DEMONSTRATING LEACHATE TREATMENT

       REPORT ON A FULL-SCALE OPERATING PLANT
          This report (SW-758) was written
by R. L. Steiner, J. D.  Keenan, and A. A. Fungaroli,
   and is reproduced as  received from the grantee.
       U.S. ENVIRONMENTAL PROTECTION AGENCY

                       1979

-------
     This project was conducted at the GROWS (Geological Reclamation
Operations and Waste Systems, Inc.) landfill in Falls Township,
Pennsylvania, with partial funding from the U.S. Environmental
Protection Agency, demonstration grant No. S-803926.

     Publication does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor
does mention of commercial products constitute endorsement by the U.S.
Government.

     An environmental protection publication' (SW-758) in the solid
waste management series.

-------
                                                            CONTENTS
                                                                Page



      Summary and Conclusions	    1




  !.   INTRODUCTION	    8




 II.   OVERVIEW OF LEACHATE TREATMENT OPTIONS	   12



      Leachate Composition	   12




      Leachate Treatment	   22



      Summary	   31




III.   LEACHATE TREATMENT SYSTEM—		—	   33



      Design  Overview	   36




           Design Flow	•	   36




           Design Leachate Characteristics	   38



           Design Concept	   4o



           Leachate  Collection  System--	•—   40



      Chemical/Physical  Section	-,-   42




           Chemical  Precipitation	   42




           Air  Stripping  of Ammonia	•	   43




           Neutralization  and Nutrient Supplementation—•	   44

-------
                                                                          w>
                                                                          V •*•
                                               CONTENTS  (Continued)

                                               ';.  •".  ::;  ;-ป  '^'' ''  Page
                                               •< ! .  ' ••! :  •*•':',•'  -, '    •    ."
     Biological Treatment Section                                 44

IV.   MATERIALS AND METHODS	--——----—-—_——_--.*'   47

     Experimental Systems	•	-—-—     47

          System 1 - Chemical/Physical fol owed  by
                   .Biological Treatment—-—-—-	—     47

          System 2 - Chemical/Physical Treatment	•	—     47

          System 3 - Biological followed by  Chemical/
                     Physical Treatment———	—--	"-     49

          System 4 - Biological Treatment	——	     49

          System 5 - Bench-Scale Testing—•	•	——     49
              1    "      '"       "        '„    ''I ..'.'-'  ,.' i';    ~ \ .   '.'      ..
     Process Monitoring	•	———	——     50

     Statistical Tests	___-________—____—_____„_     52

     Presentation of Results	;——	•—•-     54

 V.   RESULTS AND DISCUSSION		—	———     55

     Preliminary Results	     55

          Raw Leachate Quality	     55

          Lime Dosage	•	•--	—-—----"	     59

          Sulfuric Acid Dosage	     60

          Phosphoric Acid Dosage	——	—     60

     System 1 - Physical/Chemical Plus Activated Sludge	     61

          Operational, Comments	—_________:—,___.—__.	_     53

          Cost Data	•	•—-	     72
               *    '    '     .   i         ." : ']  ' „  • n , •- '    ' - f r
     Nitrification	—__-_—__	.______._	„„_„...,     75

          Summary	•	——	,-ซ-^-^—ซ—-     90

-------
 ';.   "   ,  '                                     CONTENTS (Contfnuedj




 vt-                                                            Page



     System 2 - Chemical/Physical Treatment	   90




          Operational Comments	   95




          Cost Data	:	--'	   95




     Factors Influencing Lime Treatment Performance	   95




     Systems 3 and 4 - Biological Treatment of Raw Leachate	  104



     System 5 - Laboratory Studies	  107



          Activated Carbon	'•	  107




          Additional Laboratory-Scale Studies	  115



          Leachate Treatment Plant Startup	•—•	  125



VI .   CONCLUSIONS	—		—  132



VII. REFERENCES		——		——_-_.„  136

-------
                           LIST OF TABLES

                                                                Page

 1  - Summary of System 1  Operation Data-	    6

 2 - The Strength of Raw Leachates	•	   13

 3 - Effect of Solid Waste Disposal on Groundwater Quality	   17

 k - Effect of Landfill  Depth on Leachate Composition and
     Pollutant Removal at the University of West Virginia
       iOฃE:____________________________________________________   1S
       I j3Op                                                "      IO

 5 - Theoretical Removal  of Heavy Metals During Lime
     Precipitation	•	   27

 6 - Leachate Treatability as Hypothesized by Chian and Dewalle   30

 7 - Precipitation and Average Monthly Temperature Data
     Trenton, New Jersey	   35

 8 - Summary of Effluent Criteria for GROWS Sanitary Landfill
     Leachate Treatment Facility	-•——*—   37

 9 - Design Leachate Characteristics	   39

10 - Periods of Operation of Leachate Treatment Systems	   48

11  - Routine Laboratory Chemical Analysis	   51

12 - Landfill Leachate Characteristics	——   57

13 - Effect of Equalization Pond on Raw Leachate Variability	   58

\k - System 1 Treatment Performance after Acclimation of
     Activated Sludge (August 1, 1976 - May 1, 1977 and
     July 1, 1978 - August 31, 1978)		~	   63

15 - Summary of System 1  Operation
     (8/1/76 to V30/77 and Vl/78 to 8/31/78)	   68

16 - Warm Weather Operation of System 1	ซ   70

17 - Comparison of Series and Parallel Operation of Activated
     Sludge Units					-	-—   71

18 - Operation and Maintenance Costs Incurred During the
     Operation of System 1 following Acclimation of Activated
     Sludge (8/1/76 to 5/1/77 and 7/1/78 to 8/31/78)	   73

-------
                     LIST  OF TABLES  (Continued)

                                                                 Page

19 - Ammonia Removal in Activated  Sludge  Units-"-*—-.--*—„„..„-*   78

20 - Summary of System 2 Results-—-	--—----	——,---—*~   91

21 - Summary of Effects of Chemical/Physical  Treatment-—--'"'-   93

22 - Summary of Operation  and Maintenance Costs During
     Evaluation of System  2  (11/15/75-5/1/77  and 11/1/77-
     8/31/78)	-	   96

23 - Summary of Operational Data for  Lime Treatment and
     Clarification	—	*-*-,	„   97

2k - Results of Batch Draw-and-Fi 11 Activated  Sludge
     Experiments to Determine the  Extent  of Phosphorus
     Limi tat ion------------------"•--------"*-----•*-'•"••--""•'•"•"•"""•""•'  106.

25 - System 3 Operation	•	—  108

26 - System k Operation	  109

27 - Summary of System 3 Operation Data (5/1/77-8/31/77)	  110

28 - Summary of Results of Carbon Adsorption Treatment of  Raw
     Leachate	  112

29 - Treatment of Final Effluent with Bench-Scale Activated
     Carbon Columns	  113

30 - Pilot-Scale Carbon Treatment of  Final Effluent	  114

31 - Results of Alkaline Chlorination Studies	•	  118

32 - Experimental  Protocol and Preliminary Results  in
     Evaluation of Lime Treatment Additives	  120

33 - Results of Additive Evaluation	  121

3k - Preliminary Filtration Results	  ]2k

35 - Process Loading Rates and Concentrations  Observed During
     the Period April 1, 1978 through June 30, 1978	  129

-------
                          LIST OF FIGURES

                                                                Page

 1 - Reduction in COD During Aerobic Treatment	   2k

 2 - Changes in Total Dissolved Solids (TDS) During Aerobic
     Treatment Studies	   25

 3 - Location of Leachate Treatment Plant	   3k

 k - Schematic Flow System 1 with Ammonia Stripping Lagoon	   k]

 5 - Schematic of Pilot Leachate Treatment Plant
     (Scaled Version of System 1)	   53

 6 - Raw Leachate Chemical Oxygen Demand	   56

 7 - Flow Chart for Activated Sludge in Series	   67

 8 - Effect of Temperature on Specific Oxidation Rate	   82

 9 - Substrate Inhibition of Nitrification	   85

10 - Effect of Low Concentrations of Substrate on Specific
     Oxidation Rate (R,)	   87

11 - Effect of pH on Clarifier Effluent Nickel Concentration	  101

12 - Effect of pH on Clarifier Effluent Mercury Concentration—  102

13 - Effect of Leachate Temperature on Clarifier Effluent
     Nickel Concentration	  103

\k - Carbon Breakthrough Curve	  116

-------
                    DEMONSTRATING  LEACHATE  TREATMENT
                 Report  on  a  Full-Scale Operating  Plant
               R.L.  Steiner,  Ph.D.,  P.E.,  J.D.  Keenan,  Ph.D.,
                        A.A.  Fungaroli,  Ph.D.,  P.E.
                         Summary and  Conclusions


     The  results of 3 years of operation of a  full-scale  sanitary

 landfill  leachate treatment plant are  reported.  The plant  is designed

 to provide a variety of  chemical/physical and  biological  treatment

 sequence  options.  The chemical/physical units  include equalization,

 lime precipitation, sedimentation, air stripping, neutralization and

 nutrient  supplementation.  These treatment processes are  designed to

 remove heavy metals, ammonia and organic materials, and to encourage

 subsequent biological treatment by reducing the pH and adding the

 nutrient  phosphorus.  The biological treatment process is activated

 sludge designed to provide both organic 5-day biochemical oxygen

 demand (BOD,-) degradation and nitrification.  The demonstration

 leachate  treatment plant is designed to provide operational flexibility

 in that the flow can be directed through the various unit processes and
              i
 operations in any sequence.

     The  purpose of this project was to demonstrate the efficiency of a

 number of treatment sequences.  Specifically, five modes of operation

were defined and have been investigated.   System 1  consists of chemical/

 physical   treatment followed by activated sludge; System 2, chemical/

                                  -1-

-------
physical treatment only; System 3, biological treatment followed by

chemical/physical; System 4, biological treatment only; and, System 5,

bench-scale studies, including activated carbon adsorption treatment.

     Data have been collected which can be used to characterize the

quality of raw leachate generated in an operating sanitary landfill.

These data show that the leachate from this sanitary landfill source

is high in organic matter (average chemical oxygen demand (COD)/liter

of 18,553 mg, average BOD,./! iter of 10,907 mg) and nitrogen  (average

NH^-N/liter of 1,001 mg).  At the end of the first 2 years of operation

these figures were 11,210 and 17,562; 4,460 and 10,773; and  1,503 and

1,0^7, respectively.  Thus, although influent nitrogen values have

fallen, the increase in organic strength has been extremely  large.  The

raw leachate heavy metal  concentrations are somewhat lower  than

expected, possibly reflecting the relatively high pH of the  leachate.

(Note that all data have been collected with nonfiltered samples.)

     High concentrations of ammonia in the raw leachate exceed the

plant's effluent criterion and are sufficient to inhibit the growth of

the activated sludge microorganisms.  For this reason the original

plant design was augmented with an ammonia stripping lagoon.

     System 5 studies have been conducted for a number of purposes.

Bench-scale tests have provided optimal operating data for chemical/

physical units.   In particular, System 5 has provided data for the
                            r
development of lime, sulfuric acid, and phosphoric acid dosages.

     Activated carbon adsorption has been evaluated as a treatment

method for raw leachate.  For raw leachate, carbon adsorption did not

prove to be an effective treatment procedure.  The inability to use


                                  -2-

-------
 carbon adsorption  is  the result of high suspended solids loading causing
                                                /
 pore plugging  and  the  wide range of flow variability.

      Pilot  scale data  have been, collected for System 5,  carbon
                          "**
 adsorption  of  System  1  effluent.   In this mode,  the  carbon  column would

 serve as  a  tertiary, or advanced  waste,  treatment process.   The  results

 indicate  that  the  carbon can  remove much of  the  remaining COD and heavy

 metals.   The results have been  analyzed  in terms  of  Langmuir adsorption

 isotherms and  carbon breakthrough curves.  This way  of handling  the data

 provides  preliminary full-scale design  information.

      Systems 3 and k,  those in  which raw leachate is influent to the

 biological  units,  have  received considerable  operating attention.  The

 results indicate that  the raw leachate  is  not directly treatable by

 biological  means.  Systems 3 and  k  yield an effluent which  is high in

 organic matter.  The mean  effluent  BOD from System 3 was 763 mg  per

 liter.  The performance  of Systems  3 and 4 has not been  satisfactory

 for  the treatment of this  leachate.

      Systems 1  and 2 are  those  In which  the raw leachate is  treated

 first by chemical and physical means.  The results of these  systems'

are most promising.  During the third year, these systems were preceded

by equalization.   Lime precipitation followed by  sedimentation has been

successful  in removing the heavy metals and a portion of the organic

matter.  Specifically, this sequence (System 2) has  removed about one-

quarter of the  nitrogen; one-third of the dissolved  solids; one-half

of the organic  matter; three-quarters of the suspended  solids;  and

ninety percent  of the phosphates.  The sequence has been successful in

-------
removing  the  heavy metals including one-half of the cadmium and



mercury; two-thirds of the lead, chromium, and nickel; three-quarters




of the copper; over ninety percent of the iron and zinc.



     The performance of System 2 has been studied carefully in order to




determine the treatment unit's response to a number of operational



parameters.   It was found that temperature and pH both exert an effect



on the concentration of heavy metals in the lime treatment effluent.



However, the response is not identical  for all heavy metals.  It may be



possible to use the differences in these responses in an operational



control strategy to achieve optimal removal efficiences of selected




contaminants.



     An ammonia stripping lagoon is included in the chemical/physical



treatment sequence because of the excessive ammonia levels in the raw



leachate.  During the lime precipitation/clarification/ammonia stripping



mode of operation, the following removal efficiencies have been achieved:



66 and 50 percent of the BOD and COD, respectively; approximately 60



percent of the ammonia-N and total Kjeldahl-N; approximately 75 percent



of the suspended solids; 25 percent of copper; 50 to 60 percent of



cadmium and nickel; 64 to 68 percent of lead; approximately 96 percent




of zinc; 98 percent of iron.



     The ammonia lagoon has a detention time of 1.7^ days, thereby



providing an equalizing effect.  That is, the effect of the lagoon is



to dampen the peaks and to minimize shock loadings on subsequent



treatment units.  For example, during the period in which the lagoon




was included  in the treatment sequence, the mean ammonia concentration

-------
  In  the raw leachate  was  1001  mg  per liter with a standard  deviation




  slightly  larger,  indicating  tremendous  variability.   During  the same




  period, the t]  standard  deviation  interval  for the ammonia lagoon




  effluent  was  203 to 641 mg per liter.  Thus, the equalization effect is



  significantly beneficial  in  terms  of  lessening shock  loadings.




       During the third  year, an equalization  pond was  used  to  further




  dampen the  fluctuations  in leachate quality  and  quantity.  This was




  done  to provide a more even flow to the  lime treatment unit.  The




  effect of the equalization was to  reduce  influent variability for many




  parameters, as measured  by the coefficient of  variation; and to enable



  more  uniform dosing of raw leachate with  lime.




       System 1  provided the best degree of treatment.   This sequence



  consists of equalization, lime precipitation/clarlficat ion/ammonia



  stripping/neutralization/phosphorus addition/activated sludge.  In this



 operational configuration, excellent removal efficiencies have been



 observed,  following the adaptation of the activated sludge to the waste




  (Table 1).  Except for NH^-N, BOD^ and lead, the effluent concentrations



 comply with the criteria developed by the Pennsylvania Department of



 Environmental  Resources and the Delaware River,Basin  Commission for



 discharge  to the Delaware River.   The standards for these parameters



 were not met because  of the unusually severe temperatures of the winter



 of 1976-77,  and secondarily because of the great increase in raw



 leachate strength  which began during the second year of this project.




(The  treatment  performance of  System 1  met all standards during  periods



 with relatively warm  weather.   During  the period up to January 1977




                                  -5-

-------
                              TABLE 1




                 SUMMARY OF, SYSTEM 1 OPERATION DATA
Parameter
Ammonia-N
BOD_
Cadmium
Chromium
COD
Copper
I ron
Lead
Mercury
Nickel
Zinc
8/1/76
Raw
Leachate
mc/l iter
758
11886
0.08
0.26
18490
0.40
333
0.74
0.006
1.76
19.5
to 5/1/77 and 4/1/78 to 8/31/78
Final Discharge
Effluent Percent Standard
mq/liter Removal mq/Hter
75
153
0.017
0.07
945
0.11
2.7
0.12
0.004
0.75
0.53
90.1
98.7
78.2
73.1
94.9
72.5
99.2
83.8
27.4
57.4
97.3
35
100
0.02
0.1
*
0.2
7.0
0.1
0.01
sV
0.6

No discharge standard for this parameter.
                                  -6-

-------
System 1 consistently met the effluent criteria.  Likewise, during


the third project year, the summertime performance of System  1 was

excel lent.)


      The System 1 operating data have been examined closely in order to


characterize the ammonia removal mechanisms.   In the lagoon this occurs


as the result of volatilization of the free ammonia predominant at high


pH levels.  In the activated sludge units, the principal mechanism for


ammonia removal is biological nitrification to nitrate.  The  rate of


nitrification, expressed as the specific oxidation rate, is a function


of temperature which follows the vari't-Hoff Arrhenius relationship.


The results show that the activation energy is approximately  12350 cal.
                                                         q     _i
per mole, and the Arrhenius frequency factor is 2.18 x l(r day


The data indicate that substrate inhibition due to ammonium ion

concentration occurs in this system.   This relationship has been fitted

to the Haldane inhibition model.  The maximum specific oxidation rate


is 3.5 g N oxidized per g biomass/day.  The saturation constant is 4


mg per liter,  and the inhibition constant is 36 mg per liter.
                                  -7-

-------
                   DEMONSTRATING LEACHATE TREATMENT
                Report on a Full-Scale Operating Plant


             R.L. Steiner, J.D. Keenan, and A.A. Fungaroli


                           I.   INTRODUCTION

     The potential for water pollution from sanitary landfill sites has

become recognized in recent years.  A number of studies have documented

the great pollutional strength of landfill leachates.      The quality

of this material varies with landfill age, nature and moisture content

of the wastes disposed at the site, and hydrologic and soil factors.

In spite of this variability, it can be stated that, especially for

young landfills, the values of the critical sanitary parameters of

leachate are at least an order of magnitude greater than for domestic

sewage.  The deleterious consequences following contamination of ground

and/or surface waters by leachate may be severe, and it is for this

reason that leachate treatment  is receiving attention.

     Solid waste consists of matter which can be decomposed by bacterial

or microbial action, as well  as of materials which are inert to

microbiological activity.  Some of the compounds, cellulose in

particular, are resistant to biological breakdown, but with sufficient

time decomposition will occur.  Because of this resistivity and

necessity to acclimate the biological system, the chemical characteristics

of leachate are time-dependent.  To complicate treatment, as the paper

                                  -8-

-------
decomposes, some of the Inorganic  ions which are bound to the organic




matrix are released and can be removed by water percolating through




the landfill.  The actual mechanism of removal varies with the component



but includes solution as well as colloidal transport.




     The generation of leachate in landfills is complicated and cannot




be generalized simply as surface water percolating through the sanitary




landfill.  When refuse is placed in the landfill, decomposition begins




to occur.  Some decomposition products may be water soluble whereas the




parent products might not have been.  This is especially true of




cellulose.  In addition, the inorganic constituents also must be




considered since they vary with the state of decomposition.  The amount




of water percolating through a sanitary landfill is the primary control




of leachate quality, but the chemical  characteristics of the leachate



are dependent on other parameters,  including temperature, water



composition, moisture content, time, mode of decomposition (aerobic,



etc.)  and the amount of infiltration of rainfall at the landfill.



     Recent studies have shown that leachate is produced in a sanitary



landfill when the precipitation exceeds the net evapotranspiration of



the region.  Remson, Fungaroli and  Lawrence developed a model  for




predicting the movement of leachate through a sanitary landfill.'"



Further results using this model  have substantiated the validity of the



approach and prediction of leachate generation patterns is reasonably



accurate.  Dass et_ a^. have also used  a water budget method for



predicting leachate generation.  '




     Ground and surface waters can  be protected if the landfill  is




underlain with an impervious membrane.  With proper design, leachate




                                 -9-

-------
Is then directed toward collection points.  A waste such as this, which



Is properly considered an industrial waste, must be treated prior to



surface discharge.  The leachate treatment state-of-the-art is still



embryonic, although a few small scale studies have been conducted.



These have demonstrated that neither conventional chemical treatment



nor biological treatment can achieve the high degree of treatment



efficiency expected today.  Consequently, although we know that the



pollution potential of sanitary landfill leachate can be avoided by



interception using impervious liners, we are not yet able to define



the optimum sequence of unit operations and processes required for



adequate wastewater renovation.



     The U.S. Environmental Protection Agency, Office of Solid Waste,



awarded a demonstration grant (S-803926) to investigate the



effectiveness of alternative treatment sequences as employed at the



full-scale facility in Falls Township, Pennsylvania.  A 380 liter



per minute (0.144 mgpd) plant had been constructed to treat leachate



from the GROWS (Geological Reclamation Operations and Waste Systems,



Inc.) landfill.  This project had as its primary goal the evaluation



of the technical feasibility, operational efficiency, and cost



effectiveness of four alternative treatment sequences.  These are:



(1) chemical/physical followed by biological;  (2) chemical/physical



alone; (3) biological followed by chemical/physical; and  (4) biological



alone.  The chemical/physical processing  includes precipitation of



heavy metals by lime addition, sedimentation, air stripping of ammonia,



and neutralization using sulfuric and/or phosphoric acids.  Equalization



                                 -10-

-------
of raw leachate was Initiated during the third year of the project.



Biological treatment consists of conventional activated sludge.



Additional objectives of the study were the bench-scale evaluation of



carbon adsorption on both raw leachate and unit process effluents; and



bench-scale testing to determine chemical dosage, sludge return rates,



aeration rates, and other plant operation criteria (System 5).  The



purpose of this document is to present and discuss the results of the



3 years of operation of this facility.
                                 -11-

-------
               M.  OVERVIEW OF LEACHATE TREATMENT OPTIONS

      The purpose of  this chapter  is to review the  literature regarding

  the composition of sanitary  landfill  leachates and their treatment.   In

  brief,  the character and variability  of  the  leachate dictates the types

  of treatment  systems which will be effective.  The contaminants of

  greatest concern fall  into several groups.   The first group  is the

  organic chamicals,  important primarily because they exert an oxygen

  demand  on  receiving  waters which  may  result  in a depletion of dissolved

  oxygen  deleterious to  aquatic  life.   The second major group of

  contaminants  found  in  sanitary  landfill  leachates  is comprised of the

  heavy metals. As a  group, these  elements are of concern because they

  are toxic  at  sufficiently high  concentrations.   It is conventional

  practice to chemically characterize wastewaters such as  leachate  in  terms

  of a number of other parameters.  These  are  used for a variety of

  purposes  including design, operation  control, and  evaluation of pollution

  potential.

                           Leachate Compos i tion

       In 1932, one of the first studies  indicating  that  the  disposal  of

.  solid waste could cause environmental pollution was  reported by  Calvert

  who  investigated the liquid  waste from a garbage  reduction  plant  in
               r\
  Indianapolis.   In  this process the garbage  was  cooked  and  the grease

  removed to produce  fertilizer and animal feed,  and the  liquid  waste  was

  discharged into an   impounding pit or lagoon.  An  analysis  of this liquid

  is presented in Table 2, Column 1.   Calvert  analyzed  the groundwater

  from existing wells surrounding the lagoon and  found  that  wells up to

                                   -12-

-------
                                                               TABLE 2

                                                THE STRENGTH  OF RAW LEACHATES21*'25*

Parameter
Acidity, as CaC03
Alkalinity, as CaCOj
Aluminum Oxide
Ammonla-H
Arsen c
Barium
BOD- 5
BOD -20
Cadmium
Calcium
Carbon, Total Organic
Chloride
COD
Copper
Cyanide
Dlsso ved Oxygen
Fluoride
Hardness, as CaCOj
Iron
Lead
I Hagne I urn
— * Manga ese
\jO Nlckc
| NItra e-N
Hitrl e-N + Nitrate-H
Nitrogen, Total
Organ c-N
ptl
Pltosphates, P
Phosphorus. Total, P
Potassium
Sodium
'Specific Conductance
Sulf*te
Sulfide
Suspended Solids
Total Dissolved Solids
Total Sol Ids •
Zinc

(I)
3,000-3,300
-
874
_
-
-
-
-
-
66.3

2,950
-
-
-
-
-
-
246
-
182
-
-
-
-
-
-
-
-
-
-
•
_
-
_
1,000-2. 500
-
.
-

(2)
100-9,450
25-4,000
-
-
-
-
6-7.330
-
-•
-
-
280-12,300
-
-
-
0-5.6
-
-
-
-
-
-
-
-
-
-
-
5.7-8.4

-
-
-
.
-
_
.
-
-
-

(3)
.
730-9,500
-
0.2- '180
-
-
-
21.700-30.300
-
240-2,330
-
96-2,350
-
-
-
.,
-
890-7.600
., 7-220
-
64-410
-
-
-
-
-
2-465
6.0-6.5
0.1-10
-
28-1,700
85-1,700
_
84-730
„
.
-
-
-

(4)
_
.
-
54
-
-
975
-
-
-
-
128
-
-
-
-
-
523
-
.
-
-
_
-
-
.
7

_
-
-
.
_
3
10


-
-

(5)
_
_
,,-
700
-
-
7,745
-
-
-
-
2.000
-
-
-
-
-
-
-
_
-
-
-
-
--
-
200
-
.
-
-
.
_
1,950
_
.
-
-
-
Average Ag
0.5 yr.
(6)
_
3.255

-
4.31
8.5

54,610
-
-
-
1,697
39,680
0.05 .
0.024
_
-
7,830
5,500
-
-
1.66
-
1.70
-
-
-
.
2
-
-
900

680
_
.
11,144

-
e of Fill
6 yr.
(7)
-
4.159

-
O.I
0.8
-
14,080
-
-
-
1.330
8,000
0.05
0.005
-
2
2,200
6.3

-
0.06
-
0.70
-
-
-
6.3

-
-,
810

2
-
_
6.734

0.13
Material
17 yr.
(8)
.
1.001
-
-
4.6
0.3

225
-
-
-
135
40
0.05
0.02
-
0.31
540
0.6
.
.
0.06
-
r.6o
-
_

7.0
3.0

.
74

2
.
.
1,198

0.10

(9)
.
_
"
177
-
-
.
-
-
_
.
2,340
50.715
5.0

_
_
5,500 '
1,640

_
-
0.8
-
-
.
482

43

_
3,800
_
375
-
26.500
_
43,000
129

(10)
_

•2,600-23,000
-
- '
-
41,000-180,000
-
-
-
.
-
-
-
_
_
_
.
-
_
. _
-
-
-
-
2,000-10,000
-
-
-.
-
-
-
-
-
-
-
-
J
-'

Range 1
(It)
_
0-20,850
-
0-1,106
-
-
81-33,360
-
0.03-17
60-7.200
256-28,000
4.7-2,467
40-89,520
0-9.9
-
-
-
0-22,800
0-2,820
<0. 1-2.0
17-15,600
0.09-125
-
" -
0.2-10.29
.
-
3.7-8.5
6.5-85
0-130
28-3.770
0-7,700
2,810-16,800
1-1,558
-
10-700
584-44,900
0-59,200
0-370

Range 2
(12)
_
142-3,520
-
1.4-1,028
-
-
3.9-57,000
-
-
76-3,900
70-27,700
60.2-2,467
31.1-71,680
-
.
_
-
.
0.5-2,200
_
35-1. 140

-
-
0.4-10.29
-
-
5.09-7.25
0.25-85
0.5-98
35-2.300
44-1.580
978-16,800
7.4-1,558
-
8.9-923
-
911-55,348
-

Range 3
(13)
_
560
-
155
-
-
6,300
-
-
550
3,600
470
8,000
-
-
.
-
,
<440
_
210
-
-
-
1.5
r
-
5.98
3.2
9
380
280
4,970
90
-
197
-
6,080
-
Reference No.
                                                                                   10 "
     *A1T untts are mg  per  liter except pH (pH units) and  specific conductance
(pmho/cm).

-------
 500  feet  downstream of the  site  showed  a marked  increase  in magnesium,



 calcium,  total  dissolved  solids  and  carbon  dioxide.



      Carpenter  and Setter, working at New York University  in 19^0,



 conducted one of  the  earliest  studies concerned with  landfill  leachate.



 Auger holes were  drilled  through an  existing  landfill of undetermined



 age  into  the subsoil.   Twenty-eight samples  of  leachate which were



 collected in the  bore  holes were analyzed chemically.  The range of



 concentrations  is presented in Table 2, Column 2.  These results showed



 a wide variation of concentration over  the  site, thus indicating the



 difference of filled materials at various locations, or the differences



 in the age of the refuse  at different points.  Analysis of groundwater



 in the area was not performed; therefore, the effect on the subsurface



 environment was undefined.



      The  first comprehensive research study of sanitary landfills under



 controlled conditions was conducted  at  the University of Southern


           20
 California.   Test bins,  simulating  landfill conditions, were constructed.



 Water  was added to simulate the  infiltration of 1.12 m and leachate was



 collected and analyzed.  Table 2 gives  the minimum and maximum (Column



 3) values of the  initial  (first ^5.9 liters of leachate per cu m of



 compacted refuse) leachate.  The most rapid removal (the highest



 concentrations)  occurred with the first 232 liters per cu m of refuse.



 Thus,  it was postulated that removal  would continue for many years but



at a very slow rate,  and  it was considered unlikely that all  the



constituents would ever be removed.

-------
      The  same  study  also  examined  a  field  site  consisting of  2.4 m  of




 refuse  and  0.61 m  of cover material.  The  refuse was  in  intermittent




 contact with the groundwater, analysis of  which showed  increases in all




 organic ions and a maximum biochemical oxygen demand  of  125 mg/liter.




 One conclusion of  the study was  that  the dissolved  inorganic  ions




 entering  the groundwater  through intermittent contact would decrease in




 concentration as a result of dilution and  adsorption  and travel in  the



 direction of the groundwater movement.




      The other conclusions reached in this study are  summarized as




 follows:  (1) A landfill, if located  so that it is  in intermittent  or




 continuous contact with ground water, will cause the  ground water in the




 immediate vicinity of the landfill  to become grossly  polluted and unfit



 for domestic or irrigational use;  (2) dissolved mineral matter, entering



 ground water as a  result of intermittent and partial  contact of a



 landfill with the  underlying ground water will  have its greatest travel



 in the direction of  flow, undergo a vertical  diffusion  to  a limited



 extent, and be subject to dilution, the result of which will be a




minimizing of the effect of the entering pollutant  ions; (3) a landfill,




 if located so that no portion of it intercepts the ground water, will



 not cause impairment of the ground  water for either domestic or



 irrigational use;   (4) rainfall  alone  (in the area of this study) will



not penetrate a 2.3 m thick landfill  sufficiently to cause entry of



 leachate into the underlying ground water.




     Longwel1  stated in 1957 that an appreciable proportion of refuse



could be extracted by water to produce a leachate rich in organic



                                 -15-

-------
                                             21
matter,  Inorganic  salts  (ions), and  bacteria.    The analysis of a



surface  leachate obtained  from an unnamed  landfill  is given  in Table



2  (Column 4).



      In  1961 the British Ministry of Housing and Local Government



conducted extensive research on the  placement of landfills above the



groundwater table  (which they called "dry  tipping"), and the placement



of landfills below the groundwater table (which they called "wet

          99
tipping").  In the "wet tipped" experiment the refuse was completely



submerged and the  horizontal groundwater flow rate was equivalent to



138 liters per sq  m per day.  The leachate quality  is Included in Table



2  (Column 5).  Analyses of the groundwater before and after contact with



the refuse are given  in Table 3.  These results show the considerable



extent of groundwater quality degradation due to pollution by leachate.



      In  1965, Qasim studied the seepage waters from simulated landfills


                                   23
at the University  of West Virginia.   Three concrete cylinders 0.9 m in



diameter  and 1.2, 2.4 and 3.7 m in  height were filled with municipal



refuse.  Approximately 102 cm of precipitation were artificially added



to the cylinders over a period of 6 months and leachate samples were



collected.  The maximum concentrations of certain organic and Inorganic



components in the  leachate from the  three cylinders are presented in



Table k.  Table 4 also presents the  total  weight removed per cubic



meter from each depth of fill by 102 cm of simulated infiltration.



     A summary of  results presented by Qasim demonstrates the effect of



depth on leachates generated by landfills.   Concentrations of various



pollutants were higher in leachates obtained from deeper fills.



Concentrations of various pollutants per unit depth of fill  decrease



                                -16-

-------
                    TABLE 3

EFFECT OF SOLID WASTE DISPOSAL ON GROUNDWATER QUALITY-
  GROUNDWATER QUALITY BEFORE AND AFTER INTRODUCTION
            OF "WET TIPPED" LANDFILL - 196126

Concentration (mg/liter)
Measured Quantity
Total Solids (Residue)
Chloride
Alkalinity, as CaC03
Sulfate
Biochemical Oxygen Demand (BOD^)
Organic Nitrogen
upstream or
Landfill
450
30
180
120
0
0
Downstream or
Landfill
5,000
500
800
1,300
2,500
70
                    -17-

-------
                                             Table 4


                      EFFECT OF LANDFILL DEPTH ON LEACHATE COMPOSITION AND

                      POLLUTANT REMOVAL AT THE UNIVERSITY OF WEST VIRGINIA - 1965
29
CO
I
Concentration (mg/liter)

Parameter
Alkalinity, as CaCO
Bicarbonate ^
BOD5
Chloride
Hardness, as CaCO^
Nitrogen, Total
Sodium and Potassium
Solids, Total
Sulfate
0./6 m
Fill
10,630
—
14,760
951
7,600
613
1,63'f
21,140
--
1.98 m
Fill
16,200
—
26,200
2,000
13,100
i,389
3,963
49,800
__
3.1 m
Fill
20,850
--
33,360
2,310
10,950
2,508
5,109
59,000

Pollutant Removal (kq
0.76 m
Fill
..
9.4
12.7
0.8
2.8
0.6
22.0
__
nc
1.98 m
Fill

7.1
10.6
0.6
1.8
0.6
16.6

n *
j>er cu m)
3.1 m
Fill

5.9
9.0
0.6
1.1
0.6
14.4

n o

-------
with increasing depths of refuse.  For an equal amount of influent,


shallower fills showed greater extraction rate per unit volume of fill


than deeper fills.  The bulk of the pollution was attributed to initial


leaching.


     Anderson and Dornbush conducted an extensive investigation of the

                                                                  27
groundwater leaving a landfill in Brookings, South Dakota in 1967.


An abandoned gravel pit of 160 acres with its base well below the


water table was filled with municipal solid waste.  The purpose of the


investigation was to determine which chemical parameters were the most


reliable indicators of the influence of landfi1 Is on the groundwater.


Groundwater samples from 22 wells located over the site were analyzed


for chloride, total hardness, alkalinity, sodium, pH, potassium, iron,


nitrate, and specific conductance.  A considerable increase in all


constituents measured was observed in three wells immediately downstream


of the fill area.  Although the authors did not evaluate the potential


pollution of municipal refuse, they did report an increase of up to


50 times the chloride content of native waters in the groundwater


affected by the leachate.  The major conclusion of this investigation


was that two of the most important indicators of pollution from


landfills are chlorides and specific conductance or total dissolved


solids.  Chloride ions are easily detectable, not readily absorbed by


soils, not affected by biological processes, and apparently an abundant


product of leachates.


     Disposal sites in northern Illinois were investigated in 1970 by

             28
Hughes et_ aJL   Leachate samples from three landfills of varying age



                                 -19-

-------
 were obtained  as  near  to  the base of  the refuse  layer  as  possible.   The
 results  of  these  analyses are presented  in  Table 2  (Columns  6-8).
 Although no information  is given  In the  study  as to  the composition  of
 the  solid waste  in  each fill,  and the analyses were  performed on only one
 sample,  the results do show a  decreasing trend with  time.  However,  it
 was  noted that refuse more than 15 years of age  can  still have a high
 total dissolved solids content—indicating  that  the  stabilization of
 landfills is a long process.
      The laboratory simulated  landfill or lysimeter  study conducted at
 Drexel University from 196?  to 1972 is the  only  study  reported that was
 conducted under completely controlled  laboratory conditions.  It was
 also the only  study reported  in which  the environmental conditions
 completely  simulate the existing  climatic conditions of a region, In
 this  case,  southeastern Pennsylvania.  The  refuse was  placed at as
 received moisture content  and  allowed  to  reach field capacity naturally
 through  the addition of amounts of distilled water equal to the
 precipitation  of the area m'nus the evapotranspiration.  This
 infiltration was added on a weekly basis  and varied from a rate of 8.9
 cm per month during the wet periods to zero during the dry or summer
 periods.  Approximately one year was required for the  refuse to reach
 field capacity, but small quantities of  leachate were generated before
 field capacity was  reached.  The maximum concentrations obtained in the
 first year are given in Table 2,  Column 9.
      It was concluded that this initial  leachate production came from
 the following  sources:   (l) From the refuse.  Most of the initially
generated leachate  is squeezed from the organic components of the refuse
                                -20-

-------
 by  the  compaction  and  placement  procedure.   (2)  From channeling.   Some



 of  the  water added at  the  top of the  lysimeter may find  a  direct  route



 through the refuse to  the  collection  trough,  due to any  inhomogeneities



 in  the  refuse.   (3)  From  an advanced wetting front.   The  wetting  front



 in  the  refuse  probably moves as  a  broad  band  rather than as a  single



 line  interface.  As a  result, substantial  increases in leachate will



 occur before the entire  system is  at  field capacity,   (4)  From the main



 wetting front.  This is  the leachate  which is produced when the system



 reaches field capacity.  At this time,:the input water and the output



 leachate quantities become approximately equal.



     Other studies have mentioned  the leachate problem of  refuse



 disposal in papers dealing with other aspects of the  solid waste problem.



 Leo Weaver has stated  that municipal  refuse can generate leachates high



 in organic pollutants.    Data from this study are  included in Table



 2 (column 10).



     Engineering Science in a study conducted in 196?  in southern



 California concluded that groundwater pollution, which may arise from



 refuse  leachate reaching a water source, will be shown largely as an


                                                            32
 increase in total dissolved solids and specific conductance.



     Walker in 1969 found that a sand and gravel  aquifer in Illinois



was ineffective in removing dissolved chemical ions generated by a



 landfill.    He did report that travel of leachate through a short



distance (3 to 5 m) of this aquifer will  remove organic pollutants



generated by landfills in Illinois and concluded by stating that



 inorganic pollutants constitute the greatest  source of concern.


                                 -21-

-------
Roessler noted an increase in inorganic pollutants in an industrial


water supply 2ฃ miles downstream from a refuse dump 10 years after the

                           3k
dump had started operation.


     Table 2 (columns 11 to 13)  presents a summary of values of raw


leachate composition as compiled by Chian and DeWalle."  The ranges


represent leachates examined by a number of investigators (Range 1 —


Column  II) and a variety of leachates studied at the University of


Illinois (Range 2—Column  12).  These data are the results of a recently


completed literature review.  Another recent report summarizes the


state-of-the-art with respect to ground water monitoring for leachate


contamination.^6This paper should be consulted before initiating a


monitoring program.


     The conclusion to be drawn from this review of landfill leachate


quality  (as summarized  in Table 2)  is that  its composition  is highly


variable from site to site.   In addition, the data show that even at


a given  landfill, considerable variation  is encountered with respect


to  both  space and age.  That  is, variability  is a factor within a


landfill and also over  the history  of the site.  Consequently,  it is


concluded that  landfill leachate quality cannot be predicted a  priori;


and that this quality  is  even variable  at a given site.


                           Leachate  Treatment


      Leachate treatment systems have been evaluated on a  laboratory


scale at Drexel  University.   In one study,  the purpose was  to


characterize  the biodegradation of  organic  matter both with and without


the supplementary  addition of chemicals.    The system consisted  of


                                 -22-

-------
 five aerobic units which were  treated  in  the  following manner:




 (1) control-no  treatment;  (2)  addition of sodium hydroxide  to pH 9;




 (3) addition of sodium hydroxide  to pH 11;  (A) addition of  lime;




 (5) addition of lime plus  sodium  carbonate.   Otherwise, all units were




 handled  in the  same manner.  This procedure included preparation of an




 activated sludge culture by aerating leachate.  Each experimental unit




 was seeded with this culture and was aerated  at a rate of Sk liters of




 air per gram chemical oxygen demand (COD) (1500 cu ft per Ib COD).




 During the testing, all settled solids were recycled to the aeration




 tank with no sludge wastage.  The aeration treatment systems were



 operated on a continuous basis with a hydraulic residence time of



 five days.




     The COD values decrease quite rapidly during the first six days



 and thereafter  approach a  limit (Figure 1).   The results Indicate that




 there are components of leachate which are not amenable to treatment



 in an aerobic system.  The time of adaptation of microorganisms for




 treatment of the organic fraction of leachate may be considerably longer



 than normal  sewage.  Volatile solids concentrations In these tests were



 low when compared to normal activated sludge systems.   This may be one



 reason for the  long time required for stabi1ization.




     A high variation in the concentration of total  dissolved solids



 in the treated effluent was noted (Figure 2).   The cyclic variation of



several  systems is of interest, but not all  of the systems show this




phenomenon.   Since the withdrawal  and addition of leachate was  constant,



there was no reason for the cyclic effect.  Only pretreatment with lime



                                 -23-

-------
I
NJ
           3000
FIGURE 1.   REDUCTION  IN COD DURING

           AEROBIC TREATMENT
           2000
        D)
        o  1000
        o
                                   Treatment^


                                   	No

                                   —o— Lime.NdjCO,

                                   	Lime
               0    2
                                            TIME  (DAYS)

-------
Is)
VJ1
 I
                                                 FIGURE 2.  CHANGES IN TOTAL DISSOLVED

                                                            SOLIDS (TDS) DURING AEROBIC
                                                            TREATMENT STUDIES
        
-------
gave any type of stability and COD reduction.

     Thus, neither biological waste treatment nor chemical-physical

treatment separately is able to reduce the biochemical oxygen demand

(BOD)more than eighty percent.  In fact, the efficiency of the

chemical-physical process is considerably below this level.  It is

hypothesized that two reasons exist fpr the poor removal efficiency of

each individual system: l) the large percentage of high molecular weight

organic materials, and 2) the biological inhibition caused by heavy

metal  presence.  The physical-chemical treatment is needed to remove

the metals and also to hydrolyze some of the organics, and biological

treatment to stabilize the degradable organic matter.

     In addition, biological treatment alone does not remove significant

amounts of the heavy metals.  In fact, biological units may be inhibited

due to the toxic effects of the metals.  Consequently, chemical and/or

physical processing is needed for the removal of substantial amounts of

these1materials.  Lime treatment is particularly effective in that it

creates the alkaline conditions under which the metals become insoluble.

     The removal of heavy metals during lime precipitation depends upon

the formation of  insoluble metal compounds, primarily hydroxides, at

alkaline pH.  The optimum set of conditions  is not identical for all

metals, and the  result is that it is  impossible to achieve the maximum

theoretical removals for each metal within a single tank.  In general,
                                                           •jQ
the optimum pH levels are in the range of 7-10.3 (Table 5).

Hexavalent chromium is not removed by lime addition unless It has

previously been  reduced to trivalent chromium.

                                 -26-

-------
                                   TABLE 5

       THEORETICAL REMOVAL OF HEAVY METALS DURING LIME PRECIPITATION**1
Cadmium

Hexavalent chromium

Trivalent chromium

Copper

Soluble Iron

Lead

Nickel

Zinc
                                                       Theoretical Effluent
                                    Optimum pH Range  Concentration mg/liter
   10



8.5-9.5

9.0-10.3

    7



   10
1.0
0.01
0.01
                                 -27-

-------
     These studies demonstrated that the aerobic treatment of sanitary



landfill leachate is feasible and that pretreatment may be required.



Lime precipitation appears to be the most favorable pretreatment method.



The organic fraction of leachate was found to contain substances not



readily assimilated by the microorganisms, and it was hypothesized that



chemical treatment is needed to remove these organics.


                                                                     39
     Chian and DeWalle have treated lyslmeter leachate anaerobically.



They used a completely mixed anaerobic filter with recirculation of



effluent.  The unit responded well to shock loads, produced low sludge



yields and operated without nutrient supplementation in spite of a COD:



N:P ratio as high as 4360:112:1.  Heavy metal toxicity was minimized  by



the addition of sodium sulfide.



     Chian and DeWalle have recently completed an extensive review of


                              40
leachate treatment techniques.    Their conclusion was that leachate



collected from recently leaching landfills is best treated biologically.



This is because the organic fraction of such leachate is composed



predominantly of free volatile fatty acids which are readily



biodegradable by either aerobic or anaerobic means.  On the other hand,



leachate from older landfills  is more efficiently handled by



chemical-physical processes, because these organics are more resistant



to biodegradation.  They also concluded that activated carbon and reverse



osmosis were the most efficient chemical-physical methods in terms of the



removal of organ Ics.



     A more recent study by Chian has been devoted to a detailed analysis



of the constituents of the organic fraction of grossly polluted



groundwater and of leachate collected from wells or underdrains near



                                  -28-

-------
                            k2

 solid, waste disposal  sites.    The techniques used for concentrating;




 separating, and characterizing the soluble organics were membrane




 ultrafiltration, gel  permeation chromatography and'analysis for functional



 groups and specific organics.   The free volatile fatty acids constituted




 the largest fraction  and  this  fraction became relatively smaller as  the




 age of the landfill  increased.   Increasing stability with increasing



 landfill  age was noted  for other groups of organics.   These results  tend




 to  confirm the  concept  that biological  treatment is  best suited for



 treating  leachate from  a  young  landfill,  and  that  physical-chemical



 processes are more appropriate  for older  landfills.




     The  compilation  of data presented  by Chian  and  Dewalle indicate




 the following range of  COD removal  efficiencies  for  various treatment




 methods:  0 to 98 percent  for aerobic biological; 87  to  99  percent for



 anaerobic biological; 17  to kO  percent  for  aerobic/anaerobic  biological;



 0 to kO percent  for chemical precipitation  with  alum, ferric  chloride,



 ferrosulfate  or  lime; 3k  to 3k  percent  for  activated carbon and  ion



 exchange;  0  to k8  percent  for chemical  oxidation; 56-98  percent for


                 k3
 reverse osmosis.




     As a  means  to bring order  to  the wide  disagreement  found in the



 literature, Chian and DeWalle postulated  that the age of the  landfill




affected  the character of  the leachate, and that this character is best



measured  in terms of the ratios of chemical oxygen demand to total



organic carbon (COD/TOC) or of biochemical oxygen demand to chemical



oxygen demand (BOO/COD)  (Table 6).




     A recent laboratory scale study by Uloth and Mavinic is closely



related to the present effort.     They studied aerobic biological



                                  -29-

-------
                                           TABLE 6
                  LEACHATE TREATABILITY AS HYPOTHESIZED BY CHIAN AND DeWALLE^7







COD
TOC
>2.8

2.0-2.8

<2.0

Leachate Qual i ty






BOD
COD
>0.5

D.l-0.5

<0.1







Age of
Fill
Young
(<5 yr)
Med i urn
(5 yr-10 yr)
Old
(>10 yr)
Treatment Efficiency"






COD,
mg/1
>10,000

500-10,000

<500
















(D
U
••ซ
01
o
"o
CO
G

F

P

c
o
4->
(D
r~ <4-J
(D —
0 D.
'i '5
0)  U)
tft .—
Rever
Osmos
F

G

G



•o
0>
4J
TO C
> O
— J3
4-> U
U (D
< 0
P

F

G




0)
01
c
ro
.c
C U
O X
— LU
P

F

F

O
 I
    *(G ป good; F = fair; P = poor)

-------
treatment of very strong leachates.  They were able to develop kinetic
constants fitting the Lawrence-McCarty model.    Successful operation was
attained at mixed liquor volatile suspended solids concentrations of
8000 to 16000 ppm at sludge ages in excess of ten days.
                                 Summary
     The state-of-the-art concerning the composition and treatment of
sanitary landfill leachates has been assessed.  The most obvious
characteristics of leachate are its strength and its variability.
Leachate Is generally of much greater strength than domestic sewage.
This Is especially true in terms of organic materials and the potentially
toxic heavy metals.  As important a characteristic as strength is the
variability of leachate composition.  Leachate quality not only
fluctuates from landfill site to site, but also from time to time at
one landfill.  Changes over time result from differences in seasonal
hydrology and microbiological activity.  Rainy weather may dilute the
leachate, but, at the same time, may flush out large quantities of
pollutional  material.  The typical  pattern observed over many years is
that the pollution potential  of leachate is greatest during the first
five years or so after placement,  but that leachate strength remains
significant for as long as ten to twenty years.   This sequence is
encountered because the microbiological processes responsible for the
decomposition of the solid wastes are relatively slow acting and are
first directed at the most readily biodegradable components of the
waste.
     Considerable differences are encountered in leachate quality when
comparing landfills.   In addition to the seasonal,  hydrologlc and age
                                 -31-

-------
 of  landfill  factors mentioned  above,  there are  several other  reasons



 for this observation.  The chemical nature of the wastes accepted at



 the landfill  has  a marked effect on the composition of the  leachate.



 The land disposal of  industrial  liquid and solid wastes  is  critical in



 this light.




      The variability and the strength of  leachate have important waste



 treatment  implications.  First, the sheer magnitude of the  measures of




 pollution  potential dictate the use of thorough waste treatment.  Second,



 the changes encountered from landfill to  landfill are such  that waste




 treatment  techniques applicable at .one site are not necessarily directly



 transferable  to other  locations.  That Is, It may be mandatory that each



 instance be separately engineered to achieve adequate treatment.  Third,



 the fluctuations  in leachate quality which occur over both  short and



 long  term  intervals must be accounted for in the treatment  design.  Not



only must  processes be designed to treat efficiently the waste flow from



minute to minute, but the design must also reconcile the possibility



that  treatment techniques which work well  for a young leachate may



become wholly inadequate as landfill age increases.




      It is apparent today that most landfill  leachate cannot be treated



adequately by just conventional chemical/physical treatment or



conventional biological treatment.   Rather,  what Is  needed  is a




combination of the two approaches with perhaps a supplementary form of



advanced wastewater treatment.   The purpose of this  project is to



Investigate, at both the full  and bench-scale levels of operation, the



efficiency of treatment afforded by these processes.      """""




                                 -32-

-------
                     III.  LEACHATE TREATMENT SYSTEM




     The,, leachate treatment faci 11 ty used fn this study Is located at




the GROWS landfill In Tullytown, Falls Township, Bucks County,




Pennsylvania (see Figure 3).  The plant is designed to provide maximum




operational flexibility in order to permit full-scale testing of a




variety of treatment sequences.  Plant design and treatment modes are




considered in subsequent paragraphs.




     The sanitary landfill has a surface area of 50 acres.  The landfill




will be filled with about 1,400,000 cu rn of refuse over the next several




years.  The time required to fill the landfill depends upon many unknown




factors, but it is estimated that it will probably be between 5 and 10




years.  The receipt of refuse is about 800 tons per day.  Eighty-five




percent of the refuse is from municipal sources.  The remainder is



industrial and commercial.  The landfill is also permitted to accept



sewage sludge and selected industrial liquid wastes.



     The landfill is located in the semi-humid northeastern part of the



United States.  The thirty year monthly average precipitation and



temperature data are given in Table 7.  In this region there is a net



positive infiltration of rainfall into the landfill.  As long as there




is a net positive infiltration, leachate will eventually begin to be



produced by the landfill.           *



     Because of these meteorological conditions and the site hydrologic



situation, groundwater pollution potential existed.  To alleviate this




pollution potential the Pennsylvania Department of Environmental Resources



required the landfill to be underlain by an impervious asphaltic membrane.



This membrane system was designed to collect and transport the leachate




                                  -33-

-------
                           .   LOCATION  OF LEACHATE TREATMENT.  PLANT
                                                                                        15
                 •'     PENNSBURY MANOBJ
                             STATIC PARK!
                ,  ,   .iii?ir<
                  -    ••••ja&iy-J  ^  :
                       •x%v "
ง0"  518 J83COOO FEET (PA.)
                                                                                   74ฐ45'
                                         Heavy-duty	

                                         Medium-duty
   ROAD CLASSIFICATION

             Light-duty
             PENNSYLVANIA
             Unimproved dir' = =

               O State
                                                     - S- Route

                                                           Interstate Route
           QUADRANGLE LOCATION
TRENTON WEST, PA.— N. J.
   NEM BURLINGTON 15' QUADRANGLE
        N 4007.5— W 7445/7.5
                                           -34-

-------
                      TABLE 7

'PRECIPITATION AND AVERAGE MONTHLY TEMPERATURE DATA
                 TRENTON,  NEW JERSEY**8

Month

January
February
March
April
May
June
July
August
September
October
November
December
Total
Rainfall
cm
8.87
6.58
9.75
8.15
9.19
9.14
10.62
12.12
8.89
7.21
8.03
7.29
104.85
in.
3. JO
2.59
3.o4
3.21
3.62
3.60
4.18
4.77
3.50
2.84
3.16
2.87
41.28
Temperature

0.8C
1.0
5.1
11.1
16.9
21.7
24.2
23.3
19.6
13.5
12.7
1.7


33. 4 F
33.8
41.3
52.1
62.7
71.4
76.0
74.3
67.6
56.5
45.1
35.1

                     -35-

-------
 to the leachate treatment plant.
      The treated effluent is discharged  to the Delaware Estuary.   The
 river zone is  tidal  and  flow figures  are not  available.   At  the nearest
 gage  (Trenton)  the  drainage area  is 6700 square miles  and the  projected
 low flow is  33,000  liters per second.  The discharge of treated effluent
 directly to  the Delaware River occurs  only during  the  months of December
 through  April.   During the remainder  of  the year,  the  effluent is
 returned to  the landfill.   During  the  latter  portion of the  third  year,
 a  program of land disposal  was initiated.   The landfill  has ample  storage
 capacity in  the pore space so that storage for 6 months  does not create
 any difficulties.  The effluent is spread  on  the landfill  using aeration
 nozzles.
      The treatment plant  operates under  permits  from the  Commonwealth of
 Pennsylvania Department of  Environmental Resources Water  Quality Section
 and the  Delaware River Basin  Commission.   The  effluent criteria for the
 facility are summarized  in  Table 8.
                              Design Overview
      The  purpose of  this  section Is to summarize briefly  the design
 criteria  and to  discuss the design of the  treatment facility.  The
 effluent  limits  have been mentioned above and  presented  in Table 8.
 The following paragraphs are devoted to a discussion of the leachate
quantity and quality as estimated  for  design purposes.
      Design  Flow.  The source of liquid waste  is the leachate which
 results from the degradation of refuse and percolation of rain  water
 through the  landfill.  In addition, as the treated  effluent is  recycled
to the landfill during the summer  months, and to the Delaware River
                                 -36-

-------
                     TABLE 8
         SUMMARY OF EFFLUENT CRITERIAH3 FOR
GROWS SANITARY LANDFILL LEACHATE TREATMENT FACILITY
                                      Maximum Concentration
Parameter                                   mg/liter
BOD5

Ammonia-N?trogen

Phosphate

Oi 1 and grease

I ron

Zinc

Copper

Cadmium

Lead

Mercury

Chromium
100.0

 35.0

 20.0

 10.0

  7.0

  0.6

  0.2

  0.02

  0.1

  0.01

  0.1
                          -37-

-------
during the winter, the raw leachate volume Includes this recycled
effluent.  The quantity of waste which is generated is dependent upon
many Individual factors of the landfill.  The maximum generation of
waste (Including the recycled volume) for design purposes was estimated
to be about 20 liters per sq m-week at this site.  However, the production
of leachate Is dependent upon the time cycle, both as to placement and to
the season of the year.  Leachate itself occurs as the result of the
excess of infiltration over evapotransplration and the soil moisture
deficit.  Thus, the actual generation of leachate depends upon
precipitation patterns, landfill moisture and effluent recycling.
     Since the generation of leachate is also a function of the age of
the fill, not all the expected leachate will be produced simultaneously.
There Is an initial period of operation when the landfill comes to field
capacity, followed by an extended period of leaching of contaminants,
after which there exists a period when the leachate is no longer of a
polluting nature.  It is possible that some portions of the landfill
will be  in this latter state when the final parts of the landfill are
being completed.  Hence, the maximum flow of 20 liters per sq m-week Is
a value which may never be attained for extremely strong leachate.  This
maximum flow rate was determined using the site meteorological data
presented In Table 7 and the procedure developed by Remson, Fungaroli
and Lawrence.
     Design Leachate Characteristics.  The leachate strength parameters
used for design purposes are presented in Table 9.  These were obtained
through a modest sampling program conducted during the very early stages
                                  -38-

-------
             TABLE 9


 DESIGN LEACHATE CHARACTERISTICS
 Const!tuents
                          Raw
BODg
Suspended Sol ids
Total Solids
Percent Volatile
pH, pH units
Chlorine
Iron, total
Zinc
Chloride
Organic Nitrogen
Nitrate
Sulfate
Copper
Hardness
Alkalinity
Color, standard units
Flow, mod
Temperature, F
1500
1500
3000
55
5.5
200
600
10
800
100
20
300
1
800
1100
50
.iMt
80
''All  units are mg/liter except pH,  color,
 flow,  and temperature.
                  -39-

-------
of the landfill.  However, as discussed In Chapter II, the exact




character of waste is difficult to predict for a number of reasons,




including the fact that it is subject to dilution when the infiltration



is high.  In addition, because of the on-slte variability, it is



possible that single samples do not accurately reflect the character




of the waste.



     Design Concept.  As discussed in Chapter II, a combination of




chemical/physical treatment plus biological treatment is often required




for leachate treatment.  The principle is that the chemical/physical



units can be used for the removal of refractory organics and for



pretreatment prior to the biological process.  In the latter case, the



chemical/physical processes are used for the removal of potentially



inhibitory materials such as heavy metals and ammonia-nitrogen.  The



function of the biological units Is the stabilization of organic matter



and the oxidation of ammonia nitrogen.  As a result of the findings



discussed in Chapter  II and the design leachate quality, this treatment



plant was designed to consist of lime treatment and sedimentation



followed by activated sludge and chlorination.  Equalization, air



stripping of ammonia and nutrient addition have subsequently been added



into the chemical/physical section.  A schematic of the leachate treatment




plant appears as Figure 4.



     Leachate Collection System.  The raw  leachate  is contained within




the lined landfill which was designed to allow for  the collection by



gravity of  leachate at three locations.  These locations are outfitted



with manholes from which the leachate  is pumped and transported via



pressure  lines  to the treatment facility.  The leachate entered the plant




                                 -40-

-------
FIGURE k.   SCHEMATIC FLOW SYSTEM l.WITH AMMONIA STRIPPING LAGOON
           Unit volumes are shown in  cubic meters (l  cu m = 1000
           liters = 264.2 gal  = 35-3  cu ft)
Sludge
Holding .-
21 cu m-*V
i
^x-—
Manhole flleac
""' Pump7
Equalization
Lagoon
950 cu m


 — 23.7 cu m each
75.71 cu m /I Chlorine ,
"* Aeration *J -^-, Contact /" i'b cu m
Chamber / 	 ,JU
	 <*ฃ ซ)i — 	 	 ., 	 	 /.,. 	 . River
/3.7-1 cu m ^i — * 	 	 -•' ป
	 ป Aeration ^ b . ,r , ,
Chamber ' LdndFill
L Sludge '
^_H2S0l| — J
^^"^ II M IV 1
V ft
Ammonia Stripping
Lagoon
950 cu m

-------
via a one thousand gallon holding tank In which little mixing occurred



because the flow from the individual manholes is highly variable, and



pumped sequentially.  Following start-up of the equalization pond in



the third year, the holding tank was by-passed.



                         Chemical/Physical Section



     The chemical/physical portion of the plant consists of the



following: equalization, chemical precipitation and coagulation,



sedimentation of precipitate, air stripping at elevated pH for ammonia



removal, neutralization and nutrient supplementation.  Each of these are



discussed in the following paragraphs.



     An equalization pond was put on-line during the final project year.



The volume is 950 cu m, and so the design detention time  is 1.7^ days.



This pond is aerated to maintain aerobjosis and prevent solids deposition.



A chlorinated polyethylene liner is used as the inner wall of the pond.



     Chemical Precipitation.   In the chemical treatment  phase,  the major



design goal was the removal of inorganic materials.  In particular, metals



that may  interfere with the subsequent biological treatment process are



removed; also, the metals are removed to achieve discharge standards



(Table 8).  As part of the chemical treatment, the biochemical oxygen



demand will also be reduced, and the design percentage of reduction was



30 to 50 percent, on the  basis of the experience with municipal



wastewater and with leachate as discussed  In  Chapter II.



     The chemical treatment step consists of  flash mixing followed by



quiescent conditions favorable for  coagulation as well as sedimentation



of the chemical sludge.   Lime has been the only chemical  utilized  in



                                 -42-

-------
this chemical precipitation step.  However, additional feeders and points
of injection have been provided for the use of other chemicals if
necessary.  Other chemicals which might be used include alum, ferric
chloride, synthetic polymers, and powdered activated carbon.
     This unit is an upfldw solids contact reactor clarifier.  Lime
slurry is added to cause coagulation and precipitation of the waste
materials.  The time is pumped at a rate commensurate with the rate of
leachate production.  The lime slurry is flash mixed with the Incoming
waste, and mixing, flocculation and upflow clarification occur within
a single unit.  Solids contact may be optimized by variable sludge
recycle.   The chemical treatment facIIIty Is a 3.66 m diameter, 3.66 m
deep cylinder with a hydraulic retention time of 1.7 hours at 380
liter/min flow rate.
     Sludge Is drawn off the bottom of the reactor clarifier and placed
in a common sludge holding tank with the waste activated sludge.  Sludge
return pumps are available to recirculate the sludge and mix it with the
incoming waste water to reduce the amount of chemicals that are needed
for precipitation.  (However, the practice to date has been to use fresh
lime and to not recircUlate the sludge,).  The amount of sludge that
is produced in this step depends upon the composition of the leachate.
The design projection was that approximately 5 percent of the flow will
be produced as sludge at 1 percent solids concentration.
     Air Stripping of Ammonia.  As a means of controlling excessive
levels of ammonia In the lime treated stream and in the final effluent,
a lagoon incorporating air stripping of ammonia is included in the
chemical/precipitation-clarification unit In order to take advantage of
                                 -43-

-------
the high pH of the .upflow solids contact reactor clarifier effluent and



to minimize the solids loading on the lagoon.  On occasion, sodium



hydroxide has been added to the lagoon to further elevate the pH and



force off additional ammonia.  This has been done during cold weather



periods.



     The volume of the ammonia lagoon Is 950 cu m, thus providing a



detention time of approximately 1.7^ days at design flow.  The primary



function of the lagoon is to encourage air stripping of ammonia by an



elevated Influent pH of 10,.aeration and high internal recycle.  The



lagoon is lined with chlorinated polyethylene.  In addition to ammonia



removal, the-lagoon provides equalization in terms of both flow and



sanitary parameters.



     Neutralization and Nutrient Supplementation.  Sulfuric and



phosphoric acids are added to reduce the pH of the leachate prior to



entering the biological waste treatment portion of the process.



Phosphoric acid replenishes the supply of o-phosphate, a necessary



biological nutrient, which Is precipitated and removed following the



addition of 1Ime.



                       Biological Treatment Section



     The biological treatment units consist of two aeration tanks and two



secondary clarifiers.  The units may be operated in series or parallel.



The capacity of each tank is 75,710 liter, which corresponds to a 6,6



hour detention time at the maximum flow rate of 380 liter/mln.  The



aeration chambers are provided with diffused aerators, each driven by



a 1^.2 cu m per min blower.

-------
     Depending on the actual hydraulic residence time In the aeration




tanks, the activated sludge units were designed to operate in either the




conventional or extended aeration modes.  In order to achieve this, the




mixed liquor volatile suspended solids (MLVSS) would be maintained In



the range of 3000 to 8000 mg/liter.  Because of the high organjcs loadings




experienced during the last 2 years, the MLVSS was in the 8000 to 16,000




mg per liter range.  This level Is high relative to that normally




maintained in units handlIng municipal wastewater.  However, It is




necessary because of the high BOD loading, and because of the requirement




to remove about 90 percent of the BOD remaining after chemical/physical



treatment.  The MLVSS is maintained by return sludge pumps capable of




delivering a return sludge flow equal to 200 percent of the Influent




flow.



     The waste sludge from the activated sludge units and from the




chemical treatment process is stored in the sludge holding tank.  The



capacity of this tank is 21 cu m, and sludge Is removed as required



and conveyed back to the landfill via tank truck.



     Separation of treated wastewater from the MLVSS is achieved by



gravity sedimentation in the secondary clarifiers.  The total clarifier




volume is 47,318 liters,  in two parallel independently SperSble units.



Sludge return is provided with air lifts Installed In the final settling



tanks.  A skimming device is located fn the settling basin in front of




the scum baffle to remove floating material  which is returned to the




aeration compartment.  The maximum surface overflow rate is 20.k cu m



per day per sq m (500 gpd/sq ft) based on the peak flow of 380 liter/



min.

-------
     Final effluent Is directed to the chlorine contact tank after



secondary clarification.  The chlorine contact tank provides a retention



time of 20 minutes at the 380 liter/min flow rate.  The effluent after



chlorinatlon is discharged to the Delaware River or to the land disposal



site depending upon the season of the year.  The effluent is not



chlorinated when it is recycled to the landfill.  The chlorine contact



tank is a simple baffled tank to assure mixing of the chlorine which is



provided by hypochlorlnatlon.
                                 -46-

-------
                         IV.  MATERIALS AND METHODS



                           Experimental Systems



     The leachate treatment plant', although designed for chemical /physical



treatment followed by biological treatment, was equipped with sufficient



flexibility to provide for operational evaluation of a variety of treatment



sequences.  These sequences are each defined  in the following paragraphs



with reference to Figure 4.  Note that the final effluent  is disinfected



with sodium hypochlorite prior to discharqe during the winter months.



The periods of operation of each system is summarized in Table 10.  The



principal analysis was conducted during the first two years.  Additional



trials were conducted during the third year.  The flow rate and raw



leachate qualIty was sufficiently different In the third year to



oftentimes prevent direct comparisons,



     System 1 - Chemical/Physical followed by Biological Treatment.



System 1 is the basic treatment sequence with lime treatment for metals



removal followed by ammonia stripping and conventional activated sludge.



System la refers to the use of System 1 when the air stripping was not



used, whereas System Ib signifies that the lagoon was included in the



'flow sequence.  System la was tested in the late winter of 1975-76,



and System Ib in the summer of 1976, the winter of 1976-77, the early



spring of 1977, and the spring and summer of 1978.  During the 1978 test



period, the raw leachate entered the plant via the equalization pond.



     System 2 - Chemical/Physical Treatment.  Two subsystems have been



evaluated.  These, Systems 2a and 2b, consist of lime treatment either



without or with subsequent removal of ammonia by air stripping.  The




                                 -47-

-------
                                  TABLE 10




             PERIODS OF OPERATION OF LEACHATE TREATMENT  SYSTEMS
System



  la*
  Ibt



  2a*



  2b+
   vV
 Period of Operation of Full-Scale Units



 November 15,  1975 through January 12, 1976



 June  14, 1976 through April 30, 1977



 January 1, 1978 through August 31, 1978



 November 15, 1975 through January 12, 1976



 June  14, 1976 through April 30, 1977



 January 1,  1978 through August 31, 1978



 January 12, 1976 through April 2,  1976



May 1, 1977 through August 31, 1977



January 12, 1976 through April 2,  1976



May 1, 1977 through August 31, 1977
Neither ammonia lagoon nor equalization pond used.




Ammonia lagoon used.





Both ammonia lagoon and equalization pond used.
                                -48-

-------
system without ammonia stripping (System 2a) was evaluated in the winter



of 1975-1976; and System 2b in the summer of 1976, the winter of




1976-1977, early spring of 1977, and the spring and summer of 1978.



The equalization pond was included in this treatment sequence during




the most recent,period.




     System 3 '- Biological followed by Chertical/Physical Treatment.




This is the reversal of System 1.  This system was studied during the




winter of 1976.  The results indicated poor treatment efficiency, most




likely due to heavy metal and ammonia toxicity.  However, it might have




been argued that a sufficient amount of activated sludge had not




developed.  Therefore, System 3 was reevaluated in order to test this




latter hypothesis.  This took place during the late spring and early




summer of 1977.



     System k - Biological Treatment.  This system has been tested, the



results showing poor treatment efficiency.  However, as indicated above,



the performance might have improved if a previously acclimated activated




sludge were available.  Consequently, System k was operated and tested



simultaneously with System 3 during the spring and early summer of 1977.



     System 5 - Bench-Scale Testing.  As a supplement to the full-scale



treatment processing, some smaller scale work has been undertaken.  This



effort serves two purposes.  First, it allows the operator to readily



develop operational guidelines.  For example, jar tests have been used




to determine proper chemical dosages.



     The second purpose is to provide an opportunity for evaluating



additional treatment techniques.  Specifically, bench-scale testing has




been used to evaluate activated carbon treatment of raw leachate,



                                 -49-

-------
 Granular  activated  carbon  has  been  used  fn  column  studies  to obtain


 performance  characteristics.   The  results are  discussed  In  Chapter  IV,


                               P rbces s  Monitor]ng


     An analytical  laboratory  was established  in a  trailer  located


 Immediately  adjacent  to  the  treatment  plant.   The  trailer  Is outfitted


 with the  apparatus  to perform  the analyses  indicated  below  and  Is


 environmentally controlled with a heating/air  conditioning  system.


 The need  for extensive bench-scale  testing  and  the  large number of
              i

 analyses  needed for process  control and monitoring  made the on-site


 laboratory mandatory.  The laboratory  Is operated by  the chemist-operator


 employed  specifically for this project.


     The  chemical analyses performed routinely  are  presented in Table 11.


 These have been selected on  the basis of four criteria:  they represent


 the most  common chemical parameters used in the literature  to characterize


 landfill  leachate;  they provide sufficient data to  completely evaluate the


 unit operations in  terms of  process and total system efficiency; they are


 needed for process  control; and they are required to specifically define


 the leachate.


     All  analyses are  performed In accordance with  the 13th edition


of Standard Methods, ASTM Standards pt-23,  and the  1974 edition


of EPA Methods.       The analyses are performed on total samples as


opposed to filtrate samples.   Some preparation of the raw leachate is


 required  for heavy metals determinations.


     Electrometric  techniques are used in the determination of ammonia-N,


dissolved oxygen (with periodic checks using the Azlde Modification of


                                  -50-

-------
                                TABLE 11




                 ROUTINE LABORATORY CHEMICAL ANALYSIS



Item
pH
Chemical oxygen demand
Dissolved oxygen
Mixed liquor suspended solids
Mixed liquor settleable solids
Dissolved sol ids
Volatile suspended solids
Total residue

Alkalinity
Biochemical oxygen demand
Total hardness
Kjeldahl nitrogen
Ammonia nitrogen

Phosphate
Sulfate
Chloride
Total iron
Chromium
Copper
Cadmium
Lead
Mercury
Zinc
Nickel
Ca 1 c i urn
Magnesium
Sod i uin
Potassium

Daily


Method

Dichromate reflux
Electrode
Gooch crucible
Imhoff cbne
PotentiometrJc
Gooch crucible
Drying crucible
Weekly
Titrimetric (pH 4.5)
Probe method
Titrimetric
Titrimetric
Distillation S
Potentiometric
Persulfate digestion
Gravimetric
Titrimetric
AA*
AA
AA
AA
AA
Mercury analyzer
AA
AA
AA
AA
AA
AA
Aperiodic

EPA
Storet
No.

00349
00299
70300
50086
00536
00529
00520

00410
00310
00900
00625

00610
00665
009*15
00940
—

—
--
—
—
--
—
—
—
--
—


Detec-
tion
Limit

—
—
--
-- .
—
—
™~*

_..
—
ซ


--
—
—
—
0.02
0.02
0.01
0.002
0.05
0.0002
0.005
0.005
0.003
0.005
0.002
0.005

Oil & Grease
                                 Freon extraction
•Mtomic Absorption Spectroscopy
                              -51-

-------
 the Winkler lodometric procedure), pH, and dissolved solids.   Atomic



 absorption spectroscopy is used for iron, chromium, copper,  nickel,  zinc,



 sodium,  cadmium,  lead, and potassium.



      A number of  sampling points are used in the analysis program.



 Routinely, samples are collected of (l)  the raw leachate; (2)  chemical/



 physical  sedimentation tank effluent;  (3) ammonia lagoon  effluent;



 (4)  final  effluent.   In addition,  samples are collected on an  irregular



 basis from the three landfill  manholes and directly from  the  individual



 treatment  units.   In all  cases,  every  effort is made to ensure that  a



 representative sample is  obtained.



      A pilot-scale treatment plant was constructed  in order to facilitate



 the  smaller  scale studies.   This pilot plant was built durinq  the spring



 of 1977-   The  basis  of its  design  was  to  simulate System  1 operation as



 it was apparent that this was  the  most efficient treatment sequence  for



 this  leachate.  Thus,  the pilot  plant  would  generate final effluent which



 could  be used  in  the bench-scale experiments  using  granular activated



 carbon.  A schematic of the  pilot  plant  is presented  in Figure  5,



                          Statistical  Tests



     The following notation  is used  throughout:  n,  number of data points;



x, arithmetic mean;  and s, standard  deviation.   The mean  is calculated as



                               x = 1 Vx.
                                   n •<—'  i



and the standard deviation as



                               s -   E(x-x;)2

                                 ~      n-1



where the x. are the n data points,



and the coefficient of variation is



                                  -52-

-------
                FIGURE  5.   SCHEMATIC  OF PILOT LEACHATE TREATMENT PLANT (SCALED VERSION OF SYSTEM  l)
 i
Ul
OJ
 I
                  Mixing
                   Tank
                   ?   i-
0_ - -100 ml/min
                      Clari-
                      fier
                      (10 qt
   1
Sludge
          Lime Slurry  Raw Leachate
            Storage      Storage
            (50 gal)     (50 gal)
                                      Lagoon
                                          gal)
Aeration
  Tank
(45 gal)
                                                        neutralization
                                Acid  Storage
                                  (10 qt)
                                                                                     Return  Sludge
(40 gal)

(35 gal)
(30 gal)
Clarifier
 (10 gal)

-------
The value of the coefficient of variation decreases with decreasing



variability.



                         Presentation of Results



     A note of caution is placed here for the reader.  As would be



expected in a three year study of this magnitude, a vast array of data



has been collected.  It is not appropriate to present the entire data



set in this report.  Consequently, many of the tables presented represent



averages taken over certain specified time periods.  Thus, values given



In different tables are not necessarily comparable.  This is especially



true If the time periods are not identical.  The reader is asked to pay



particular attention to the time periods and to avoid making comparisons



when these periods differ.
                                  -54-

-------
                        V.  RESULTS AND DISCUSSION




                            Preliminary, Results




     Raw Leachate Quality.  A summary of actual leachate quality is shown




In Table 12.  These data are a summary of the entire set of results.  As




is evident from a comparison of Tables 9 and 12, there are significant




differences between the two.  These changes toward Increased leachate




strength are seen mainly between the design and actual raw leachate




organic matter, dissolved solids, pH and ammonia.  The biodegradable




organics concentration is three times the design level.  Dissolved




solids are an order of magnitude greater, caused by increased hardness,




organic matter and chloride.  The ammonia concentrations actually



observed have been extremely high and have been a source of operating




problems especially in the biological units.  The factors influencing



this difference between the projected and observed leachate quality



have been discussed in Chapter II.




     Considerable variability in the raw leachate quality has been noted



on a day-to-day basis.  The influent COD data are presented in Figure 6



to show this variability.  (An additional indication is provided by the



coefficient of variation data provided In Table 13, columns 1 and 2.)




It was felt that the raw leachate variability had an adverse Impact on




the efficiency of the subsequent treatment units.  This was because it



was impossible to fine tune the operational  controls as quickly as the




influent quality changed.  This was especially true for the lime dosage



in the heavy metals removal  step.  For this reason, the equalization



pond was constructed.   The effect of the pond on dampening of the




                                 -55-

-------
                   FIGURE 6.  RAW LEACHATE CHEMICAL OXYGEN DEMAND

                              (Note change In scale on ordinate)
  !_

  <0
 o
 o
 o
ff\
 o
 X

 J_

 
-------
                                             TABLE 12
                                 LANDFILL LEACHATE CHARACTERISTICS*
Item
Biochemical oxygen demand
Chemical oxygen demand
Suspended sol ids
Dissolved sol ids
pH, pH units
Alkalinity, as CaCO,
Hardness, as CaCO_
Calcium . ^
Magnesium
Phosphate
,Ammonia-N
Kjeldahl-N
Sulfate
Chloride
Sod i urn
Potassium
Cadmium
Chromium
Copper
Iron
Nickel
Lead
Zinc
Mercury
11/15/75-9/1/76
(5-day) 4,460
11,210
1,994
11,190
7.06
5,685
5,116
651
652
2.81
1,966
1 ,660
114
4,816
1,177
969,
0.043
0.158
0.441
2^5
.531
.524
8.70
.0074
Concentration*
9/1/76-9/1/77
13,000
20,032
549
14,154
6.61
5,620
4,986
894
454
2.61
724
760
683
4,395
1,386
950
0.09
0.43
0.39
378
1.98
0.81
31
.0051
9/1/77-8/31/78
11,359
21,836
1,730
13,181
7.31
4,830
3,135
725
250
2.98
883
611
428
3,101
1,457
968
0.10
0.22
0.32
176
1.27
0.45
11.0
0.012
11/15/75-8/31/78
10,907
18,553
1,044
13,029
6.85
v • v j
5,404
4,652
818
453
~ *> J
2 74
mm + ง —f
1 ,001
984
^^* i
462
4,240
1,354
961
0.086
0.28
0.39
312
1.55
0.67
21
0.007
 These values represent the arithmetic mean of raw leachate- data collected during the indicated  intervals.
+A11 units mg/Uter except pH.

-------
                       TABLE 13




EFFECT OF EQUALIZATION POND ON RAW LEACHATE VARIABILITY

Coefficients of Variation

Unequal ized
Raw Leachate
Equal ized
Raw Leachate
Project Year

Alkal inity, as CaCO_
Ammonia-N
BOD
Cadmium
Calcium
Chloride
Chromium
COD
Copper
Dissolved Sol ids
Hardness, as CaCO,
Iron
Kjeldahl-N
Lead
Magnesium
Mercury
Nickel
PH
Phosphates
Potassium
Sodium
Sqlfate
Suspended Sol ids
Zinc
First Second
0.23 0.21
1.14 0.21
0.60 0.72
0.39 0.80
0.31 0.57
0.58 0.16
0.56 0.32
0.40 0.66
0.81 1.00
>0.24 0.58
1.02 0.48
0.65 0.77
1.23 0.19
0.60 0.68
0.47 0.32
0.90 1.04
0.33 0.97
0.64 2.33
0.63 0.74
0.32 0.21
0.32 0.31
0.90 2.10
0.65 1.32
0.65 0.62
Third
0.26
0.33
0.50
0.73
0.57
0.30
0.49
0.25
0.67
0.16
0.41
0.33
0.39
0.53
0.29
1.73
0.52
--
0.84
0.19
0.16
0.71
0.64
0.68
                      -58-

-------
Influent quality Is Impossible to ascertain because raw leachate samples

were not collected after pond startup•, •  Rather, the Influent samples were

taken from the pond.  As an approximate Indicator of the degree of

attained equalization, Table 13 was constructed.  This compares the

coefficients of variation for each parameter during the three project

years.  As shown In Table 13, the variability decreased for one-half

of the parameters and was approximately the same for several others.

This is taken as indirect evidence of the efficiency of equalization.

     Lime Dosage.  Jar tests were carried out In the laboratory in order

to determine proper dosages for the 1Jme treatment unit.  In the first

series of tests, three types of 1 ime were monitored for their ability

to raise the pH of raw leachate to 10,0,  The 1Jmes used were high

magnesium lime, high calcium quick 1|me  and high calcium hydrated lime.

The results may be summarized as:
                                                   Dosage
                                            lb/1000 gal    kg/cu m
          High Magnesium Lime
          High Calcium Quick Lime
          High Calcium Hydrated Lime
125
 52
 50
15
 6.2
 6.0
It is economically impractical to use the high magnesium because its

properties are such that to raise the pH to 10.5 requires 30 kg per

cu m (250 lb per 1000 gal).

     Required dosages to obtain pH 10,0 are nearly identical for both

types of high calcium lime.  For pH greater than 10,0, the high calcium

quick lime becomes more efficient and hence (s desirable economically.

However, the slaking characteristics of the quick lime have caused

problems with pumping the resultant slurry so that this lime cannot be

                                  -59-

-------
used with the available lime feed system.  On the other hand, the hydrated



lime does not offer such problems, and consequently, the high calcium




hydrated lime is being used.



     Sulfuric Acid Dosage.  The amount of sulfuric acid required to lower



the clarifier effluent pH to 6.5 has been determined.  To do this,



approximately 0.6 ml of concentrated sulfuric acid per liter leachate




(0.6 gal/1000 gal)  is required.  The actual dosages used are presented




later in this chapter as part of treatment costs.



     Phosphoric Acid Dosage.  The heed for a phosphoric acid supplement




became apparent from three  lines of evidence:  (a) very low phosphate



levels in the chemical/physical effluent;  (b) unrealistically low values



obtained In the biochemical oxygen demand test; and  (c) poor biological



treatment performance following the chemical/physical process.  These



points all Indicated that the chemical/physical treatment effluent is



phosphorus deficient, and that, if biological treatment is to follow,  it



must be supplemented with phosphorus.  Additional evidence was collected



by performing a series of BOD,, tests  in which a variable amount of



phosphorus supplement was added to the bottles.   It was observed that



the BOD increased with the  amount of phosphorus.  In addition, bench



scale tests Indicated greater activated sludge production when o-phosphate



was added.  Thus,  it has been concluded that orthophosphate, as



phosphoric acid,  should be  added to form a nutrient  supplement.




     Additional experiments were performed to address the phosphate



Issue.  These were  conducted using the pilot scale unit shown in Figure



5.  The unit was  operated with and without an addition of phosphorus.



                                  -60-

-------
 Raw leachate COD  was  4528 mg per liter during  the no addition  run,  and
 4823 and  21406 mg per liter during  the two runs  receiving  a  phosphate
 supplement.   The  lime clarifier  effluent  COD values  for  these  runs  were'
 3926,  4062 and 18121,  respectively,  and the average  fractions  of  COD
 removed during biological  treatment  were  0.711,  0.811, and 0.762,
 respectively.   These  results indicate  that the removal efficiency of
 acclimated activated  sludge can  be  improved by the addition  of phosphorus,
 and  that  the removals  were  acceptable  when the effluent  phosphate
 concentration  is  1  to  2 mg  per liter as P.
     The  preliminary  calculation of  phosphoric acid  dosage has been made
 on  the basis of providing a ratio of BQD:N;P of  100:5:1.  This is
 approximately  4-6  liter  (1-1.5. gal)  phosphoric acid  per  day.   More
 recently, however,  the criterion is  to add phosphoric acid so  that
 there  is  measurable o-phosphate  in the bio-unit  effluent.  This amounts
 to about  3.8 liter  (l  gal)  of phosphoric acid per day.
           Sy s tern  1 -  Phys I ca 1 /C hem i ca 1 Plus Act i va ted S1 udge
     System  1  consists of chemical/physical treatment followed by
 activated sludge.  The early attempts  (winter and spring 1976) to develop
 an activated sludge culture were not successful.   As discussed  In
 connection with Systems 3 and 4,  phosphorus limitation and ammonia
 toxicity  inhibited these efforts.  These two difficulties were overcome
 by the addition of phosphoric acid as a neutralizing agent for the lime
 treatment effluent and by the use of air stripping of ammonia.   System
 1  was successful  only after the  Implementation of these measures and,
as a result,  this discussion Is limited to the time period after
 implementation.
                                 -61-

-------
       The BOD, COD, and ammonia-N data showed a dramatic improvement  in



  treatment efficiency during and after August, 1976.   Approximately four



  weeks had been needed to develop the activated sludge microorganisms  to



  the point where they were capable of rapid growth at the expense  of the



  leachate substrate.  A similar time sequence was observed in  1978



  during a re-test of System 1.   Table 14 shows the results following the



  successful adaptations of the  activated sludge.   The starting dates for



  analysis of these data were chosen at the points.at  which the activated



  sludge had become fully acclimated In terms of ammonia-N, BOD and COD



  removals.  The time periods are August 1, 1976 to May 1, 1977 and July



  1,  1978 to August 31, 1978,



       The results presented in  Table H demonstrate the high  level  of



  treatment efficiency attainable with System 1,   This treatment system



  achieved removals during the 1976-77 trial of approximately eighty-nine



  percent or more for ammonia, BOD, COD, suspended solids, and  iron; and



  greater than two-thirds for alkalinity, hardness,  kjeldahl nitrogen,



  copper, chromium, magnesium, cadmium, lead, and  zinc.   Relatively poor



  removals of mercury were achieved with System #1,  The fairly low removals



•  of  nickel may be related to the relatively low hydrogen ion concentrations.



  Based on theoretical  considerations of the solubility of nickel hydroxide,



  the nickel concentration in the effluent  should  be on the order of 0,01


      35
  mg/1   , as opposed to observed average of 0,75 mg/1.   This observation



  Is  perhaps also due to the formation of nickel complexes with unknown



  chelating agents within the landfill.  Thus,  we  are  able to see removals



  as  a function of clarifier pH, and this opens the  possibility that the



                                   -62-

-------
                                                            TABLE 14

                              SYSTEM 1* TREATMENT PERFORMANCE AFTER  ACCLIMATION OF ACTIVATED SLUDGE
                               (August 1, 1976 - May 1,  1977  and July 1,  1978  - August  31,  1978)
 i
ON
(JO

August 1 ,
1976 - May 1,
1977
Concentration

Parameter
Suspended solids
Dissolved sol ids
COD
BOD
Alkal inity
Hardness
Magnesium
Calcium
Chloride
Sulfate
Phosphate
Ammonia-N
Kjeldahl-N
Sodium
Potassium
Cadmium
Chromi-uir;
Copper
Iron
Nickel
Lead
Zinc
Mercury
Flow, gpd
*ft
Influent
686
13563
18M8
12468
51(79
5331
499
929
1(264
645
2.15
70S
748
1310
906
0.08
0.23
0.44
376
1.91
0.82
22
0.006
21034

Effluent**
101
5693
939
118
685
1314
107
347
2592
951
13.7
80
102
821
524
0.01
0.07
0.10
3.0
0.76
0.12
0.57
.004

Percentage
Removal
97.4
58.0
94.9 •
99.1
87.5
75.4
78.6
62.6
39.2
—
—
88.7
86.4
37.3
42.2
87.5
75.0
77-3
99.2
60.2
85.4
97.4
28.9

July 1,
1978 - August
31, '978
Concentration
;VA
Influent
1655
13091
18505
8143
5262
2504
275
653
8578
178
1.39
1076
—
1248
872
0.06
0.16
0.'20
9.7
0.88
0.30
3.38
.0.003
10,010
**
Effluent
478
7244
1008
464
1496
1456
105
113
2254
836
17.2
6.3
—
1145
743
0.04
0.04
0.16
0.71
0.67
0.11
0.16
0.002

Percentage
Removal
71.1
44.7
94.6
94.3
71.6
41.9
62.0
82.7
73.7
—
—
99.4
—
8.3
14.8
33-3
75.0
25.0
99.3
24.1
64.6
95.3
33.3


* 1_. . C I • JJ. • J. .... ...
             This system consists of  lime addition, sedimentation, air stripping, neutralization, nutrient supplementation
             and activated  sludge.
            **mg/l

-------
 operator can control pH as a method of differentially affecting
 effluent heavy metals concentration.
      The results observed with phosphates and sulfates should be noted.
 The concentration of phosphates and sulfates Increase during the course
 of treatment because of the addition of sulfuric and phosphoric acids
 as neutralizing agents.   Initially, both acids were used In excess in
 order to encourage the growth of the activated sludge microorganisms.
 That is, the goal was to provide a very favorable environment in terms
 of both pH and the nutrient phosphorus,-  However, following the
 successful  acclimation of the activated sludge,  the addition of sulfuric
 acid practically stopped while that of the phosphoric acid  was
 drastically cut back.  Neutralization  was  not needed because of the
 recarbonation  effect  of  aeration  in the lagoon.   The final  criterion
 for phosphoric  acid addition  was  to provide  just  enough  to  satisfy
 the microorganisms' demand  as indicated by an effluent concentration  of
 about  1  mg/Uter.  That  is, the criterion  was to  add  enough  H-PO,  so
 that there  is  residual phosphorus  (l mg/llter)  in  the effluent.  This
 level  is one to  two orders of magnitude greater than  the amount  In
 the  lime clarifier effluent.
     The difficulties in obtaining  a healthy  culture of activated sludge
were overcome.  The operating experience indicated that the earlier
problems were in fact due to ammonia toxicity and phosphorus limitation.
The ammonia stripping lagoon maintained the concentration of this
Inhibitor below toxic levels.   The mean and standard deviation of the
lagoon effluent ammonia concentration were such that 95 percent of the
                                 -64-

-------
time, the feed to the activated sludge unit was less than 423 mg




NH,-N/liter.  The corresponding raw leachate concentration was 1072




mg NH--N/1iter.  (Similar results were obtained during the other test



period.  Pooling all the data, the -1  standard deviation interval  for



the lagoon effluent was 203 to 6k] mg per liter, whereas, for the raw




leachate, the mean was 1,001 mg NH,"N per liter and the standard



deviation, 1049 mg per liter.)   Thus, the lagoon functioned to minimize




the shock loading effect of inhibitory ammonia concentrations.  This in




turn provided an opportunity for the development of microorganisms




capable of extracting carbonaceous BOD,  As this group became established,




organic concentrations in the mixed liquor were reduced and this created




conditions suitable for the development of nitrifying organisms.  Growth




of these groups of microorganisms has resulted in the low effluent




concentrations of both BOD and ammonia.



     As seen in Table 14, a considerable change in alkalinity occurs



during biological treatment.  There are two main mechanisms by which




this occurs.  First, the aeration causes some removal of gaseous



carbon dioxide, resulting in a shift of the carbonate equilibria and a



change in total bicarbonate alkalinity,  |t is probable, however, that



in this case, nitrification has a more profound errect on alkalinity.



As a result of nitrification, alkalinity is consumed and carbon dioxide




is produced.  Neglecting the effect of bipmass synthesis, the



theoretical value is 7.14 mg alkalinity as CaCO, destroyed per mg



NH.-N oxidized.  In this study a ratio of 4,46 mg alkalinity per mg




NHi-N removed was observed after the development of the activated sludge




culture.  This is in excellent agreement with the theoretical value if



                                 -65-

-------
one considers  that  the observed value  Includes  the  effects of  bfomass



growth and ammonia  stripping  in the  bio-units as well as  shifting



chemical equilibria  in addition to those of  nitrification.  These



conclusions are drawn from  the data  presented in Tables 15 (Column 7)



and 20 (Column 12) which  represent activated sludge effluent and



influent values,  respectively.



     System 1 was re-tested from January 1,  1978 through  August 31, 1978,



following completion of the equalization pond.  Cold weather,  the extreme



strength of the process influent, and  the high  flows Inhibited



development of an activated sludge until April, 1978.  The sludge did not



become acclimated until July  1, 1978,  following a process change from



parallel  to series deployment of the aeration tank-secondary clarifler



system.  A schematic ?s shown in Figure 7.  The change was undertaken



in an effort to gain greater use of  the available aeration capacity.



This was deemed necessary because of the tremendous process loadings



observed in 1978  (Table 35).



     The acclimated activated sludge performed well.  The average flow



into the biological- units was 10,000 gpd, whereas, during the same



period, the raw leachate flow was 50,000 gpd.  The difference  Is due



to the extreme strength of the leachate which inhibited the activated



sludge so that only a portion of the flow could be treated biologically,



(The entire flow was treated in the chemical/physical section of the



plant during this period.)



     The  data for the re-test of System 1 are presented in Table ]k,



and a summary of both tests appears  in Table 15,  The results Indicate




                                 -66-

-------
                         FIGURE 7.   FLOW CHART FOR ACTIVATED  SLUDGE  IN  SERIES
I
ON
           AMMONIA
           LAGOON
ACTIVATED
  SLUDGE
  UNIT #1
CLARIFIER
    #1
ACTIVATED
  SLUDGE
  UNIT #2
ฃLARIplER
    #2   X-    ,
         EFFLUENT
                                                       -I

                                                      SLUDGE
                                  I	
                                                                                               SLUDGE

-------
oo
 i
                                                              TABLE 15

                                                    SUMMARY OF SYSTEM  1  OPERATION

                                             (8/1/76 to 4/30/77 and 4/1/78  to  8/31/78)

Alkalinity, as CaCO.
Ammonia-N
BOD
Cadmium
Ca 1 c i urn
Chloride
Chromium
COD
Copper
Dissolved Sol ids
Hardness, as CaCO-
Iron *
Kjeldahl-N
Lead
Magnesium
Mercury (ppb)
Nickel
PH
Phosphate
Potassium
Sod i urn
Sulfate
Suspended Solids
Zinc

51*50
758
11886
0
888
4161
0
18490
0
13516
5054
333
748
0
465
5
1
6
2
903
1301
577
555
19
.*
X



.078


.26

.40




.74

.52
.76
.74
.05




.5
Raw
cv
0.26
0.32
0.78
0.91
0.57
0.23
0.77
0.71
1.06
0.63
0.49
0.85
0.25
0.80
0.34
1.15
1.13
1.32
0.63
0.20
0.29
2.77
1.46
1.14
n
53
56
52
54
54
53
54
172
53
171
51
52
44
54
54
48
55
197
53
47
53
48
167
52
X
1878
350
3930
0.025
293
2616
0.07
6892
0.31
5995
1211
3.24
294
0.17
107
3.95
0.61
8.60
0.12
514
785
363
193
0.63
Lagoon
cv
0.43
0.05
0.63
1.37
0.61
0.08
1.19
0.57
1.33
0.30
0.54
1.00
0.24
0.74
0.45
1.29
1.18
1.09
3.89
0.22
0.38
0.68
1.13
1.69
n
44
53
46
45
44
43
44
165
48
175
38
42
35
44
43
38
43
181
43
38
43
41
165
44
X
803
75
153
0.017
314
2544
0.07
945
0.11
5824
1327
2.71
102
0.12
107
4.01
0.75
7.61
14.1
535
862
937
133
0.53
Effluent
cv
1.02
1.01
1.75
1.64
0.52
0.33
0.59
0.80
0.44
0.25
0.55
1.04
1.46
0.65
0.36
2.70
0.78
__
0.50
0.15
0.39
0.76
1.42
1.15
n
48
74
69
55
57
57
66
223
56
213
54
70
49
69
57
61
56
237
66
50
56
55
209
70
Percent
Removal
85.2
90.1
98.7
78.2
64.6
38.9
73.1
94.9
72.5
56.9
73.7
99.2
86.4
83.8
72.0
27. 4
57.4

__
40.8
33.7

76-. 0
97.3
              vt-
               x
                   mean,  mg  per  1iter
               cv = coefficient  of variation
               n  = number of data  points

-------
 that  System  1  Is  able  to  handle  large  loading  rates,  and  to  achieve



 high  removal efficiencies.   Data  collected during  the warm weather



 months  (August  1,  1976 to December  1,  J976 and July  1,  1978  to August



 31, 1978) have  been analyzed separately.-  These data  show that System  1



 can produce an  effluent of sufficient  quality to meet the final effluent



 criteria  (Table 16),



      Approximately equivalent results  are obtained using  the biological



 reactors  in either the parallei  (1976"77) or the series (1978) mode



 (Table  17).  In terms  of  percent  removal efficiency,  the  series and



 parallel operations provided approximately equal performance with



 respect to BOD  and COD.   Removals of ammonia-N were much  greater with



•the series operation.  This  is due  to  the localization of NhV-N removal



 in the second tank following BOD oxidation fn the first,  (Note that the



 average flow during the series test was 10010 gpd, whereas during the



 parallel test,  the flow was 21034 gpd  (Table 14),



      Operational Comments.   Operating problems were  encountered In the



 biological treatment unit.  The most serious of these was a tendency of



 solids to float in the secondary clarifier.  The result of this has been



a decreased ability to achieve the expected level of solids separation.



The presence of the floating sludge has been investigated and is



characterized as being the result of three separate and distinct causes;



flotation, turbulence and denitrification.  It was apparent that there



was some carryover of floating materials to the clarifier from the



aeration tank.   The leachate contains considerable amounts of surface



active materials capable of flotation,  and this contributes significantly



                                 -69-

-------
                             TABLE 16




                WARM WEATHER OPERATION OF SYSTEM 1

Parameter
Ammonla-N
BOD
Cadmium
Chromium
Copper
Iron
Lead
Mercury
Zinc
Effluent Concentration
8.7
75.8
0.02
0.06
0.10
1.22
0.11
0.0045
0.37
Permit Standard
35.0
100.0
0.02
0.10
0,20
7.0
0.10
0.01
0.60

Time periods included are 8/1/76 to 12/1/76 and 7/1/78 to 8/31/76.




All units are mg per liter.
                                -70-

-------
 I
•-J
                                                   TABLE 17



                     COMPARISON OF SERIES AND PARALLEL OPERATION OF ACTIVATED  SLUDGE  UNITS
Parallel Operation
(8/1/76 - 5/1/77)

Parameter
Alka Unity, as CaCO-
Ammonia-N
BOD-5
Cadmium
Calcium
Chloride
Chromium
COD
Copper
Dissolved sol ids
Hardness, as CaCO-
Iron *
Kjeldahl-N
Lead
Magnes i urn
Mercury
Nickel
Phosphates
Potassium
Sodium
Sulfate
Suspended sol ids
Zinc
*mg/ liter
a.
Influent
1604
286
356k
0.021
308
2508
0.08
6481
0.34
5710
1242
3.6
294
0.18
105
0.004
0.61
0.041
487
712
384
147
0.72


ซ
Effluent
685
80
118
0.01
347
2592
0.07
939
0.1
5697
1314
3.0
102
0.12
107
0.004
0.76
13.7
524
821
951
101
0.57

Remova 1
Efficiency
57.3
72.0
96.7
52.4
-12.7
-3.3
12.5
85.5
70.6
0.3
-5.8
16.7
65.3
33.3
-1.9
0.0
-24.6
—
-7.6
-15.3

31.3
20.8

Series Operation
(7/1/78 - 8/31/78)

n
Influent
3127
738
5966
0.04
227
2933
0.06
9873
0.16
8635
1007
1.85

0.12
116
0.003
0.62
0.53
743
1160
258
663
0.24

JL
Effluent
1496
6.3
464
0.04
113
2254
0.04
1088
0.16
7244
1456
0.71

0.11
105
0.002
0.67
17.2
743
1145
836
478
0.16

Remova 1
Efficiency
52.2
99.1
92.2
0.0
50.2
23.2
33.3
89.0
0.0
16.1
-44.6
61.6

8.3
8.7
33.3
-8.1

0.0
1.3

27.9
33.3


-------
to the carryover phenomenon,.  The original scum control  device was not


capable of handling the unexpectedly large amount of these materials.


At the same time, an excessive amount.of turbulence existed in the



secondary clarifier.


     The reduction of solids separation efficiency was compounded by a


propensity of the activated sludge to become anaerobic and to rise in


the clarifier due to denitrification.  This is seen clearly when one


follows closely the settleable solids test.  At first, the sludge


settles properly with a dense sludge layer overlain by a clear


supernatant containing little turbidity, so that at the end of 30 to 45


minutes the settleable solids are about 300 mg/liter.  If the test is


continued for another hour, the sludge comes to the surface in typical


rising sludge fashion.  This is not a case of filamentous bulking as


indicated by microscopic examination, the clear supernatant observed


in the settleable solids test, and the sludge volume  index of approximately


80 ml/g.  Another indication of anaerobic denitrification as the cause


of the floating sludge is the repeated observation of very low DO levels


in the clarifier.  This problem was accentuated when a portion of the


plant aeration capacity was diverted to the ammonia-stripping lagoon.


     These problems were resolved by minor design  and operational changes.


The  installation of a more efficient scum  removal  system has ameliorated


the  floating sludge problem.  Additional  baffles were used to control


the  turbulence  in the clarifier.  Extra aeration capacity was  installed


to maintain appropriate concentrations of  dissolved oxygen,


     Cost'Data.  Costs incurred during the operation  of the biological
  t

units are  indicated  in Table 18.  The operation and maintenance  costs


                                  -72-

-------
                                                 TABLE  18

                OPERATION AND  MAINTENANCE  COSTS  INCURRED DURING THE  OPERATION OF SYSTEM 1
                            ,    FOLLOWING  ACCLIMATION OF ACTIVATED SLUDGE
                                 (8/1/76 to  5/1/77 and  7/1/78  to 8/31/78)
 Characteristics

Total Flow, gal
cu m
Lime, Ib
lb/1000 gal
kg/ cu m
Sulfuric acid, gal
gal/1000 gal
1 i ter/cu m
Phosphoric acid, gal
gal/1000 gal
NaOH, gal
gal/1000 gal
1 i ter/cu m
NaOCl, gal
gal/1000 gal
1 i ter/cu m
8/1/76- 5/1/77
5332635
20186
132600
24.9
3.0
407
0.076
0.076
80.25
0.015
733
0.137
0.137
571.5
0.107
0.107
7/1/78 - 8/31/78 Total
620620*
2368
84300
27.2
3.3
380
0.122
0.122
35
0.011
0
0
0
330
0.532
0.532
5953255
22554
216900
25.7
3.1
787
0.093
0.093
115.25
0.014
733
0.123
0.123
901.5
0.151
-0.151
Costs, $/1000 gal
     Power
     Lime
     HjPOj,
     NaOH
     NaCl

     Total
1.92
0.75
0.06
0.04
0.09
0.08

2.94
1.92
0.82
0.10
  03

  40
                                                                          3.27
1.92
0.77
Q.07
0.04
0.08
0.11

2.99
      During this time period, the average flow through  the chemical/physical  section of the plant
was 50,000 gpd.^whereas that through the biological  units  was  10,010 gpd.   The dosages in this table
reflect these different flows.

-------
are shown for the operational periods following-acclimation: August 1,



1976 to May 1, 1977 and July 1, 197B to August 31, 1978.  The costs



Include those for NaOH, added to the ammonia lagoon to enhance ammonia



stripping, and NaOCl, added to the final effluent to provide disinfection



prior to discharge and to provide ammonia oxidation during the cold



months.
         >

     The data Indicate a cost of $2,99 per thousand gallons treated.



The high power costs reflect the demand for electricity for leachate



pumping, effluent pumping, and maintenance of the laboratory in addition



to the requirements for actual treatment.



     The costs during the re-test of System 1 warrant discussion.  During



this period,  the flow through the System 2 section of the plant was



approximately 50,000gpd, while the flow through the activated sludge



units was  10,010 gpd.  The dosages of chemicals and the costs of treatment



shown  in Table 18 take account of  these differences.  Thus the dosage



for lime  is based on 50,000 gpd, whereas that for NaOCl is  based on 10,010



gpd.  The  costs during the re-test are  somewhat higher than anticipated.



This results  from two causes.  First, the final effluent was chlorinated



during the entire 1978 test  period prior to  land disposal,  The effluent



was chlorinated for a  relatively small  portion of the time  during  the



earlier  test.  The  second reason  is that lime and sulfuric  acid dosages



were greater  during  the  re-test.  This  was because more lime was added



to enhance ammonia  removals  in the ammonia lagoon, and, as  a consequence,



sulfuric acid dosages were also high.   The labor  requirement  is



approximately 20 man-hours per week,



                                 -Ik-

-------
                                Nitrification



     Nitrification is becoming a standard and widely used wastewater



treatment process.  It is the aerobic microbiological conversion of



ammonia nitrogen to nitrate nitrogen.  As such, nitrification is



applied In situations where the pollution potential of ammonia is



severe, especially in comparison with nitrate nitrogen.  The adverse



impacts associated with high concentrations of ammonia are promotion



of eutrophication; toxicity, especially as a function of pH, to aquatic



organisms; Interference with chlorinatlon due to reactions leading to



the formation of chloramines; and, depletion of dissolved oxygen fn



receiving streams concomitant with the oxidation of ammonia to nitrate,



As a result of considerations such as these, effluent ammonia standards



are often established by regulatory agencies,



     The nitrifying bacteria, typified by Nitrosomanas and Nitrobacter,



oxidize ammonia to nitrite and nitrite to nitrate, respectively.  These



microorganisms are chemoautotrophs, and the nitrogen oxidation reactions



provide the bacteria with a source of energy.  A stoichiometric



relationship can be written for the overall synthesis of biomass and


                                            56
oxidation of ammonia and nitrite as follows:
NH4+ +1.83



1.88 H2C03
1.98 HC0
0.020
1.041
                                                          0.98 N0
                                                                      (1)
Nitrifying bacterial biomass  is represented as C,-H7N02.  Eqn  (1)



indicates that the theoretical cellular yield is 0.16 g biomass per g



NH.-N completely oxidized,   (Also evident from Eqn  (l) is the destruction



of 7.1 g alkalinity per g NH.-N oxidized, as discussed previously,)



                                  -75-

-------
      Much  of  the literature concerned  with  waste  treatment  applications



 of  nitrification has  been  devoted  to municipal  sewage.   A  recently



 published  design manual  has summarized this literature.     The ammonia



 content  of municipal  sewage is  characteristically on  the order of twenty
 to  forty  mg  per  liter.   Occasionally,  however,  industrial wastes with

                                            rQ ฃ1

 much  higher  concentrations  are  encountered,        |n  this section of



 the report,  dqta  are  presented  gnd  discussed showing  that nitrification



 is  inhibited at high  substrate  concentrations.   Substrate inhibition



 such  as described herein must be  considered during  both  the design and



 operation of biological  treatment plants  intended for nitrification.



      Boon and  Laudelot have investigated  the nitrite  oxidizing bacterium,


                         62
 Nitrobacter  winogradskyi.    They have shown that the rate of nitrite



 oxidation,which  is inhibited by nitrite,  is a function of the nitrite



 concentration  which can  be  expressed as
                                                                     (2)
                                       1
                                         -1
where y  is the specificgrowth rate, time   ; y, the maximum specific




growth rate  in the absence of inhibition, time  ; S, the limiting



substrate concentration, mass per unit volume; K , saturation constant,



numerically equal to the lowest concentration of substrate at which y


               A,

is one-half of y, mass per unit volume; K., inhibition constant,



numerically equal to the highest concentration of substrate at which

                 A

y is one-half of y, mass per unit volume.




     The assumption of neglecting substrate inhibition Is not particularly



misleading at low substrate concentrations.  This is because the third




                                  -76-

-------
term in the denominator of Eqn  (2) becomes much less than the second

term.  As a result, as the value of 5  is reduced, Eqn  (2) approaches

the more conventional Monod expression
                                „ „ TJL_                          {3)
                                         3

This equation  is a reasonable description of nitrification at low values

of S, as are encountered in municipal sewage treatment applications.

Equation (2) becomes increasingly invalid as S increases, as in the

treatment of leachate.  The purpose of this section is to consider

nitrification at high ammonia concentrations, and to discuss the results

in terms of designing and operating nitrification and leachate treatment

faci1ities.

     The data were collected during the period August 1, 1976 to May 1,

1977.  The fraction of nitrifying bacteria in the biomass has been

calculated directly from the amounts of BOD and NH.-N oxidized,  and from

the estimated yields of the two groups of oxidizers.  The assumed yield

values are 0.15 g biomass per g nitrogen oxidized-day, and 0.55 g biomass

per g BOD oxidized-day.

     More detailed results of the removal of ammonia in the bio-units

are given in Table 19.   The concentration of NH.-N is calculated from
the equation,
                                H_0   NH^  + H20
using the observed temperature to determine the equilibrium constant,

the observed pH and the analytical results of the sum of NH.-N plus

NH -N (denoted as NT in Table 19).  Three removals are given in Table 19

                                  -77-

-------
                                                                TABLE 19

                                              AMMONIA REMOVAL IN ACTIVATED SLUDGE UNITS*
00
 I
Date Influent Data
NT
5/27/76 549
6/10 241
6/1 it 277
6/22 283
6/30 325
7/15 308
7/21 350
7/27 290
8/3 322
8/10 255
8/20 305
8/23 364
8/24 364
8/25 364
8/26 360
8/27 356
8/28
8/29
8/30
9/3 369
9/10 302
9/13 364
9/15 364
9/23 322
BOD
2020
— — i.
2970
3550
3670
2499
2049
2555
1763
1949
2760
—
—
3780

—
—
—

—

—
2905
2637
COC
4863
_•*_
4961
4047
4351
3984
3915
3960
3672
3621
4843
5078
5175
5080
5039
3347
5118
—
5381
4220
5061
6370
5118
4264

Temp.
24.6
?<; n
t-J * V
25.2
26.0
25.0
26.3
26.5
27.1
24.5
26.5 .
24.5
28.8
27.5
24.4
26.5
26.5
28.2
27-3
24.5
21.3
20.8
23.7
22.0
24.5

NT
. 736
___
496
392
330
367
302
261
297
161
36
15
10
14
17
6
1.
1.
1.
1.
2.
8.
3.
6.
Bio-Reactor
NH4+-N
— ._

...
—
—
—
— •
—
—
—
35

—
14
—
—
2
5
5
9 1.9
7 2.6
8
6 3.5
0 5.8
Data
PH
8.04

7.59
7.37
7.80
7.36
7.57
7.34
7.74
7.42
7.45
7.43
7.31
7.35
7.42
7.22
7.82
—
7.59
7.3
7.7
7.5
7.6
7.7

BOD
2980

2325
2200
3710
1152
180
1181
111
30
48
12
36
27
24
14
35
25
35
43
310

37
123

COD
9725

4186
1556
4656
717
775
3280
664
621
768
406
420
384
526
474
299
396
565
546
1053
710
599
620
Remova 1 s
"I
0

0
0
0
0
.14
.10
.08
.37
.88
.96
.96
.96
.95
.98

...
	
.99
.99
.98
.99
.98
R3










.61


.63





.53


.50
.92
"4










.026


.012





.015


.022
.036
                                                                             .
                        NT refers to the analytically determined value of NHj,-N plus NHj,  whereas NH/,-N is the calculated
                  ammonia concentration.  R|  is the removal  of NT based on influent and effluent concentrations, R2 is the
                  fraction of NT oxidized, and R3 is the specific oxidation rate,  g NT oxidized/g biomass-day.  Units are
                  mg  per liter unless otherwise noted.

-------
TABLE 19 (Cent.)
Date
9/29/76
10/7
10/13
10/19
10/28
11/3
11/13
11/18
11/23
11/24
12/2
12/4
12/6
12/11
12/13
12/14
12/22
12/27
1/4/77
1/10
1/12
1/14
1/24
1/25
2/1
2/9
2/17
2/23
3/2
3/3
3/10
3/17
3/24
3/30
4/6
4/14
4/21

Influent Data
NT
367
255
218
240
227
235
275
296
319
—
319
—
400
—
342
353
361
—
442
308
251
233
232
231
230
176
162
164
244
283
160
289
316
296
255
247
246

BOD
2660
1788
1590
1850
—
1750
1830
1880
2730
—
3020
—
—
—
—
6630
4435
—
3885
—
—
1630
—
1620
1725
1570
1490
1906
—
5360
9870
8990
9750
9440
2560
4090
5010

COD
4360
3068
2-791
3310
3170
3454
3262
3461
5910
6423
6902
8055
8716
9154
8550
9077
7116
7871
6798
3152
3178
2675
—
—
2936
2761
2526
2557
—
8635
15271
16351
1,5683
15895
3881
6853
8651

Temp.
22.7
23.2
18.9
15.5
12.5
14.7
11.2
12.2
10.2
12.9
6.9
4.4
4.2
11.0
8.0
3.2
4.1
3.0
5.0
4.0
3.0
0.0
0
0
1.0
3.0
6.6
9.0
11.3
12.3
15.5
12.6
5.8
11.0
15.0
15.0
20.9

Bio-Reactor Data
NT
i.
13.
5.
5.
1.
3.
3.
110
63
65
7*(
33
123
95
30
15
83
185
151
161
—
145
—
146
151
144
156
153
165
174
140
195
—
196
202
187
164
i
NH^-N pH
5 1.5 '7.6
2 12.9 7.5
2 5.2 7-0
8 5.7 7.4
0 1.0 7.6
4 3.4 7.6
2 3.2 7.4
109 7.6
58 8.6
— 8.5
71 8.4
— 8.4
— 8.4
— 7.9
— 7.9
10 7.8
83 7.2
— 8.0
143 8.6
— 8.4
— 8.5
139 8.6
— —
140 8.6
145 8.6
140 8.4
151 8.3
149 8.1
162 7.8
170 7.9
135 8.0
183 8.3
7.9
188 8.2
197 7.9
184 7.7
154 8.1

BOD
83
31
52
46
—
53
39
29
97
—
78
38
—
—
—
19
100
—
61
—
—
25
—
37
43
47
86
84
—
94
666
587
834
478
983
364
686

COD
581
299
271
295
268
321
286
296
400
382
392
501
. 564
518
398
410
645
864
2830
730
—
582
—
560
550
538
586
565

917
1559
2321
3842
1713
2560
1820
1787

Removals
Rl
.99
.95
.98
.98
.99
.99
.99
.63
.80
—
.77
—
.69
—
.91
.96
.77
—
.66
.48
—
.38
—
.37
.34
.18
.04
.07
.32
.39
.13
.33

.34
.21
.24
.33

R3
.52

.57
.43
.27
.65
.48
.35
..52

.50





.70

.34

	 r
.38

.17
.09
0
0
0

.15
.05
.12
—
.33
.14
.19
.28

R4
.013

.019
.014
.007
.022
.011
.010
.015

.015





.020

.008


.006

.001
.001
0
0
'o
	
.0.15
'.0.08
0.45
—
0.98
0.08
.68
.68


-------
The first, R.,  is the  percent  removal efficiency  based on  influent and



effluent concentrations; the second, R2,  is  the fraction of N_ which



is oxidized  in  the aeration tank; and R-,  is the  specific  oxidation



rate.  The units of R, are g nitrogen oxidized per g biomass-day, and



it is calculated as the daily  mass of nitrogen oxidized divided by the



biomass.  R,  is determined using the fraction of  nitrifiers in the



biological solids, expressed as volatile  suspended solids.



     As indicated in Table 19  the overall  removal of ammonia  in the



activated sludge units was considerable.   Although this Is primarily



due to the microbial process of nitrification, the physical stripping



of ammonia caused by aeration  accounts for some or the removal.   It Is



the purpose of  this discussion to consider the fgctors affecting  the



ammonia converted by nitrification.  It is apparent from Table 19 that



a profound inhibitory effect was placed on the specific oxidation rate.



as a result of extreme winter  temperatures.  Since the result was to



decimate the nitrifying population, the following discussion  is limited



to the data obtained up to 25  January 1977.  Data are not  included from



the third year of operation because of changes in flow rates and  loading



rates.



     No simple pattern could be perceived  which related nitrification



measured as the specific oxidation rate (R-, Table 19) with the



concentration of organic matter.  Nitrification did not occur when the



reactor BOD,, exceeded one gram per liter and when the COD exceeded three



grams per liter.  However, subsequent to the formation of a culture of



acclimated activated sludge,  capable of nitrification and organics



oxidation, there was no consistent relationship between the specific



                                  -80-

-------
oxidation rate and the concentration of organic matter  in the reactor,


Nitrification proceeded even though the activated sludge Influent


concentration of BQD_ and COD averaged 3564 and 648] mg per  liter,


respectively.  This was true when the ammonia content of the bio-un?t


influent did not exceed 300 mg per liter.  At higher ammonia levels,


the toxicity of aromonig predominated and very little oxidation of either


organics or ammonia occurred.


     It is reasonable to expect some inhibition due to the elevated


levels of BOD,, and COD recorded in Table 19,  The nitrlfers, such as


NItrosomonas and N|trobacter, are chemolJthotrophs, i.e., they are


autotrophs, and their carbon source is Inorganic.  These bacteria are


especially sensitive to organics In pure culture.  Certain heterotrophlc


bacteria, fungi, and actinomycetes are capable of performing


nitrification,  '   although it is generally believed that the rate


of heterotrophic nitrification is much less than that of the autotrophic


nitrifiers.    The sensitivity of these organisms to organics is


presumably much less than it is for the autotrophs.  Wild et al.


observed no effect of BOD, in the range of 5"!10 mg per liter,  on the


rate of nitrification in activated sludge processes.


     The effect of temperature on the specific oxidation rate Is shown


in Figure 8.   This curve agrees well  with curves developed and  presented

                                                68
recently by the Environmental Protection Agency.    The curve has been


fitted in the least squares sense and the equation of the line  is



                           R3 P. Xe"E/RT                             (5)
                                  -81-

-------
oo
to
i
                          FIGURE 8.  EFFECT OF TEMPERATURE  ON  SPECIFIC OXIDATION

                                     RATE.  THE LEAST SQUARES  LINE OF BEST FIT

                                     SHOWN.  THE NOTATION 1  REFERS TO TEMPERATURE, K:

                                     AND R  IS SPECIFIC OXIDATION RATE.
                                                                      R,ซ 2.1754 * 109,xp |- I2347/I.99T)
                                           10             20


                                      Tempergture, K

-------
                                                  "
where, R, * specific oxidation rate, day


                      9 '   "1
            2,175 x 10 .day   s Arrhehiu



            activation energy * 123^7 cal/mple
                                9 '   "1
                 X  =f 2,175 x 10 .day   s Arrhehius frequency factor
                 R  s gas constant ป 1 ,99 cal/mole"K



                 T  = temperature, K



The curve of Figure 8 is somewhat lower than the theoretical one



developed by the EPA,   and this is most likely due to an inhibitory



effect of some wastewater fraction such as heavy metals, organ fcs, and



substrate.




     A number of values for E have been reported in the 1 iterature.



Wong-Chong and Loehr presented data showing the variation of E with



pH.'   Their results indicated that for the ammonia oxidation, the value



of E ranges from 16 to 21.6 K cal per g mole,  Sutton et/a_l_. state that



E is a function of the treatment mode, and in particular of the staging



of the biological units, and the sol ids retention time.    For single



stage units, such as this one (August 1 , 1976 - May 1, 1977), they report



that the values of E are 25.1, 21.25, and 11,9 k cal/g mole for solids



retention times  (SRT) of 4, 7, and 10 days, respectively.  The figures



refer to the overall nitrification process.  The last number is of



particular importance since an SRT of ten days is considered to be the



minimum needed for thorough nitrification7  (although Hutton and LaRocca7


                                                        7k
considered thirty days to be more reasonable, and others   consider three



to four days as the minimum needed to ensure successful nitrification).
The value of 12350 cal per g mole obtained here is in good agreement with



that of Sutton et al.
                                  -83-

-------
     A number  of factors  were Investigated  for potential  Inhibition  of

 nitrification.   As  discussed  above,  there  Is  no consistent  observed

 effect of  organic matter,  measured as  BOD,,  and COD,  on  nitrification.

 However, based  on the  literature.    It Is probable that some of  the

 Inhibition noted above results  from  the relatively high BOD and  COD

 concentrations  in the  aeration  tanki   The heavy metals  are  another

 likely source of inhibition,  although  an obvious relationship was not

 observed.   Average  heavy  metals concentrations in the aeration tank

 are  listed in Tables 14 and 15 as effluent  values.

     The clearest reason  for  the observed inhibition  is that increasing

 concentrations  of substrate are the cause.  This is  shown in Figure  9,

 where  the  data  show a  reasonable fit to the classic  substrate inhibition

 model.  The model is based on the Haldane   mechanism for the substrate

 inhibition of enzymes, and is expressed here as
                                  1 * Ks * S
                                      S    K,
                                                                     (6)
where R_ is the specific oxidation rate, day " ; R_ is the maximum

specific oxidation rate which would be obtained in the absence of

substrate inhibition, day~ ; K  is the saturation constant; and, K, is

the inhibition constant.  Kg and K, are numerically equal to the higher

and lower substrate concentrations, respectively, at which R, equals
            A
one-half of R,.

     The curve in Figure 9 has been developed using a quasiซleast squares

procedure and LIneweaver-Burke plots to evaluate the kinetic constants,

                                  -84-

-------
          3r-
 I
CO
vn
                           FIGURE 9.  SUBSTRATE INHIBITION OF NITRIFICATION.

                                      R  IS THE SPECIFIC OXIDATION RATE
                                                                                              U—4 -i-
                                                                                                  S   36
                                                                                                          200
                                         NH^-N, mg/1iter

-------
The equation used to generate the line Is
                                  3,50
                                                                    (7)
Indicating that the value of R. Is 3.5 day  , Kg is 4.0 mg per liter,



and K. Is 36 mg per liter.



     Figure 10 shows a Michaelis-Menton plot of the data obtained at



the lower concentrations.  These values correspond to the situation



encountered in the treatment of domestic sewage, and this figure therefore


                                           78
resembles the type curve usually presented,7   In the treatment of



Industrial wastes where the Influent ammonia-N may be much higher, it



will be necessary to consider the substrate inhibition of nitrification.



It should be mentioned that the nature of the situation, I.e., a full



scale plant treating a variable Influent, with differences in pH,



temperature, and other operating variables, contributes to the scatter



of the data observed in Figures 9 and 10,



     In the excellent review prepared by Focht and Chang, much of the


                                                       79
literature regarding nitrification has been summarized.    They report



that the saturation constant for the Michaelis-Menton model Is In the



range of one to ten mg N per liter for ammonium oxidation, and five to



nine mg N per liter for the conversion of nitrite.  Poduska and Andrews



state that the saturation constants for N11 ro somona s and N11 r6bac te r



are approximately one mg per liter for full scale activated sludge



systems.    The value of four mg N per 11ter reported here Is consistent



with these data.



                                 -86-

-------
i
00
                                    FIGURE 10,  EFFECT OF LOW CONCENTRATIONS OF

                                                SUBSTRATE ON SPECIFIC OXIDATION
                                                RATE (R )
   +        20
NH.-N, mg/llter
                                                                     30

-------
     Focht and Chang also state.that the fIrst observation of substrate
 Inhibition in nitrification was made in tne early studies of Meyerhof
who showed that Nitrosomonas and Nitrobacter are inhibited at
concentrations of NH.-N and NOj-N exceeding 60 and 350 mg per liter,
             8l
respectively.    Wild ฃฃฃ]_. found no evidence for inhibition due to
                                                  82
ammonia in the range of six to sixty mg per liter.    Focht and Chang
also point out that end product inhibition has been demonstrated with
                                  83
both Nitrosomonas and Nitrobaeter,    This refers to the situation in
which accumulated nitrate or nitrite inhibits further nitrification,
     A number of Investigators have reported that inhibition of
                                                                OJ. OO
nitrification is due to ammon!a-N and to unionized nitrous acid,
Verstraete eฃ a\_. recommend that startup of nitrification units treating
highly nitrogenous wastes must consist of a gradual  increase in the
nitrogen loading In order to avoid the deleterious effects of these
                                  89
undissoctated species of nitrogen.    The data shown in Table 19 have
been examined in this light in order to assess nitrification inhibition
which could be attributed to the free ammonia concentration.  When
this is done, however, it is observed that there is no relationship
between the specific oxidation rate and the concentration of free
ammonia.  In this respect, the results of this study are In conflict
with those cited above, and the question merits further study,
     Kholdebarin and Oertli have reported recently their studies of
                           90                                    4>
batch growth nitrification.    They observed that the Ionized NhV-N has
a stimulating effect on nitrite oxidation,  |n the present study,
NH.-N was observed to exert an inhibitory effect.  As shown in Figure 9,
                                  -88-

-------
 the effect  is not profound until the concentration exceeds ten to



 twenty mg NHj^-N per liter.  The ammonium  ion  levels used by Kholdebarin



 and Oertli were 2,8 mg N per  liter, and in one experiement, 28 mg N per



 liter.    Thus, assuming that Nt^-N inhibits  nitrification, the



 experiments of Kholdebarin and Qertlj must be performed at higher



 concentrations in order to be comparable with the results presented here.



     The approach which has been tollowed here has been to attempt to



 simplify an exceedingly complex mjcrobial process.  The conversion of



 ammonia has been considered the rate limiting step, and has received



 primary attention.  The maximum growth rate of Nltrobacter is larger



 than that of Nitrosomonas and the value of the saturation constants



 is approximately the same.  Thus, nitrite is oxidized more quickly than



 is the ammonia, and this forms the basis of the above assumption.



 Certain aspects of the inhibition of nitrification have not been



 considered.  Examples are product inhibition due to nitrite and nitrate,



 and the non-competitive inhibitory effect of nitrous acid.   A third area



 of simplification is that the environmental factors are quite variable



 at the leachate treatment plant, due to influent quality and quantity



 changes.



     It is contended that this approach Is justified because of the



 complexity of full-scale operation.   The value is that nitrification is



 viewed from the overall perspective, and it is seen that substrate



 inhibition is a phenomenon operating fn this system.  This  knowledge can



 serve as a gu(ae ?n both the design  and operation of nitrification

                                                               QO
 systems, and can ultimately be incorporated within steady-state   and


       93
dynamic   models of the process,



                                 -89-

-------
     Summary.  The rate of nitrification, as expressed as the specific



oxidation rate, follows the van't-Hoff Arrhenius relationship which



indicates that the activation energy Is approximately 12350 cal  per mole,


                                                    9    ~1
and that the Arrhenius frequency factor is 2,18 x 10  day  .  The data



Indicate that substrate inhibition due to ammonium ion concentration



occurs In this system.  This relationship has been expressed as  a



Haldane inhibition model in which the maximum specific oxidation rate



is 3.5 g N oxidized per g biomass day, K  Is 4 mg per liter, and KI is



36 mg per 1Iter.



                  System 2.  Chemical/Physical Treatment



     This discussion  is presented In two parts.  The first consists of



the results associated only with lime treatment, and the second includes



the ammonia stripping lagoon.  Full-scale data were collected for System



2a without the lagoon during the periods November 15, 1975 to January



12, 1976 and June 14, 1976 to April 30, 1977, all dates inclusive.  The

                                \

results of this phase of the treatment plant operation are summarized



In Table 20, columns  1-7.  The distinction between the two periods permits



an assessment of the  performance of the ammonia lagoon.



     Table 20 summarizes the changes in each parameter attributable to



the lime treatment.   In very approximate terms, the 1ime precipitation/



clarification sequence, System 2a, removed  (see Column 7, Table 20)



one-quarter of the nitrogen; one-third of the dissolved solids;  one-half



of the organic matter, hardness and alkalinity; three-quarters of the



suspended solids; and ninety percent of the phosphates.  The removal of



heavy metals was over one-half of the mercury and cadmium; two-thirds



                                 -90-

-------
                                                  TABLE  20

           SUMMARY OF SYSTEM 2 RESULTS.   EACH SYSTEM CONSISTS  OF LIME TREATMENT AND CLARIFICATION
              ., SYSTEM 2a and 2b ARE WITHOUT AND WITH AIR STRIPPING  OF AMMONIA, RESPECTIVELY.*
Parameter

Alkalinity, as CaCO,
Ammonia-N
BOD-5
Cadmium
Calcium, as CaC03
Chloride
Chromium
COD
Copper
Dissolved Sol ids
Hardness, as CaC03
Iron
Kjeldahl-N
Lead
Magnesium, as CaCO?
Mercury
Nickel
PH
Phosphates
Potassium
Sod i urn
Sulfate
Suspended Sol ids
Zinc
System 2a
Influent
n
(l)
65
66
65
62
63
75
63
212
62
274
76
61
72
60
70
66
63
300
63
75
78
72
224
61
X CV
(2) (3)
5668 0.23
1167 1.25
10356 0.86
0.07 0.87
863 0.54
4590 0:43
0.25 0.75
16618 0.74
0.46 0.93
12652 0.56
5257 0.41
350 0.74
1157 1.28
0.75 0.73
562 0.42
5.69 1.30
1.58 1.21
6.80 1.28
2.40 0.63
941 0.27
1284 0.35
• 409 3.23
843 1.14
19 1.06
Effluent
n
(4)
43
48
46
41
43
45
43
151
52
218
49
43
50
43
43
40
43
242
42
46
48
46
183
43
X CV
(5) (6)
3052 0.30
890 1.26
5265 1.03
0.03 0.49
696 0.57
3516 0.74
0.09 0.58
7188 0.67
0.10 0.47
7972 0.32
2461 0.50
3.8 3.05
867 1.39
0.24 0.85
209 .90
2.85 1.22
.57 1.44
8.46 8.04
.26 .83
613 .36
830 .45
426 .89
239 1.15
0.61 1.35
R
(7)
46.2
23.7
49.2
57.1
19.4
23.4
64.0
56.7
78.3
37.0
53.2
98.9
25.1
68.0
62.8
49.9
63.3
—
89.2
34.9
35.4
-4.2
71.6
96.8
System 2b
Influent
n
(8)
79
88
77
80
80
78
80
261
79
232
76
74
61
80
80
74
81
394
84
61
78
79
255
86
X
(9)
5316
785
11668
0.086
841
3927
0.25
18566
0.43
10456
4645
300
739
0.68
421
0.021
1.60
6.93
2.52
944
1366
512
967
15.9
CV
(10)
0.78
0.31
0.71
0.82
0.57
1.49
0.71
0.62
0.94
1.04
0.49
0.82
0.27
0.76
0.42
2.43
1.06
1.56
0.75
0.21
0.30
2.46
1.11
1.15
Effluent
n
(11)
70
86
71
71
70
68
70
254
73
236
64
68
52
67
69
63
67
383
74
52
66
70
253
70
X CV
(12) (14)
2374 0.28
412 0.52
3600 0.84
0.035 1.09
424 0.70
2669 0.15
0.08 0.94
8793 0.63
0.27 1.27
4650 1.09
1587 0.53
5.61 1.28
349 0.41
0.23 1.31
117 0.62
0.0101.96
0.73 0.93
866 1.79
0.27 2.07
572 0.25
956 0.40
525 0.67
288 0.93
0.85 1.39
R

55.3
47.5
69.1
59.3
49.6
32.0
68.0
52.6
37.8
55.5
65.8
98.1
52.7
65.6
72.1
52.4
54.3

89.3
39.4
30.0
-2.6
70.2
94.6
x - mean, mg/liter.  R stands for percent removal.   System 2a effluent  is  the  lagoon  influent.

-------
of the chromium, nickel, and  lead; three-quarters of the copper and



over ninety percent of the  iron and zinc.  The  increase in sulfate is




due primarily to contaminants  in the chemicals, although oxidation of




sulfides may contribute somewhat,  in other words, this section of the



system performed as expected  in pre-tfeating the leachate prior to



biological treatment.




     The results of the overall chemical/physical section including the



ammonia lagoon  (System 2b) are listed in Table 20 (Columns 8-1*0 which




shows the basic statistical relationships.  Treatment performance in



terms of percent removal efficiency of the lagoon alone and in



conjunction with lime treatment are also seen in Table 20.  The primary



goal of the lagoon was achieved as the concentration of ammonia-N was



reduced to a level which was found to be tolerable for purposes of



biological waste treatment.  A splash plate, which was installed to



promote air/water contact, did not produce an appreciable effect on



lagoon ammonia removals, and was therefore removed.



     Many parameters other than ammonia were altered while in the lagoon



(see Table 21).   There was some stabilization of organic matter as shown



by the reductions in BOD, COD and dissolved solids.   This was mediated



by biochemical processes and the increase in suspended solids is related



to the growth of microorganisms.  The reduction in alkalinity is due to



aeration effects although nitrification reactions may partially contribute



to the observation.   The reduction of hardness, calcium and magnesium are




related and may be explained by the formation of calcium and magnesium



carbonates.  In this form, these would not be detected by the usual




tests.   Most of the other changes noted in the lagoon effluent vs. lagoon



                                 -92-

-------
                               TABLE 21

         SUMMARY OF EFFECTS OF CHEMICAL/PHYSICAL TREATMENT*




Suspended Solids
Dissolved Sol ids
COD
BOD5
Alkal inity
Hardness
Magnes ium
Calcium
Chloride
Sulfate
Phosphate
Ammonia-N
Kjeldahl-N
Sod i um
Potassium
Cadmium
Chromium
Copper
1 ron
Nickel
Lead
Zinc
Mercury
pH

4.
Influent
1044
13029
18553
10907
5404
4652
453
818
4240
462
2.74
1001
984
1354
961
0.086
0.28
0.39
312
1.55
0.67
21
0.007
6.85

Lime Treatment
Effluent*
239
7972
7188
5265
3052
2461
209
696
3516
426
0.26
890
867
830
613
0.03
0.09
0.10
3.8
0.57
0.24
0.61
0.003
8.46
Ammon i a
Lagoon
Effluent+
288
4650
8793
3600
2374
1587
117
424
2669
525
0.27
412
349
956
572
0.04
0.08
0.27
5.6
0.73
0.23
0.85
0.010
8.66

*The influent data are those collected during the entire operational
 period, whereas the effluent figures refer to those periods when the
 specific units were operating.

+A11 units are mg/liter except pH which is expressed in pH units.

                                 -93-

-------
Influent comparison are due to the.1 Imitations of the experimental



techniques or to the radically variable nature of the raw leachate.



     The overall treatment efficiency .of the complete chemical/physical



section Is summarized In Table 20,. These data do not Include the effect



of neutralization.  The values In the last column (Column 14) represent



removal efficiencies for the 1ime precipltation/sedimentat Ion/ammonia



stripping sequence.  In terms of organic matter, 69.1 and 52.6 percent



of the BOD and COD are removed, respectively.  Approximately fifty to



sixty percent of the ammonia-N, total nitrogen, suspended solids,



alkalinity and hardness are removed.  The removal of metals was as



follows: 38-54 percent of copper, nickel and mercury; 59~68 percent of



chromium, cadmium, and lead; 95 percent of zinc; and 98 percent of iron.



     Chlan and Dewalle have formed an hypothesis, which is summarized in


                                                     Oil

Table 6, concerning the treatability of raw leachate.    The BOD/COD



ratio observed in this study (Column 2 of Table 20) of the leachate was



0.62 and the average COD was 16618 mg/liter.  This is also shown in



Column 9 of Table 20, in which it  is seen that the ratio is 0.63, and



the average COD, 18566 mg/liter.  Thus, according to Chian and DeWalle,



the leachate treatment efficiency obtainable with lime should be fair.



In this study (Column 7, Table 20) the lime treatment efficiency for



BOD and COD has been about fifty percent.  Hence, in terms of the removal


                                                           95
of organ Ics, the Chian and DeWalle hypothesis is supported.    However,



It must be mentioned that their hypothesis did not include the removal



of heavy metals, and that the lime treated heavy metal removals have been



good to excellent at this facility.



                                   -94-

-------
     An additional effect of the ammonia stripping lagoon Is the



equalizing effect which, as noted by LaGrega and Keenan, can be measured


                                                           96
in terms of both flow variability and quality fluctuations.    The



presence of the lagoon has allowed the operator to control the flow



leaving the lagoon by control of the pump settings.  This has provided



additional flow equalization to the biological units.



     Operational Comments.  The primary operational factor has been the



chemicals required for precipitation and neutralization.  A summary of



these is presented in Table 22.  The rows labeled "average applied dose"



have been calculated by omitting those days on which chemicals could not



be added because of equipment malfunctions.



     Cost Data.  The cost of materials and electricity is a part of Table



22.  The units are given in terms of dollars per one thousand gallons of



leachate treated.   The data indicate that the cost of System 2 with, and



without, operation of the lagoon has been $2.37 and $2.35 per thousand



gallons treated, respectively.   The total cost figures have been obtained



as the ratio of total costs to total volume of liquid treated.   The power



costs are quite high, reflecting energy consumption not only for chemical



treatment, but also for leachate pumping, air compressors and the



laboratory.   Manpower costs for operation and maintenance are approximately



twenty hours per week.



             Factors Influencing Lime Treatment Performance



     The principal operating characteristics of the lime treatment system



are presented in Table 23 for data collected during the period  June 14,



1976 to April  21,  1977.   These  data only are included as this was the


                                 -95-

-------
                                                       TABLE 22

                        SUMMARY  OF OPERATION AMD MAINTENANCE COSTS DURING EVALUATION OF SYSTEM 2
                                         (11/15/75-5/1/77 and 11/1/77-8/31/78)
us
During Operation
Without Lagoon
11/15/75-5/1/77

Flow,



Lime,




NaOH,




NaOCl




Costs






average gpd
Ipd
total gal
total cu m
average dose,
lb/1000 gal
kg/cu m
total Ib
kg
average dose,
gal/1000 gal
1/cu m
total gal
total liter
, average dose,
gal/1000 gal
1 cu/m
total gal
total liter
, $/1000 gaJ
Power
Lime
NaOH
NaOCl
Total

22,805
86,326
8,004,483
30,300

29.7
3.57
238,055
108,077

0
0
0
0

0
A
0
0

1.1.8
.89
0
0
2.37
Total
38,618
146,170
16, 798, 842
63,584

19,
2,
326,050
147, 894

0
0
733
2,926

0
0
901
3,413

1
0


2
During Operation with Lagoon
(Ammonia)
6/14/76-5/1/77 ฃ 11/1/77-8/31/78






.40
.33



.044
.044



.054
.054
.5


.70
.58
.03
.04
.35
6/14/76-8/1/77
21 .391!
80,987
5,348,622
20,247






36.6
4.39
195,650
88,825

0
0
733
2,926

0
0
571
2,163

1
1
0
0
2



.137
.137



.107
.107
.5


.70
.10
.09
.08
.97
1/1/77-8/31/78
17,224
65,183
11,450,220
42,337

11.
1.
130,400
59,069

0
0
0
0

0
0
330
1,250

1
0
0
0
2





39
37








.03
.03



.70
.34

.02
.06

-------
                                                               TABLE 23


                                   SUMMARY OF OPERATIONAL DATA FOR LIME TREATMENT AND-CLARIFICATION*
--g
 i
Cadmium
Date
6/14
6/22
6/30
7/15
7/21
7/27
8/3
8/10
8/20
8/25
9/3
9/15
9/23
9/29
10/7
10/19
11/13
11/18
11/23
12/14
12/22
C
.020
.028
.01
.025
.018
.020
.025
.02
.05
.05
.05
.019
.015
.017
.015
.021
.009
.012
.03
.009
.01
R
.231
.282
.700
.50
.617
.615
- .667
.091
.444
.286
.375
.525
.400
.'(52
.50
.222
.710
.707
.333
.95
.90
Chromium
C
.01*
.04
.06
.06
.06
.08
.Ok
.03
.09
.10
.09
.07
.06
.09
.05
.08
.03
.03
.11
.07
.08
R
.50
.60
.25
.684
.739
• 714
--
.25
.571
.286
.625
.632
.625
.654
.500
.529
.625
.667
.500
.854
.333
Copper
C
.08
.10
.09
.12
.26
.2
.01
.18
.078
.064
.10
.1
.11
.13
.05
.08
.02
.03
.01
.04
.06
R
.771
.895
.763
.90
.807
.796
.808
—
.926
.559
.545
.919
.931
.859
.643
.75
.90
.786
.909
.934
.455
Iron
C
.48
2.72
2.54
2.22
1.52
2.89
.1
.92
.68
.58
.85
1.09
1.15
.42
1.74
2.59
1.98
2.13
2.43
2.99
1.78
R
.994
.982
.98}
.964
.996
.995
.997
.997
.997
.997
.997
.997
.996
.999
.983
.994
.966
.981
.995
.997
.996
Lead
C
__
--
—
.07
.05
.07
.02
.03
.11
.04
.48
.22
.35
.-
, .
.4
.
.
t
--
.2
R
__
—
—
.873
.889
.879
.80
.864
.929
.871
.392
.728
.646
.815
.667
.20
.875
.981
.787

.333
Magnesium
C
312
312
323
87
43
36
22
123
257
250
225
45
25
54
71
92
212
136
115 -
97
101
R
.448
.525
.441
.858
.936
.860
.900
.836
,471
.520
.595
.912
.947
.894
.862
.793
.668
.809
.818
.883
.789
Mercury
C
1.17
1.06
6.5
2.96
1.72
8.5
3.1
3.4
3.5
3.
5.5
2.1
3.9
7.5
1.8
1.78
.09
.21
.15
.22
.15
R
.957
.960
.606
.80
.204
.444
.852
—
.364
.857
.633
.781
—
.074
--
.282
.930
.756
.865
.808
.934
                 For each chemical species, the effluent concentration (C) Is given in mg/liter except mercury  (tig/liter),
           pH (pH units), lime dosage (ib), Temp. (C), and flow (gpd), and R represents the fraction of material  removed
           by this treatment section.

-------
       CM    o mm — UM/MAJT rno u\r*c* IAPJ eg eg CM —

       u\vDu\vo\or~-a-vo^r-a- — CM -a- CM m tn •— CM O m.3•
       eMCMCMCMCgcMegcMCMCMCMCMCMCMCM— — — —
                        OOOOOOOOOOOOOO
                        *s22aas

       in1 eg rwrrnrn'in' — CAOO — in-^
                         " """    ^ ^ *"™ *™

               ง"ซ S IN SSS i
       oocr\
Q)
T3
          ~ o cooo •—  i  f\ tr\vO — in un CT\ro en I
       3-ocor~.r~.cn  i  o^mr~.tr\o->oieor~.<^i
                     , o o o eg -3* moo en o o o o o
             __   _>r~.cocM—-a-cMcocMinomcMoovOi
         > \o eg CMCO eg      CM—              eg eg — —
 u> o:
 u
 4J
 ra
jr
 Q.
 Ul
 O
JC
0. O
cgco inclines eg r~.mtno^o^eg\o rno^^r r~.vo r~.vp
ooooooooooooooooooooo
          ^i- ™. uii.iป_f   o — o vo en   CM in cr\ r— m 
-------
period of primary analysis, and also because of the change in the
             !

leachate which occurred during the third year.


     The function of the lime treatment/clarification section of the


plant is to pre-treat the leachate prior to biological treatment, i.e.,


to remove a portion of the organic matter and toxic substances.  The


latter include ammonia as well as the heavy metals.  The lime


precipitation process has consistently produced an effluent which meets


the standards except for cadmium and lead,  (it should be noted that the


final effluent met these standards up to the onset of cold weather, and


that the probable mechanism for the additional removal is adsorption


onto the surface of the biological floe.)   It Is apparent that,



considering the variability of the incoming waste, equalization would


improve process efficiency and effluent quality.   The influent changes


occur so rapidly that the lime feed mechanism cannot maintain a constant


dose, with the result that there are occasions when the lime dosage is


inefficient and/or inadequate.  As a result of this consideration, the


raw leachate equalization lagoon was constructed.


     A desirable goal is to be able to operate the lime treatment process


selectively to Improve the quality of the effluent with respect to one or


more of the heavy metals.  As the solubilities of the metals vary as

                         18
different functions of pH  , it is not possible to optimize the removal


of all metals at a given pH, and the pH becomes an operational indicator


the operator can use to achieve differential removals.  Thus, one can fine


tune the operation to provide an effluent of suitable characteristics.


With this goal in mind, the data of Table 23 were examined to discover a



                                  -99-

-------
relationship between pH, temperature, and the concentration of heavy



metals In the clarjfier effluent.



      In order to develop a relationship, which could be used as an



operational tool, between pH and the removal of heavy metals, the




effluent concentrations were examined as possible functions of pH.



Figures 11 and 12 show the relationship between pH and nickel and mercury




effluent concentrations, respectively,  [n these figures, the circled



data points were collected when leachate temperatures were less than




16 C, whereas the others represent higher tempratures (see Figure 13).



     The effects of pH and temperature on the nicKel content of the



clarffier effluent are shown in Figure 11.  Process efficiency is not




a strong function of pH over the range of 8.9 to 11.7, and jt does not



matter whether one measures nickel in the effluent or the fraction



removed.   Effluent concentrations of nickel  centered about 0.20-0.30



mg/1 over the entire pH range, except at low temperatures.  For liquid



temperatures less than 16 C, the effluent nickel  level increased



substantially.  This is shown In Figure 13.



     Iron concentration decreased with increasing pH with the lowest



concentrations resulting at pH 10.3-12.2,   The effect of low temperatures



(<16 C) was to increase the amount of iron in the clarifier effluent.  In



terms of the Iron removal efficiency, the fraction of iron removed was



independent of pH and temperature, as It was always greater than 0,98.



     The response of chromium, lead and phosphate to pH was more or less



flat with no apparent relationship.  In each case, temperature did not



Influence the effluent concentrations.  There was a tendency for lower




effluent  chromium concentrations In the pH 9-10,5 range, and for better




                                -100-

-------


s_ i n
d) ' •"-
i ^Z
0 "ฃ,
UJ
Z ;5
H
LU
_l
U.
fc -25
h
' 1 MVI >iป i i u.p | |_u
THE C
LESS "




ฎ

x x ฎ
— 	 1 	
8.5
9.0
                 EFFECT OF pH ON CLARIFIER EFFLUENT NICKEL CONCENTRATION.
                 THE CIRCLED VALUES REPRESENT WASTEWATER TEMPERATURES OF
                                  x      x
10.0
                                       pH
                                                      11 .0
                                                                   12.0

-------
  10
          FIGURE 12.   EFFECT OF pH ON CLAR.IFIER EFFLUENT MERCURY CONCENTRATION.
                      THE CIRCLED VALUES REPRESENT WASTEWATER TEMPERATURES  OF
                      LESS THAN 16ฐC.       -
1
o
I
r   6
CTl
a:
o
LU
z:
t-
LU
 LU
        	L
     8.8   9
                                       10
                                          ฉ
                                                                               <ฃ>
11
12
12.5
                                            PH

-------
                 FIGURE  13.   EFFECT OF  LEACHATE. TEMPERATURE  ON CLARIFIER EFFLUENT

                              NICKEL CONCENTRATION
         1.0--
o
10
       o
       UJ
          .-5
                                       10                    20

                                               TEMPERATURE, K
30

-------
removals of lead at pH levels below 10.1, . However, these effects are



certainly not dramatic.




     The concentration of zinc In the clarifier effluent decreased with



Increasing pH.  Minimum concentrations occurred at pH 10.k to 12.2.  The



effect of low temperatures was to increase the zinc content of the




effluent.



     Chromium, copper and mercury concentrations show a tendency toward



a U-shaped response to pH.  The curve for mercury best illustrates this



relationship and is presented in Figure 12..  For each of these metals,



the effect of low temperatures is to reduce effluent concentrations.



The best pH values for chromium, copper and mercury removal are 10.2-



11.2,' 10-11, and 9.8-10.8, respectively,




          Systems 3 and k - Biological Treatment of Raw Leachate



     These treatment sequences were tested in full-scale during the late



winter and early spring of 1976.  Severe problems were encountered in



achieving successful treatment.   The primary reason underlying these



problems was the inability to develop a healthy activated sludge.



Approximately eight weeks were allocated to attempts to adapt a sewage



activated sludge culture to the law leachate.  After this did not succeed,



an investigation revealed that growth of activated sludge was not possible



because of ammonia inhibition and phosphorus limitation.   The problems



were demonstrated by the observations that the average concentrations in



the biological units during this time were 9^0 mg/Hter of ammonia-N and




less than one of phosphorus.  The data thus indicated that in the aeration




tanks, the ratio of BOD:N:P was 6620:760:1 which is in marked contrast




to the usual recommendations which are in the range of 90-150:5:1.



                                -104-

-------
     The phosphorus  limitation was  Investigated  in two ways.  First,




replicate BOD tests were set up with varying additions of phosphate




buffer.  It was found that the BQD_ increased with the phosphorus




addition up to an upper level, indicating that, within this range,




phosphorus was limiting.  As a result of this finding, the BOD




procedure was modified by the addition of sufficient phosphorus to



overcome the limitation.




     Second, a bench test was initiated to evaluate the hypothesis that



phosphorus limitation was the reason for the poor development of




activated sludge (see Table 24).  The tests consisted of once daily




batch draw-and-fi]1 experiments in which the increase in settleable




solids was used to monitor the growth of activated sludge.  The control




reactor received raw leachate only, whereas the sample reactor received



raw leachate plus seven ml  of BOD phosphate buffer per liter of raw



leachate.  Thus, in the sample reactor, the BOD:N:P ratio was about



118:13.5:1.   The results are summarized in Table 24.   It' is seen that




over the short-term, there was an apparent positive impact upon the



production of activated sludge and the utilization of COD.  However,



when the tests were continued for several  weeks, it became obvious




that there was no effect of phosphorus addition on either the development



of activated sludge or the removal of organics.




     The results of these experiments have  been interpreted in the



following manner.   First,  the biochemical  oxygen demand tests, and




the chemical  analyses showed that the leachate was severely phosphorus



limited.  This problem became more serious when biological treatment




                                -105-

-------
                              TABLE 24

RESULTS OF BATCH DRAW-AND-FILL ACTIVATED SLUDGE EXPERIMENTS TO DETERMINE
 THE EXTENT OF PHOSPHORUS LIMITATION.   RESULTS SHOW GROWTH OF ACTIVATED
     SLUDGE AS ML SETTLEABLE SOLIDS PER LITER, AND COD AS MG/LITER.

\*)D:N:P

time, days^


0
1
2
3
4
5
6
7
8
9
10
11
12
13

15
16
17
18
19
Control
6620:760:1

Influent effluent
COD COD SS
12813
7704
9339
--
8388 26
12868 10698 20
10193 15
11603
NR
9912
—
5963
9012 7115
13174
8606
8221
NR
—
6349
5625
Sample
118:13.5:

influent
COD
12813
7704
9339
—
8388
12868
10193
11603
--
9912
— —
5963
9012
13174
8606
8221

--
6349


1

effluent
COD SS




40
7597 ^0
35





7115






5469

                              -106-

-------
  followed  lime addition because of  the precipitation of calcium phosphate



  salts  !n  that unit.  Secondly, the batch draw-and-fil1 experiments showed




  that alleviatlon of  the phosphorus limitation alone Is not enough to




  encourage the growth of activated  sludge microorganisms.   It was concluded




  it would  be necessary to  reduce ammonia concentrations to a non-inhibitory



  level  before successful biological treatment could be achieved.




  Consequently, the ammonia-stripping  lagoon was started up prior  to



  evaluating System 1.




      Systems 3 and k were re-evaluated during the spring and summer of




  1977.   The change-over from System 1 occurred on May  1, 1977, and data



  were collected until August 31, 1977.  Summaries of the operating results



  for Systems 3 and k are presented  in Tables 25'and 26, respectively.




  System  k  did not operate well enough to recommend its further use.



      System 3, biological treatment  followed by chemical/physical



  treatment, achieved very good removal efficiencies, as shown in Table 25:



  approximately three-quarters or more of nitrogen, all  the heavy metals,



  and suspended solids; and, ninety  percent or more of the organic matter.



  However,  as shown in Table 27, the effluent quality does not approach



'  the standards placed on it, in terms of ammonia-N and organic matter.



  In addition, the standard for lead has not been met with System 3.  In



  sharp contrast, System 1 met all standards during the warmer months of



  late summer and early Fall, 1976 (see Table 16).  .




                       System 5 - Laboratory Studies



      Activated CarDon.  The preliminary evaluation of this system



  (System 5) has been carried out for  raw leachate treatment.  These data




                                 -107-

-------
                                TABLE  25

                           SYSTEM  3  OPERATION5'

Alkalinity, as CaCOj
Ammonia-N
8005
Cadmium
Calcium
Chloride
Chromium
COD
Copper
Dissolved Sol ids
Hardness, as CaC03
I ron
Kjeldahl-N
Lead
Magnesium
Mercury
Nickel
pH
Phosphates
Potassium
Sodium
Sulfate
Suspended Sol ids
Zinc
influent
n
12
12
13
13
13
13
13
38
13
39
13 '
13
12
13
13
13
13
39
12
13
13
14
39
13
X
5087
649
12649
0.11
937
4178
0.48
21152
0.27
14742
4463
348
708
0.76
350
0.007
2.0
7-6
2.3
1076
1536
658
1136
40
cv
0.20
0.17
0.22
0.52
0.17
0.19
0.33
0.21
0.38
0.20
0.21
0.39
0.16
0.28
0.22
0.51
0.50
0.56
0.57
0.13
0.09
0.23
0.47
0.42
effluent
n
8
8
9
10
10
9
10
32
10
33
10
10
8
10
10
10
10
32
8
10
10
9
32
10
X
1178
153
763
0.02
287
1496
0.08
2257
.07
5353
924
1.02
180
0.15
48
.002
0.27
10.20
0.56
476
719
513
180
0.51
cv
0.83
0.85
1.42
1.15
0.51
0.73
0.36
1.06
0.60
0.42
0.43
1.15
0.76
0.75
0.58
1.08
0.78
4.38
1.26
0.56
0.68
0.87
1.38
0.51
Percent
Remova 1
76.8
76.4
94.0
81.8
69.4
64.2
83.3
89.3
74.1
63.7
79.3
99.7
74.6
80.3
86.3
71.4
86.5
__
75.7
55.8
53.2
22.0
84.2
98.7
      Data collected from May 1, 1977 through August 31,  1977.  System
3_is biological treatment followed by chemical/physical treatment.
(x ** mean, mg/liter)
                                  -108-

-------
                             TABLE 26

                        SYSTEM 4 OPERATION'


AI ka Unity, as CaC03
Ammonia-N
BOD 5
Cadmium
Ca 1 c i urn
Chloride
Chromium
COD
Copper
Dissolved Solids
Hardness, as CaCO-j
Iron
Lead
Magnesium
Mercury
Nickel
pH
Phosphates
Potassium
Sod i urn
Sulfate
Suspended Solids
Zinc
Kjeldahl-N

n
12
12
13
13
13
13
13
38
13
39
13
13
13
13
13
13
39
12
13
13
14
39
13
12
influent
5
5087
649
12649
0.11
937
4178
0.48
21152
0.27
14742
4463
348
0.76
350
0.007
2.0
7.6
2.3
1076
1536
658
1136
40
708

cv
0.20
0.17
0.22
0.52
0.17
0.19
0.33
0.21
0.38
0.20
0.21
0.39
0.28
0.22
0.51
0.50
0.56
0.57
0.13
0.09
0.23
0.47
0.42
0.16

n
1
1
2
I
1
1
1
37
11
39
11
1
1
1
1
1
38
10
11
11
12
36
11
16
effluent
* - " cv
2788 0.13
312 0.35
2150 0.72
0.08 0.46
573 0.66
3778 0.23
0.37 0.39
468o 0.77
0.22 0.22
10081 0.20
2805 0.48
195 0.79
0.50 0.52
242 0.46
.0070.47
1.29 0.37
8.55 0.78
4.6 0.84
996 0.14
1412 0.09
853 0.44
1322 0.68
19 0.65
347 0.38
Percent
Kemova 1
45.2
51.9
83.0
27.3
38.8
9.6
22.9
77.9
18.5
31.6
37.1
44.0
34.2
30.9
0
35.5
—
—
7.4
8.1
—
--
52.5
51.0
   "Data collected from May 1
is biological  treatment only.
1977 through August 31
(x = mean;  mg/liter)
1977.   System
                               -109-

-------
                               TABLE 27

                  SUMMARY OF SYSTEM 3 OPERATION DATA
                            (5/1/77-8/31/77)



Parameter
AmmonJa-N
BOD5
Cadmium
Chromium
COO
Copper
Iron
Lead
Mercury
Nickel
Zinc
Raw
Leachate
mg/1
649
12649
0.11
0.48
21152
0.27
348
0.76
0.007
2.0
40
Final
Effluent
mg/1
153
763
0.02
0.08
2257
0.07
1.02
0.15
0.002
0.27
0.51

Percent
Remova 1
76.4
94.0
81.8
83.3
89.3
74.1
99.7
80.3
71.4
86.5
98.7
Discharge
Standard
mg/1
35
100
0.02
0.1
*
0.2
7.0
0.1
0.01
*
0.6

*No discharge standard for this parameter.
                               -110-

-------
are presented in Table 28.  These tests have been performed with an



upflow column of depth 0,3 m and diameter 0,46 m, containing 15.9 kg



of granular activated carbon.  The influent flow was 38 liter/min, thus



providing a hydraulic loading rate of 232 liter/min-sq m.  As shown in



Table 28, no appreciable treatment cgn be attributed to the carbon



treatment.  It should be noted that excessive suspended solids loading



and influent variability contributed to this finding.  The effect of,the



solids is to cause blockages and hence reduce process efficiency.  The



influent was not constant during any of the tests because it was drawn



from the actual plant influent.  Therefore ft is impossible to calculate



removal efficiency.  However, it Is evident from Table 28 that no



renovation is occurring in the carbon columns.  Hence, it is concluded



that carbon adsorption is not appropriate when applied to raw leachate,



although, as mentioned below, it may be suitable for final effluent



polishing.



     System 5 has also been assessed as an advanced waste treatment



technique.  This has been done using the effluent from the pilot



facility shown in Figure 5.  A summary of the data is given in Tables



29 and 30.  The data in Table 28 have been developed using small carbon



columns and a flow rate of 10 ml per mln,  |n each case, a 25 ml sample



was collected at each sampling Interval,  Although batch data are more



amenable to such mathematical treatment, a preliminary analysis has



been conducted to describe the results of Table 29 In terms of a Langmu.ir



adsorption isotherm.  The Langmuir Isotherm results from assuming



reversible adsorption and an adsorbed monolayer;




                                -111-

-------
                                        TABLE 28



           SUMMARY OF RESULTS OF CARBON ADSORPTION TREATMENT OF RAW LEACHATE*
Parameter

Gal Ions Treated
PH
Suspended Solids
Dissolved Solids
Total Volatile Solids
Chemical Oxygen Demand
Experiment Number

0
8.31
2880
10570
4240
4450
1
50 250
7.90 8.11
2280 3060
10490 9130
4040
3840 5760

0
___
980
10600
5120
8645
2
50
—
640
10450
3950
7217

320
—
20
11150
4230
9864

0

820
11100
5220
9530
3
100
___
540
11000
4930
9440

370
___
430
10970
5000
9841
Experiment Number

Gal Ions Treated
PH
Suspended Sol ids
Dissolved Solids
Total Volatile Solids
Chemical Oxygen Demand

0
—
520
10060
4530
9840
4
60
—
420
10690
5330
10784

0
__-
770
10460
5010
10013
5
20
—
500
10130
4360
9170

260
—
740
11040
5020
10040

0
___
1120
10040
6440
8833
6
60
___
1020
9400
6170
9723

300
___
1040
10230
6 180
9960
All units are mg/liter except pH and gallons treated.

-------
                                        TABLE 29




          TREATMENT OF  FINAL  EFFLUENT WITH BENCH-SCALE ACTIVATED  CARBON  COLUMNS*

Sample

1
2
3
k
5
6
1
2
3
k
5
6
7
8
9
10
Carbon
Volume, cm3

270
270
270
270
270
270
14.5
14.5
26.2
26.2
40.7
40.7
52.5
52.5
77.8
77.8
Weight Flow
of Carbon, g ml

100
100
100
100
100
100
5
5
10
10
15
15
20
20
30
30
0
100-125
225-250
350-375
475-500
600-625
725-750
250-275
500-525
250-275
500-525
250-275
500-525
250-275
500-525
250-275
500-525
COD
2088
354
657
830
748
827
869
1858
1818
1660
1760
1620
1660
1500
1620
1386
1500
Cu Fe
0.015 5.10
0 2.58
0 3.98
.015 4.80
.30 5.50
.68 1.60
.12 2.30










Zn Cr NJ
0.99 0.30 1.46
.30 .15 .08
.58 .20 .15
.53 .16 .26
.56 .15 .22
.61 .19 .26
.55 .18 .25










Ca Mg
174 142
0.74 0.89
1.72 2.00
2.29 2.59
2.05 3,47
5.86 8.6
>10 >10v











*A11  units are mg per liter,  unless  otherwise  stated.

-------
                          TABLE 30

       PILOT-SCALE CARBON TREATMENT OF FINAL EFFLUENT
Volume of Carbon Treated
  Final Effluent, liter

            0
           95
          189
          289
          379
          416
          454
          492
          530
          568
          606
          662
          719
          795
          871
Sorbate Concentration Following
Carbon Treatment, mg COD/liter

            1900
              25
              35
             395
            1312
            1539
            1695
            1773
            1828
            1859
            1875
            1852
            1891
            1883
            1848
                          -114-

-------
                           -
                           m
     (— )  H value of —when monolayer has been completed
        x •=" moles of sorbate adsorbed
m
            weight of carbon
       C  =i equilibrium molar concentration of sorbate


        b B adsorption coefficient


The analysis Indicated that (2~)ฐ is 9,709 mg per nig, and that b is 5,79


1 1 ters per mg.


     The data shown in Table 30 have been collected from a larger column


charged with 120 Ib of carbon at 3 flow rate of 4 gal per min.  These


results have been plotted as a carbon breakthrough curve In Figure 14.


The cross-hatched area of Figure \k can be used to determine the


fractional capacity, f, of the adsorption zone.  The value of f is


estimated at 0.31.   From this, it is estimated that the depth of the


adsorption zone Is 8.99 ft.  These values can be used as the basis of


the design of a full-scale unit.


     Additional Laboratory-Scale Studies.  A number of smaller scale


studies were undertaken during the third year.  These were designed to


evaluate a number of possible technologies which might be applied to more


efficiently treat the leachate.  In addition to the activated carbon


treatment discussed above, the following were also assessed: alkaline


chlorination, effluent filtration, and effluent breakpoint chlormation.


In addition, the use of additives in the lime treatment process was


investigated as a means of increasing the compaction of the lime sludge,


Each of these topics is discussed in subsequent paragraphs.


                                -115-

-------
                                   -9U-
       EFFLUENT SOLUTE  CONCENTRATION,  COD,  mg/liter
    K>
    vn
   ui
   o
< -^i
o vn
s;  o
>  o

H
m

;5

m  —
73  to
    vn
rn
to
01
   vn
   o
    CO
    o
    o
    N>
    vn

-------
      Alkaline Chlorination,   During  a  part of the third  year,  difficulties



 were encountered  in  the efficiency of  the Iime treatment process.   It  was



 believed  that the problems were  possibly due to excessive hydraulic




 loadings.   In this light, the use  of compacting agents as discussed




 below was  studied, and  minor  design  changes  were Instituted.   One other




 explanation  for the  difficulties encountered In the  removal of heavy metal




 is  that the  metals may  have formed complexes with organic matter or with




 cyanides.   In this chemical form,  the  metals resist  removal by lime




 treatment.   It might  have been possible  to improve the efficiency of lime



 precipitation of  heavy  metals by adding  chlorine during  this step.




 Alkaline chlorination has the effect of  disrupting heavy metal  complexes,



 particularly those involving  cyanides.   In this  manner,  the metals are



 freed from the complex ing agents,  after  which  they precipitate under the



 alkaline pH  conditions.




     Preliminary  evaluations  of  alkaline chlori-nation were conducted.



 The  results  were  negative as  shown In  Table  31.   The samples of raw



 leachate were supplemented with  the  chemicals  indicated  in Table 31,



 mixed for 5  min with a magnetic  stlrrer,  and  then  allowed to settle for



 30 min.  The volume of  sludge produced was recorded, and  analyses for




 heavy metals  were  performed on the clarified  supernatant.  The addition



 of NaOCl had  no significant effect on  sludge volume, nor on the




 concentration of copper and cadmium,  may  have  been associated with



 reduced concentrations of zinc,  nickel  and lead, and was definitely




 associated with reduced concentrations  of  iron and chromium.  The results



were deemed  negative, because the addition of small amounts of chlorine




                                -117-

-------
                                                   TABLE 31
                                   RESULTS OF ALKALINE CHLORINATION STUDIES
00

Sample
#
1
2
3
4
5
6
7
Lime
Addition*
0
19
19
25
30
19
19
NaOCl
Addition*
0
0
5
10
15
10
15
PH
—
11.3
11.2
11.1
11.3
10.5
10.0
Sludge
Volume*
—
100
120
150
120
>90
>90
Fe+
7.5
2.28
0.3
0.263
0.275
0.313
0.425
Cu+
0.25
0.213
0.275
0.216
0.293
0.21
0.209
Cr+
0.175
0.153
0.141
0.134
0.138
0.104
0.099
Pb+
0.25
0.3
0.25
0.24
0.24
0.15
0.18
U- +
Ni
1.095
0.831
0.769
0.691
0.668
0.69
0.668
Cd+
0.076
0.071
0.061
0.073
0.056
0.04
0.059
Zn+
10.6
0.09
2.08
1.27
1-95
0.09 '
0".05 ".

     millillters
     mi 11igrams per  1iter

-------
did not significantly affect the ability of the process to meet effluent
                         r   h

standards.  Very high concentrations of NaOCl may be helpful jn attaining

the chromium standard (see Table 1),

     Compaction of Lime Sludge,  The second possible cause of poor lime

treatment performance to be investigated was that of inferior

settleabl1ity of the sludge,  A number of Jar tests have been conducted

in an effort to evaluate the effect of additives on the efficiency of

heavy metal precipitation and on the compaction of the resultant lime

sludge.  These tests have been performed to evaluate the performance of

lime, a commercial aluminate preparation, and sodium hydroxide.  The


results are presented in Tables 32 and 33,  The first of these tables

shows the experimental  protocol whjch was, fn brief, to treat each flask

with lime to give a final pH of 11,7,  Aluminate was added at the rate of

300 mg/1 as recommended by the manufacturer's representative.  Sodium

hydroxide was added at the rate of 125 ppm in accordance with the

observation of the leachate treatment plant operator who noted a decided

improvement in process efficiency at this dosage of caustic.  |t should
                                                     t
be noted that the use of aluminate did not result in an especially clear •

supernatant and that it did result in a relatively large volume of

sludge.  On the other hand, the addition of NaOH provided a very clear

supernatant, at least equal to that obtained with lime alone, and a

relatively small  sludge volume.  The results, in terms of the removal
           ป
of heavy metals,  are presented  in Table 33,  None of the additives were

an improvement over lime alone, in terms of meeting the standards (see

Table 1).  In absolute terms,  neither additive,  alone or in combination,

provided improved treatment across the board.  In fact, in several  cases,


                                -119-

-------
                             TABLE 32

            EXPERIMENTAL PROTOCOL AND PRELIMINARY RESULTS
              IN EVALUATION OF LIME TREATMENT ADDITIVES*
Preliminary Results
Experimental Protocol
Flask
Number
1
2
3
4
5

Lime, ml
0
18
18
12
15
Additives
Aluminate, ml
0
0
10
0
10
Supernatap
NaOH, ml ClarityJ-
0
0 ++
0
0.1 ++
0.1 +
t Sludge
Volume, ml
0
110
120
80
120
      The experimental flasks each contained raw leachate plus the
additives as indicated.

     $The clarity of the supernatant was evaluated qualitatively as
follows: —, very turbid; -, turbid; +, clear; -H-, very clear.  Also
the volume of sludge (in ml) produced was noted.
                                -120-

-------
                              TABLE 33


                   RESULTS OF ADDITIVE  EVALUATION'
Experimental Protocol
Parameter
Analyzed
Iron
Copper
Cadmium
Chromium
Lead
Zinc
Nickel
Mercury
Raw
Leachate
Control
52.0
0.2k
0.0k
0.15
0.19
6.40
0.71
0.002
Lime
Only
0.45
0.14
0.06
0.09
0.16
0.03
0.56
0.002
Lime
+
Aluminate
1.21
0.16
0.04
0.07
0.19
0.12
0.52
0.002
Lime
+
NaOH
0.77
0.14
0.07
0.08
0.16
0.10
0.58
0.001
Lime + NaOH
-i-
Aluminate
1.71
0.24
0.03
0.09
0.15
0.08
0.53
0.002
     *
      All  units are mg per liter in the supernatant following
indicated  protocol.
                              -121-

-------
the removal decreased due to the additives.  The conclusion is that
neither NaOH nor aluminate, when added to lime and leachate at a final
pH of 11.7, results in improved removal of heavy metals.  However, the
addition of 125 ppm NaOH provides a significant reduction in sludge
volume, a fact which would improve the efficiency of the lime clarifier
operation.  Sludge handling and disposal will also be rendered more
efficient.  Therefore, the addition of NaOH was recommended.
     An added benefit  is a higher pH  in the ammonia lagoon which
increases the ammonia  stripping capacity of the lagoon.  This is
illustrated by the observation that the supernatant pH where NaOH was
added had fallen to only 11.5, whereas, in the flasks not receiving
NaOH, the supernatant  pH had dropped  to less  than  11.  This  is a
demonstration of the strong buffering properties attributable to  the
addition of NaOH.
     An additional  recommendation was made which  improved the efficiency
of the  lime clarifier. This was that the  depth of the  sludge blanket
be kept minimal.   This prevented solids carryover  which  occurred  on
occasion  in the  past.   Sludge  blanket depth  may be minimized by
sem?-continuous  withdrawal  of  settled sludge. An  automatic timing
mechanism was  installed to ensure  sludge  removal  semi-continuously on
a 2k hr basis.
      Effluent  Filtration.   The physical  process of filtration has
 several  potential  applications in  leachate treatment.   Filtration of
 activated sludge effluent  would be needed before  activated  carbon or
 ion exchange treatment, if these operations are necessary.   This  is
                                 -122-

-------
 required to reduce the solids loadings on these units,  and hence,  to



 minimize problems  due to clogging.




      A second  application of filtration would  be to use it between the




 ammonia stripping  lagoon and the  activated sludge process.   The  purpose




 of  doing this  is to remove residual  heavy metal  precipitates as  well  as



 calcium carbonate  which may form  during lagooning or which  may escape



 the lime clarifier unit.   The benefits which would accrue  include



 lower heavy metals loadings and solids loadings  on the  biological



 treatment units.




      The first step in  evaluating the  efficacy of filtration  is  a




 number  of small scale units.   The purpose of these would be  to provide




 a preliminary evaluation  of the removal  of heavy metals and  particulates




 from  the ammonia lagoon  effluent, and  to determine the compatability  of



 a filtered  effluent  from  the  activated  sludge process with granular



 activated carbon.




      The principal   design  parameters which must  be considered are  filter



 configuration; method of  flow control;  terminal  headless (ft of water);



 filtration  rate (gal/min-sq ft); filter media, sizes and depths;  and,



 backwashing requirements.  The basic filter configurations are upflow



 filtration  through a relatively deep, coarse filter medium; using a



 filtered water collection device within the filter medium and bringing



water in from both  the bottom and the top; dual or mixed media with



conventional downflow; and single medium downflow filtration.




     Preliminary testing of filtration has been conducted using filter



paper (Whatman No.  40).  The  results are presented  in Table 3k,  and




they show that additional removals of most of the heavy metals are



                                -123-

-------
IS)
-t-
                                                 TABLE  34

                                                                     j

                                      PRELIMINARY  FILTRATION. RESULTS'

Metal
1 ron
Copper
Cadmium
Chromium
Lead
Nickel
Mercury
Zinc
Unfil
256.
0.
0.
0.
0.
1.
0.
23.
Raw Leachate
tered Filtered

48
13
36
47
46
003
4
194
0
0
0
0
1
0
17
.
.40
.08
.36
.35
.29
.004
.2
Ammon
Unfil
32.
0.
0.
0.
0.
1.
0.
4.
ia Lagoon Effluent
tered Filtered

13
06
18
23
22'
003
63
18.
0.
0.
0.
0.
1.
0.
3.

20
08
15
19
08
0025
36
Final Effluent
Unfil tered Filtered
10
0
0
0
0
1
0
1
•
.26
.05
.12
.12
.06
.004
.28
5.6
0.25
0.04
0.09
0.10
0.88
0.0025
1.70.

             All  units are mg per liter.

-------
achieved by filtration at any intermediate step of the treatment




sequences.



     Effluent Breakpoint ChlorinatJon.  Although nitrification is an




extremely efficient process for the removal of ammonia-N, it is a



biological treatment method.  As a result, nitrification can be




temperamental and prone to upset.  As noted earlier, the nitrifying




organisms are very sensitive to temperature.  Also, the ammonia




stripping process is less effective during the winter.  Consequently,




the effluent in the colder months is very high in ammonia-N.  The




removal of residual ammonia in the final effluent may be accomplished




by breakpoint chlorination.  This procedure oxidizes the NH.-N to




gaseous end-products including N2-  Breakpoint chlorination has been



practiced at the GROWS treatment facility and is a proven technology



for meeting the effluent criterion for ammonia (see Table l).



     The chlorine demand, and therefore chlorination operating costs,



are expected to decrease as the degree of organics removal  increases,




and this savings can affect some portion of the carbon costs.  An



economic tradeoff can be made here vis-a-vis the costs of activated




carbon treatment of the final effluent.



     Leachate Treatment., Plant Startup.  The results of this study have




indicated clearly that Systems 1 and 2 are very effective in the



treatment of leachate.  The best results were obtained with a process



train consisting of raw leachate equalization, lime precipitation and




clarification, ammonia stripping lagoon, activated sludge providing



for carbonaceous and nitrogenous oxidation, sedimentation,  and




                                -125-

-------
effluent  chlorination.  As  noted  in  Tables  H,  15,  20 and  21,  the



removal efficiencies observed with Systems  1  and  2  are extremely high,



and provide  the  best opportunity  for meeting  the  effluent  criteria.



     Nevertheless,  in spite of  the success  of Systems 1 and  2, some



problems  were encountered during  startups.  Some  of these  have been



discussed elsewhere.  These include  the nutrient  deficiency  due to



phosphate precipitation in  the  lime  treatment unit;  and substrate



inhibition resulting from excessive  ammonia-N concentrations.



     Excessive loadings have been received  by the plant in terms of



ammonia,  metals, and organlcs.  For  the past  year,  raw leachate flows



have been in the 50,000 to  80,000 gpd range,  which  when coupled with




the extremely high  influent concentrations  yields process  loadings in



excess of those which had been experienced  during the initial



operational phase of the plant.  The combination of  excessive



concentrations and  loadings is the primary  factor inhibiting rapid



successful process  startup.



     There are a number of  secondary reasons  for  the poor startup



performance.  The first of  these  is that the  third year operational



effort was initiated in mid to late winter.  As a result,  the



development of a healthy activated sludge culture was inhibited due to



low temperatures, as well  as to the presence of high concentrations of



metals and organics.  These high concentrations resulted not only



because of the leachate strength, but also because of operational




problems.   Examples of these problems include instances where raw



leachate has by-passed the chemical/physical section, and  insufficient



air has been provided to the bio units.




                                -126-

-------
      Other secondary factors  underlying the startup problems include




 those relating  to plant operation  and  the behavior of the sedimentation




 tanks.   Frequent  problems  were encountered which  affected adversely the




 performance of  both  the chemical/physical  and  the secondary clarifiers.




 Eddy  currents were occasionally noted  in the secondary clarifier  and



 these resulted  in carry-over  of biological  solids.   The high rate of




 internal  recycle  in  the activated  sludge undoubtedly contributed  to this




 condition.   In  the case of the chemical/physical  sedimentation  tank,




 some  reduction  in clarification efficiency  was observed  as  a result of




 excessive  sludge  volumes.   Lime sludge  production  had  increased as  the



 acidity and  the rate of generation of 1eachateJncreased.




      Many  of the  problems  mentioned above are essentially different



 aspects of plant  reliability.   Each section  of the  plant needs a  high



 level of reliability.   This is  especially true of  the  lime  treatment



 section.   It is recommended that,  in the future, as  these plants  become



 more  sophisticated, greater consideration be given  to automatic process




 control techniques.  As an example consider  the lime slurry  system.  At



 the present time,  if this  system malfunctions such  that the  1ime  feed




 stops, then raw 1eachate passes to the ammonia lagoon and thence  to the



 bio units where pronounced  inhibitory effects are observed.  These



 problems could be avoided  by an automatic valving system which would




 recycle the lime clarifier effluent to the equalization pond whenever



 the effluent pH dropped below some set point such as pH 10 or pH  10.5.




 Process reliability is  important at plants such as this because they



normally operate at very close to inhibitory concentrations of



                                -127-

-------
 materials such as ammonia, metals, and organics.  Thus, small
..perturbations can result in process upset followed by an extensive
 period during which the discharge standards are contravened.
      Most of the startup problems, however, are related directly to the
 increased leachate flow and strength.  Higher flows have not diluted
 the leachate, but have been associated with increased concentrations of
 pollutants.  This, in turn, is partially a result of the introduction
 of industrial liquid wastes.
      An indication of the extremely high loadings is provided in Table
 35-  The data show process loadings applied to System 2 (lime treatment/
 ammonia lagoon) and to System 1  (activated sludge).  These data are for
 the period April 1, 1978 through June 30, 1978.  This represents the
 startup period for the most recent evaluation of Systems 1 and 2.  For
 comparison, the original design loadings, derived from Table 9, are
 included.  It can be seen that the more recent loadings are very high
 (except for the metals and suspended sol ids).  When this is coupled
 with the high concentrations of inhibitory substances, the basis for
 the poor startup performance becomes clear.
      The role of high concentrations relates to a previous discussion.
 As noted in Figure 9, the phenomenon of substrate inhibition must be
 considered.  Even substrates such as ammonia-N or BOD which are vital
 requirements at low concentrations become inhibitory at higher
 concentrations.  The significance for leachate treatment is that the
 system operates at influent concentrations high enough to be inhibitory.
 Slight perturbations result in process instabi1ity and/or poor startup.
                                 -128-

-------
                       TABLE 35

  PROCESS LOADING RATES AND CONCENTRATIONS  OBSERVED
DURING THE PERIOD APRIL 1,  1978 THROUGH JUNE 30,  1978.
Concentration
of Raw Leachate
mg/1 iter
Observed Design
Alkal inity, as CaCO,
Anwion i a-d
BOD
Cadmium
Calcium
Chloride
Chromium
COD
Copper
Dissolved Solids
Hardness, as CaC03
1 ron
Kjeldahl-N
Lead
Magnesium
Mercury
Nickel
Organic-N
Phosphates
Potassium
Sodium
Sulfate
Suspended Solids
Total Solids
Zinc
4915
990
13695
0.09
895
2965
0.24
19864
0.24
13225
3750
165
860
0.37
275
0.253
1.45
=0
1.98
1000
1455
515
1510
14735
11.44
1100
—
1500
—
—
800
--
ซ
1
--
800
600
—
—
—
—
—
100
—
__
__
300
1500
3000
10
Process Loading Rates
Ib per day
Raw Leachate
Observed
2710
546
7550
0.05
493
1634
0.13
10950
0.13
7290
2065
90
475
0.20
150
0.14
0.80

1.09
550
800
285
830

6.31
Design
1320
—
1800
--
	
960
—
—
1.20

960
720

.-
__
_ซ-•
__

__
__
__
360
.1800

12
Appl led to
Lime Treat-
ment Section
1730
350
4830
0.03
315
1045
0.08
7000
0.08
4660
1320
60
305
0.13
95
0.09
0.51

0.80
350
515
180
530

4.03
Appl led to
Activated
Sludge System
210
40
560
0.003
40
155
0.007
880
0.009
605
135
0.42
45
0.02
7.8
0.01
0.07

0.02
55
80
55
20

0.06

-------
Thus, the principal source of process startup problems are the



combination of excessive loadings and concentrations near the upset



threshhold.



     The activated sludge system has been operated most successfully at



a mixed liquor volatile suspended solids (MLVSS) concentration of



6000-12,000 mg per liter, depending upon the influent BOD concentration.



During periods of satisfactory operation the food to microorganism



ratio has been in the 0.12 to 0.32 day"  range, as calculated by
            where



                              L - F:M, Ib BOD per Ib MLVSS-day



                              0_ ~ Process hydraulic loading, gpd



                             C  = Process influent BOD, mg per liter



                              V = Aeration tank volume, gal



                             X  = MLVSS, mg per liter
                              3


The operation of the activated sludge units in series  (see Figure 7)



was accomplished by maintaining 12,000 mg MLVSS per liter and 6000 mg



MLVSS per liter, in the first and second stages, respectively.



     The following recommendations are made in an effort to reduce



future startup problems:



1.  Raw leachate equalization is a valuable aid in dampening peak



    concentrations of materials which are inhibitory to subsequent



    biological processes; and, in controlling chemical dosages in



    the lime precipitation units.



                                -130-

-------
 Ammonia  removal ,by stripping via  aeration reduces  concentrations
 to  a  level  below  the  point  of inhibition.
 Nutrient:  supplementation  may be necessary.   This  is  especially true
 for the case of a  high carbon waste with  biological  treatment  units
 preceded  by lime precipitation.
 Excessive loadings may necessitate treatment of only a portion  of
 the total waste flow during  startup.  At  these times, the balance
of the flow should be recycled to the landfill.
With a waste containing very  high levels of ammonia-N, startup
should occur during warm weather months.  Otherwise, nitrification
will be inhibited, and the resulting ammonia concentrations will not
permit biological  activity.
                           -131-

-------
                         VI.  CONCLUSIONS



The GROWS landfill leachate is characterized by high organic strength



and by large day-to-day variations.  Much of the variability can be



dampened by raw leachate equalization.



Considerable experience has been gained in the operation of activated



sludge units on raw leachate and on leachate which has received



chemical/physical treatment.  It has been concluded that this raw



leachate must be pre-treated in order to render it amenable to



activated sludge processing.  The results indicate that the raw



leachate inhibits the growth of the activated sludge microorganisms.



System 3 (biological treatment of raw leachate) has yielded an



average effluent BODj. of 763 mg per liter, a concentration which is



clearly not acceptable.



The operation of the chemical/physical units (System 2) has continued



for some time to gain experience under a wide variety of operating



conditions and sufficient data have been collected to provide an



evaluation of this method of treating raw leachate.  Lime treatment



alone  (System 2a) provides  removal efficiencies of approximately 50



percent of the organic matter, 75 percent of suspended solids,



one-half of mercury and cadmium, and at least two-thirds of the



other heavy metals.
                           f


The complete chemical/physical treatment sequence consisting of



lime precipitation/sedimentation/ammonia stripping (System 2b)



achieved the following removals of efficiency: kS to 69 percent



of the organic matter, ammonia-N and total kjeldahl-N; 70 percent



of the suspended solids, and 50 percent or better of the heavy



                            -132-

-------
    metals except copper, for which,fhe removal efficiency was 37.8.




    percent.  In the final year of  the project, System 2b was augmented




    with an initial raw leachate equalization step.




5.  Temperature and pH have an effect on the concentration of heavy




    metals in the lime treatment effluent.  However, the response is




    not identical for all heavy metals.  If the differences were more



    thoroughly characterized, it might be possible to use them in an



    operational  control procedure.




6.  Activated sludge treatment of the effluent from the chemical/physical




    units has been extremely successful (System 1).  It is apparent that




    the reduction in ammon'ia-N afforded by the ammonia stripping lagoon




    provides conditions suitable for the growth of activated sludge




    microorganisms.  The ammonia lagoon, in conjunction with the balance



    of System 2, provides ammonia removals of approximately 50 percent



    resulting in activated sludge influent concentrations of 1^9-423 mg




    NH-N/liter (95 percent confidence interval).  During the final



    year of the project, these concentrations increased somewhat due



    to the maintenance of a lower pH in the lagoon.  Under this condition,



    the activated sludge quickly adapted to the leachate with the result




    that effluent BOD- concentrations were consistently low except  during



    cold weather.  Nitrifying organisms  developed and produced a



    nitrified effluent with very low concentrations of ammonia.  Cold



    winter weather inhibited the biochemical  oxidation of organics  and



    ammon i a.




7.  Activated sludge treatment has been effective in both the series




    (Figure 7) and parallel (Figure 4) modes of operation.   That is, the



                               -133-

-------
    plant configuration has had the aeration tanks in either parallel
    or series, with approximately equal results.
8.  Overall, the treatment sequence consisting of chemical/physical
    (lime precipitation, sedimentation, ammonia stripping, and
    neutralization) followed by activated sludge (System 1) has
    produced an excellent final effluent with the following
    characteristics:
         a.  Organic matter has been reduced to 153 mg BODg/liter.
             This is a 99 percent removal.  The corresponding COD
             removal efficiency is 95 percent.  The effluent BOD
             co COD ratio is 0.16.
         b.  The effluent ammonia concentration is 75 mg/liter,
             representing 90 percent removal.
         c.  Heavy metals are found in the effluent at the following
             levels (percent removals are shown in parenthesis):
             0.017 mg cadmium/liter (78.2 percent); 0.07 mg
             chromium/liter (73.. 1 percent); 0.11 mg copper/liter
             (72.5 percent); 2.7 mg iron/liter  (99.2 percent);
             0.12 mg lead/liter (83.8 percent); 0.004 mg mercury/
             liter (27.4 percent); 0.75 mg nickel/liter (57.4
             percent); 0.53 mg zinc/liter (97.3 percent).
9.  The kinetics of nitrification have been followed during System Ib
    operation.  The rate of nitrification, expressed as the specific
    oxidation rate, follows the van't-Hoff Arrhenius relationship.
    The data indicate that the activation energy is approximately
                                -134-

-------
     12350 cal per mole, and that the Arrhenlus frequency factor is



     2.18 x 109 day"1.



10.  The data show that substrate Inhibition due to ammonium Ion



     concentration occurs in System Ib nitrification.  This relationship



     has been expressed as a Haldane inhibition model in which the maximum



     specific oxidation rate is 3.5 g N oxidized per g biomass day, K



     is k mg per liter, and Kj  is 36 mg per liter.



11.  Pilot scale operation of System 5 shows that activated carbon



     treatment of System 1 effluent is an effective way of removing



     much of the remaining organic matter.  In addition, considerable



     removals of heavy metals occur in the carbon columns.



12.  A bench-scale evaluation of alkaline chlorlnation as a means of



     improving heavy metals removal  Indicated that  the technique is not



     appropriate.



13-  Studies of lime sludge compaction and supernatant clarity reveal



     that the addition of sodium hydroxide (125 mg  per liter)  improves



     both factors.   The addition of a commercial aluminate preparation



     (300 mg per liter) had an  adverse impact on compaction and clarity.



14.  Preliminary testing of filtration shows that additional  removals



     of most heavy metals can be achieved.



15.  Full-scale studies of breakpoint chlorlnation  have shown  that this



     method can be used to attain the ammonia effluent criterion.
                                 -135-

-------
                        VII.  REFERENCES
 1.
2.
3.
4.
5.
6.
7.
8.
'9.
Remson,  I., A. A. Fungaroli,  and A. W.  Lawrence.   Water move-
     ment  in  an unsaturated  sanitary  landfill.  Journal of
     the Sanitary Engineering Division,  Proceedings  of  the
     American Society of Civil  Engineers,  94(SA2):307-317,
     Apr.  1968.

Calvert, C. K.  Contamination of ground  water by  impounded
     garbage  waste.  American Water Works  Association
     Journal, 24:266-270,  1932.

Carpenter, L. V., and L. R.  Setter.   Some  notes on sanitary
     land-fills.  American Journal of Public Health  and the
     Nation's Health, 30(1):385-393,  Apr.  1940.

Merz, R. C.,  and R. Stone.   Factors controlling utilization
     of  sanitary landfill  site; final report to Department of
     Health,  Education, and  Welfare,  National Institutes  of
     Health,  United States Public Health Service,  May 1,  1960
     to May 31, 1963.  Los Angeles, University of  Southern
     California, [1963].   126 p.

Longwell, J.  The water pollution aspect of refuse disposal.
     Paper no. 6261.  Institution of  Civil Engineers
     Proceedings, 8:420-424,  1957.

Great Britain Ministry of  Housing and Local Government.   Pol-
     lution of water by tipped  refuse; report of the
     Technical Committee on  the Experimental Disposal of
     House Refuse in Wet and Dry Pits.   London, Her Majesty's
     Stationery Office, 1961.   141 p.

Qasim, S. R.  Chemical characteristics of  seepage  water  from
     simulated landfills.  PhD  Dissertation, West  Virginia
     University, Morgantown, W. Va.,  1965.  145 p.

Anderson, J. R., and J. N. Dornbush.  Influence of sanitary
     land,fill on ground water quality.   American Water Works
     Association Journal, 59(4):457-470, 1967.

Hughes, G. M., R. A. Landon, and R. N. Farvolden.  Summary
     of findings on solid waste disposal sites in
     northeastern Illinois.  Environmental Geology Notes no.
     45.  Urbana, 111., Illinois State Geological Survey,
     Apr. 1971.  25 p.
                           -136-

-------
10".  Steiner, R. L.  Chemical and hydraulic characteristics of
          milled refuse.  PhD Dissertation, Drexel University
          College of Science, Philadelphia, Pa., 1973.  319 p.

11.  Weaver, L.  Refuse disposal, its significance.  In Ground
          Water Contamination; Proceedings of the 1961 Symposium,
          Cincinnati, Apr. 5-7, 1961.  Robert A. Taft Sanitary
          Engineering Center Technical Report W61-5.  Cincinnati,
          U.S. Department of Health, Education, and Welfare,
          Public Health Service, Bureau of State Services,
          Division of Water Supply and Pollution Control.
          p.104-110.

12.  [Engineering-Science, Inc.] -In-situ investigation of
          movements of gases produced from decomposing refuse;
          final report.  Publication no. 35.  Sacramento, Calif.,
          State Water Quality Control Board, 1967.  116 p.

13.  Walker, W. H.  Illinois ground water pollution.  American
          Water Works Association Journal, 61(1):31-40, Jan. 1969,
14.
15.
Roessler, B.  Beeinflussung des Grundwassers durch Mull- und
     Schuttablagerungen.  Vom Wasser, 18:43-60, 1950/51.
Chian, E. S. K., and F. B. DeWalle.
     leachates and their treatment.
                 Sanitary landfill
                 Journal of the
          Environmental Engineering Division, Proceedings of the
          American Society of Civil Engineers, 102(EE2):411-431,
          Apr. 1976.
16.
17,
Remson, I., et al.
     landfill.
Water movement in an unsaturated sanitary
18,
19,
Dass, P., G. R. Tamke, and C. M. Stoffel.  Leachate production
     at sanitary landfill sites.  Journal of the Environmental
     Engineering Division, Proceedings of the American
     Society of Civil Engineers, 103 ( EE6 ): 9-81-988, Dec. 1977.

Calvert, C. K.  Contamination of ground water by impounded
     garbage waste.

Carpenter, L. V., and L. R. Setter.  Some notes on sanitary
     land-fills.
20.  Merz, R. C., and R. Stone.  Factors controlling utilization of
          sanitary landfill site.

21.  Longwell, J.  The water pollution aspect of refuse disposal.

22.  Great Britain Ministry of Housing and Local Government.  Pol-
          lution of water by tipped refuse.
                         -137-

-------
 23.


 24.



 25.



 26.



 27.


 28.


 29.



 30.


 31.

 32.


 33.

 34.


 35.


36.


37.
 Qasim, S. R.  Chemical characteristics of seepage water from
      simulated landfills.

 Steiner, R. L.  Chemical and hydraulic characteristics of
      milled refuse.  Adapted from fable 18 on p.55-60 et
      passim.

 Chian, E. S. K.,  and F. B. DeWalle.  Sanitary landfill
      leachates and their treatment.  Adapted from data in
      text.

 Great Britain Ministry of Housing and Local Government.   Pol-
      lution of water by tipped refuse.  Adapted from data on
      p.116, 128.

 Anderson, J. R.,  and J. N. Dornbush.   Influence of sanitary
      landfill on  ground.water'quality.

 Hughes,  G.  M.,  et al.   Summary of findings on solid waste
      disposal sites in northeastern Illinois.

 Qasim, S.  R.   Chemical characteristics of  seepage  water  from
      simulated  landfills-.   Adapted from Table 14 on p.81  et
      passim.

 Qasim, S.  R.   Chemical characteristics of  seepage  water  from
      simulated  landfills.

 Weaver,  L. ,  Refuse  disposal,  its  significance.

 [Engineering-Science,  Inc.]   In-situ  investigation of
     movements of gases produced  from decomposing  refuse.

Walker, W. H.  Illinois ground water  pollution.

Roessler, B.  Beeinflussung des Grundwassers  durch Mull- und
     Schuttablagerungen.

Chian, E. S. K.,  and F. B. DeWalle.   Sanitary landfill
     leachates and their treatment.

Dass, P., et al.  Leachate production  at sanitary landfill
     sites.

Schoenberger, R. J., et al.  Treatability of leachate from
     sanitary landfills.  In Proceedings; Fourth Mid-Atlantic
     Industrial Waste Conference, Newark, Nov. 18-20, 1970.
     University of Delaware,  p.411-421.
                           -138-

-------
 38.
39.
40.
41.
42.
43.
44.
45.
46.
Weston, R. F.,  and R. A. Morrell.   Treatment  of water  and
     wastewater for  removal  of heavy metals,   In  J.  A.
     Borehardt  et al.,  eds.   Viruses and  trace contaminants
     in water and wastewater.  Ann  Arbor, M-ich.,  Ann Arbor
     Science, 1977.  p.167-182.

Chian, E. S. K., and F. B. DeWalle.  Treatment of high
     strength acidic wastewater with a  completely mixed
     anaerobic  filter.  Water Research, 11(3):295-304, 1977.
Chian, E. S. K., and F. B. DeWalle.
     leachates  and their treatment.
                                           Sanitary  landfill
Weston, R. F., and R. A. Morrell.  Treatment of water and
     wastewater for removal of heavy metals.  Adapted from
     Table 3 on p.171.
Chian, E. S. K.
     leachates.
Stability of organic matter in landfill
Water Research, ll(12):225-232, 1977.
Chian, E. S. K., and F. B. DeWalle.
     leachates and their treatment.

Chian, E. S. K., and F. B. DeWalle.
     leachates and their treatment.
                    Sanitary landfill
                    Sanitary landfill
Uloth, Vป C., and D. S. Mavinic.  Aerobic bio-treatment of
     a high-strength leachate.  Journal of the Environmental
     Engineering Division, Proceedings of the American
     Society of Civil Engineers, 103(EE4);647-661, Aug. 1977.

Lawrence,, A. W., and P. L. McCarty.  Unified basis for
     biological treatment design and operation.  Journal of
     the Sanitary Engineering Division, Proceedings of the
     American Society of Civil Engineers, 96(SA3):757-778,
     June 1970. '
47.
48.
49.
Chian, E. S. K., and F. B. DeWalle.
     leachates and their treatment.
     p.427.
                                          Sanitary landfill
                                          Adapted from table on.
Personal communication.  Trenton, N.J. Weather Bureau to
     R. L. Steiner, Applied Technology Associates, 1976.
     Data calculated for 30-year period ending Dec. 1975.

Industrial waste permit no. 0974203.  Issued by Commonwealth
     of Pennsylvania, Department of Environmental Resources
     to Geological Reclamation Operations and Waste Systems,
     Inc., 1975.  [3 p.]
                          -139-

-------
 50.
 51.
Remson, I., et al.
     landfill.
Water movement in an unsaturated sanitary
 52.
 53.
 54.
,55.
 56.
 57.


 58.



 59.



 60.
Standard methods for the examination of water and wastewater.
     14th ed.  Prepared for and published jointly by American
     Public Health Association, American Water Works
     Association, and Water Pollution Control Federation.
     Washington, American Public Health Association, 1976.
     1193 p.

Annual book of ASTM standards; Part 23:  Water, atmospheric
     analysis.  Philadelphia, American Society for Testing
     and Materials, 1973.  1108 p.

Manual of methods for chemical analysis of water and wastes.
     [2d ed*]  Environmental Protection Publication
     625/6-74/003.  Cincinnati, U.S. Environmental Protection
     Agency, Office of Technology Transfer, 1974.  298 p.

Steiner, R. L., and A. A. Fungaroli, eds.  Analytical pro-
     cedures for chemical pollutants; research project on
     pollution of subsurface water by sanitary landfill.
     Series 1, no. 8.  Philadelphia, Drexel Institute of
     Technology, June 1968.  27 p.

Weston, R. F., and R. A. Morrell.  Treatment of water and
     wastewater for removal of heavy metals.

[Brown and Caldwell, Inc.]  Process design manual for
     nitrogen control.  Cincinnati, U.S. Environmental
     Protection Agency, Office of Technology Transfer, Oct.
     1975.  [423 p.]  (Distributed by the National
     Technical Information Service, Springfield, Va.,-as
     PB-259 149/3BA.)

[Brown and Caldwell, Inc.]- Process design manual for
     nitrogen control.

Prakasam, T. B. S., and R. C. Loehr.  Microbial nitrification
     and denitrification in concentrated wastes.  Water
     Research, 6(7):859-869, July 1972.

Button, W. C., and S. A. LaRocca.  Biological treatment
     of concentrated ammonia wastewaters.  Water Pollution
     Control Federation Journal, 47(5):989-997, May 1975.

Adams, C. E., Jr., and W. W. Eckenfelder, Jr.  Nitrification
     design approach for high strength ammonia wastewaters.
     Water Pollution Control Federation Journal,
     49(3):413-421, Mar. 1977.
                              -140-

-------
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
Verstraete, W., H. Vanstaen,  and  J.  P.  Voets.   Adaptation
      to nitrification of activated sludge  systems  treating
      highly nitrogenous waters.   Water  Pollution Control
      Federation Journal, 49(7):1604-1608,  July  1977.

Boon, B., and H. Laudelout.   Kinetics of nitrite oxidation  by
      Nitrobacter winogradskyf:.  Biochemical Journal,
      85:440-447, 1962.
 [Brown and Caldwell,  Inc.]
     nitrogen control.
Process design manual for
Eylar, 0. R., Jr., and E. L. Schmidt.  A  survey of hetero-
     trophic microorganisms from soil  for ability to  form
     nitrite and nitrate.  Jour-nal of  General Microbiology,
     20:473-481, 1959.

Hirsch, P., L. Overrein, and M. Alexander.  Formation of
     nitrite and nitrate by actinomycetes and fungi.  Journal
     of Bacteriology, 82(3):442-448, Sept.  1961.

Brezonik, P. L.  The dynamics of, the nitrogen cycle in natural
     waters.  PhD Dissertation, University  of Wisconsin,
     Madison, Wis., 1968.  265 p.

Wild, H. E., Jr., C. N. Sawyer, and T. C. McMahon.  Factors
     affecting nitrification kinetics.  Water Pollution
     Control Federation Journal, 43(9):1845-1854, Sept. 1971.,
[Brown and Caldwell, Inc.]
     nitrogen control.

[Brown and Caldwell, Inc.]
     nitrogen control.
Process design manual for
Process design manual for
73.
Wong-Chon, G. M., and R. C. Loehr.  The kinetics of microbial
     nitrification.  Water Research, 9:1099-1106, 1975.

Button, P. M., et al.  Efficacy of biological nitrification.
     Water Pollution Control' Federation Journal,
     47(11):2665-2673, Nov. 1975.

Stover, E. L., and D. F. Kincannon.  One- versus two-stage
     nitrification in the activated sludge process.  Water
     Pollution Control Federation Journal. 48(4):645-651,
     Apr. 1976.

Button, W. C., and S. A. LaRocca.  Biological treatment of
     concentrated ammonia wastewaters.

-------
74.  Wild, H. E., Jr., et al.  Factors affecting nitrification
          kinetics.

75.  Button, P. M., et al.  Efficacy of biological nitrification.

76.  [Brown and Caldwell, Inc.]  Process design manual for
          nitrogen control.

77.  Haldane, J. B. S.  Enzymes.  London, Longmans, Green and Co.,
          1930.  235 p.  (Monographs on biochemistry.)

78.  [Brown and Caldwell, Inc.]  Process design manual for
          nitrogen control.
                                     s
79.  Focht, D. D., and A. C. Chang.  Nitrification and denitrifica-
          tion processes related to waste water treatment.
          Advances in Applied Microbiology, 19:153-186, 1975.

80.  Poduska, R. A., and J. F. Andrews.  Dynamics of nitrification
          in the activated sludge process.  Water Pollution
          Control Federation Journal, 47(11):2599-2619, Nov. 1975.

81.  Focht, D. D., and A. C. Chang.  Nitrification and denitrifica-
          tion processes related to waste water treatment.

82.  Wild, H. E., Jr., et al.  Factors affecting nitrification
          kinetics.

83.  Focht, D. D., and A. C. Chang.  Nitrification and denitrifica-
          tion processes related to waste water treatment.

84.  Prakasam, T. B. S., and R. C. Loehr.  Microbial nitrification
          and denitrification  in concentrated wastes.

85.  Boon, B., and H. Laudelout.  Kinetics of nitrite oxidation by
          Nitrobacter winogradskyi.

86.  Boullanger, E., and L. Massol.  Etudes sur les microbes nitri-
          ficateurs.  Annales  de 1'Institut Pasteur  [Paris],
          18:181-196, T904~:[

87.  Aleem, M. I. H.  The physiology and chemoautotrophic metabo-
          lism of Nitrobacter  agilis.  PhD Dissertation, Cornell
          University, Ithaca,  N.Y., 1959.  125 p.

88.  Anthonisen, A. C.,  et al.  Inhibition of nitrification by
          ammonia and nitrous  acid.  Water Pollution Control
          Federation Journal,  48(5):835-852, May 1976.
                             -142-

-------
89.  Verstraefce, W., et al.  Adaptation to nitrification of acti-
          vated sludge systems  treating highly nitrogenous waters.

90.  Kholdebarin, B., and J. J. Oertli.  Effect of pH and ammonia
          on the rate of nitrification of surface water.  Water
          Pollution Control Federation Journal, 49(7):1688-1692,
          July 1977.

91.  Kholdebarin, B., and J. J. Oertli'.  Effect of pH and ammonia
          on the rate of nitrification of surface water.

92.  Adams, C,. E., Jr., and W. W.  Eckenfelder, Jr.  Nitrification
          design approach for high strength ammonia wastewaters.

93.  Poduska, R. A., and J. F. Andrews.  Dynamics of nitrification
          in the activated sludge  process.

94.  Chian, Eซ S. K., and F. B. DeWalle.   Sanitary landfill
          leachates and their treatment.

95. .Chian, Eซ S. K., and F. B. DeWalle.   Sanitary landfill leach-
          ates and their treatment.

96.  LaGrega, M. D., and J. D. Keenan.   Effects of equalizing
          wastewater flows.  Water Pollution Control Federation
          Journal, 46(1);123-132,  Jan.  1974:

97.  Dass, P., et al.  Leachate production  at sanitary landfill
          sites.
                                                              pal 764
                                                              SW-758
               ftU.S.GOVWKtซtIfซllniHSOfFK!t:W7l -626-301/1541
                                -143-

-------

-------
                           EPA  REGIONS
U.S. EPA. Region 1
Solid Waste Program
John F. Kennedy Bldg.
Boston, MA 02203
617-223-5775

U.S. EPA, Region 2
Solid Waste Section
26 Federal Plaza
New York, NY 10007
212-264-0503

U.S. EPA, Region 3
Solid Waste Program
6th and Walnut Sts.
Philadelphia, PA 19106
215-597-9377

U.S. EPA, Region 4
Solid Waste Program
345 Courtland St., N.E.
Altanta, GA 30308
404-881-3016
U.S. EPA, Region 5
Solid Waste Program
230 South Dearborn St.
Chicago, IL 60604
312-353-2197

U.S. EPA, Region 6
Solid Waste. Section
1201 Elm St.
Dallas, TX 75270
214-767-2734

U.S. EPA, Region 7
Solid Waste Section
1735 Baltimore Ave.
Kansas City, MO 64108
816-374-3307
U.S. EPA, Region 8
Solid Waste Section
1860 Lincoln St.
Denver, CO 80295
303-837-2221

U.S. EPA, Region 9
Solid Waste Program
215 Fremont St.
San Francisco, CA 94105
415-556-4606

U.S. EPA, Region 10
Solid Waste Program
1200 6th Ave.
Seattle, WA 98101
206-442-1260

-------
A
                              *
    o
            •
                               \

-------