption was verified  by observation
of the medium upon dismantling the units.  For example, the unit contained
dark columns of bacteria extending from the filter bottom  to the surface.
The columns were approximately 3 inches in diameter with no evidence of
                      50    60   70   80    90    100
                        DAYS  OF  CONTINUOUS OPERATION
  FIGURE 6 - TOTAL NITROGEN REMOVAL AND INFLUENT PRESSURE VERSUS TIME
    FOR SAND FILLED 18-INCH DIAMETER FILTER WITH ONE HOUR DETENTION
                                   -25-

-------
bacterial growth outside of the dark areas, thus indicating that the
flow was short circuiting.

The sand-filled filter operated at a one-half hour detention time did not
have a pressure build up.  This was attributed to the higher flow-through
rate of the filter which led to short circuiting and did not allow a
bacterial mass to accumulate.  This absence of bacteria accumulation was
demonstrated by the low nitrogen removal rate (Table 5),  and short
circuiting determined by tracer analysis.

Although it took longer to develop, the pressure required to force
influent into the filter containing activated carbon medium had a pressure
variance similar to the sand filter.  Pressure variations characteristic
of this filter are plotted in Figure 4.  The influent pressure normally
was above 40 psig after 300 days of operation, and on one occasion
exceeded 70 psig.  An efficiency-pressure relationship also existed in
this filter, although not as dramatic as with the sand filter.

At the higher pressures the activated carbon bed was forced above its
normal level after 425 days of operation, thereby, blocking the effluent
system and creating a large void at the filter bottom.  To alleviate this
problem and to make possible the continued operation of the filter, one
quarter of the medium was removed and surge of water was applied to loosen
the bacteria mass that was causing the blockage.  The medium which had
been removed was replaced and the filter was then restarted.  This dis-
placement of medium was again repeated on the 511th day.   However, instead
of a single large void forming at the bottom, the entire medium bed had
expanded leaving pockets of voids.  During the initial pressure variations
and the occurrence of operational problems, the nitrogen removal rate of
the filter did not suffer greatly.  It continued to remain at a high level
until the second media displacement.  At that point effluent nitrogen
concentrations increased to an average 4.6 mg/1, most likely due to a loss
of bacterial population and a decline in water temperature.  Due to the
long-term operational problems encountered and the high cost of activated
carbon, this filter was abandoned after 594 days of continuous operation.

The filters containing the 0.375-inch aggregate and 0.62-inch volcanic
cinders were not characterized by low nitrogen removal rates.  However,
the high influent pressures required for operating these units as com-
pared to those for the 1-inch media led to the decision not to use these
two media for further experimentation.  Another reason for the elimination
of volcanic cinders is their poor durability.  The cinders left a powdery
residue when transported or handled.  This slow disintegration could lead
to operational troubles within a filter especially if back washing were
required.  For this reason, all volcanic cinder media were eliminated.

As a result of the work described in the preceding paragraphs, media less
than 1-inch in diameter were eliminated from consideration.  The deter-
mination of the upper limit of medium size was then investigated.  Two
18-inch diameter and two 36-inch diameter filters containing rounded
aggregate with a size range of 1.5- to 3-inch diameter were used for this
purpose.  The filters were operated continuously for one year.  Although
                                   -26-

-------
no pressure problems were encountered, the nitrogen removal efficiencies
for the filters were never as high as for those containing 1-inch diameter
media (cf. Tables 4 and 5).  It was observed that a larger mass of bacteria
was present at the surface of the filters containing the larger media than
at the surface of those containing the smaller media.  This indicates that
the bacteria were not contained as effectively in the filters having the
larger sized media as those containing 1.0-inch diameter and smaller.  As
a result, the density of the bacterial population in the former was not as
great as in the latter.

The SURFPAC was also eliminated for the same reason.  It was observed that
a mat of bacteria was always present at the top of this filter, and although
left in operation over 450 days it was obvious no substantial bacterial mass
could accumulate, mainly due to the open design of the medium.

Summary of Media Evaluation.  It was concluded that the 1-inch diameter
rounded aggregate would be the most effective in the denitrification of
tile drainage in up-flow anaerobic filters.  The smaller diameter media
proved to be conducive to short circuiting and high influent pressures
after long periods of continuous operation.  These problems were due to
the large masses of bacteria retained within the media beds.  The larger
media on the other hand did not retain adequate cultures and, therefore,
required longer hydraulic detention times for efficient operation.  SURFPAC
was not suitable as a medium for the anaerobic filter process because its
design did not permit retention of a bacterial population sufficiently
large to accomplish efficient nitrogen removal.
Effect of Temperature and Nitrogen Loading on Nitrogen Removal

Two major factors affecting the required detention period needed to
meet a specific effluent nitrogen criterion are water temperature and
nitrogen loading.  Predicted temperatures and expected nitrate-nitrogen
concentrations of the agricultural wastewater are shown in Figure 7.  The
predicted temperatures are based on recorded temperatures of the Delta-
Mendota Canal inasmuch as the agricultural wastewater drain will be exposed
to the same climatic conditions as the canal.  Nitrate concentrations are
based on extensive nitrogen sampling studies of tile drainage systems
throughout the San Joaquin Valley (30).  To date only preliminary studies
of the effect of temperature and nitrogen loading on required detention
time have been completed.

The preliminary studies were conducted in two 18-inch diameter filters
filled with 1-inch diameter aggregate and operated at 1- and 2-hour
detention times.  Initially, the filters were used to determine the effect
on temperature and nitrogen removal while receiving a constant influent
nitrate-nitrogen concentration of 20 mg/1.  As seasonal water temperature
dropped from those prevailing in the summer of 1968 to the temperatures
of 1968-69 winter seasons, the nitrogen concentration of the effluents
from both filters met the 2 mg/1 total nitrogen criterion until water
temperatures fell below 14°C.  At that point, only the filter  on the
                                  -27-

-------
2-hour detention  time met  the  criterion.
Figure 8.
These data are plotted in
The only observed difference  between warm and cold weather operation  of
the filters with 1-  and  2-hour detention times during this period of
experimentatipn was  the  amount of filter required to achieve a given
efficiency.  Profile samples  for nitrate and nitrite taken at the quarter
points of the medium beds  showed that as the temperature dropped, the
percent of the filter bed  required to achieve the same degree of treatment
as obtained at the higher  temperatures increased (Table 6).

At temperatures above 16°C,  the filter operated at the 2-hour detention
time was able to reduce  the  20 mg/1 nitrate-nitrogen in the influent
to less than 2 mg/1  of nitrate and nitrite by the time the waste had
reached the first profile  sample location (18 in.)  above the influent
line.  With the detention  time at 1 hour the same degree of treatment was
attained by the time the second profile location (36 in.) was reached.
In both cases the elasped  time was within 0.5 hours after introduction of
the waste into the unit.   At  a temperature range of 12-16°C the "2-hour"
filter required three-quarters of the medium depth (54 in.) to produce an
effluent containing  2 mg/1 of total nitrogen.  The "1-hour" filter required
the entire medium depth  to achieve a 2 mg/1 total nitrogen concentration
in the effluent at a temperature range of 14-16°C.

On the basis of these encouraging results the influent nitrate concentration
to these filters was increased to 40 mg/1 nitrogen.  This increase was
started into the "2-hour detention time" filter in early March 1969 (after
   50'
   40-
 < ^n-
 et 3L)1
 z
 o
 o
 z 20-
    10-
                                                       WATER TEMPERATURE
       NITRATE-
        NITROGEN
                                                                      30
                                                                      -25
                              tu
                            20 9=
                              v-
                              <
                              
-------
                                                                      TABLE  6

                                        COMPARISON OF PROFILE NITRATE PLUS NITRITE NITROGEN  CONCENTRATIONS  AT
                                                VARIOUS  TEMPERATURES  FOR SHORT AND LONG TERM OPERATION

                                                Data Obtained With  The Use of  18-Inch  Diameter  Filters
                                                 Containing 1-Inch  Diameter  Rounded Aggregate as  Media
                                                          N03  + N02 mg/l-N
NO3 + NO2 mg/l-N
I
to
TEMPERATURE
RANGE
°C
22-24
20-22
18-20
16-18
14-16
12-14
DETENTION
TIME
HOURS
1.
2.
1.
2.
1.
2.
1.
2.
1.
2.
1.
2.
0
0
0
0
0
0
0
0
0
0
0
0
JUNE 1968 THROUGH MAY
SAMPLE LOCATION
18 36 54

1.
6.
0.
2.
0.
8.
3.
7.
2.
NO
NO
NO
03
75
73
45
74
43
22
20
65
DATA AT
DATA AT
DATA AT
0.45
4.65
0.07
0.55
0.10
4.92
1.75
5.17
2.73
TEMPERATURE
TEMPERATURE
TEMPERATURE
0.22
3.45
— Q-
0.15
0.21
3.21
0.3
3.32
1.01
1969 JUNE 1969 THROUGH DECEMBER 1969
ABOVE INFLUENT - INCHES
EFF. 18 36 54 EFF.

0.27
0.07
0.07
0.68
0.03
0.60
0.47
2.87
1.06
7.21
7.22
9.88
7.48
8.74
6.85
7.70
12.60
3.81
3.97
2.53
3.14
3.17
3.68
5.45
5.92
1.61
1.84
3.14
3.10
4.70
4.44
NO REPRESENTATIVE DATA
6.67 6.99
NO REPRESENTATIVE DATA
6.54 6.13
5.26
3.87
3.13
4.27
1.08
.58
3.60
2.40
3.64
3.09
6.5
4.07
3.13
3.60
                         NOTE:   Data are based on an influent of 20 mg/1 nitrate-nitrogen.

-------
240 days of operation) and to the "1-hour detention  time" filter in early
April 1969 (after  200 days of operation).  The total nitrogen concentration
of the effluent during the period of the higher nitrogen dosage is plotted
in Figure 8.   It is  apparent from the data that at the high nitrate con-
centration neither filter was able to produce an effluent meeting the
2 mg/1 or less total nitrogen criterion.  It was concluded from these
results that longer  detention times would be necessary to treat waters
having high nitrate  concentrations during periods of low temperature.

A further investigation  of the effect of temperatures was conducted with
the use of a 4-inch  diameter filter operated in a temperature controlled
environment.   It was operated with & theoretical hydraulic detention time
of 8 hours, and received an influent containing a constant 20 mg/1
nitrate-nitrogen.  It was operated at temperatures of 7.5°C, 15°C, 20°C,
and 25°C.  The relationship developed by Van't Hoff  and Arrhenius was
             IO.CH
             8.0-
E
i

LU
O
QC
         UJ

         U_
         U.
         LU
             6.0-
4.0-
             2.0-
                                   I-HOUR  DETENTION TIME
                      vss   NOTE: 40 mg/1 N03-N  INFLUENT
                         \
                         \
                         \
                         \
                         \
      2-HOURS          \ \
      DETENTION  TIME
                           \
                           \
                        DETENTION TIME

                         NOTE: 20 mg/1 N03-N INFLUENT
                           2-HOURS DETENTION TIME
               10
             "is"
                             —r~
                             20
~25
                         WATER TEMPERATURE

   FIGURE 8 -EFFLUENT  NITROGEN VERSUS WATER TEMPERATURE
                                  -30-

-------
used in analyzing the data gathered.  With  the  relationship  developed
from this work and upon  the  removal  efficiencies  achieved in the field
units, a curve (Figure 9) was  established that  predicts  the  detention
time required to produce an  effluent containing 2 mg/1 total nitrogen at
the predicted annual temperature ranges  and nitrogen concentrations shown
in Figure 7.  There is an obvious  difference between the predicted deten-
tion periods and those indicated by  the  results obtained in  the  preliminary
field experiments in the spring of 1969.  The predicted  detention times
of 1 to 2 hours for the  months of .March  and April correspond to  the tem-
perature and nitrogen ranges prevailing  at  the  time  the  experiments were
conducted.  In the experimental runs  a 2 mg/1 total  nitrogen effluent
could not be achieved.   The  difference between  predicted and actual results
may be due to two reasons.   One could be operational control.  The field
units were exposed to more operational upsets than the temperature con-
trolled filter.  The second  reason and probably the  most critical may be
long-term hydraulic changes within the filters, which might  have reduced
the theoretical detention times.   This latter reason is  discussed in more
detail in a later section.   Further  studies on  the effect of temperature
and nitrogen loading are being performed in the operational  studies of
1970.  The predicted values  for temperature and nitrate  loading  are being
duplicated and emphasis  is being placed  on  better operational control
to determine the actual  detention  time necessary  to  meet a 2 mg/1 total
nitrogen criterion.
Effect of Long-Term Operation

As the feasibility studies in the 18-inch diameter  filters entered  their
second year, it became apparent from the nitrogen removal data  that


  4.0i
CO
cc.
13
o
I
i  3.0-j
co
p 2.0-
UJ
I—
LU
O
LU
tr
O
u
I
I-
     —I	1	1	1	1	1	1	1	1	1	1	1	1—
     DEC   JAN   FEB   MAR   APR   MAY   JUN   JUL   AUG   SEP   OCT   NOV   DEC
                                    MONTHS

   FIGURE 9 - PREDICTED DETENTION TIMES FOR PROJECTED SEASONAL VARIATIONS OF
          NITROGEN AND PROJECTED MINCMUM WATER TEMPERATURES


                                   -31-

-------
efficiencies were not as high as during the previous year.  Table 7
contains a summary of the effluent nitrogen concentrations as related
to temperature for the time periods involved.

As stated in the preceding section, the only difference between high and
low temperature operation was the percentage of filter depth required to
achieve the 2 mg/1 effluent criterion.  Total effluent ammonia plus
organic nitrogen during this period averaged approximately 0.8 mg/1.
However, during the period of July 1969 through December 1969, duplication
of the earlier results was not obtained.  At temperatures above 22°C con-
centrations of total nitrate plus nitrite-nitrogen' in the effluent at the
1- and 2-hour hydraulic detentions averaged 1.08 mg/1 and .58 mg/1,
respectively.  However, total effluent ammonia plus organic nitrogen for
this period averaged over 1.60 mg/1 from both filters, increasing the
total nitrogen concentration for both filters to more than 2 mg/1.  Further,
at temperatures below 22°C, neither the "1-hour" nor the "2-hour" filter
was able to produce an effluent containing 2 mg/1 total nitrate plus nitrite.
Profile samples, as shown in Table 6, had considerably higher concentrations
than were obtained in the coldest temperatures of the 1968-69 winter season.
These differences showed that with long-term operation major operational
problems were occurring which needed to be controlled.  Variations of
parameters which may have caused these differences are discussed in the
sections which follow.

Total and Volatile Suspended Solids.  The concentrations of total suspended
and volatile suspended solids in the effluent of the filters remained rel-
atively low throughout their operation  (Table 8).  The higher concentrations
were detected in late September of 1969 and only for a period of two to
three weeks.  The low concentrations of effluent volatile solids indicate
that a low concentration of bacteria is emitted from the filter.  The nitro-
gen that could possibly be added to the waste by mineralization of such
bacteria is accounted for in the total effluent nitrogen values given for
each process in their respective tables of data.  Preliminary investigations
into the types and number of bacteria contained in the effluent of filters
and ponds indicated that the determination of these parameters would be very
difficult to achieve satisfactorily.  Because this type of information was
considered of secondary importance for the field research being conducted,
the answers to these questions were not pursued.

Effluent Organic and Ammonia Nitrogen.  The data in Table 7 show that
the total organic and ammonia nitrogen in the filter effluent increased
in the summer of 1969 to more than twice that recorded in the earlier
studies of 1968.  In Figure 10, total ammonia plus organic nitrogen of the
effluent is plotted as a function of "1- and 2-hour detention time" filters
with 1-inch aggregate.  Initially, this increase in Kjeldahl nitrogen was
thought to be an indication of a seasonal slough-off of bacteria similar
to that which occurs in a trickling filter.  However, as discussed in the
preceding section, the variations in filter effluent solids were not sub-
stantial.  Also, the occurrence of the peak concentrations of total Kjeldahl
did not correspond to the occurrence of higher solids concentrations.
                                   -32-

-------
                                             TABLE 7

                          COMPARISON OF EFFLUENT NITROGEN CONCENTRATIONS
                    AT VARIOUS TEMPERATURES FOR SHORT AND LONG TERM OPERATION

                      Data Obtained With The Use of 18-Inch Diameter Filters
                    Containing 1-Inch Diameter Rounded Aggregate as Media
                              JUNE 1968 THROUGH MAY 1969
JUNE 1969 THROUGH DECEMBER 1969
TEMPERATURE
RANGE °C
22-24
1
to
' 20-22
18-20
16-18
14-16
12-14
DETENTION
TIME HOURS
1.0
2.0
1.0
2.0
1.0
2.0
1.0
2.0
1.0
2.0
1.0
?-0
DAYS AT
TEMPERATURE
0
13
0
12
20
20
14
34
33
78
74
NO +NO
tng/l-U
NO DATA
0.25
NO DATA
0.27
0.55
0.07
0.68
0.03
0.60
0.47
2.12
1.06
TOTAL N
mg/1
0.91
0.86
1.26
0.83
1.25
0-74
1.42
1.45
2.87
1.77
DAYS AT
TEMPERATURE
63
63
21
21
23
25
0
34
0
16
37
36
NO-+NO
mg/l-N
1.08
0.58
3.60
3.39
3.64
3.09
6.58
4.02
6.44
3.60
TOTAL N
mg/1
2.83
2.26
4.88
4.92
4.70
4.17
NO DATA
9.83
NO DATA
4.65
7.11
4.24
NOTE:  Data based on an influent of 20 mg/1  nitrate-nitrogen.

-------
                               TABLE 8
               TOTAL  AND VOLATILE SOLIDS EFFLUENT DATA
                                 "FOR
           FILTERS  FILLED WITH ONE-INCH DIAMETER AGGREGATE
                                 TOTAL
DETENTION   NUMBER OF     SUSPENDED SOLIDS
  TIME     MEASUREMENTS   MEAN   RANGE   STD
  HOURS                   mg/1   mg/1    DEV
      VOLATILE
  SUSPENDED  SOLIDS
MEAN    RANGE     STD
mg/1    mg/1     DEV
1 43
2 40
4.
4.
88
99
2
2
.0-9.9
.4-11.2
2.
2.
57
43
2.
2.
54
56
1.
0.
0-6.6
5-7.4
1.47
1.78
The major part  of  the increase in total ammonia plus  organic nitrogen was
due to an increase in dissolved ammonia nitrogen, most  likely generated
from the degradation of excess bacterial growth within  the filter.   The
concentration of organic nitrogen in filter effluents remained relatively
constant at  approximately 0.6 mg/1.  The seasonal change  in organic
nitrogen averaged  only 0.2 mg/1; however, ammonia nitrogen concentrations
were as much as one milligram per liter greater in warmer seasons than in
cooler seasons.  Ammonia concentrations ranged from approximately 0.1 mg/1
to 1.2 mg/1.  It is likely that the peak concentrations of Kjeldahl nitrogen
reached in July 1969 (Figure 10) were the result of a higher rate of bac-
terial decay caused by warmer water temperatures.  As the temperature
   2.CH
   1.5-
 u. 0.5-
 bj
                    NOTE: I-INCH DIAMETER AGGREGATE FILLED FILTERS
      AUG  SEP  OCT  NOV  DEC  JAN  FEB  MAR  APR  MAY  JUN  JUL  AUG  SEP  OCT  NOV  DEC
            1968                                 1969

        FIGURE 10-TOTAL EFFLUENT AMMONIA PLUS ORGANIC NITROGEN VERSUS TIME
                                  -34-

-------
dropped in the autumn, the concentrations of ammonia nitrogen decreased
to  the levels found in 1968.  The rise and fall in ammonia indicates that
the variance in this nitrogen form may be cyclic.  Further study of the
variations and methods of controlling the amount of biomass within the
filters are receiving emphasis in the operational studies of 1970.

Changes in Hydraulic Patterns Within Filters.  To determine the hydraulic
pattern within the filters and their actual hydraulic detention times,
tracer studies were performed at intervals throughout the experimental
period.  Results of several of the tracer studies involving the 18-inch
diameter filters containing 1-inch diameter aggregate are plotted in
Figure 11.  The studies were performed approximately two months after the
filters were started.  They indicate an excellent agreement between
theoretical and actual hydraulic detention times.  Furthermore, the mixing
pattern within the filters can be regarded as  being equivalent to that of
two vessels in series, one comprising 75 percent of the filter volume
           X
           Ul
           cc
           H-
           z
           LJ
           2
           LJ
           _l
           UJ
           CC
           UJ
           cc
           t-
           u.
           o
           o
           cc
           u.
              0.
                         THEORETICAL DETENTION TIMES
FIGURE II  - RESULTS OF HYDRAULIC TRACER STUDIES ON ANAEROBIC FILTERS
              CONTAINING ONE INChi  DIAMETER AGGREGATE
                                   -35-

-------
and having perfect plug flow  and  the second comprising the remaining 25
percent and functioning as  a  completely mixed vessel.  As long-term
evaluation continued,  it became evident from tracer results that the
hydraulic patterns within the filters were changing.  The results of tracer
studies conducted approximately one year after start-up are plotted in
Figure 12.  The actual hydraulic  detention times decreased significantly
and the mixing pattern underwent  considerable change.  The actual detention
times for all of the filters  were less than the theoretical.  The previously
verified 2-, 1-, and 1/2-hour detention times were reduced by 25 percent
33 percent, and 57 percent, respectively.  This reduction was attributed
to the accumulation of excess bacteria, which in turn resulted in reduction
of available void volume.  The original mixing pattern, the greater part of
which was plug flow, had changed  to the extent that complete mixing
           z
           z
           LJ
           ce
           i-
           z
           LU
           2
           LJ
           _J
           IU
           CC
           LU
           0
           <
           o:
           I-
           IJL
           O
           z
           o
           h-
          o:
                  0
 0.5         1.0        1.5
THEORETICAL  DETENTION TIMES
2.0
  FIGURE 12- RESULTS OF HYDRAULIC TRACER STUDIES ON ANAEROBIC FILTERS
                CONTAINING ONE INCH  DIAMETER AGGREGATE
              AFTER 12-14 MONTHS OF CONTINUOUS OPERATION
                                   -36-

-------
predominated, and plug flow was at a minimum.   In  the  "1-  and  2- hour
detention  time" filters, 50 percent of  the filter  contents were completely
mixed, while 33 percent and 25 percent  of the contents were in stagnant
zones, respectively.  The remaining part of  the volume was characterized
by plug flow.  The filter operated at 1/2-hour  detention time contained
37 percent of its volume as completely  mixed while 57 percent was in
stagnant zones.

The tracer studies showing a high percentage of the filters'volume as
completely mixed zones are confirmed by the nitrogen profiles shown in
Table 6.  Under completely mixed hydraulic conditions, the concentration
of nitrogen should be uniform throughout the filter.  Data on the profiles
for the long-term operating filters do  show a fairly even concentration of
nitrate-nitrite at all levels, thus demonstrating  completely mixed charac-
teristics .

The changes in hydraulic mixing pattern and the reduction in detention times
are the major causes of discrepancies between results of 1968 and late
1969.  It is believed that the bacterial growth within the filters can be
limited so that the original hydraulics of the filters would be maintained,
which in return would allow a high nitrate reduction rate.
Removal of Bacteria Mass from Filters

After recognizing the operational problems attendant with excess bacterial
mass, attempts were made to dislodge and flush out the excess growth from
several of the 18-inch diameter filters.  The attempts consisted in forcing
water upward through the filter at rates varying from 5 gal/ft^/min to
70 gal/ft^/min.  Variations of flushing procedures have included use of
compressed air at the lower hydraulic rates, and alternate flushing upward
and draining the filters several times in succession.  These methods are
patterned after those discussed by Hamann and McKinney (31) on upflow
filtration in sand media.

Attempts to dislodge the bacteria by draining without first breaking up the
bacterial mass failed.  However, this procedure was attempted on a filter
which had been in operation for over one year without any backwashing.
Periodic dislodgement and a means of removing the bacteria (i.e., draining)
before a large mass becomes established would perhaps be more effective.  To
date the most effective method found for displacing heavily clogged units is
the application of high rate hydraulic loadings.  Preliminary cost analyses
of the volume and flow-through rate show this method is not economically
feasible.  Other methods for removal of bacterial mass are being investi-
gated.
Possible Operational Problems

Two studies were conducted which dealt with possible operational problems
associated with the use of anaerobic filters for denitrification.  These
were the effect of algal growth in -the agricultural waste prior to treatment
                                    -37-

-------
and the effect of scale-up on the hydraulics of filters.

Algal Growth.  It has been predicted (32) that there will be from 20 to 60
mg/1 of algal cells in the influent to any plant treating tile drainage
exported from the San Joaquin Valley.  Apparatus was constructed in July
1969, to feed a controlled amount of algae to the influent of an operating
18-inch diameter filter containing 1-inch diameter aggregate.  The filter
was operated with a theoretical hydraulic detention of 2 hours.  A data
summary for this filter is presented in Table 9.
                                TABLE 9

                    EFFLUENT NITROGEN CONCENTRATIONS
                  AT VARIOUS TEMPERATURES FOR AN ALGAE
                      SUPPLEMENTED FILTER INFLUENT

      Data Obtained with the use of an 18-Inch Diameter Filter
       Containing 1-Inch Diameter Rounded Aggregate as Medium
WATER DETENTION
TEMPERATURE TIME
*C HOURS
22-24
22-24
20-22
18-20

16-18
14-16
12-14
4
2
2
2

2
2
2
INFLUENT
ALGAL
CONC.
mg/1
25
25
20
No Algae
Feed
10-20
50
50
DAYS
OF
OPERATION
34
33
5
10

27
18
31
EFFLUENT
mg/1
0.47
0.80
1.97
0.58

1.75
2.26
2.35
NITROGEN CONCENTRATION
NH3 + ORG N
mg/1
1.83
0.95
2.10
0.88

0;85
1.16
2.61
TOTAL N
mg/1
2.30
1.75
4.07
1.46

2.60
3.42
4.96
         Note:  Data based on an influent of 20 mg/1 nitrate-nitrogen.
During the first three months of operation the filter received an influent
algal concentration of approximately 20 mg/1 with no apparent operational
problems.  Nitrate-nitrite reduction during these months was equal to the
1968-69 performance of the 2-hour detention time filter  (cf. Table 7).
Total nitrogen removal generally was lower due to higher concentrations
of ammonia and organic nitrogen in the effluent.  Results of effluent
volatile solids analyses during this period average approximately 3 mg/1.
Therefore, it was obvious that little, if any, of the algal cells were
passing through the filter.  This retention of the cells and their subsequent
degradation accounted for the higher effluent concentrations of Kjeldahl
nitrogen.
                                   -38-

-------
When the algae concentrations were increased to 50 rag/1, it became
obvious that algal cells were passing through the filter inasmuch
as the effluent volatile solids increased to over 10 mg/1 and became
blackish-green in color.  This substantiated that the filter was clogging,
and that algae were decomposing and/or being sloughed off.  Analyses of
filtered and unfiltered effluent ammonia and organic nitrogen samples
showed that of an average total Kjeldahl nitrogen concentration of 4.23
mg/1 only 19 percent was dissolved while the remaining 81 percent was in
the form of algal and bacterial suspended solids passing through the filter.

The nitrate plus nitrite concentrations at temperatures below 16°C average
approximately 2.3 mg/1.  This concentration was greater than that of the
nitrate plus nitrite concentrations characteristic of the effluent from
"2-hour detention time" filter operated without algae feed in the 1968-69
experimental period (Table 7).  Probably, the decrease in nitrate-nitrite
removal rate was due at least in part to hydraulic changes in the filter.
Although it was apparent that at the higher water temperatures the algae-
laden influent did not affect denitrification, it would be necessary to
remove any algae from the effluent in order to meet the 2 mg/1 total
nitrogen criterion.  Furthermore, a system would have to be provided for
flushing out the algae retained within the filter so that the hydraulics
of the filter would remain unaffected.

Pilot Scale Filter.  In the spring of 1969, a 10-foot square filter was
constructed to study the effects of scale-up hydraulics on the performance
of an anaerobic filter.  The filter was started in Hay 1969 by allowing
it to remain stagnant with an initial methanol concentration of 100 mg/1.
The nitrate concentration was reduced to less than 2 mg/1 within three
days.  The flow-through operation was begun with theoretical detention
time of 7 hours.  A summary of the operational changes and nitrogen
removal data obtained with the filter is given in Table 10.

The filter required a long start-up period for two reasons.  First, the
stagnant period probably was too short; and secondly, as is often the case
in experimentation, the start-up of the unit was plagued with mechanical
problems.  The data show that from days 66 to 89 of continuous operation,
the concentration of the effluent total nitrogen averaged less than 2 mg/1
when the unit was operated at a theoretical 5.25-hour detention time.  At
a detention time of 2 hours (days 90 through 193) the effluent nitrogen
concentration increased to approximately 6 mg/1.  No improved trend in
performance was observed until the detention time was lengthened to 3 hours
on the 194th day of operation.  That change was followed by a decrease in
effluent total nitrogen concentration to an average of 3.5 mg/1.  Pre-
liminary results in 1970 show that at nitrate loadings exceeding 30 mg/1,
an effluent containing 2 mg/1 or less total nitrogen could be produced at
a theoretical detention time of 5.5 hours.  Further studies are being
undertaken to determine the proper operating procedures for the filter.

Results of tracer studies performed on the large filter are plotted in
Figure 13.  Data analysis revealed that the theoretical and actual
hydraulic detention times for the first two studies performed on the 58th
                                   -39-

-------
                                TABLE 10

                OPERATIONS AND EFFLUENT NITROGEN SUMMARY
                         FOR PILOT SCALE FILTER
DAYS
OF
OPERATION
0-39
40-65
66-89
90-134
135-147
148-186
187-193
194-224
DETENTION
TIME
HOURS
7.23
5.25
5.25
2.0
2.0
2.0
2.0
3.0
TEMPERATURE
°C
18-20
20-22
20-22
20-22
18-20
16-18
14-16
12-14
EFFLUENT
NITROGEN CONCENTRATION
NO + N02 TOTAL NITROGEN
ffig/1 me/1
2.49
3.43
0.73
5.38
6.00
4.72

2.71
3.20
4.35
1.51
5.76
6.87
5.41
No Data
3.53
      Note:  Data based on influent of 20 mg/1 nitrate-nitrogen.
and 161st day of operation showed excellent agreement.  As with smaller
filters, analysis indicated that the filter volume could be likened to
two vessels operating in series, one having a plug flow and the other
being completely mixed.  The first analysis of detention time indicated
the existence of a 17 percent zone of plug flow, a 77 percent zone of
complete mixing and a 6 percent stagnant zone.  According to the second
analysis, the zone arrangement was 38 percent plug flow, 62 percent com-
pletely mixed, and no stagnant zone.  The third tracer study performed
after 311 days of operating revealed a distinct hydraulic change.  According
to Figure 13,  the actual detention time was 4.1 hours instead of the
theoretical 5.5 hours.  The hydraulic characteristics of the filter were
estimated to be 40 percent plug flow, 42 percent completely mixed, and 18
percent stagnant zones.  The decrease in detention undoubtedly was due to
an increase in the bacterial mass within the filter, which caused the
stagnant zones similar to the smaller filters.  The biomass increase has
been monitored throughout the life of the filter by measuring the pressure
within the filter with the use of the in-filter sample extraction tubes.

Several pressure profiles of the filter are shown in Figure 14 as well as
their increase with time.  It is obvious that the greater pressure is in
the bottom layers of the medium bed as would be expected.  The increase
in pressure from day 156 to day 188 and the relative stability of the
pressure through day 294 indicated a build up of bacterial mass which
accounted for some of the formation of stagnant zones observed in the
previous tracer study.  An unexpected characteristic of the hydraulic
pattern changes was the continuing increase in the percentage of volume
of plug flow observed with each successive tracer analysis.  This trend is
not comparable to that in the 18-inch diameter filters, in which the amount
                                    -40-

-------
 CO
 UJ
 tr
 UJ
  1.0

  0.9

  0.8

  0.7


  0.6



  0.5



  0.4

0.368



  0.3
 o:
 UJ
 o

 or
     0.2
o

cc
U_
     O.I
           X
               \
        ANALYSIS^
         AFTER 58
         DAYS OF
         OPERATION
                 A
                 z\\\
                           ANALYSIS AFTER 161 DAYS
                         j,	OF  OPERATION	
                           \
                            \
                             \
                                 \
          ANALYSIS  AFTER 311
            OF OPERATION
                                   \
                                    \
                                     \
                                      \
                                                2.0
                 0.5         1.0         1.5
               THEORETICAL  DETENTION TIMES

FIGURE 13 -  RESULTS OF HYDRAULIC TRACER STUDIES
       PERFORMED  ON PILOT SCALE FILTER
                       -41-

-------
of plug flow decreased as  the biomass  increased.  No operational change or
other possible cause was observed  that may  have reversed  the decreasing
plug flow trend observed in the smaller filters.

It was concluded from results and  observations made in  the first eight
months of operation of the pilot scale unit that no particular or unusual
problem was encountered.  It was realized  that operational control methods
of the unit must be refined in order to eliminate variations in nitrogen
reduction efficiency.  A reduction in theoretical detention time may be
expected with long-term operation  if biomass controls within the filter
are not taken.  Again, such controls are expected to be developed for the
filter during the second year of study.
u
1.0
2.0
t-
Ul
LU
U-
1
* 3.0
CL
LU
Q
<
O
LU
5 4.0
5.0
e. n



^\
*v\
^
M
\\ ^-
: \
: \
'; \
\
: \
4




^
\"^
XNXX
L
	 	 68 [
	 97 [
___ _ IOC
., i "? K r
	 188
	 294



" -^
Xv
EGEUD
)AYS OF OPEF
)AYS "
DAYS OF OPE
3AYS " "
DAYS OF OPE
DAYS " "



s^ ^ ^^^
^•^
?ATInN
RATION
It
RATION
It



"^
                 0         5         10         15         20        25
                              PRESSURE - INCHES OF WATER

      FIGURE 14-RESULTS OF PRESSURE PROFILES FOR PILOT SCALE FILTER

                                    -42-

-------
                          Pond Denitrification

The pond denitrification studies were begun in June of 1967 with the use
of the pilot scale deep ponds.  The work is described in a report entitled
"Field Evaluation of Anaerobic Denitrification in Simulated Deep Ponds"
(18).  During the experiments the nitrate-nitrogen content of the influent
was maintained at 20 mg/1.  Results of the studies indicated that in
uncovered simulated ponds, removal efficiencies would not exceed 50 to
60 percent.  Furthermore, it appeared that removal efficiency in the open
ponds was almost independent of detention time.  Data obtained with the
uncovered ponds indicated that nitrogen removal rates remained the same
at detention times ranging from 5 to 14 days.  Experiments with covered
simulated ponds indicated nitrogen removal efficiencies of 90 percent at
about a 10-day detention time or 80 percent at about 5-day detention
were possible.  Based on these initial feasibility results, large-scale
experiments involving uncovered and covered ponds were begun in early
1969.
Uncovered Pond

The uncovered pond was filled with irrigation return water in late
February of 1969 and was operated for a total of 222 days.  Methanol was
added until an in-pond concentration of 100 mg/1 was reached.  The pond
was continuously mixed by means of the recirculation line but flow-through
operation was not started.  These procedures were followed in an attempt
to develop a bacterial culture.

Nitrogen Removal Performance.  Results obtained with the uncovered pond
and operational changes are tabulated in Table 11.  During the early
weeks of operation practically no nitrogen reduction took place.  The
little reduction that did occur probably can be attributed to an algal
bloom which occurred in the pond.  Apparently the bloom delayed the
development of anaerobic conditions and the establishment of a denitrifying
bacterial population by keeping the dissolved oxygen concentration at or
near the saturated level of 10-15 mg/1.  After 67 days of stagnant operation
an herbicide (simazine) was applied to the pond surface in an attempt to
eliminate the algal bloom.  This resulted in a decrease in the algal cell
count from 124,000 cells/ml to less than 2,000 cells/ml.  The dissolved
oxygen concentration decreased to undetectable levels at the pond bottom,
and anaerobic conditions were established.  Algal cells were not detected
in any significant concentration during the remainder of the operation of
the pond.

Anaerobic conditions were allowed to remain undisturbed for an additional
12 days after the herbicide was applied.  During this time the in-pond
nitrogen concentration decreased from approximately 15 mg/1 to 10 mg/1.
At the end of 12 days, flow-through operation was started at a theoretical
detention time of 20 days.  As indicated by the data in Table 11, effluent
nitrogen concentrations declined through day 200 except for a short period
following a reduction in detention time from 20 days to 10 days.  As
                                   -43-

-------
                                TABLE 11

                         OPERATION AND NITROGEN
                        CONCENTRATION SUMMARY FOR
                         THE UNCOVERED DEEP POND
OPERATION
0-35
36-73
74-80
81-135
136-155
156-166
167-200
201-221
222-224
225-260
THEORETICAL
HYDRAULIC
DETENTION
TIME DAYS
Infinite
Infinite
Infinite
20
20
10
10
10
10
10
TEMPERATURE
°C
14-18
18-20
20-22
22-24
24-26
24-26
22-24
20-22
18-20
16-18
EFFLUENT OR IN-POND
NITRATE + NITRITE
mg/1
16.1
14.8
10.7
7.26
3.08
4.90
1.90
4.23
6.64
12.7
CONCENTRATION
TOTAL NITROGEN
mg/1
16.9
15.5
12.0
8.76
4.63
6.40
3.39
5.52
7.85
14.0
        Note:  Data based on an influent of 20 mg/1 nitrate-nitrogen.
temperature dropped, nitrogen removal efficiencies decreased significantly.
The major nitrogen form present in the effluent throughout the pond
operation was nitrate.  Nitrite was present in insignificant concentrations
averaging approximately 0.6 mg/1 at the 20-day detention time, and 0.35 mg/1
at the shorter detention time.  Moore (12) also noted the appearance of
low levels of nitrite in a continually mixed vessel.  Total Kjeldahl
nitrogen averaged 1.51 mg/1 at the 20-day detention time and 1.37 mg/1
at the 10-day detention time.  It is significant that at no time, even
during the warmest temperatures, did the pond produce an effluent having
a total nitrogen content as low as 2 mg/1.  According to the data,  the
highest efficiency that may be expected with the influent nitrogen
concentration at 20 mg/1 would be approximately 85 percent nitrogen
removal under favorable environmental conditions.   The nitrogen removal
rates as attained in the uncovered pond indicate that the 2 mg/1 total
nitrogen criterion for the effluent would not be consistently met throughout
the year.

Hydraulics.  A determination of pond hydraulics was made to learn of-their
effect on pond performance.  Results of a tracer study performed when the
pond was at a theoretical detention time of 10 days are plotted in Figure
15.  The actual detention was calculated to be nine days.  The mixing
pattern was estimated to be 28 percent plug flow,  61 percent completely
mixed, and 11 percent stagnant zones.  Although the actual detention time
was less than the theoretical, the difference was small and, therefore,
                                   -44-

-------
o
z

z
<
s
UJ
cc
UJ
s
Ul

UJ

cr
UJ
o

cc
H-

U.
O
o
<
cc
u.
l.U
0.9

0.8
07

0.6


0.5

0.4
0.368

0.3



0.2






n i
\
\

\

\


































\
\



















V^iO DAYS THEORETICAL
\ DETENTION TIME
\
»
\
\

v_

V






















.

\
\
\
\

\
\
\
\






















\
\
\
                0.5        1.0        1.5

             THEORETICAL  DETENTION  TIMES

        FIGURE 15- HYDRAULIC TRACER  RESULTS

          FOR THE  UNCOVERED DEEP POND
2.0
                        -45-

-------
only a small proportion of the pond's poor nitrogen removal was attributed
to the short detention time.

Suspended and Volatile Suspended Solids.  Table 12 shows a summary of the
total and volatile suspended solids results in the uncovered pond taken
at various sample locations within the pond after flow-throueh oceration
was begun.

Exclusive of the influent, the samples were essentially equal in solids
concentration.  The uniform distribution of solids throughout the pond
further affirms the results of the tracer analysis that the pond volume
was mostly completely mixed.  The essentially equal values of suspended
solids concentrations in the recirculation line and effluent indicated
that seeding of the influent could be done by recycling a portion of the
effluent rather than using a separate recycle from the pond bottom.
Facilities should be present for draining the ponds to remove flocculant
organisms which might accumulate in the sludge layer.

Temperature Variance.  The temperature of the surface and bottom levels
of the pond was studied to detect the occurrence of a daily or seasonal
"overturning" of the pond.  It was found that a temperature differential
always existed between the surface and bottom.  The extent of the differ-
ential was seasonal.  During cooler months, the differential was as much
as 10°C and the surface temperature alone had a daily variance of 5°C.
Despite the daily temperature variances, no evidence of overturning of
the pond was observed.

                                TABLE 12

         DATA SUMMARY OF THE TOTAL AND VOLATILE SUSPENDED SOLIDS
                  FOR THE UNCOVERED ANAEROBIC DEEP POND

SAMPLE
LOCATION


NO. OF
MEASUREMENTS


TOTAL
SUSPENDED SOLIDS
MEAN
mg/1
RANGE STD.DEV.
mg/1

VOLATILE
SSUPENDED -SOLIDS
MEAN
ra^/1
RANGE STD.DEV.
mR/1
Influent        25

Surface         29

Bottom          26

Recirculation   28

Effluent        24
 8.34  1.5-15.0   4.5

12.1   7.4-23.2   3.0

12.9   9.3-30.6   3.2

12.4   6.9-45.9   3.3

13.0   7.9-18.0   2.7
3.50  0.3-7.0    2.2

7.50  3.3-14.5   2.7

6.60  3.8-15.3   2.0

7.50  4.6-27.6   1.5

6.90  3.2-12.3   1.7
                                   -46-

-------
Covered Pond

The covered pond was started on a continuously mixed basis in early March
of 1969.  At the time, the in-pond nitrogen concentration was 10 mg/1.
Methanol was added to bring its concentration to approximately 100 mg/1.
Within 7 days bacterial population was active inasmuch as the nitrogen
concentration decreased to less than 2 mg/1 so flow-through operatioas
were begun.

Performance.  The initial detention time used for flow-through operation
was a theoretical 20 days.  As time progressed a series of reductions in
hydraulic detention were made, depending on the extent of nitrogen removal
and upon the prevailing environmental conditions.  The data are summarized
in Table 13.

Except for a period in which mechanical problems affected the covered
pond's nitrogen removal (days 63-97) a theoretical detention period of
fifteen days was found to be long enough to produce an effluent containing
a concentration of 2.0 mg/1 or less total nitrogen in a water temperature
range of 14° to 22°C.  The theoretical detention time was reduced to ten
days on the 125th day of operation.  Water temperatures were approximately
                               TABLE 13

         OPERATING AND EFFLUENT NITROGEN CONCENTRATION SUMMARY
                       FOR THE COVERED DEEP POND
 DAYS OF
OPERATION
THEORETICAL
 HYDRAULIC
 DETENTION
 TIME DAYS
   WATER
TEMPERATURE
   RANGE
    °C
                                       EFFLUENT NITROGEN  CONCENTRATION
 NITRATE + NITRITE
	mg/1	
 TOTAL NITROGEN
	mg/1	
  0-27
 28-39
 40-62
 63-97
 98-124
125-166
   &
187-197
167-186
198-218
219-249
250-260
261-268
    20
    14
    15
    15
    15

    10

     7.5
    10
    10
    10
    10
  14-16
  14-16
  16-18
  18-20
  20-22

  20-22

  20-22
  18-20
  16-18
  14-16
  12-14
       1.05
       0.51
       0.35
       1.08
       0.54

       0.29
       2.48
       1.39
       3.03
       2.44
       6.83
      2.00
      1.96
      1.42
      2.58
      1.80

      1.49

      3.79
      2.57
       ,19
       ,73
      7.79
                                  -47-

-------
20-22°C.  This detention period was also capable of meeting the 2 mg/1
total nitrogen criterion.  These results verified on a large scale basis
the conclusions reached in the feasibility studies.  A further verification
of the feasibility results was made when the pond was operated at a
theoretical detention time of 7.5 days from day 167 through day 186 in a
water temperature range of 20-22°C.  At the 7.5 day theoretical detention
time the average total nitrogen concentration of the effluent was approx-
imately 4 mg/1.  This performance is in agreement with the efficiencies
predicted on the basis of the field feasibility studies, i. e., 80 percent
nitrogen removal with the influent nitrate concentration at 20 mg/l-N.
Upon returning the pond to a theoretical 10 day detention time on day 187
(water temperatures 20-22°C) the nitrogen removal rate of the pond recovered
to produce an effluent containing a total nitrogen concentration of less
than 2 mg/1.  As temperatures dropped to below 20°C it was found that 10
days were not sufficient length of time to meet the 2 mg/1 total nitrogen
criterion.  Nitrogen removal efficiency decreased as water temperature
decreased.  At the end of this operational period the total nitrogen
concentration within the pond exceeded 7 mg/l-N.  It is believed that by
increasing the detention time to a suitable period the effluent criterion
can be met.

Observed nitj.wgci.1 J.WL.UIO w<=i.ti ^JLULJ^J-CUL i-u uinj&c j. cpvj i. ucu uy riuujLc: \__L^./ cttiu
those found in the uncovered pond.  Nitrite was rarely detected at con-
centrations exceeding 1 mg/1 of nitrogen.  In the case of the ponds,
organic nitrogen was the predominant form of the ammonia plus organic
nitrogen component.  Organic nitrogen usually remained within a range of
0.8 to 1.5 mg/l-N, while the total ammonia plus organic nitrogen varied
between 1.0 to 2.0 mg/l-N.  This indicated an obvious difference between
the make-up of the effluent Kjeldahl nitrogen of the ponds and that of
the filters.  As stated above, the pond's Kjeldahl nitrogen was almost
entirely organic nitrogen, while the filter effluent Kjeldahl consisted
mostly of ammonia.  This was as expected on the basis of the physical
make-up of the two processes, since the ponds allow the bacteria to flow
out with the effluent, while the filters retain them.  Thus, in the filter
effluent, the ammonia would be most prominent due to bacterial decay, and
in the pond effluent the organic nitrogen would be the prevailing form due
to the greater concentrations of bacteria.  When nitrogen removal of the
pond decreased, for example at the theoretical detention time of 7.5 days
or at the colder temperatures, the most prevalent form found was nitrate.
The other monitored forms remained relatively constant in concentration.

Temperature Effect and Variance.  As noted in the previous section, the
nitrogen removal rate in the covered pond decreased in proportion to the
decline in water temperature.  Although the criterion of an effluent
nitrogen concentration of 2 mg/1 or less was not met, the seasonal dis-
charge of agricultural wastewater from the San Joaquin Valley is such that
in practice it will be possible to lengthen the detention periods in the
winter to produce the required effluent nitrogen concentration.  A pro-
jection of the expected detention time required in operating a covered pond
to produce an effluent having a maximum nitrogen concentration of 2 mg/1
                                  -48-

-------
 is  shown in Figure 16.   The projection is based on empirical results and
 the dashed  portion of the curve has not as yet been verified.

 Temperature variance  within the pond was minimal.   In thirty-six hour
 studies,  in which  temperatures  were measured at various points within the
 pond, it  was found that  the differential between any portion of the pond
 volume and/or daily variance of the average pond temperature was less than
 0.2°C.  In  addition to the  studies,  continuous records made with the use of
 8-day temperature  recorders placed  at the pond bottom and surface showed no
 noticeable  variation  between the temperatures of these locations.  These
 observations  indicate that  the  styrofoam cover had a definite effect on
 temperature  control within  the  pond.

Hydraulics.  Tracer analyses  of the  pond mixing patterns were performed
when the  pond was  operated  at theoretical detention times of 10 and 7.5
days.  The  analyses were made on the 140th and 183rd days.

Fluorescence readings of cross-sectional samples  taken as a part of the
analysis  of  the theoretical 10-day  detention time  indicated that the
pond was  completely mixed inasmuch as  the dye was  uniformly distributed
 throughout  the pond.  The tracer response curve shown in Figure 17  is
verification of the observed  distribution of dye.   The mixing pattern of
the pond may be classified  as being  82 percent completely mixed with 18
percent of the pond volume  considered  stagnant.  About 4 percent of the
flow through the pond vas short-circuited, while the  remaining  96 percent
  50
< 3o^
UJ
s
o
tu
Q
   10-
              \
              \                                                     '
               \                                                    I
               \                                                   I

                \
\
\
\
 \
 \

     DEC    JAN  FEE)  MAR   APR   MAY   JUN   JUL   AUG   SEP   OCT   NOV  DEC

   FIGURE 16- PREDICTED DETENTION TIME FOR TREATMENT OF AGRICULTURAL RETURN WATERS
                       BY COVERED POND DENITRIFICATION
                                  -49-

-------
was passed into  the mixed zone.
by the tracer was 8.2 days.
The actual  detention time as  indicated
A final tracer analysis was performed when  the pond was being operated
on a 7.5-day theoretical detention time.  At  the flow-through rate
required by this detention time, the response curve and the cross-section
samples indicated the  existence of a definite change in the mixing pattern.
The initial in-pond  samples showed that the dye formed a uniform  layer
across the bottom of the pond.  This layer  gradually rose to the  surface
becoming somewhat diffused at the influent  end because of the recirculation.
I.U
00
.y
Oft
tn r>-7
*-J U.I
z
z 06
— VJ.O
<
•^
UJ
DC f\f\
"- U.D
H
Z
UJ
S r>4
uj °-4
_J /-) •2CO
• j L/.OOO
cc
UJ
o r»^
< *J-^
cc
>-
u_
o
z
o
1— f*\ O
§
cc
u_
n i
\
\ \
\ \
\ \
\\
\








^7.5 DAYS
\
\
\ J
X^v^
v\\
>
\
\
\
\
\
\
\




THEORETICAL

DAYS THEOR!
^-DETENTION


\
\ \
\
\
\
\
\
\
\
\
\



DETENTION '

;TICAL
TIME



\
                                                            IME
                        0.5          1.0         1.5         2.0
                     THEORETICAL  DETENTION TIMES
               FIGURE  17 - HYDRAULIC  TRACER  RESULTS
                     FOR  THE COVERED DEEP  POND
                                 -50-

-------
These samples suggested a pond which was partially plug flow and partially
completely mixed.  The effluent concentration curve (Figure 17) indicates
that this was the case.  The mixing pattern was 27 percent plug flow, 44
percent completely mixed, and 24 percent stagnant zones.  The short-cir-
cuiting noted in the previous analysis was not observed in this study.
Because of the relative size of the stagnant zones, instead of functioning
on a theoretical detention time of 7.5 days, the pond was actually
operating on a 5.3-days detention time.

Total and Volatile Suspended Solids.   A summary of the total and volatile
suspended solids concentrations in the covered pond is shown in Table 14.
                               TABLE 14

        DATA SUMMARY OF SUSPENDED AND VOLATILE SUSPENDED SOLIDS
                  IN THE COVERED ANAEROBIC DEEP POND
TOTAL
SAMPLE NO. OF SUSPENDED SOLIDS
LOCATION MEASUREMENTS MEAN RANGE STD.
mg/1 mg/1
Influent
Surface
Middepth
Bottom
Recirculation
Effluent
47
143
120
138
47
47
6.
18.
18.
19.
17.
17.
7
2
3
5
0
2
1
4
5
4
3
4
.5-14.
.1-44.
.2-44.
.7-36.
.1-32.
.3-40.
2
4
2
0
5
7
4.
8.
9.
11.
8.
10.
DEV.
6
7
1
9
1
8
VOLATILE
SUSPENDED SOLIDS
MEAN RANGE STD.DEV
mg/1 mg/1
3.3
10.2
9.3
9.7
8.8
9.7
0-5.7
2.5-29.5
1.3-24.0
1.2-22.7
1.1-24.0
2.4-25.5
2.6
5.4
4.4
4.7
4.5
6.6
Values for the surface, middepth, and bottom represent samples taken from
nine locations at each depth.  The high range does not represent differences
due to the locations at which the samples were taken, but rather to seasonal
differences in solids concentration.  In comparison to the summer of 1969,
lower volatile solids concentrations were seen during the 1969-70 winter
operation when the 10-day detention time proved to be too short and the
bacterial population was washed out faster than it could reproduce.

The uniform concentration of solids of the in-pond, effluent and recirculation
samples indicated that the pond was mostly completely mixed, and thus confirms
the findings of the tracer studies.  Since the concentrations of solids of
the effluent and recirculation line .were essentially equal, it is feasible
that mass inoculation of the influent could be done directly from the efflu-
ent line, thus eliminating a separate recirculation line for each pond.  A
                                  -51-

-------
drain for removal of any accumulated sludge should be included in the
design.
                 Consumptive Ratio for Field Processes

As previously stated, the consumptive ratio assumed in Equation 10 for
the denitrification of the tile drainage was 1.3.  To insure an adequate
supply of carbon it was standard procedure to inject from 10 to 50 mg/1
more methanol than was indicated by Equation 10 when starting filters or
ponds.  The start-up methanol feed rate was cut back by degrees when an
excess of methanol appeared in the effluent and nitrogen removal effi-
ciencies became acceptable.  In this way, a practical lower limit of
65 mg/1 of methanol for 20 mg/1 nitrate-nitrogen and 8.0 mg/1 dissolved
oxygen was determined to be suitable in the experimental work.   As contin-
uous operation of the filters and ponds progressed, a consumptive ratio
of 1.47 with a standard deviation of +.36 was determined for the units
being operated at the Inteizagency Agricultural Wastewater Treatment Center.
The high standard deviation more likely is due to inherent difficulties in
conducting field studies than to fluctuations in system requirements.  The
ratio 1.47 was used as the consumptive ratio in Equation 10 when calculating
the amount of methanol to be used when making cost estimates for the pro-
jected nitrogen loadings.
                           Regrowth Studies

A series of laboratory experiments were performed to determine whether or
not removal of nitrogen from tile drainage would effectively reduce any
biostimulatory effect caused by the drainage prior to treatment.  These
experiments are described in detail in a report entitled "Effects of
Agricultural Wastewater Treatment on Algal Biossay Response" (33).  It
was found from the experiments that invariably mixtures containing waste-
waters treated by bacterial filter systems had lower bioassay responses
than did untreated wastewater.  "Respiking" the bacterial filter sample
with nitrate-nitrogen resulted in bioassay responses equal statistically to
those of untreated wastewater.  It was concluded in the report that under
the environmental conditions imposed in the regrowth experiments, tile
drainage which had undergone treatment by bacterial denitrification with
subsequent removal of nitrogen to a 2 mg/1 or less concentration did not
have a biostimulatory effect when added to San Joaquin river water.
Furthermore, it was apparent from the studies that nitrogen was the nutrient
required to create a biostimulatory response in waters receiving the tile
drainage.
                           Botulism Studies

A special study was made by the California Department of Fish and Game
on  the possibility of outbreaks of Type C Botulism in water fowl having
contact with full scale denitrification filters or ponds.  Details of
the experiments used in the study are available in a report published by
                                  -52-

-------
the Department of Fish and Game (34).  On the basis of their research,
members of the Department concluded that a botulism potential would exist
in filters and ponds if large invertebrate populations developed in them.
Many invertebrates contain botulism bacteria or spores in their digestive
tracts.  The bacteria multiply and the spores develop into "vegetative"
forms as the invertebrate carcass decays.  Another foreseable hazard
which could occur in deep ponds would be the occurrence of decaying animal
carcasses on which fly maggots can thrive.  The maggots could be ingested
by waterfowl and a botulism outbreak could occur.  However, since the water
surface area would be relatively small, any botulism outbreak would not be
considered serious because the birds could be excluded by a mechanical
barrier or avian scaring device.

The Department of Fish and Game recommended that construction of any ponds
for use as water impoundment, such as an anaerobic denitrification ponds,
should include in their design steep-sloped levees, a minimum of shoreline
area, and a water depth as deep as possible.  These recommendations are in
keeping with the basic design of the ponds used for anaerobic denitrification.

Among the management practices of anaerobic filters and deep ponds recom-
mended by the Department of Fish and Game would be the burning of all
vegetation prior to flooding of the pond and removal of vegetation from
all filter surfaces and from pond levees during operation.  Removal of the
vegetation would eliminate food sources for invertebrates and, therefore,
toxic bacteria which use dead invertebrates as a substrate.  Moreover,
removal of levee or filter vegetation would insure that an outbreak of
botulism could not go undetected by being obscured.  Another recommended
practice was to maintain the water at a constant level since fluctuations
of water levels are accompanied by invertebrate die-off.  Ponds should not
be allowed to become stagnant for long intervals since this also may lead
to invertebrate die-off because of the development of adverse conditions
in the water.  Animal carcasses found in ponds or on filter surfaces
should be removed immediately and disposed of in a sanitary manner.
                        Process Cost Estimates

A second objective of the feasibility studies at the Interagency Agricultural
Wastewater Treatment Center was to develop preliminary cost estimates for
the processes found to be technically feasible.  Of the anaerobic denitri-
fication processes, only the anaerobic filter and covered pond met the
criterion of technical feasibility.  Hence, cost estimates were made only
for these two processes.  In making the estimates it was necessary to deter-
mine the critical design period of the year.  The predicted seasonal varia-
tions of tile drainage flow and nitrogen concentrations in tile drainage from
the San Joaquin Valley are shown in Figure 18.  These data show that April
is the critical month in which both high irrigation return flows and high
nitrogen concentrations are expected to occur.  Adding these facts to the
effects of the prevailing environmental conditions for the month (low water
temperatures, etc.) indicates that cost estimates should be based on the
detention times required in April.  In making the estimates, it was assumed
                                  -53-

-------
that the total flow in the San Luis Drain will be treated at a plant near
the Kesterson Reservoir  (Gustine,  California).   Within 30 years, drainage
flows from the San Luis  unit service  area are expected to reach a maximum
annual quantity of over  50 billion gallons.
Basis for Estimates

In addition to the above  stated  elements,  there are other factors which
are common to the estimates.

     1.   All costs are based  on January  1970 dollars.

     2.   Debt service is calculated for 50 years at 5 percent.

     3.   Costs  per million gallons treated are calculated by dividing
          the total annual cost  (debt service plus operations and main-
          tenance costs)  by the  design capacity.

     4.   An engineering  and contingency factor of 57 percent was assumed
          for capital cost items whose design and operation were based on
          the experimental data  obtained at the treatment center.  A
          lower  engineering and  contingency factor was selected for items
          which  were  common waste treatment facilities or the price of
          which  was obtained from the manufacturer.  (See footnotes in
          Tables 16 and 19).

     5.   All land is acquired at the start of the project at $500 per
          acre.
         JAN   FEB  MAR  APR  MAY   JUN   JUL   AUG   SEP   OCT   NOV  DEC
 FIGURE 18-PREDICTED SEASONAL VARIATION OF TILE DRAINAGE FLOW 8 NITROGEN CONCENTRATIONS
                         FROM SAN JOAQUIN VALLEY, CALIFORNIA
                                   -54-

-------
     6.   The laboratory, office building,  maintenance,  and storage areas
          are basically  the same as  those at the San Jose,  California
          Sewage Treatment Plant.  Capital costs for these  facilities are
          based on information  published  by Guthrie (35).

     7.   Sedimentation  tank  loadings  and separation chemical additions
          for denitrification pretreatment are:   900 gpd/ft^ and 10 mg/1
          Fe  (SO/)., plus 0.5  mg/1  cationic polyelectrolyte.  Capital costs
          for sedimentation are from figures published by  Smith (36).

     8.   Electric power costs  are calculated on the basis  of Ic/kw-hr.

     9.   General plant  operation  and  maintenance (O&M)  costs are based on
          curves published by Smith  for  trickling filter O&M (36).  It
          is assumed  that the general  plant O&M  costs include replacement
          and power costs for all  plant  operations except  for the items
          indicated in Tables 17 and 20.

    10.   The cost of reaeration was taken from  figures published by Smith
          (36).
Filter Denitrification Design Criteria  and  Cost  Estimates

The schematic diagram for  the filter  denitrification process is presented
in Figure 19.  Sedimentation tanks  are  included  in the design to remove
any suspended solids (i.e. algae) which would be likely to clog the filters
and/or by degradation release excess  ammonia or  organic nitrogen.   Flow
through the sedimentation  tanks  is  by gravity, after which the water is
pumped through the filters.  Flow from  the  filters through the aerators
(reaeration step) is also  by gravity.   Design criteria for the filter
denitrification plant are  presented in  Table 15.

Methanol feed pumps are provided for  each of the filters.  The filter
boxes are fabricated of tilt-up  panels  connected by way of columns.  For
        Fe2(S04)3 6
       POLYELECTROLYTE
         ADDITION
                                        METHANOL
                                        ADDITION
  PLANT
INFLUENT
                SEDIMENTATION
                   TANKS
           L_
                                                               PLANT
                                                             EFFLUENT
                           FILTER  DENI-TRIFICATION

                     FIGURE  19   SCHEMATIC DIAGRAM
                                  -55-

-------
ease of operation, maintenance, and construction the box is subdivided by
walls similar to the external box walls into sections 100 feet by 100 feet
on a side.  The floor of the box is made of a 12-inch thick reinforced
concrete slab.

Filter washing (removal of excess bacterial cells)  is accomplished by:
(1) closing a sliding effluent gate, (2) pumping compressed air into the
bottom of the filter while the influent line is still open, and (3) when
the total water depth in the filter box reaches 13 feet the influent is
stopped and two 24-inch drain lines are opened rapidly.  The water and
bacterial solids flow by gravity to the washwater lagoons.   Once the solids
settle, the washwater is decanted and pumped back to the headworks of the
plant.  After denitrification has taken place the dissolved oxygen
concentration of the treated waste is brought up to 4 mg/1 by mechanical
aeration.  The capital costs for the filter plant having a 228 MGD
capacity are itemized in Table 16.
                               TABLE 15

                FILTER DENITRIFICATION DESIGN CRITERIA

     Ultimate Design Capacity                       228 MGD
                                                              o
     Sedimentation Tank Surface Loading             900 gpd/ft

     April Filter Hydraulic Detention Time          3 hours

     Filter Medium Depth                            6 feet

     Filter Box Wall Height                         14 feet

     Average Annual Methanol Concentration          68 mg/1
                                                    o
     Washwater Lagoons                  2,690,000 ft /phase

     Reaeration               Plant effluent D.O. = 4 mg/1
                                 -56-

-------
                                TABLE 16

                CAPITAL COSTS FOR FILTER DENITRIFICATION
                        DESIGN CAPACITY - 228 MGD
NUMBER
       ITEM
            CAPITAL EXPENDITURE
              1970 DOLLARS
(Engineering and Contingency Included)
  2
  3
  4
  5
   7
   8
Pretreatment

  Separation Facilities

Nitrogen Removal

  Filter Construction
  False Bottoms
  Pumps
  Wash Water System
Post Treatment

  Reaeration Equipment

Other

  Land Acquisition
  Buildings
                                                              $5,400,000
          $28,100,000
           10,000,000
            1,100,000
              600,000

          $39,800,000
                                                              39,800,000
                                                               1,500,000
              100,000
              700.000
                                                 $800,000
                                                        800,000
                            GRAND  TOTAL
                                                    $47.500,000
         NOTE:   Items Numbers  1,3,4,6,  and  8  received a  25 percent
                 engineering  and  contingency factor.

                 Items Numbers  2  and  5 include a 57 percent engineering
                 and contingency  factor  furnished by the U.S.  Bureau of
                 Reclamation  in its "reconnaissance" estimates.

                 No engineering and contingency factor was assigned to
                 the land cost.
                                   -57-

-------
The unit costs for the filter treatment plant are presented in Table
17.  These costs are based on the assumptions stated in the previous
section.  From the data in the Table, the cost for treatment by filter
denitrification would be approximately 92 dollars per million gallons
for a plant operated at full capacity.
                              TABLE 17

             TREATMENT COSTS FOR FILTER DENITRIFICATION
                      DESIGN CAPACITY - 228 MGD
        ITEM                              COST/MG         PERCENT OF
	(1970 DOLLARS)     TOTAL COST

CAPITALIZED COSTS

  Pretreatment

    Separation Facilities                   6                 6

  Nitrogen Removal

    Filter Construction
    False Bottoms
    Pumps                                  43                47
    Wash Water System

  Post Treatment

    Reaeration                              2                 2

  Other

    Land, Buildings, and Miscellaneous      1                 1

ANNUAL O&M COSTS

    General O&M                            21                23
    Methanol                               19                21
                        TOTAL             $92/MG            100 Percent
Pond Denitrification Design Criteria and Cost Estimates

The schematic for covered pond denitrification is presented in Figure
20,  In this configuration, any algae present in the influent to the
treatment plant are removed in standard sedimentation tanks.   The essen-
tially algae-free waste is pumped into covered ponds designed according
to the specifications listed in Table 18.  Six covered ponds grouped about
a common effluent line and one recycle pumping station are provided per

                                 -58-

-------
phase of construction.  The ponds have 3 feet of freeboard, and levees
having sides with a 1:1 slope and a crest 15 feet wide.  To prevent
seepage into or out of the ponds, 14 percent of the internal area is
sealed with soil cement.  The remaining 86 percent is sealed by com-
paction of the native soil.  The pond covers used for this cost estimate
were assumed to be of the same type used at the treatment center (DOW
STYROFOAM - WT).  The cost of covering the ponds is based on information
provided by the DOW Chemical Company.  Individual methanol feed pumps
are provided for each pond.  The 25 percent recycle is taken from the
common effluent line and pumped around the six ponds as a group.
Reaeration is accomplished by mechanical aeration.  The capital costs
for the covered ponds are itemized in Table 19.
                              TABLE 18

                POND DENITRIFICATION DESIGN CRITERIA
Ultimate Design Capacity

Sedimentation Tank Surface Loading

April Pond Hydraulic Detention Time

Pond Water Depth

Pond Length  to Width Ratio

Recycle

Average Annual Methanol Concentration

Reaeration
          228 MGD

          900 gpd/ft2

          15 days

          15 feet

          5

          25 percent

          68 mg/1

Plant effluent D.O. = 4 mg/1
                                                              PLANT
                                                            EFFLUENT
Fe2 (SO
POLYELE
ADD
PLANT '
INFLUENT
4)3
:TR
TIO

a
OLYTE
M
SEDIMENTATION
TANKS



PUMPING
STATION

METf
ADD
4ANO
TION

L
COVERED
PONDS




REAERATX*

                          POND DENITRIFICATION

                   FIGURE 20   SCHEMATIC DIAGRAM
                                  -59-

-------
                              TABLE 19

             CAPITAL COSTS FOR DEEP POND DENITRIFICATION
                      DESIGN CAPACITY - 228 MGD
                                              CAPITAL EXPENDITURE
NUMBER         ITEM                             197° DOLLARS
                                 (Engineering and Contingency Included)

         Pretreatment

  1        Separation Facilities                        $ 5,400,000

         Nitrogen Removal

  2        Pond Construction            $ 9,000,000
  3        Pond Covers                   19,000,000
  4        Pumps                          1.000.000

                                        $29,100,000      29,100,000

         Post Treatment

  5        Reaeration Equipment                           1,500,000

         Other

  6        Land Acquisition             $   800,000
  7        Buildings                        700,000

                                        $ 1,500,000       1,500,000

                         GRAND TOTAL                    $37,500,000
         NOTE:  Items numbers 1,4,5, and 7 received a 25 percent
                engineering and contingency factor.

                Item number 3 received a 20 percent engineering
                and contingency factor.

                Item number 2 includes a 57 percent c.i-gi'weering and
                contingency factor furnished by the U.S. Bureau of
                Reclamation in its "reconnaissance" estimates.

                No engineering and contingency factor was assigned to
                the land cost.
                                 -60-

-------
The unit costs for a covered pond denitrification plant are presented
in Table 20.  The cost of 88 dollars per million gallons is based on
the construction assumptions listed previously.
                               TABLE 20

               TREATMENT COSTS FOR POND DENITRIFICATION
      ITEM
   COST/MG
(1970  DOLLARS)
PERCENT OF
TOTAL COST
CAPITALIZED COSTS

   Pretreatment

     Separation Facilities

   Nitrogen Removal

     Pond Construction
     Pond Covers

   Post Treatment

     Reaeration

   Other

     Land, Buildings and Miscellaneous

ANNUAL O&M COSTS

     General O&M
     Methanol
     Pond Cover O&M

                            TOTAL
    11
    20
    21
    19
     7

   $88/MG
   12
   23
   24
   22
  100 Percent
                                  -61-

-------
Treatment Costs of San Joaquin Valley Drain Flows

The actual costs of treatment of agricultural tile drainage by either
anaerobic filters or covered ponds will depend on the time of staging
plant construction.  It is expected that the unit cost of each process
will be higher than the values given in Tables 17 and 20.  It is believed
that plant construction most likely will be staged so that the 30 year
period of increasing flows can be treated in an economical manner.  With
staged construction there would occur a time period in the early part of
each stage when a portion of the treatment facilities will not be in
use.  Because of this excess capacity, the cost per million gallons treated
will be higher than if the unit cost was based on design capacity.
                                  -62-

-------
                                SECTION V

                             ACKNOWLEDGMENTS
This phase of the field investigations concerned with bacterial
denitrification of tile drainage was performed under the joint direction
of Messrs Percy P. St. Amant, Sanitary Engineer, Environmental Protection
Agency; Louis A. Beck, Sanitary Engineer, California Department of Water
Resources; and Donald G. Swain, Sanitary Engineer, United States Bureau
of Reclamation.

The field work was the responsibility of Messrs Thomas A. Tamblyn and
Bryan R. Sword, Sanitary Engineers, Environmental Protection Agency.  Mr.
Tamblyn was the major contributor to the theoretical section of this report
and also developed the basic process designs and cost criteria upon which
the economics were based.

The cooperation and assistance given by the interagency staff of the
treatment center was a major contribution to the success of the field
studies.  These personnel were:

Robert G. Seals 	 Chemist, Environmental Protection Agency
William R. Lewis  .  .  .   Chemist, California Department of Water Resources
Norman W. Cederquist  ........ Technician, US Bureau of Reclamation
Gary E. Keller  	 Technician, US Bureau of Reclamation
Gary L. Rogers  	 Technician, US Bureau of Reclamation
Mathew C. Rumboltz	Technician, US Bureau of Reclamation
Clara P. Hatcher.  .  .  .Laboratory Aid, California Dept. of Water Resources
Elizabeth J. Boone.  .  .Laboratory Aid, California Dept. of Water Resources
Betty A. Smith	Clerk-Typist, Environmental Protection Agency
                       Consultants to the Project

Dr. Perry L. McCarty	Stanford University,  Palo Alto
Dr. William J. Oswald 	  University of California, Berkeley
Dr. Clarence G. Golueke 	  University of California, Berkeley
                           Report Prepared by:

Bryan R. Sword	Sanitary Engineer, Environmental Protection Agency
                                   -63-

-------
                                SECTION VI

                               REFERENCES
 1   Effects of the San Joaquin Master Drain on Water Quality of the
     San Francisco Bay and Delta.   U.  S.  Department of the Interior,
     Federal Water Pollution Control Administration,  Southwest Region,
     San Francisco, (January 1967).

 2   Sawyer, C. N., Chemistry for  Sanitary Engineers,  McGraw-Hill Book
     Company, Inc., New York (1960).

 3   Schroeder, E. D., "Dissimilatory  Nitrate Reduction by Mixed
     Bacterial Populations," a dissertation presented to Rice University
     at Houston, Texas in 1966, in partial fulfillment of the requirements
     for the degree of Doctor of Philosophy.

 4   Hutchinson, G. E., A Treatise on  Limnology-Volume 1, John Wiley and
     Sons,  Inc., New York (1957).

 5   Sherman, V. B. D., and McRae, K.  C.,  "The Influence of Oxygen
     Availability on the Degree of Nitrate Reduction  by Pseudomonas
     denitrificans," Canadian Journal  of  Microbiology, _3, 3,  505-530
     (1957).

 6   Kefeuver, M., and Allison, F. E., "Nitrate Reduction by Bacterium
     denitrificans in Relation to  Oxidation-Reduction Potential and
     Oxygen Tension,"  Journal of  Bacteriology. 73, 8-14 (1957).

 7   Strickland, L. H., "The Reduction of Nitrates  by Bact.  coli."
     Biochemical Journal. 25, 98 (1931).

 8   Thimann, K. V., The Life of Bacteria, 2nd Edition, The MacMillan Co.,
     New York, 412 (1966).

 9   Taniguchi, S., Sato, R., and  Egami,  F.,  "The Enzymatic Mechanism of
     Nitrate and Nitrite Metabolism in Bacteria," A Symposium on
     Inorganic Nitrogen Metabolism,  edited by W. D.  McElroy and B.  Glass,
     Johns  Hopkins Press, Baltimore (1956).

10   Heredia, C. F., and Medina, A., "Nitrate Reduction and Related
     Enzymes in Escherchia coli,"  Biochemical Journal. 77, 24 (1960).

11   McCarty, P. L., Beck,  L. A.,  and  St.  Amant, P. P., "Biological
     Denitrification of Wastewaters  by Addition of  Organic Materials,"
     Proceedings, 24th Annual Purdue Industrial Waste Conference,
     Lafayette, Indiana (May 1969).
                                   -64-

-------
12   Moore, S. F., "An Investigation of the Effects of Residence Time
     on Anaerobic Bacterial Denitrification," a thesis presented to the
     University of California at Davis, California in 1969, in partial
     fulfillment of the requirements for the degree of Master of Science.

13   Tamblyn, T. A., and Sword, B. R., "The Anaerobic Filter for the
     Denitrification of Agricultural Subsurface Drainage," Proceedings,
     24th Annual Purdue Industrial Waste Conference, Lafayette, Indiana
     (May 1969).

14   Hoover, S. R,, and Forges, N. , "Assimilation of Dairy Wastes by
     Activated Sludge," Sewage and Industrial Wastes. 24, 306-312 (1952).

15   Johnson, W. K., and Schroepfer, G. J., "Nitrogen Removal by
     Nitrification and Denitrification," Journal of Water Pollution
     Control Federation, 36, 1015 (1964).

16   Stern, G., "Literature Review-Removal of Nitrogen from Wastewaters,"
     unpublished Federal Water Pollution Control Administration report
     (August 1966).

17   McCarty, P. L., Feasibility of the Denitrification Process for
     Removal of Nitrate-Nitrogen from Agricultural Drainage Waters,
     Appendix California Department of Water Resources Bulletin No.
     174-3 (June 30, 1966).

18   Brown, R. L., Field Evaluation of Anaerobic Denitrification in
     Simulated Deep Ponds, California Department of Water Resources,
     Bulletin No. 174-3 (May 1969).

19   Finsen, P. 0., and Sampson, D., "Denitrification of Sewage Effluents,"
     Water Wastes Treatment Journal, 7, 298-300 (1959).

20   Parkhurst, J. D., Dryden, F. D., McDermott, G. N., and English, J.,
     "Pomona Activated Carbon Pilot Plant," Journal of the Water
     Pollution Control Federation, 39, R70-R81 (1967).

21   Speece, R. E., and Montgomery, R. G., Nitrogen Removal from Natural
     Waters, New Mexico State University, Engineering Experiment Station
     Technical Report 48, Las Cruces, (September 1968).

22   Bringmann, G., and Kuhn, R., "Rapid Denitrification Process with
     Automatic Redox Control," Gesundheits-Ingenieur, 83, 333-34 (1962).

23   Chlotte, A., Blanchet, J.,  and Cloutier,  L.,  "Performance  of Flow
     Reactors at Various Levels  of Mixing,"  Canadian Journal of Chemical
     Engineers, 38,  (1960).
                                   -65-

-------
24   Milbury, W.F., "A Development and Evaluation of a Theoretical Model
     Describing the Effects of Hydraulic Regime in Continuous .Microbial
     Systems," a dissertation presented to Northwestern University at
     Evanston, Illinois, in 1964, in partial fulfillment of the
     requirements for the degree of Doctor of Philosophy.

25   Standard Methods for the Examination of Water and Wastewater,
     American Public Health Association, Inc., New York, 12th Edition
     (1965).

26   Boos, R. N., "Quantitative Colormetric Microdetermination of Methanol
     with Chromotropic Acid,"  Analytical Chemistry, 20, 964-965 (1958).

27   Synthetic Methanol, Commercial Solvents Corporation,  Industrial
     Chemicals Department (1960).

28   Beyer, G. B., "Spectrophotometric Quantitative Determination of
     Methanol and Distilled Spirits,"  Journal of the Association of
     Agricultural Chemists. 34, 745-755 (1951).

29   Summary Report-The Advanced Waste Treatment Research Program-January
     1962 through June 1964, U. S. Department of Health, Education and
     Welfare, Public Health Service, Division of Water Supply and
     Pollution Control, PHS Publication No. 999-WP-24 (April 1965).

30   Glandon, L. R., Nutrients from Tile Drain Systems, California
     Department of Water Resources, Bulletin No. 174-6 (To be published).

31   Hamann, C. L., and McKinney, R. E., "Upflow Filtration Process,"
     Journal American Water Works Association, 60, 9, 1023-1039
     (September 1968).

32   Goldman, J. C., Arthur, J. F., Oswald, W. S., Beck, L. A., "Combined
     Nutrient Removal and Transport System for Tile Drainage from the
     San Joaquin Valley."  American Geophysical Union National Fall
     Meeting, San Francisco, California (December 15-18, 1969).

33   Effects of Agricultural Wastewater Treatment on Algal Bioassay
     Response, Environmental Protection Agency, 13030 ELY-9 (To be published)

34   Hunter, B.F., Clark, W. E., Perkins, P. J., Coleman,  P. R., Applied
     Botulism Research Including Management Recommendations, California
     Department of Fish and Game, (January 1970).

35   Guthrie, Kenneth M., "Capital Cost Estimating ," Chemical Engineering,
     76, 6, March 24, 1969.

36   Smith, Robert, "A Compilation of Cost Information for Conventional
     and Advanced Wastewater Treatment Plants and Processes ,"USDI, Federal
     Water Pollution Control Administration, Cincinnati, Ohio, December
     1967.
                                  -66-

-------
                              SECTION VII

                             PUBLICATIONS

SAN JOAQUIN PROJECT,  FIREBAUGH,  CALIFORNIA

1968

"Is Treatment of Agricultural Waste Water Possible?"
     Louis A. Beck and Percy P.  St. Amant, Jr.  Presented at Fourth
     International Water Quality Symposium, San Francisco,  California,
     August 14, 1968; published  in the proceedings of the meeting.
1969

"Biological Denitrification of Wastewaters by Addition of Organic
 Materials"
     Perry L. McCarty, Louis A. Beck, and Percy P.  St.  Amant,  Jr.
     Presented at the 24th Annual Purdue Industrial Waste Conference,
     Purdue University, Lafayette, Indiana.  May 6, 1969.

"Comparison of Nitrate Removal Methods"
     Louis A. Beck,  Percy P. St. Amant, Jr.,  and Thomas A.  Tamblyn.
     Presented at Water Pollution Control Federation Meeting,  Dallas,
     Texas.  October 9, 1969.

"Effect of Surface/Volume Relationship, C02 Addition, Aeration,  and
 Mixing on Nitrate Utilization by Scenedesmus Cultures  in Subsurface
 Agricultural Waste Waters".
     Randall L. Brown and James F. Arthur.  Proceedings of the
     Eutrophication-Biostimulation Assessment Workshop, Berkeley,
     California.  June 19-21, 1969.

"Nitrate Removal Studies at the Interagency Agricultural Waste Water
 Treatment Center, Firebaugh, California"
     Percy P. St. Amant, Jr., and Louis A. Beck. Presented at 1969
     Conference, California Water Pollution Control Association,
     Anaheim, California, and published in the proceedings of  the
     meeting.  May 9, 1969.

"Research on Methods of Removing Excess Plant Nutrients from Water"
     Percy P. St. Amant, Jr., and Louis A. Beck. Presented at 158th
     National Meeting and Chemical Exposition, American Chemical
     Society, New York, New York.  September  8, 1969.

"The Anaerobic Filter for the Denitrification of Agricultural
 Subsurface Drainage"
     T. A. Tamblyn and B. R. Sword.  Presented at the 24th Purdue
     Industrial Waste Conference, Lafayette,  Indiana.  May 5-8,  1969.
                                  -67-

-------
SAN JOAQUIN PROJECT, FIREBAUGH. CALIFORNIA (Continued)

1969

"Nutrients in Agricultural Tile Drainage"
     W. H. Pierce, L. A. Beck and L. R. Glandon.  Presented at the
     1969 Winter Meeting of the American Society of Agricultural
     Engineers, Chicago, Illinois.  December 9-12, 1969.

"Treatment of High Nitrate Waters"
     Percy P. St. Amant, Jr., and Perry L. McCarty.  Presented at
     Annual Conference, American Water Works Association, San Diego
     California.  May 21, 1969.  American Water Works Association
     Journal.  Vol. 61. No. 12.  December 1969.  pp. 659-662.

The following papers were presented at the National Fall Meeting
of the American Geophysical Union, Hydrology Section, San Francisco,
California.  December 15-18, 1969.  They are published in Collected
Papers Regarding Nitrates in Agricultural Waste Water.  USD!, FWQA,
#13030 ELY December 1969.

"The Effects of Nitrogen Removal on the Algal Growth Potential of
San Joaquin Valley Agricultural Tile Drainage Effluents"
     Randall L. Brown, Richard C. Bain, Jr. and Milton G. Tunzi.

"Harvesting of Algae Grown in Agricultural Wastewaters"
     Bruce A. Butterfield and James R. Jones.

"Monitoring Nutrients and Pesticides in Subsurface Agricultural
 Drainage"
     Lawrence R.Glandon, Jr., and Louis A. Beck.

"Combined Nutrient Removal and Transport System for Tile Drainage
 from the San Joaquin Valley"
     Joel Goldman, James F.Arthur, William J. Oswald, and Louis
     A. Beck.

"Desalination of Irrigation Return Waters"
     Bryan R. Sword.

"Bacterial Denitrification of Agricultural Tile Drainage"
     Thomas A. Tamblyn, Perry L. McCarty and Percy P. St. Amant.

"Algal Nutrient Responses in Agricultural Wastewater"
     James F. Arthur, Randall L. Brown, Bruce A. Butterfield, Joel
     C. Goldman
                                        . ir. s. i;ovi;nNMi.NT IWNTIM; OK KICK  1072—(.
                                   -68-

-------
1
/Irces.sion Ninnhcr
w
5
r\ 1 Subject Fii'lcJ & Group
05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
Organization
Environmental Protection Agency
             Office of Research  & Monitoring
             Robert S. Kerr Water Research Center,  Ada, Oklahoma
    Title
   DENITRIFICATION BY ANAEROBIC  FILTERS  AND PONDS
IQ Author(s)
Sword , Bryan
R.
16

21
Project Designation
EPA Project Number
13030 ELY
Wo/c
22
    Citation
   BIO-ENGINEERING ASPECTS  OF AGRICULTURAL DRAINAGE
   Report Number  13030ELY04/71-8
   Pages-68.  Figures-20.  Tables-18,  References-36
23
    Descriptors (Starred First)
    *Agricultural Wastes,  *Denitrification, *Irrigation Water, *Return Flows,
    *Nitrate,  *Anaerobic Treatment
25
    Identifiers (Slurred
*San Joaquin Valley, California, Bacterial Denitrification,  Anaerobic
Filters, Anaerobic Ponds
27
                   of nitrogen  from  tile  drainage by means of bacterial reduction
was investigated at the Interagency  Wastewater  Treatment Center near Firebaugh,
California.  The major nitrogen form in tile  drainage is nitrate (approx.  98%).
The process required that an organic carbon source be added to the waste to
accomplish reduction of the nitrogen.  The bacterial process was used in two
configurations; anaerobic filters  and  anaerobic deep ponds.  It was found that
with the addition of 65 rag/1 «xi«»K of methanol, 20 mg/1 nitrate-nitrogen
could be reduced to 2 mg/1 or less of  tutal nitrogen within one hour of treatment
by filter denitrification at water temperatures as low as 14*C.  The same
removal was achieved at 12"C in a  filter  operating at a detention time of two
hours.  A covered deep pond required an actual  detention time of eight days at
water temperatures of approximately  22*C  and  a  theoretical detention time of
15 days at temperatures of approximately  16°C to accomplish the same removal.
An uncovered pond was not able  to  achieve the same results at theoretical detention
 times as long as 20 days.  The projected costs for both processes are
approximately $90 per million gallons.

        This report was submitted  in fulfillment of Project 13030 ELT under the
sponsorship of the Office of Research & Monitoring of the Environmental
Prnf^rtion Agency.
                              '"•s"'""°"
                                      Environmental Protection Agency
 WR:t02 (REV. J U L. V !9G9»
 •V R SI C
                        SEND WITH COPV OF DOCUMENT. TO: WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                                 U.S. DEPARTMENT OF THE INTERIOR
                                                 WASHINGTON. O. C. 20210

                                                                          * GPO: IS7U-389-630

-------
ENVIRONMENTAL PROTECTION  AGENCY
   Publications Distribution Section
   Route 8, Box 116, Hwy. 70, West
   Raleigh,  North Carolina  27607
     POSTAGE AND FEES PAID
ENVIRONMENTAL PROTECTION AGENCY
        Official Business
                                    If your address is incorrect, please change on the above label;
                                    tear off; and return to the above address.
                                    If you do not desire to continue receiving this technical report
                                    series, CHECK HERE Q  ; tear off label, and return it to the
                                    above address.

-------