EPA-600/2-76-213C
October 1976
Environmental Protection Technology Series
STATE-OF-THE-ART:
MILITARY EXPLOSIVES AND
PROPELLANTS PRODUCTION INDUSTRY
Vol. Ill - Wastewater Treatment
industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-76-213c
October 1976
STATE-OF-THE-ART: MILITARY EXPLOSIVES
AND PROPELLANTS PRODUCTION INDUSTRY
VOLUME III WASTEWATER TREATMENT
By
James Patterson, Norman I. Shapira, John Brown
William Duckert, Jack Poison
American Defense Preparedness Association
Washington, DC 20005
Project Officer
Richard Tabakin
Industrial Environmental Research Laboratory - Cincinnati
Edison, New Jersey 08817
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental
Research Laboratory, Cincinnati, of the U. S. Environmental Protection
Agency, and approved for publication. Approval does not signify that
the contents necessarily reflect the views and policies of the U. S.
Environmental Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
11
-------
FOREWORD
When energy and material resources are extracted, processed,
converted, and used, the related pollutional impacts on our
environment and even on our health often require that new and
increasingly more efficient pollution control methods be used.
The Industrial Environmental Research Laboratory - Cincinnati
(lERL-Ci) assists in developing and demonstrating new and improved
methodologies that will meet these needs both efficiently and
economically.
This project, "State-of-the-Art: Military Explosives and
Propellants Production Industry", was undertaken as part of
Environmental Protection Agency's Miscellaneous Chemical Industries
program to establish a baseline of information concerning the
military explosives industry, the magnitude of its waste problems,
and the -adequacy of the industry's treatment technology. The
results of the study have indicated that many of the wastes do
present significant problems of toxicity and/or resistance to
treatment, in addition to problems unique to explosives. Although
some treatment systems in use do protect the nation's waterways
from contamination, others are inadequate, generate secondary air
or solid waste problems, or are not widely used due to budgetary
limitations. Further research effort is needed by EPA and/or
Department of Defense to control pollutants generated by certain
sectors of the industry. The data and results of the investigation
have been used extensively by EPA's Office of Water Programs in
developing standards for the explosives industry. It will also
allow engineering staffs at several commercial military manufac-
turing facilities to examine their wastes and compare control
technology with that being used or developed at other installations.
Finally, it will enable EPA to determine our own research efforts
in this industry and how they would relate to other programs.
Questions or requests for additional information should be directed
to the Industrial Environmental Research Laboratory - Cincinnati,
Field Station - Edison, New Jersey.
David G. Stephan
Director
Industrial Environmental Research Laboratory
111
-------
ABSTRACT
This study, contained in three volumes, addresses the wastewater
effluents of the military explosives and propellants production
industry. Both manufacturing and LAP (Load, Assemble, and Pack) ac-
tivities are covered. Volume I describes the industry, as well as
the production processes and technology. Volume II details the
wastewater effluents of manufacture and LAP operations by product,
process, and military installation, to the extent that data are
available. Volume III describes and evaluates the effectiveness of
various treatment technologies for water pollution abatement now in
use or under investigation by product, process, and military in-
stallation.
A comprehensive long-term effort has been underway by the
Department of Defense for a number of years for the purpose of
modernizing munitions production plants. Pollution abatement is an
integral part of the modernization program. Although extensive study,
research and development investigations have been undertaken, and
although significant water pollution abatement and water management
plans have been developed, implementation is generally in only the
initial stages at selected military facilities. Major Government
emphasis and very substantial funding are essential to: the contin-
uation of necessary pollution abatement research and development; the
demonstration of promising new treatment technologies; and the imple-
mentation of effective and economical treatment system construction
programs. Recommendations are set forth in detail in Volume I.
iv
-------
VOLUME III
CHAPTER VI
WASTEWATER TREATMENT
The reader of this report is advised that it consists of six
chapters, contained in three volumes, each addressing separate aspects
of the explosives and propellants wastewater effluents and treatment
situation, and that duplication and repetition among these chapters has
been kept to a minimum. Thus, the reader is cautioned that the use or
interpretation of statement or evaluations taken out of context from
the study in its entirety could lead to serious misunderstandings and
incorrect assessments.
-------
VOLUME III
TABLE OF CONTENTS
CHAPTER VI - WASTEWATER TREATMENT
Page
SECTION I - Introduction 1
SECTION II - Trinitrotoluene 3
IIA - Red Water 3
IIB - Pink Water/LAP 13
SECTION III - Acids 35
IIIA - Acetic Acid and Acetic Anhydride 35
IIIB - Nitric Acid Manufacture and Concentration 45
IIIC - Sulfuric Acid Manufacture, Concentration
and Recovery. 62
SECTION IV - Nitrocellulose 69
SECTION V - RDX and HMX 83
SECTION VI - Nitroglycerin 103
SECTION VII - Sellite 114
SECTION VIII - Propellants 120
VIIIA - Solvent Propellants 120
VIIIB - Solventless Propellants 140
SECTION IX - Miscellaneous LAP Activities 146
IXA - Cast Propellants 146
IXB - Pressed Explosives 149
IXC - Plastic Bonded Explosive 152
APPENDIX I - References 153
-vii-
-------
LIST OF FIGURES
Page
Figure 1 - Acidification of Red Water 8
Figure 2 - Recycle of Red Water Wastes 9
Figure 3 - Solubility of TNT in Water 16
Figure 4 - Pink Water Treatment for Loading, Assembling, and
Packing Operations 18
Figure 5 - Relationship of Color and Carbon Adsorptive
Capacity for TNT 19
Figure 6 - Results from Multiple Regeneration Experiments 29
Figure 7 - Acetic Acid and Acetic Anhydride Manufacture 36
Figure 8 - Solubility of Calcium Sulfate, K = 2.4 x 10~5
sp
Solubility of Barium Sulfate, K - 1.1 x 10~10 65
Figure 9 - Current NC Water Use (gpd) at Radford AAP - One Line
NC Capacity; 144,000 Ib/day (Pulp), or 120,000 Ib/day
(Linters) 70
Figure 10 - Water Balance for Continuous NC Line 73
Figure 11 - Boiling Tub Pit Water Nitrocellulose Fines 74
Figure 12 - Proposed NC Water Balance 77
Figure 13 - Competitive Adsorption of RDX and TNT Using
Filtrasorb 400 at Keyport Test Site 95
Figure 14 - Sellite Manufacture at Joliet AAP 115
-V111-
-------
LIST OF TABLES
Page
Table 1 - Plants With Pink Water Waste 14
Table 2 - Reported TNT Concentrations in Pink Wastewaters, mg/1. 15
Table 3 - Performance of LAP Wastewater Treatment System at
Joliet AAP 20
Table 4 - Carbon System Effluent at Radford AAP 22
Table 5 - Comparison of Red Water Incinerator Condensate and
LAP Pink Water at Joliet AAP 25
Table 6 - LAP Wastewater Character for NAD Crane Rockeye
Facility 27
Table 7 - Summary of Results: TNT Combination Wastes
Treatability Runs - Color Removal 31
Table 8 - Wastewater Volumes from Acetic Acid Recovery and
Acetic Anhydride Manufacture 39
Table 9 - Combined Wastewater Characteristics from Acetic Acid
Recovery and Acetic Anhydride Manufacture 40
Table 10 - Biodegradability of Acetic Acid Recovery Decanter
Sweetwater 44
Table 11 - Holston AAP Wastewater Volumes 47
Table 12 - Wastewater Characteristics of Holston AAP Magnesium
Nitrate-Based NAG Process 48
Table 13 - Comparison of Characteristics of Combined Wastewaters
from AOP plus NAG Facilities 50
Table 14 - Current Treatment Practices at AOP/NAG Facilities 51
Table 15 - Proposed Modifications to Current Treatment 53
Table 16 - Ammonia Bionitrification Processes and Performance
Data 56
Table 17 - Summary of Nitrate Treatment Methods 57
Table 18 - Waste Characteristics of Sulfuric Acid Manufacture
and Concentration 63
-ix-
-------
Page
Table 19 - Wastewater Volumes from NC Production at Radford AAP.. 72
Table 20 - Wastewater Characteristics of Radford AAP NC
Processes ............................................. '-*
Table 21 - Performance Data on Radford AAP Facility for NC and
Acid Area Wastes ...................................... ^9
Table 22 - Reactants for RDX and HMX Manufacture ................. 84
Table 23 - Selected Solubilities of RDX and HMX .................. 86
Table 24 - Pollutant Discharge from RDX Manufacture at Holston
AAP [[[ 87
Table 25 - Pollutant Discharge from HMX Manufacture at Holston
AAP [[[ 88
Table 26 - Discharges to Catch Basins at Holston AAP ............. 89
Table 27 - RDX and HMX Wastewaters Resulting from LAP Activities. 91
Table 28 - Performance of Joliet AAP Treatment System..... ....... 93
Table 29 - Performance of Individual Units of Joliet AAP
Treatment System ........... . .......................... 94
Table 30 - Treatment for TNT and RDX by NAD Hawthorne Pilot
Plant .................... . ............................ 99
Table 31 - Pilot Treatment Plant Efficiencies, NAD Hawthorne ..... 100
Table 32 - Wastewater Characteristics of Batch Nitroglycerin
Production .................. ..... ..................... 104
Table 33 - Combined Wastewater Characteristics of Radford AAP
Continuous NG Nitration and Spent Acid Buildings
9463 and 9466 ................................. . ....... 106
Table 34 - Wastewater Characteristics of Radford AAP NG Store
Houses 9471 and 9472 .................................. 107
Table 35 - Treatment Chemicals Cost for Nitroglycerin
Manufacturing Wastewater Treatment .................... 112
Table 36 - Sellite Wastewater Characteristics Before Treatment... 116
Table 37 - Sellite Wastewater Treatment System Effluent
-------
Page
Table 38 - Sellite Lagoon Effluent Characteristics 119
Table 39 - Comparison of Lagoon Influent and Effluent Sellite
Wastewater Characteristics 119
Table 40 - Wastewater Volumes for Solvent Propellant Manufacture
at Radford AAP 123
Table 41 - Wastewater Characteristics of Single Base Propellant
Manufacture at Radford AAP 124
Table 42 - Wastewater Characteristics of Multi-Base Propellant
Manufacture at Radford AAP 125
Table 43 - Characteristics of Water-Dry Waste from Radford AAP... 126
Table 44 - Characteristics of Propellant Conveyance Wastewater
at Radford AAP 126
Table 45 - Characteristics of Solvent Recovery Still Bottoms at
Radford and Badger AAP's 128
Table 46 - Process Water Use in Ball Powder Manufacture 129
Table 47 - Solvent Levels in Selected Ball Powder Process
Effluents 130
Table 48 - Smokeless Powder Wastewater Treatment Plant Operation
and Average Waste Concentration 133
Table 49 - Results of Reverse Osmosis Treatment of Solvent
Wastewater 135
Table 50 - Ozone Oxidation Treatment of Solvent Wastewater 135
Table 51 - Treatment Results for Dissolved Organic Propellant
Ingredients 137
Table 52 - Wastewater Characteristics from Solventless
Propellant Manufacture at Radford AAP 141
Table 53 - Badger AAP Rocket Area Water Usage 143
Table 54 - Pilot Plant Treatment of Ammonium Picrate 151
-xi-
-------
CHAPTER VI
SECTION I - INTRODUCTION
1. This chapter is devoted to: an evaluation of current treat-
ment technologies applied to manufacturing and LAP wastewaters associ-
ated with military explosives and propellants; an assessment of the
effectiveness of current methods of treatment as well as those under
development; and an identification of alternate treatment procedures
which might: (1) be applicable where none are now in use; (2) be more
effective than those now in use; or (3) provide the opportunity for
recovery of materials currently discharged without treatment or lost
through treatment.
a. The conclusions and recommendations resulting from the
evaluations made in this Chapter are summarized in Chapter I, Volume I.
b. The chapter is divided into nine principal sections as
follows:
I. Introduction
II. Trinitrotoluene
IIA. Red Water
IIB. Pink Water/LAP
III. Acids
IIIA. Acetic Acid and Acetic Anhydride
IIIB. Nitric Acid Manufacture and Concentration
IIIC. Sulfuric Acid Manufacture, Concentration and
Recovery
IV. Nitrocellulose
V. RDX and HMX
VI. Nitroglycerin
VII. Sellite
VIII. Propellants
IX. Miscellaneous LAP Activities
IXA. Cast Propellants
1KB. Cast Explosives
IXC. PBX
c. Each Section briefly describes pertinent product manu-
facturing or handling procedures, and, to the extent that information
is available, identifies significant waste sources, and wastewater
volumes and characteristics. This is followed by a discussion of water
management techniques, where significant reductions in waste volume, or
product recovery by process modifications appear feasible. Thereafter,
each section describes and evaluates current wastewater treatment prac-
tices and modifications to current treatment practices under development.
The next major sub-section is devoted to an evaluation of the applica-
bility of alternative treatment technologies, beyond those currently in
use or proposed. Where current or projected air pollution abatement
-1-
-------
procedures are expected to Impact on wastewater volumes or character-
istics, the impact of air pollution control is also discussed. Finally.
each section closes with a brief summary.
d. The conclusions and recommendations resulting from the
evaluations of treatment technologies currently in use or under de-
velopment are presented in Chapter I, Volume I.
e. The contents of the preceding two volumes are as follows:
(1) Volume I. Introduction to The Military Explosives
and Propellants Industry. Executive Summary,
Conclusions and Recommendations.
(2) Volume II. Characterization of The Wastewater
Effluents of The Military Explosives and Propellants
Industry, including Detailed Data Tables.
f. The evaluations of treatment technology presented in
Chapter VI (Volume III) , as well as the conclusions presented in Chapter
I (Volume I) are based upon a wide range of data, data collection meth-
ods, data collection conditions and analytical procedures, as well as a
wide variety of data sources, many of which are incomplete, and some of
which are conflicting. As a result, the conclusions and recommendations
assessing the effectiveness of treatment technologies represent the best
professional judgments of the authors, within the constraints and lim-
itations of available factual data. In some instances, lack of data in
published form on current investigations have precluded incorporation
of an adequate assessment of these investigations. Data sources are
cited in detail throughout Chapter VI, and are listed in Appendix 1,
"References."
g. The evaluations presented in Vol. Ill are based upon the
most recent information available at the time of final editing of this
Report. Data in Vol. Ill may be in conflict with data presented in
Vol. II, which are based primarily on AEHA monitoring and survey studies,
some of which are several years old.
-2-
-------
SECTION II - TRINITROTOLUENE
2. Introduction
There are three basic wastewaters associated with the manufacture
and loading of TNT into munitions. These are: spent acids and acid
washes; red water; and pink water. Red water results from the sellite
purification step in TNT manufacture. Acidic wastes result from the
toluene nitration steps. Pink water results from TNT contact with plant
clean-up and scrubber water during manufacturing and LAP operations, and
as a condensate from red water evaporative concentration and incineration.
Pollution abatement for acidic wastes will be discussed separately
in Section III of this chapter, since acid wastes are characteristic of
many products, including TNT, nitroglycerin and nitrocellulose.
SECTION II-A - RED WATER
3. Waste Sources
a. The purification process in TNT manufacture entails the
use of sellite, a concentrated (16%) solution of sodium sulfite (Na^SO-,).
Crude TNT is washed with sellite and the unwanted isomers of TNT react
with it, leavinga-TNT. The sellite solution, together with subsequent
rinse waters, constitutes the red water. Currently, some TNT manufac-
turing plants have a market for this red water in the paper industry.
However, the market is only a short-term solution to the disposal of red
water, since the paper industry through process modifications will even-
tually negate the need for red water. Thus, the Army must become inde-
pendent of an outside market, insofar as the disposal of red water is
concerned.
b. Several approaches have been investigated to treat the red
water generated in the sellite TNT purification step:
(1) Evaporative concentration and/or incineration of
red water.
(2) Regeneration of sellite precursors from red water
liquor or ash by either the fluidized bed reduction
technique or "Tampella" process.
(3) Carbon adsorption to remove the organic constituents
of the red water.
(4) Flocculation and filtration to remove the color
material by proper coagulants and coagulant aids.
(5) Acidification of the red water to recover the nitro-
bodies for reuse in making TNT.
-3-
-------
Approach 1 above is currently used to concentrate red water prior to sale
to the paper industry and, where there is no market, to convert the red
water to dry ash for land disposal. Red water incineration has been
practiced at Joliet, Radford and Volunteer AAP's. Newport AAP, now
undergoing reconstruction for modernization, will also utilize red water
incineration if necessary. The remainder of the approaches described
above are under investigation as potential replacements for incineration.
c. The incineration ash contains as its major components
sodium sulfate (79 percent) sodium carbonate (14 percent) and sodium
sulfite (4 percent). Ash production at the modernized Newport AAP is
projected at 1500 Ibs/hour at a TNT production rate of 15 million lb/
month (3f). The evaporation condensate, while essentially distilled
water, contains sufficient nitrobody color to be classified as pink
water. Data from Joliet AAP show the nitrobody content of the conden-
sate to be about 15 mg/1 (3f). Color development occurs at TNT levels
as low as 1.5 mg/1, while a-TNT is soluble up to 200 mg/1 (3i).
d. Crude TNT purification by sellite washing removes selec-
tively the three to five percent of unwanted TNT isomers containing a
nitro group in the three or five position. In the modernized continuous
TNT lines, the sellite wash (red) water contains about 25 percent dis-
solved solids. Typically, this might include 17.3 percent organics,
2.3 percent sodium sulfite, 0.6 percent sodium sulfate, 3.5 percent
sodium nitrite and 1.7 percent sodium nitrate (15zp). Red water from
the older, batch process is more dilute, but represents about 1.6 times
the mass of impurities per ton of crude TNT manufactured by the continu-
ous process (15zp). The batch process red water normally contains re-
sidual spent nitrating acids, which in the continuous process are washed
out separately- In the continuous process, water wash of the TNT, yield-
ing "yellow water" then precedes the sellite wash. This water wash re-
moves the residual acidity and the resulting yellow water contains nitric
and sulfuric acids plus nitrobodies. Schulte, et al. (16h) have reported
that the production of 100 Ibs of TNT by the continuous process requires
47 Ibs toluene, 210 Ibs sulfuric and 125 Ibs nitric acids, 6 Ibs sodium
sulfite, 1 lb soda ash and 640 Ibs water. The resulting waste volume is
about 100 gallons.
e. There are considerable differences in both the waste volume
and character between the batch and continuous TNT processes. For ex-
ample, in the batch process there is a neutralization wash with soda ash
solution, prior to the sellite wash (lg(3)). This is not used in the
continuous process. Instead, a water wash precedes the sellite wash and
this acidic "yellow water" is returned into the toluene mono-or di-
nitration step. The acid wash is then followed by two sellite washes.
The result of this modification is both a reduction in wastewater volume,
and reduced discharge of constituents associated with the batch neutral-
izing wash. Excess yellow water, above that volume that can be returned
to the nitration stages, is combined with other process wastes for treat-
ment by either incineration or neutralization (3f). It has been reported,
for example, that at Radford AAP the total process water use is 240 gal/ton
-------
TNT (3u). Of this, 22 gal/ton is yellow water which is returned to
the continuous nitration line, and 106 gal/ton is red water from the
sellite wash. The balance is scrubber water.
4. Effects of Water Management
a. As a result of various water management programs, par-
ticularly those associated with the continuous TNT manufacturing pro-
cess, waste streams will be both reduced in volume and segregated ac-
cording to their character. For example, Radford AAP proposes to reduce
process wastewater discharge from each of their continuous lines from
53,000 down to 8,500 gpd (3za). This reduction will result from: re-
ducing the number of wet scrubbers required for acid fume control;
return of acidic yellow water to the first or second nitration steps;
recycle of transfer water between TNT purification and finishing stages
and other recycle flows. After implementation of these water manage-
ment programs, the TNT lines will have a tank car wash water flow of
15,000 gpd, which will be recycled (3za). In addition, an estimated
4,200 gpd of red water per TNT line (84 gal/ton) will be sold off-plant,
or disposed of by incinerator (3za). The incinerator will also receive
excess yellow water, after neutralization and blending with red water,
and water from the post-nitration acid wash (ln(3), 15x).
5. Current Treatment Practice
a. Red water resulting from sellite purification is first
concentrated in rotary kilns to about 35-40 percent solids (3f). The
concentrated liquid is then either sold to the paper industry, or in-
cinerated to a dry residue. This technique was reported by Schott, et al.
in 1943 (16i) and patented by Hales, et_ al. in 1945, as a method to avoid
violent ignition of the residue (16j). It remains the only proven method
for disposal of large quantities of red water (16k).
b. At Radford AAP, ash residue from red water incineration
was disposed of to a sanitary landfill from 1969 through 1971 (3f).
Thereafter, the red water was concentrated and sold (ln(3)). The red
water incinerator consisted of four single-chamber rotary kilns. A re-
ported 24,000 tons of red water were incinerated per year, using natural
gas at 700,000 ft3 per month as the primary fuel (4c). The ash yield was
90 tons per year. Joliet AAP stockpiles red water incinerator residue on
open land. It has been reported that at Joliet AAP, 0.179 Ibs of ash is
produced per Ib of TNT manufactured (lg(3)). Approximately 200 million
pounds of ash are now stockpiled. The residue is reported to be 90 per-
cent soluble, and some is washed away by rainwater (15zb). The red water
disposal plant, built in 1965 at a cost of $8 million, uses fuel oil.
c. The Volunteer AAP red water incinerator complex utilizes
twelve rotary kilns to incinerate red water produced in excess of that
sold to the paper industry. Newport AAP, out of production while the
plant undergoes modernization, will continue to use rotary kilns, as in
the past. Pink and yellow water streams will be neutralized, concentrated
-5-
-------
and blended with red water for incineration disposal. Ash production
at the modernized plant is estimated at 1500 Ibs/hour, at a TNT pro-
duction rate of 15.0 million Ib/month. Ash will be landfilled, with the
landfill sites lined with protective membranes to prevent ground water
contamination (3f). It has also been reported that water runoff from
the ash pile will be collected and lime-treated (3f). This is presum-
ably intended to partially control sulfate discharge by precipitation
of gypsum.
d. In each plant, the evaporation condensate represents a
disposal problem. The condensate contains sufficient nitrobody color
to be classified as pink water. For example, condensate from the Joliet
AAP has a typical nitrobody content of 15 mg/1. In no instance is this
condensate treated before discharge. The potential treatment technology
applicable to this condensate will be discussed in a subsequent section
of this chapter.
e. The existing evaporation/incineration treatment process
suffers from four deficiencies: (1) the condensate does not comply with
accepted standards for nitrobody content; (2) energy requirements for red
water concentration and incineration are high; (3) the ash residue pre-
sents a major solids disposal problem, with significant potential for
surface and groundwater contamination; and (4) there is a potential air
pollution problem associated with the incinerator operation. The U- S.
EPA has reported 50 tons per year particulate matter, 5 tons per year
SO and 648 tons per year NO emitted from the Radford AAP red water
incinerator (4c).
6. Modifications to Current Treatment
a. Because of the deficiencies of the present red water dis-
posal program, the Army currently has under investigation several modi-
fications. Modifications which have or are being studied include: (1)
reverse osmosis as the red water concentration step prior to incinera-
tion, and to recover recycle water; (2) red water acidification to re-
move one sulfonate group from the TNT isomers, and allow recycle of DNT
into the nitration process or the recovery of useful products such as
diaminotoluenes; (3) a fluidized bed reduction system to recover useful
by-products from the incinerator ash; and (4) the "Tampella process"
which would recover useful products directly from the red water.
b. Concentration of the red water by reverse osmosis has been
studied at Volunteer AAP, using a two-stage pilot unit. Pretreatment
with sulfuric acid was required to lower the pH from above 8 to 7 or be-
low, in order to prevent scale formation in the RO unit. Pilot tests
indicated that direct concentration of red water from 5 percent to the
35 percent strength required for incineration, was impractical. The RO
unit could concentrate to 14-15 percent solids, although significant
leakage of TNT through the RO membrane occurred (3h, 3i). Reverse os-
mosis thus appears impractical, except as a water recovery step, followed
by further concentration in evaporators.
-6-
-------
c. The red water acidification process is aimed at desulfona-
tion, and recovery of DNT (3h). Unsuccessful attempts were made at Rad-
ford AAP to desulfonate 2,4- dinitrotoluene-5-sulfonic acid, and addi-
tional efforts are planned (15zp). Figure 1 is a concept schematic for
a complete recycle acidification process. The process depends, however,
upon a development of capability to successfully produce DNT in the hy-
drolysis reaction,
d. The fluidized bed reduction system involves a chemical
conversion of red water incineration ash. The ash is ground for optimum
fluidization, and is reacted with a reducing gas which at the same time
is used to fluidize the solids. Water and CC>2 are added to carbonate
the reduced products. The products expected from the process are Na2C03
and H2S, which can then be used to remanufacture sellite, thus recycling
the wastes. A process schematic is shown in Figure 2. The reducing
gases are H2 and CO, in N2 (3h). A technically feasible fluidized bed
process has recently been investigated at bench scale level (3z). How-
ever, problems with maintaining fluidization at high temperatures and
slow second order kinetics have resulted in abandonment of the process
(15zp).
e. The most promising process appears to be the Tampella pro-
cess, which has been successfully used in the paper industry (3f). The
Tampella process uses the concentrated red water liquor, in a reduction
atmosphere, to also produce Na2S which can be carbonated to yield Na2C03
and I^S, which can then be recycled to produce sellite. The initial
pilot work on this process has been completed, and proved successful.
Further study is required to assess the effect of impurities and pro-
cess economics (15y). Investigation of alternate concentration proce-
dures is warranted.
7. Alternatives to Current Practice
a. Alternative treatment technologies considered for red water
include: (1) activated carbon; and (2) chemical destruction. A limited
amount of work has been carried on with each of these technologies.
b. Schott, et_ al. (16i) reported in an early study of red
water that activated carbon dosages up to 2000 mg/1 (8.34 tons carbon/MG*
red water) produced only 15.8-36.5 percent reduction in the organic con-
tent of the waste, while even higher dosages (up to 10 gm/1) were required
for effective color removal. The authors concluded that activated carbon
treatment, although somewhat effective, requires excessive quantities of
carbon and would be prohibitive in cost.
c. In a later, and more extensive study of carbon treatment of
red water, Schultz, ^t_ aJL. (16h) reported that in addition to TNT, other
compounds present in the red water, (some formed by side reactions in the
sellite purification step), include hexanitrodibenzyl,tetranitromethane,
cyanic acid, trinitro benzoic acid, dinitrocresol, phenol and various
nitrotoluene sulfonic acids. In their carbon work, they studied the
treatability of red water from a continuous TNT line.
*MG = million gallons
-------
so,
RED
WATER
I
HYDROLYSIS
AND
REARRANGEMENT
H2S°4
2,4DNT Ca(OH)2
NEUTRALIZATION
TNT
NITRATION
CRUDE
TNT
NaSO,
PURE TNT
ABSORTION
TOWER
SO,
Figure 1 - Acidification of Red Water
-8-
-------
CRUDE TNT
SODIUM SULFITE
SO
SODIUM
CARBONATE
ABSORTION
TOWER
'SODIUM SULFITE
RED WATER
PURE TNT
FLUID
BED
REDUCTION
Figure 2 - Recycle of Red Water Wastes
-9-
-------
d. Among other results, these researches found that more
acidic waste pH increased the adsorption ofa-TNT by the carbon. Thus
at pH 1.4, "n" from the Freundlich isotherm plot was 1.6 while at pH
7.0, "n" was measured at 5.2. Generally the value of "n" is inversely
proportional to the efficiency of a particular carbon. Further, the
adsorption capacity of the carbon was somewhat enhanced with acidic raw
wastes, although yielding less than half the measured capacity for pure
a-TNT in distilled water of 0.675 Ibs TNT/lb carbon. The greater capac-
ity of 0.675 Ibs/lb is frequently cited, however, in discussing the use
of carbon to treat TNT wastes such as pink water, which theoretically
is predominately ^-TNT.
e. In short detention time (2.5-3.0 minutes) carbon column
studies, for pure a-TNT in distilled water, efficient TNT removal oc-
curred up to a total throughput of 2150 bed volumes. However, carbon
column treatment of raw waste (red water) resulted in almost instantan-
eous breakthrough of both Silas-Mason reactable substances and color
(16h). With longer detention time in the columns (15-25 minutes) im-
proved TNT removal occurred, although color breakthrough was again very
rapid. Carbon capacity for TNT at these detention times and pH 1.5 was
0.124-0.175 Ibs/lb. Gas chromatography analyses of the carbon column
effluent indicated complete removal of DNT and TNT, even after the
Silas-Mason test showed breakthrough of Silas-Mason reactable substances.
Thus, the carbon appears effective in removing DNT and TNT, but not
effective in removing some of the other constituents of the red water,
and its application for red water treatment appears doubtful. Further,
the sum of TNT and DNT concentration measured in the raw waste by gas
chromatography (GC) exceeded the concentration by the Silas-Mason pro-
cedure (52.6 vs 46.5 mg/1). Other substances detected by GC did not
show up by the Silas-Mason technique (16h). These results raise serious
questions concerning the specificity and sensitivity of the Silas-Mason
procedure relative to monitoring red water quality. Chromatography
methods such as gas, liquid-liquid, and thin layer, for detecting ex-
plosives in wastewater, many of which were developed at NSWC White Oak,
appear more reliable and sensitive than the Silas-Mason procedure.
f. Edwards and Ingram (161), in exploring chemical treatment
of TNT wastes, found that reducing chemicals such as sulfur dioxide and
^28,0^ were ineffective in reducing the color of red water. However,
the use of chemical oxidants appears more promising. Schott, et al.
(16i) reported that chlorination at dosages of 5 to 20 mg/1 reduced color
in red water 25 percent or more, while bromination at equivalent dosage
was twice as effective as chlorination. Ozonation as a method of treat-
ment was disappointing, with up to 24 hours of ozonation yielding only
30 percent reduction in color. Recent success with UV-catalysis in
ozonation of refractory organics (16m) suggests that ozonation treatment
might benefit from the application of this technique and indeed, there
is preliminary evidence to indicate that TNT wastewater can be success-
fully treated by UV-catalyzed ozonation (15z, 3z). UV ozonation gener-
ates a singlet oxygen, a reactive chemical species which is undoubtedly
responsible for destruction of the nitrobodies. Other methods of gen-
erating singlet oxygen may become more economical than UV-ozone (15zq).
-10-
-------
>< g. Although there has been no effort directed toward land
disposal of red water, work is being done at NAD Crane on soil disposal
and soil microorganism biodegradation of waste TNT (14j , 14t). The tox-
icity of TNT to some organisms is well documented, and its presence in
soil may interfere with normal soil microbiological activity. Indeed,
there is already evidence that the presence of TNT in soil impedes the
degradation of paper, indicating that toxic side effects may preclude
this method of disposal. Soils have been reported as effective in ad-
sorbing TNT, up to 0.1 percent by weight (16q). If concentration on the
soil proceeds more rapidly than degradation, localized TNT deposits
could build up to hazardous levels in the soil.
8. Summary
a. The current practice of evaporative concentation of red
water, with either sale to the paper industry or incineration and land
disposal of ash, is unsatisfactory for several reasons. Concentration
and incineration is costly in terms of energy requirements and, in fact,
the proposal by Newport AAP to also concentrate and incinerate pink and
yellow water along with red water is difficult to understand. Pink water
and yellow water are extremely dilute solutions as compared to red water,
and can be treated by processes described subsequently in this chapter.
b. Further, with the disappearance of the paper industry red
water market, the need to incinerate all red water and dispose of the
large quantities of resulting ash will increase the energy demand, as
well as represent a potential hazard for ground water and surface water
contamination by leaching of ash stockpiles. Avoidance of this problem
would require containment and treatment of ash pile runoff, although
certainly the simple lime treatment proposed for Newport AAP would not
be adequate. Incineration is a terminal process, negating any possi-
bility of by-product recovery. Both evaporation and incineration yield
a pink water condensate which should receive further treatment, despite
the current practice of discharge without treatment. In addition, there
are severe air pollution consequences associated with incineration, in-
cluding the emission of particulates plus oxides of nitrogen and sulfur.
c. The lack of success shown in studies of red water treat-
ment by activated carbon, reverse osmosis, acidification and fluidized
bed reduction leave only two viable alternate treatment technologies
under investigation. These are the Tampella process and UV-catalyzed
chemical oxidation. Neither has been proven at pilot scale. Major ef-
forts should immediately be directed toward an intensive evaluation of
each of these procedures while, as a back-up to their possible lack of
success, preliminary studies should be undertaken to identify and assess
other potential treatment methods.
d. Special attention should be given to the ineffectiveness
of reverse osmosis treatment, failing both in the capability to achieve
acceptable levels of concentration and to prevent leakage of pollutants
-11-
-------
across the membrane. This is a deficiency characteristic not only for
explosives wastewaters, but for many other industrial wastes (16zy).
The deficiency lies in the membranes available for reverse osmosis sys-
tems and, despite the early promise of reverse osmosis, its use remains
severely restricted. A major research effort is warranted to develop
more reliable and acceptable membranes.
-12-
-------
SECTION II-B - PINK WATER/LAP
9. Waste Sources
Pink water comes from both TNT manufacturing plants and LAP's.
That from manufacturing plants arises from: nitration fume scrubber
discharges; red water concentration distillates; finishing building hood
scrubber and washdown effluents; and possibly spent acid recovery wastes
(2b). An additional source of pink water is from demilitarization wash-
out of munitions containing TNT. However, demilitarization is not a
part of the scope of this study. The first two sources of pink water
above may contain TNT isomers and dinitrotoluenes. Pink water from
LAP's contains essentially a-TNT, frequently contaminated with RDX, HMX,
wax or other additives. It has been reported that the effluent of the
TNT spent acid recovery plant at Radford AAP contains a nitrobodies dis-
charge of 35-371 Ibs/day (4c, 8c). Table 1 summarizes the plants having
pink water waste. Laundry waste waters have also been reported to con-
tain TNT. Table 2 summarizes available data on the nitrobody content
of pink waters from various sources. In general, evaporator condensate
is relatively low in TNT, although Neal (3f) reported that this waste-
water contains 150 mg/1 TNT. LAP wastewater is higher in TNT, which
may exceed 100 mg/1. If the data for scrubber water (Holston AAP and
NAD McAlester) are representative, TNT content from this source is also
high. Figure 3 shows the solubility of TNT in water as a function of
temperature (13a). Despite the relatively low TNT concentration of
evaporator condensate, the mass discharge may be substantial. For ex-
ample, at full TNT production, the condensate discharge for Joliet AAP
is projected at 325 gpm (8d). At a TNT concentration of 4 mg/1, this
represents a daily discharge of 15.6 Ibs TNT.
10• Current Treatment Practice
a. At present, only pink water from some LAP washdown and
scrubber operations is treated. Processes in use include prefiltration
plus activated carbon adsorption at Joliet and Iowa AAP's, and evapor-
ative and leaching ponds at Cornhusker, Milan, Louisiana and Lone Star
AAP's, and NAD McAlester. One unique disposal plan is to be employed
at Newport AAP, where pink water will be sent to the red water evapor-
ative system. Excess water will be vaporized and the concentrated
solution treated along with concentrated red water, by incinerator (3f).
In all cases, the pink water flows through sumps prior to treatment, to
remove settleable solids. These sumps are mucked out at regular inter-
vals, and the sludge taken to burning grounds for disposal.
b. Most Army and Navy plants actively developing pink water
pollution abatement facilities propose to use activated carbon treat-
ment, since this technique has been proven effective, whereas possible
alternative technologies such as polymeric resins and biological treat-
ment are still in the investigative stages.
-13-
-------
Table 1. Plants With Pink Water Waste*
Army
Navy
Holston
Iowa
Joliet
Kansas
Lone Star
Louisiana
Milan
Newport
Radford
Volunteer
Crane
Hawthorne
McAlester
Yorktown
*See Volume I for detailed information
on plant activities.
-14-
-------
Table 2. Reported TNT Concentrations in Pink Wastewaters, mg/1
Plant
Joliet AAP
NAD Crane
Rockeye
Plant A
Plant B
Hols ton AAP
Radford AAP
NAD McAlester
Iowa AAP
Milan AAP
Louisiana AAP
Cornhusker AAP (inactive)
Evaporator
Condensate
1.4
3.7-6.1
4.4
16
7.3
Clean Up
LAP
178.3
40.1
5.8-11.2
20.2
3-9*
90-175
75
30.6-38.4
86.9
<1
80
57
Scrubber
Water
2-22
30-80
Laundry
Wastewater
2.9
25.4
2.7
Reference
lg(3)
8d
3v
8e
3f
14f
14f
14 f
5d
6d
3w
2b
14u
15za
13a
13c
2b
2b
2b
*From incorporation of TNT into Composition B,
-------
0.15
0.01
20
30
40" 50 60
TEMPERATURE - °C
70
80
90
100
Figure 3 - Solubility of TNT in Water (13a)
-------
c. Two carbon systems are currently in operation, one at
Joliet and one at Iowa AAP. Joliet uses a two column series downflow
system; and Iowa a two column expanded bed parallel upflow system. In
both installations, the carbon beds are preceded by sumps to remove
settleable solids, and diatomaceous earth filters to remove suspended
solids carried over from the sumps. The basic layout is the same for
both systems, as represented in Figure 4, a schematic of the Iowa in-
stallation (3i). Both installations are used to treat LAP wastewater,
Because of the explosive material trapped in the diatomaceous earth
filters, the filter material is used only until clogged, and then
burned. Exhausted carbon is also destroyed by burning, since there is
not yet an accepted way to regenerate it (15zb).
d. The successful use of carbon to treat TNT wastes was re-
ported more than 20 years ago. Ruchhoft, et_ al_. (16q) studied the
treatment of both pure o^-TNT (colorless) and TNT wastewaters in which
color had developed. For the colorless solution, at a TNT concentra-
tion of 121.8 mg/1, they found up to 99.5 percent removal of TNT was
possible with powdered carbon, and that powdered carbon was more effec-
tive than granular. Much less effective TNT removal (40.0-48.7%) was
found for the colored TNT wastewater. The effect of color development
on carbon treatability has been confirmed by Nay, _et_ _§_!. (16r). These
authors concluded that as the waste color intensity increased, the
waste became more toxic to microorganisms and more refractory to bio-
degradation, and treatment by any process, including carbon adsorption,
became more difficult. While activated carbon was effective in remov-
ing both color and TNT, the development of color significantly decreased
the carbon adsorptive capacity for TNT. Their results are presented in
Figure 5 (16r). These workers also reported that among other factors
which caused color intensification was adjustment of wastewater pH from
an acidic to near-neutral value, with exposure to light enhancing color
development. Spano, et_ al. (16s) reports that pink wastewaters are
typically initially acidic (pH 3-4) and colorless.
e. Experience with carbon treatment of LAP pink water at
Joliet AAP has been good. Wastewater passing through the system is sub-
jected to two-step treatment. First, it is filtered in a diatomaceous
earth filter to remove gross and suspended solids. Then, it is pumped
downward through two carbon columns, to remove dissolved explosives con-
taminants (TNT and RDX). The columns are connected in series and are
piped so that the columns may be used in either order. This allows the
final effluent always to leave the column most recently charged with
fresh carbon. Each column has dimensions of 9-foot height by 30-inch
diameter. The carbon charge is 960 Ibs. of Filtersorb II, changed
monthly. The average wastewater volume treated by the system is 6165 gpd
(lg(3)). Table 3 summarizes operating data for the system, presenting
range and average results from eleven samples collected by AEHA personnel
in 1972 during a two-week study of Joliet AAP (lg(3)).
f. The data of Table 3 indicate a raw wastewater of somewhat
alkaline pH, high in solids and explosives. It is noteworthy that the
-17-
-------
TREATED WATER RETURN TO PROCESS
00
80 GPM
WASH WATER
CHEMICAL
FEED
FILTER SUPPLY
MONITOR
IATOMIT
TANK
TURBIDITY
MONITOR
PINK
WATER
CARBON \ / CARBON
COLUMN COLUMN
STORAGE
TANK
FILTER \ / FILTER
TANK J I TANK
3
DRAIN TO SUMP
Figure 4 - Pink Water Treatment for Loading, Assembling, and
Packing Operations
-------
10,000
a;
o
_l
o
(J
} "10 100
CARBON ADSORPTIVE CAPACITY (mgTNT/gms CABON )
Figure 5 - Relationship of Color and Carbon Adsorptive Capacity for TNT
-19-
-------
Table 3. Performance of LAP Wastewater Treatment System at Joliet AAP (lg(3))
Parameter*
PH
Total Solids
Suspended Solids
TOC
Kjeldahl-N
TNT
RDX
Color
LAP
Influent
Range
6.8-8.4
903-1790
220-336
93-188.4
10.3-25.4
156.2-235.0
87.5-180.0
-
Average
7.9
1401.5
138.5
121.1
17.0
178.2
145.2
-
Effluent From:
Diat. Earth Filter
Range
7.6-8.6
959-1796
26.0-271.0
100.0-162.0
8.9-23.0
143.5-213.0
87.5-185.0
-
Average
7.9
1418.5
108.6
121.1
15.3
175.7
148.9
-
1st Carbon Column
Range
7.6-8.2
794-1791
0.0-40.0
5.9-64.0
4.4-8.6
0.0-111.2
0.0-77.5
5.0-20.0
Average
7.8
1138.3
8.4
24.3
7.2
14.7
30.1
10.0
2nd Carbon Column
Range
6.7-8.0
762-1497
0.0-7.0
7.8-20.9
4.0-4.9
0.0-25.0
0.0-46.4
0.0-20.0
Average
7.7
1069.8
1.2
12.1
4.6
3.7
19.5
8.0
Average
Percent
Change
-
23.7
99.1
90.0
72.9
97.9
86.6
-
o
I
*A11 units except pH and color in mg/1. Color in PTU.
-------
diatomaceous filter accounts for only 21.5 percent removal of suspended
solids, with the bulk of removal occurring in the first carbon column.
The filter also accounted for no significant removal of explosives.
g. The AEHA*study team suggested that filter efficiency could
be improved by implementing an effective maintenance schedule. They
further suggested that if the wastewater were maintained slightly acidic,
most of the TNT would remain as a-TNT, and thus be more readily removed
by carbon adsorption. The final effluent is of general good quality,
with suspended solids, TNT and RDX averages of 1.2, 3.7 and 19.5 mg/1,
respectively. The range of data in Table 3 suggests that by close pro-
cess control, effluent levels of less than 0.1 mg/1 of TNT and RDX can
be achieved. However, a recent claim by Rosenblatt, et_ a.L. (2b) , that
the Joliet system achieves an average effluent of less than 1 mg/1 is
not substantiated by the data of Table 3.
h. In studies to determine the comparative cost and efficiency
of various carbons for the Iowa AAP treatment system, it was found that
the carbon required per 1000 gallons wastewater treated ranged from 18.3-
40.9 Ibs. (13b). The loading criterion was grams of carbon needed to
reduce TNT from 65 to less than 3 mg/1. The Iowa system employs paral-
lel upflow columns, preceded by a diatomaceous earth filter. The columns
are recharged with fresh carbon when the effluent TNT level reaches 5
mg/1 (13c). Effluent TNT levels from 0.1 up to 180 mg/1 have been re-
ported, however (13c). The system treats about 25,000 gpd of pink water
(2b).
i. Data available from Radford AAP for pilot carbon treatment
of pink water over the period 6 September, 1973-28 February, 1974 and re-
ported in Table 4. The effluent pH is even more basic than reported for
Joliet in Table 3. Effluent total solids are comparable, suspended sol-
ids much higher, and TNT significantly lower at Radford. However, in-
fluent data were not available.
j. An extensive pilot study of a carbon system was recently
carried out at NAD Hawthorne (15zc, 14zb). Several alternatives to the
diatomaceous earth filter were investigated, due to its tendency to cake
up when the wastewater contains wax. Alternatives studied were multi-
media (sand and coal) pressure filters and air flotation. Both multi-
media and diatomaceous earth filtration performed comparably. In one
series of runs, influent TNT averaging 1018 mg/1 was reduced by the com-
bination of air flotation and sand filtration to 414 mg/1. The same
waste stream was reduced to 411 mg/1 TNT by air flotation plus diatoma-
ceous earth filtration (14zb) . Experiments with TNT removal from waste-
waters of Composition A-3, Composition B, HBX-1 and Picratol showed
equally comparable removal efficiencies for sand filtration versus dia-
tomaceous earth filtration. Thus, selection of best filtration tech-
niques can be made on the basis of cost-effectiveness and ease of oper-
ation. Cooling of the wastewater prior to filtration was also investi-
gated. This pretreatment was intended to precipitate wax , TNT and other
soluble components, which would be removed prior to the carbon, and thus ^
extend the carbon life. Cooling was successful, although some difficulties
were encountered with explosive precipitate coating the cooling system.
* U.S. Army Environmental Hygiene Agency
21
-------
Table 4. Carbon System Effluent at Radford AAP (6d)
Parameter
pH
Total Solids
Suspended Solids
TNT
Range
7.7-9.3
552-1870
2.0-125
0.0-2.4
Average
-
951.3
38.7
0.91
-22-
-------
k. One of the most serious deficiencies of the carbon system
is the lack of current capability to regenerate the exhausted carbon.
As a result, the carbon is disposed by incineration after a single use.
This lack of regenerative capability results in three disadvantages for
carbon treatment: (1) the economics of carbon use are mugh higher than
normal for other industrial wastewater carbon systems, (2) there is no
potential to recover and reuse the a-TNT waste product captured by the
carbon, and (3) the incineration of the carbon is a potential source of
air pollution.
1. Two regenerative techniques have been studied, thermal and
solvent extraction. Eliassen (16t) reported in 1944 that activated car-
bon would treat TNT wastewater, but if the carbon is regenerated in a
closed cell it is liable to explode. Recent tests by Rosenblatt, e_t^ al.
(3c) confirm this in lab tests. They report that as the degree of~car^
bon saturation with TNT increases from x/m of 0.217 to 0.675 (the satu-
ration value), an exothermic reaction occurred within the thermal re-
generation furnace. At the higher degree of saturation, explosion
occurred in both an air and nitrogen atmosphere. At the lower degree of
saturation (x/m = 0.217), there was no detectable exothermic reaction,
and up to 80 percent of the TNT in the spent carbon was removed safely
in 30 minutes at a furnace temperature of 900°F. Although the carbon
was not retested in this study to determine its regenerated capacity,
other sources report that the regenerated carbon recovers only about 60
percent of its original capacity (15y). The off-gas from the regenera-
tor furnace was reported to contain water, carbon dioxide, nitrogen ox-
ides and methane. Successful applications of thermal regeneration to
spent activated carbon would thus produce air pollutants. Additionally,
the low degree of saturation acceptable for thermal regeneration indi-
cates that the carbon would be used only to one-third capacity before
regeneration. The Army proposes to undertake pilot plant thermal re-
generation studies, to determine equipment design criterion for possible
GOCO plant implementation (3i).
m. Acetone, methanol and toluene have been investigated as
chemical solvents for carbon regeneration (3c, 15s, 13f). The potential
advantages of the method are: (1) regeneration takes place within the
fixed bed; (2) TNT contaminated solvent can be reused as regenerant or
as raw material in TNT manufacture; and (3) the carbon may be more com-
pletely saturated with TNT than is possible for thermal regeneration
(3c). Spano, et al. (16s) in lab column studies, found that only 22.3
percent of the TNT was removed by two bed volumes of acetone. These
preliminary and limited results are in contrast to pilot results with
acetone regeneration of carbon, described below (13f). Rosenblatt, et al.
(3c) also in lab scale tests on TNT saturated carbon, used a toluene wash,
followed by acetone. The toluene removed 67 percent of the TNT, while
the sequential wash with acetone removed only trace amounts of TNT. After
stripping with toluene, the carbon was reused to remove TNT. An amount
of TNT equal to that recovered in the initial toluene wash was again
removed by the carbon. A second toluene wash removed 92.4 percent of
the TNT adsorbed during the second carbon use. Thus it appears likely
-23-
-------
that the sorptive capacity approaches some constant level on repeated
use, followed by toluene regeneration of the carbon.
n. In separate pilot plant studies at Iowa AAP, Brinck (13f)
investigated solvent stripping of TNT-laden carbon by use of acetone,
methanol, and toluene. The pilot study, involving in-place regeneration
of carbon columns, yielded 92 percent TNT removal by acetone, 34 percent
by methanol and 28 percent by toluene. Although acetone effectively
removed the TNT, the carbon did not recover capacity equivalent to the
TNT removed. The carbon was reactivated only to the extent of 43 per-
cent, 33 percent and 31 percent for acetone, methanol and toluene, re-
spectively (13f). Although the results of Brinck (13f) and Rosenblatt,
et al. (3c) are somewhat conflicting relative to the efficiency of TNT
stripping by acetone, both studies indicate only partial recovery of
adsorptive capacity upon regeneration with either toluene or acetone.
o. An economic feasibility study of the regeneration of car-
bon by chemical means is underway at Edgewood Arsenal (3i) .
p. With the exception of a demonstration scale biological
treatment system to be constructed at NAD McAlester, all current con-
struction plans for pink water abatement call for the use of carbon sys-
tems. Both up-flow and down-flow systems are under consideration. The
up-flow system is easier to pump, but results in a lower carbon loading
with TNT. Unless one of the carbon regeneration techniques now under
investigation proves technically and economically feasible, a down-flow
system will be used since the carbon can be loaded to saturation before
disposal (15y). A down-flow fixed bed system with two columns in series
appears to be optimal, as compared with other down-flow multi-column
systems (3c).
q. At Joliet AAP, where carbon systems are already in use for
LAP pink water, the same system is proposed to handle red water evapor-
ator plant condensate (lg(3), 15zb). The proposed system will include
a diatomaceous earth filter and two down-flow columns in series. The
characteristics of the condensate are sufficiently similar to that of
LAP wastes (Table 5), that a carbon system should prove effective. In
general, the condensate represents a more dilute waste than the LAP pink
water, particularly with regard to suspended solids and TNT. Indeed, it
is noteworthy that Joliet AAP proposes to install a treatment system for
a waste more dilute in TNT (0.1-2.6 mg/1) than the design TNT effluent
for other proposed carbon systems.
r. NAD Crane proposes a recycle approach to handle fume
scrubber water from their Mine Fill A Line. The scrubber water will be
recirculated, through a settling basin and. filter. Since the water
should quickly become saturated with TNT, subsequent captured TNT would
remain in the particulate form, and be removed by sedimentation and
filtration (15ze). The recirculation system should negate the need for
carbon treatment, unless blowdown from the recycled scrubber water be-
comes necessary.
-24-
-------
Table 5- Comparison of Red Water Incinerator Condensate and
LAP Pink Water at Joliet AAP (lg(3))
Parameter*
Temperature °F
PH
Conductivity
Acidity
Alkalinity
Color, PCU
Total Solids
Suspended Solids
Dissolved Solids
Total Volatile Solids
Total Organic Carbon
Kjeldahl-N
Nitrates
Sulfates
TNT
RDX
Condensate
Range
162-176
8.1-9.2
190-440
0.0-8.0
52.0-244.0
5.0-6.0
201.0-317.0
1.0-16.0
185.0-302.0
119.0-250.0
31.7-39.0
15.1-24.0
2.4-4.6
26.0-44.0
0.1-2.6
-
Average
170.2
8.9
326.2
1.6
180.0
5.8
250.0
11.7
238.5
192.0
35.3
18.0
3.4
37.2
1.4
-
LAP Pink Water
Range
-
6.8-8.4
1364.0-9000.0
0.0-32.0
164.0-356.0
-
903.0-1790.0
22.0-336.0
881.0-1482.0
426.0-954.0
93.0-188.4
10.3-25.4
-
-
156.2-235.0
87.5-180.0
Average
-
7.9
4082.1
17.6
289.3
-
1401.5
138.5
1264.8
547.6
121.1
17.0
_
-
178.2
145.2
*A11 mg/1 except pH, temperature, conductivity and color.
-25-
-------
s. NAD Crane will, however, use carbon treatment for clean-
up water from their LAP operations (15ze). Table 6 summarizes the waste
characteristics of one of the flows to be treated (the Rockeye LAP area).
A concept engineering study for control of this TNT waste at NAD Crane
recommends the use of carbon at two loading sites, over other alterna-
tives, in light of its proven capability (14f). The wastewater to be
treated will contain approximately 95 mg/1 TNT, with a flow of 90-95
gpm at each site. Each carbon system will include a flow equalization
tank of 100,000 gallons capacity, two diatomaceous earth leaf filters
operating in parallel, and three parallel down-flow carbon columns.
Flow will be equally distributed between the three columns. The use of
down-flow parallel columns is a departure from the system used success-
fully at Joliet AAP (series down-flow) and Iowa AAP (parallel up-flow).
It likely represents a compromise between the two types of systems,
which will yield a more highly saturated carbon than the Iowa system,
but less effective removal than a series system such as Joliet's. How-
ever, if a high suspended solids removal occurs, such as is experienced
at Joliet, the proposed columns may well clog before the carbon is ex-
hausted. To counter this possibility, the NAD Crane systems will in-
corporate carbon column backwash capabilities. A feasibility study
report submitted to the concept engineer, nevertheless recommended use
of the Joliet type system (14f).
t. At Yorktown Naval Weapons Station, a two column series
carbon system has recently been installed at their pilot LAP facility.
The system includes a diatomaceous earth filter, and is sized for a 20
gpm flow. Additional 20 gpm carbon systems are under construction at
four other LAP facilities at the station. However, these latter four
systems will consist only of a filter plus a single up-flow carbon
column. The use of a single column is a departure from practice at other
Army and Navy installations, but possibly reflects the design TNT efflu-
ent criteria of 5 mg/1 or less (14r). This is a considerably higher TNT
effluent than is achieved with the Joliet or Iowa AAP dual carbon
columns.
u. Evaporative disposal is also a treatment technique cur-
rently applied to pink water. The successful use of evaporative treat-
ment is dependent upon soil and climatological conditions. NAD Hawthorne
(Nevada) successfully uses sumps, followed by evaporative ponds, to dis-
pose of pink water resulting from fume and dust control (15zc). After
evaporation, the residue is burned.
v. NAD McAlester (Oklahoma) has four lagoons to receive
wastewater from various areas of the facility. Although all are de-
signed to overflow, the regional net annual evaporative loss of 10
inches/year has, to date, resulted in no overflow from the lagoons (15za),
Plant personnel expectation is that>at present wastewater discharge vol-
umes, there should be no overflow, with the possible exception of the
November-December rainy season. Wastewater sources include fume and
dust scrubber water and clean-up water. A recent report on the impact
of modernization of NAD McAlester indicated that a carbon treatment sys-
tem will be installed to handle the TNT waste flows from the "A" plant
-26-
-------
Table 6. LAP Wastewater Character for NAD Crane Rockeye Facility (14f)
Parameter*
PH
Temperature °C
Suspended Solids
Dissolved Solids
COD
TNT
RDX
HMX
Oil and Grease
NH3-N
N03-N
Kjeld-N
Phosphorus
Chloride
Calcium
Sodium
Average
7.7
10.3
15.0
140.0
44.0
40.1
0.0
8.5
70.0
1.5
0.8
2.7
0.3
18.0
33.0
67.0
Range
7.3-8.2
5.0-17.9
0.0-110.0
119.0-755.0
12.0-84.0
16.8-87.2
0.0
2.4-16.0
6.0-153.0
0.3-3.1
0.0-3.9
1.1-3.5
0.0-1.2
8.0-51.0
22.0-45.0
5.0-275.0
*A11 mg/1 except pH and temperature.
-27-
-------
(14m). Installation of a demonstration biological oxidation ditch is
planned for the "B" plant, allowing cost comparisons.
w. Both Louisiana and Lone Stone AAP's have been reported to
use evaporative/leaching ponds for TNT wastewater (2b). At Louisiana
AAP the wastewater (138,000 gpd) containing approximately 80 mg/1 TNT,
is trucked to leaching ponds. At times of heavy rainfall the ponds
overflow (2b). Lone Star AAP wastewater is reported to have a volume
of 20,000 gpd with TNT content of approximately 50 mg/1. This is trucked
to holding ponds, which are reported to overflow due to heavy rains (2b).
The experience of these two plants emphasizes the need to properly bal-
ance seasonal net evaporative rate with pond capacity, in designing
evaporative ponds for final disposal.
11. Modifications and Alternatives to Current Treatment
a. Several alternatives to carbon adsorption and evaporative
disposal have been investigated for pink water. These include: solvent
extraction; reverse osmosis; adsorption onto fly ash; resin adsorption;
ozonolysis; and biological treatment. Ozonolysis was discussed in the
preceding section on red water treatment.
b. Solvent extraction using toluene in a bench scale (300
ml/min.) two stage continuous countercurrent extraction system was in-
vestigated at Iowa AAP (13a). Treating wastewater containing 87 mg/1
TNT, the process was found to remove and recover up to 97 percent of the
TNT. Using a 27:1 water to solvent ratio, the unit achieved an effluent
TNT level of less than 3 mg/1. However a full scale unit for the Iowa
AAP was estimated to require a 12,000 gph treatment capacity, with oper-
ating costs of $1.15 (1964 basis) per 1,000 gallons of water treated.
c. Treatment by reverse osmosis has been studied at Volunteer
AAP (3f). Although the process is effective in concentrating the pink
water, excessive TNT leakage through the available membranes has re-
sulted in abandonment of further study of this process (15y).
d. The successful use of fly ash to adsorb TNT has been re-
ported (16u, 16v), and fly ash was used at one time to reduce TNT in a
creek at Iowa AAP (13a). However, Meek (13d) has reported that fly ash
has a very low adsorption capacity for TNT, at 1.08 gm TNT per Ib fly ash
(0.0024 gm/gm ash). When compared to the capacity of carbon for TNT
(0.4-0.7 gm/gm) the quantity of fly ash required becomes disproportion-
ate, and ultimate disposal remains a problem, as with carbon.
e. The use of polymeric ion exchange resins has been inves-
tigated for TNT removal from LAP waters (3h, 14zb). One resin has been
shown to be an effective adsorbent, with easy chemical regeneration and
a long life cycle. Figure 6 presents results from multiple regeneration
experiments with this resin, using acetone (3h). The resin, which is a
copolymer of styrene and divinyl benzene, is sold as Rohm and Haas
Amberlite XAD-4 (16s). Comparison studies of XAD-4 (16s, 14zb, 3za)
-28-
-------
N5
^O
I
v3*fO
QOQfl
£.7\)
_J OT>
__, -j- 232
52
t/> -*- 1 "7 A
UJ H- 1/4
oil
> x 11A
Q
-------
with activated carbon indicate that the two adsorbent materials have
equivalent TNT capacity to the breakthrough point in a column. After
breakthrough, carbon removal efficiency exceeds the resin. However,
the possibility of multiple regeneration and reuse of the resin may
offset this disadvantage. Another, and perhaps more significant dis-
advantage is that the resin column does allow some color components to
pass through, while carbon is more effective in that respect. This
suggests the possibility of a resin first stage followed by a carbon
second stage series adsorption system.
f. In pilot studies at Iowa AAP, influent TNT of 81-116 mg/1
was reduced to below 1 mg/1 by resin adsorption. Saturation capacity
of the resin is about 0.11-0.15 gm TNT/gm resin, with TNT breakthrough
above 1 mg/1 occurring at 40 percent of saturation capacity (16zzz) .
Acetone regeneration of the resin could be accomplished within two bed
volumes of the resin column.
g. A disproportionate degree of effort is currently being
directed toward evaluation of biological treatment processes for TNT
wastes, considering the rather poor prognosis indicated by a review of
historical records. For example, an early study by Schott, e± ai_. (161)
indicated that neither activated sludge nor trickling filters could ef-
fectively degrade TNT wastes and that, in fact, treatment performance
fell off rapidly with increased wastewater TNT concentration, indica-
ting toxic effects upon the biological processes. Edwards and Ingram
(161) in another early study, likewise reported TNT wastes very resist-
ant to biological attack. Results such as these led Rudolfs (16n) to
conclude that TNT wastes cannot be oxidized biochemically, and purifica-
tion cannot be expected in streams.
h. Despite such early evidence there is a great deal of
literature (16w, 16z), and reports of on-going research, which indicate
that TNT wastes are susceptible to biological treatment. Neal (3f) re-
ports that because of the toxic nature of some of the organics in red
water to microorganisms, it is necessary to dilute this waste with an-
other waste for biological treatment. He further states that biological
treatment, following neutralization, is possible. However, treatment
for removal of color and dissolved inorganics should be by physical-
chemical methods. Nay et^ al. (8e) confirms this requirement, at least
for color removal, by other than biological treatment. In their studies,
color removal by biological treatment with an activated sludge system
was evaluated, but essentially no color removal was observed. Further,
some of their data suggested that color development was closely related
to bio-toxicity during treatment (16r). Table 7 reports the results of
their activated sludge experiments. Unsatisfactory results with both
activated sludge and trickling filters have also been reported in a
recent treatment feasibility study for NAD Crane (14f). The concept
engineering report resulting from these studies concluded that bio-
degradation of the TNT wastewater was not a technically feasible treat-
ment method. In substantiation of the toxicity effect, it has been report-
ed that nitrobody wastes from ct-TNT manufacture are toxic to the environment
when released into streams at concentrations of TNT greater than 2.5 mg/1
-30-
-------
Table 7. Summary of Results:
Color Removal (16r)
TNT Combination Wastes Treatability Runs -
Run
1
2
3
4*
5
6
7
8
9
10
11
12
13
14
15
TNT In
(mg/1)
5.0
5.0
5.0
10.0
10.0
10.0
15.0
15.0
15.0
20.0
20.0
20.0
25.0
25.0
25.0
Color In
103
145
131
101
121
120
168
167
194
271
274
278
284
316
280
MLSS
(mg/1)
1740
1500
1680
1860
2070
1740
1910
2070
2050
1680
1600
1750
1650
1580
1600
Detention Time
(hours)
14
7
3
14
7
3
14
7
3
14
7
3
14
7
3
TNT Out
(mg/1)
0.8
1.4
1.4
3.0
3.0
4.0
4.2
5.1
5.5
6.6
10.0
10.0
6.7
10.3
10.3
Color Out
(mg/1)
76
61
65
132
129
110
240
264
272
210
270
225
172
319
264
Color Removed
(%)
26
58
50
-31
-13
8
-43
-58
-40
23
1
19
39
-1
5
ave.
3%
*R-4 was repeated 12 times to investigate MLSS saturation by TNT. Color removal
during these 12 additional runs averaged 18% with a range of -5% to 30%.
-------
(3h). Unless acclimation occurs in a biological treatment process, this
would indicate a need for great dilution of the wastewater, prior to
treatment. The successful application of biological treatment would also
require that toxic by-products not be formed in the biodegradation of the
waste.
i. Despite these factors, there is recent evidence thata-TNT,
at least, is biodegradable and that, perhaps, other nitrobodies of the
wastewater may also be susceptible to biological treatment. Claims of
successful biodegradation of pure a-TNT, with addition of supplemental
organic nutrients, have recently been made (14z). In none of those re-
ports is there evidence of successful degradation of a-TNT without sup-
plemental nutrients, and in no case has color been allowed to develop in
the TNT solution prior to the degradation experiments. For example,
Southgate (16w) reports successful treatment of TNT at concentrations up
to 40 mg/1 in domestic sewage, with trickling filtration. A 1970 Army
study with bench scale activated sludge revealed that TNT could be re-
duced from 29 mg/1 to 8-15 mg/1, when mixed with a glucose/peptone sub-
strate of COD 802 mg/1 (Is). This is typical of more recent results,
which indicate that very high supplemental organic nutrient resources •
are required in order to achieve TNT biodegradation (14k, 16o, 16p).
Nay, et_ _aJL. (16p, 16r) have reported that the development of color makes
the waste more resistant to biological treatment. The U. S. Navy is
currently conducting a pilot scale study of an Oxidation Ditch at NOL
White Oak (la(3)), and is planning on installing a facility using this
process for the treatment of pink water at NAD McAlester. In the current
NOL White Oak study, cornsteep liquor is used as the supplemental nutri-
ent. Although there is strong evidence that ct-TNT, at least, is biode-
gradable, many factors indicate that biological treatment is not desir-
able. These factors include: the recognized toxicity of even 1-2 mg/1
of TNT; the lack of definition of degradation by-products and their pos-
sible toxicity; the high supplemental nutrient demands for successful
biological treatment; and the increased resistance of pink water, after
color development, to biodegradation.
12. Impact of Air Pollution Control
Many Army and Navy facilities propose to install water scrubber
systems as part of their air pollution abatement programs. When in-
stalled in TNT handling or LAP buildings, these scrubbers will generate
additional pink water flows. Their TNT content should be similar to
values shown in Table 2, and their treatability comparable to present
pink water wastes.
13. Summary
a. Although the wastewaters of LAP activities are well character-
ized, pink water originating from other sources such as red water evap-
orative condensate, spent acid recovery, nitrator fume scrubbers, and
laundry wastewater are not adequately characterized. Wastewater char-
acterization data for these "secondary" pink water sources are required.
-32-
-------
b. Although activated carbon treatment is being used success-
fully, and its use will be expanded at both Army and Navy plants, design
and performance criteria are not well established. For example,'there is
preliminary evidence that low pH, and avoidance of development of color,
will enhance treatment efficiency. Further study is required to assess'
the effect of such factors on carbon performance. A variety of upflow
and downflow, series and parallel, single and multiple carbon column sys-
tems are in use or under construction. Design criteria for carbon (or
any treatment) system must reflect uniform performance goals and efflu-
ent requirements. These have not been established. Uniform design cri-
teria should be defined, reflecting such factors as: whether the carbon
will be used once and disposed, or regenerated; required effluent color
and TNT levels; effects of pretreatment, such as filtration, on carbon
performance; and similar factors. The use of single columns, such as
proposed at NAD Crane and NWS Yorktown, is questionable, in light of the
present lack of definition of design criteria.
c. The current practice of use of carbon to saturation, with
disposal by incineration, has several deficiencies including: preclusion
of recovery of TNT; high cost for replacement carbon; and air pollution
from carbon incineration. In light of the proposed widespread use of
carbon systems, an intensive study of carbon regeneration is warranted.
The potential for regeneration, particularly with solvents, is signifi-
cant, and can overcome the disadvantages of the present system. Since
final carbon system design must reflect whether or not the carbon can be
regenerated and reused, systems now in the design stage should incorpor-
ate the flexibility for regeneration. Experience with acetone regener-
ation yielded high percentage recovery of TNT from spent carbon, but only
partial reactivation of the carbon. This, however, may have resulted
from saturation of the carbon active sites with acetone, which could in
fact be readily removed.
d. Polymeric resin adsorption has potential as an alternate
to, or in combination with carbon. Its major disadvantage is low TNT
capacity, compared to activated carbon. However, it is capable of being
regenerated, and of sustaining through multiple regeneration a constant
level of performance. If satisfactory carbon regeneration techniques
cannot be developed, a resin system, or a resin first stage with carbon
second stage appear to be viable alternatives. For the second alterna-
tive, the second stage .carbon would still require incineration, although
there would be much less carbon required, as the second stage would act
only as a polishing unit.
e The present use of evaporative ponds is an economical
method of pink water disposal. However, unless such ponds are sealed
significant wastewater loss may occur by leaching. A better understand-
ing is required of the effects of such leaching on both soil and ground
water quality. Further, the experiences of both Louisiana and Lone Star
AAP's with pond overflo^ during rainstorms emphasizes the need to balance
seasonal, rather than annual net evaporative loss with pond capacity.
-33-
-------
f. Biological treatment of pink water, while superficially
attractive, raises several serious questions. TNT is both bio-refractory
and toxic. Such wastes are not typically susceptible to standard bio-
logical treatment methods. Experience to date suggests that TNT can be
reduced in concentration in wastewaters treated biologically, providing
large quantities of supplemental nutrient are added. An important ques-
tion which must be answered is what biotransformations occur, and what
is the nature and impact of these by-products in the receiving water?
Further, biological treatment will apparently not reduce color and it is
conceivable that tertiary treatment such as ozonation, ion exchange or
activated carbon would be required for effluent from a biological waste
treatment system.
-34-
-------
SECTION III - ACIDS
14. Introduction
The three major acids associated with military explosives and pro-
pellants manufacture are acetic, nitric and sulfuric. Acids are not
involved in LAP operations. Concentrated nitric and sulfuric acid mix-
tures are used in explosives intermediate nitration processes, as for
TNT, nitrocellulose and nitroglycerin. Acetic acid is used in the prep-
aration of acetic anhydride, and the manufacture of HMX and RDX. Treat-
ment technology for both acetic acid and acetic anhydride wastewaters
will be discussed in this section. Acetic acid is recovered and puri-
fied for reuse, but not manufactured by military explosives plants. Weak
nitric acid is manufactured by the ammonia oxidation process (AOP), and
processed to yield concentrated nitric acid. Weak sulfuric acid is man-
ufactured or purchased, concentrated and fortified to produce oleum.
Spent sulfuric acid is regenerated and reused at most manufacturing
plants.
SECTION III-A - ACETIC ACID AND ACETIC ANHYDRIDE
15. Introduction
These products are manufactured only at Holston AAP. Acetic acid
and acetic anhydride, together with nitric acid and ammonium nitrate,
are used in the nitration of hexamine to produce a slurry of crude RDX
or HMX. The slurry is vacuum filtered, and washed to remove residual
acidity (Figure 7). Initial filtrates contain 10% acetic acid, 2-3%
nitric acid and some RDX or HMX. This filtrate is distilled to recover
weak acetic acid. Acetic acid is separated by distillation to 20% re-
sidual filtrate volume, and soluble explosive is recovered from the still
liquor by crystallization at 30°C. The liquor is then treated with so-
dium hydroxide to convert residual ammonium nitrate to sodium nitrate
and ammonia (which is recovered by vaporization and condensation).
Acetic acid is neutralized to sodium acetate. The neutralized solution
is currently stored in holding ponds at Holston AAP, but will be pro-
cessed into fertilizer when a processing facility is completed (2c).
Recovered weak acetic acid is concentrated to glacial (99%) acetic acid
by azeotropic distillation. A portion is reused directly in HMX and RDX
manufacture, while the remainder is used in the manufacture of acetic
anhydride. This latter product is produced by cracking the glacial
acetic acid in furnaces, with the water of reaction removed by conden-
sation. Uncracked acetic acid from the reaction is removed by azeotropic
distillation and recycled to the cracking furnaces. Ketene gas, formed
in cracking, is reacted with additional acetic acid to yield crude acetic
anhydride. The crude product is refined by distillation, and used in HMX
and RDX manufacture.
-35-
-------
D - Bldg
E-Bldg
Explosives
nexamme
Acetic Acid """
Nitric Acid
Amm . Nitrate
Acetic
Anhydride
Nitration
Reactors
Slurry
i
Dilution
Liquor
Slurry:
Sed. & Vac.
Filtration
(Initial , and
wash with water)
Slurry i
•» Water tc
first Acid Filte
( Washwater Filtrate)
B - Bldg
Acetic
Acid)
Crystallized Explosive
Neutralized and
to Storage Lagoons
Recovered by
Condensation for use in AOP
Sludge
Primary
Distillation
1 . Neutralization
2. Evaporation
NH + Water Vapor
o
AREA A
Acetic
Acid
(60$
Condensed)
AREA B
Crud«
Acetic Anhyride
Acetic
Anhydride
Manufacture
To Area B
Refining
(Distillation)
Bldg 7 and 8
Glacial Acetic Acid
Glacial (99$)
Acetic Acid
Acetic Acid
Concentrator
(Azeotropic
Distillation)
Bldg 2
Bldg 6
Acetic
Anhyride
•*• To Area B
Figure 7 - Acetic Acid and Acetic Anhydride Manufacture
-36-
-------
16. Waste Sources
a. As shown in Figure 7, spent acetic plus nitric acid is
recovered in the first filtration step of RDX or HMX, and sent to pri-
mary distillation. A wash with water, to remove residual acidity, is
recovered and used as dilution liquor in the hexamine nitration procedure.
b. The primary distillation units receive contaminated weak
acetic acid, and separate it into dilute acetic acid, ammonia and sludge
fractions. Wastewater from the units averages 11.2 MGD* and consists
primarily of cooling water mixed with some steam condensate containing
acetic acid and ammonia, and equipment and building washdown water (ld(2))
This combined flow is reported to contain a COD of 20-56 mg/1 (ave. 32),
and ammonia-nitrogen of 0.7-5.6 mg/1 (ld(2)). However, another source
has reported an ammonia-nitrogen concentration of 900 mg/1 in the con-
taminated fraction of the waste flow, with dilution by cooling water re-
ducing this to 5 mg/1 (3f). This same source reports that the waste
also contains BOD, acetic acid and explosive, although concentrations
are not given. Based upon individual discharge volumes (ld(2)), it is
estimated that only approximately one percent of the total flow is con-
taminated process water, with the remaining 99 percent being cooling
water.
c. The acetic acid azeotropic distillation facility concen-
trates dilute acetic acid received from the primary distillation line
(Figure 7). The azeotroping agent added is n-propyl acetate. Water
and n-propyl acetate are removed by evaporation, recondensed, and phase
separated. The water from the phase separation decanter (sweet water)
is passed through a flash column to recover as much n-propyl acetate as
possible, before discharge to waste. The 99+ percent glacial acetic
acid is removed from the column bottoms, for explosives manufacture re-
use or for acetic anhydride manufacture.
d. During azerotropic distillation, there is an undesirable
buildup of solids in the distillation column. Sludge bleedoff is nec-
essary, with this material being sent to a sludge heating operation to
distill additional acetic acid. The sludge is then discharged. Spent
sludge may contain high concentration of heavy metals (Cr, Cu, Fe, and
Mn) from corrosive destruction of distillation columns (ld(2), 4a).
In addition to the flash column and sludge heater discharge (312,000
gpd) cooling and condenser water are discharged at 24 MGD (4a). Pro-
cess effluents range in PH from 2.8 to 3.8 and contain nitromethane,
methyl nitrate, acetic acid, n-propyl acetate, nitric acid and trace
amounts of explosives (4a). Although the wastes are not well character-
ized, as a result of the high percentage of dilution water, available
data strongly suggest that the process effluents are extremely high
in COD, complex organics and possibly heavy metals.
*million gallons/day
-37-
-------
e. In the acetic anhydride manufacturing process, glacial
acetic acid is vaporized and fed to a cracking furnace together with
triethyl phosphate, a reaction catalyst. The vapor stream, containing
ketene, is sequentially scrubbed by four recycling glacial acetic acid
scrubbers to yield crude acetic anhydride. Vapor leaving the fourth
scrubber goes through a once-through water scrubber, which discharges
to waste. The vapor from the water scrubber goes to a barometric con-
densor. Non-condensibles vent to the atmosphere, and condensate is
discharged as wastewater. The process effluent is reported as 0.5-0.55
MGD (4a), which is combined with a cooling water discharge of approx-
imately 2.3 MGD (ld(2)). The process effluent contains acetic anhydride,
acetic acid, acetaldehyde, acetonitrile, methyl nitrate, ethanol, meth-
anol, ethyl acetate, propanol, propyl acetate and other organics. The
water scrubber effluent BOD is reported as 743 mg/1 (4a).
f. The final step in acetic acid recovery and acetic anhydride
manufacture is refining of the crude acetic anhydride (Figure 7). Re-
fining is by distillation in two-stage columns heated in the lower stage.
Refined anhydride is withdrawn from the lower stage, sent to a second
small column for removal of color bodies, and stored for use in explo-
sives manufacture. The vapors from the top of the two-stage anhydride
column contain acetic acid and acetic anhydride, plus some impurities.
This vapor is condensed and sent to a stripping column where it is sep-
arated into acetic acid, anhydride bottoms (returned to the refining
column), and acetic acid vapors. These vapors are concentrated to
glacial strength by azeotropic distillation.
g. Sludges from the refining column are heated under vacuum
to recover additional acetic anhydride. The residual sludge is dis-
charged as waste (70,000 gpd). Sludge from the azeotropic stills is
acidified with sulfuric acid to break down acetamide, heated to recover
acetic acid, and discharged (8,100 gpd). There is also a 280 gpd flow
from a flash column involved in acetic acid concentration. This flow
is reported to contain acetone, ethyl acetate, and acetonitrile (methyl
cyanide)(4a). These discharges are combined with cooling and condensing
water, to yield a total discharge of about 9.5 MGD.
h. Table 8 summarizes available information on process and
total waste discharge volumes, and Table 9 available information on
combined discharge waste characteristics. Dilution water exceeds 98
percent of total discharge for all except acetic anhydride manufacture,
for which it represents about 80 percent. Despite the high degree of
dilution, kjeldahl nitrogen is notably high in the acetic anhydride re-
fining effluent, as are phosphorus and sulfate. High COD's are indicated
in the primary distillation effluent, and in Outfall 1 (Table 9) dis-
charge from acetic acid concentration. Acetic acid is present in all
wastewater where-monitored, at less than 3 mg/1.
17. Effects of Water Management
The data of Table 8 indicate that strict segregation of cooling
water from process waste will result in a ten to fifty-fold reduction
-38-
-------
Table 8. Wastewater Volumes from Acetic Acid Recovery and Acetic Anhydride Manufacture
Process
1. Primary Distillation
2. Acetic Acid Concentration
3. Acetic Anhydride
Manufacture
4. Acetic Anhydride
Refining
TOTAL, MGD
Process Effluent
Source
Steam Con dens ate
Sludge & Flash Column
Scrubber & Con dens ate
Sludge & Condensate
Volume, gpd
120,000*
312,000
500,000-550,000
78,380
1.01-1.06
Cooling
Water,
MGD
11.1
23.7
2.3
9.4
46.5
Total
Flow,
MGD
11.2
24.0
2.8-2. 85
9.5
47.5
Percent
Cooling
Water
99.1
98.7
80.7-82.1
98.9
97.9%
*Estimated.
-------
Table 9. Combined Wastewater Characteristics from Acetic Acid
Recovery and Acetic Anhydride Manufacture (ld(z))
Parameter*
Temperature °F
PH
N02 + NO,,-N
NH3-N
Kjeld-N
Sol. Phosphorus
Total Solids
Susp. Solids
Diss. Solids
COD
TOC
BOD
Sulfate
Phenol
Acetic Acid
Primary
Distillation
68.9
7.7
1.24
1.72
1.73
< 0.03
319
11.7
307
32
10
-
_
-
~"
Acetic Acid
Con cen t r at i on
Outfall 1
82.1
7.3
1.60
-
0.8
0.72
132
7.6
124
103
35.0
>67
16.0
< 0.05
< 2.0
Outfall 2
61.5
7.8
1.29
-
0.5
0.04
137
5.4
131
16
6.4
1.0
15.3
<0.05
2.9
Acetic Anhydride
Manufacture
Outfall 1
70.8
7.7
1.2
-
0.12
0.04
122
6.0
116
11
6.5
2.0
17.5
0.05
2.0
Outfall 2
70.5
7.3
0.93
-
<0.5
0.06
126
9.1
117
10
5.8
<1.0
16.8
<0.05
<2.0
Acetic Anhydride Refining
76.3
7.8
1.04
-
10.5
34.6
319
6.7
121
17.1
10.5
11.3
31.2
<0.05
1.8
I
-e-
o
i
*A11 as mg/1 except temperature and pH.
-------
in contaminated wastewater volume, yielding 1.0-2.4 MGD wastewater. It
has been suggested that increased recycling, reuse, and recovery of pro-
cess flows, together with cooling water segregation, could hold the total
process wastewaters in the last three processes of acetic acid recovery and
anhydride manufacture to 1 MGD or less (4a). Available data indicate
that these process effluents would be extremely high in BOD and complex
organics (including acetic acid), and perhaps toxic due to both toxic
organics and heavy metals (from anhydride sludges). Segregation of pro-
cess from cooling waters has been scheduled for the acetic acid concen-
tration process, with process effluent neutralization and treatment in
an aerated lagoon (Id(4)). Segregation of cooling and process wastes for
all facilities is reported due by early 1976 (ld(4)). However, current
expectation is that while implementation of waste segregation will pro-
ceed at Holston AAP on schedule, no actual treatment will be implemented
until operation of the lagoon commences in February, 1976 (4a).
18. Current Treatment Practice
Pollution from acetic acid use results from the acid recovery and
acetic anhydride manufacture processes. All spent acid is recovered,
with dilute acid wash waters recycled as dilution liquor to RDX and HMX
nitration processes. Sludge from primary distillation of acetic acid
filtrate is neutralized with sodium hydroxide and held in storage lagoons.
When a sludge processing plant is completed, this sludge will be con-
verted to fertilizer, and sold. Wastewater from primary distillation,
at 11.2 MGD (99.1% cooling water), is discharged without treatment. All
wastewaters and sludges from acetic acid concentration and acetic anhy-
dride manufacture and refining are currently discharged without treatment.
19. Modifications and Alternatives to Current Practice
a. By flow segregation, process wastewaters can be reduced
to 120,000 gpd in Area B (primary distillation) and approximately 1 MGD
in Area A (acetic acid concentration, and anhydride manufacture and re-
fining) . In addition it appears likely that in Area A best treatment
will dictate separate collection and handling of the sludge and liquid
fractions of the process effluents. Separate sludge handling has al-
ready been implemented for the primary distillation process in Area B.
b. Current plans for the liquid waste from primary distilla-
tion are to combine it with other liquid wastes from explosives manu-
facture and treat in an aerated lagoon (3f). However, bench scale
attempts to biologically treat and carbon adsorb the wastes have been
unsuccessful (5c). Biological treatment by activated sludge and trick-
ling filtration resulted in effluent COD concentrations of 55 and 87 mg/1
respectively, for influent COD values of about 350 and 450 mg/1. Dif-
ficulties with sludge bulking were also experienced in the activated
sludge treatment. Denitrification experiments were successful in ni-
trate removal, although buffering of the system was required to maintain
a pH below 9- Carbon adsorption experiments on both raw waste and
biologically treated effluent yielded unfavorable adsorption isotherms,
-41-
-------
and a carbon column effluent of 16 mg/1 BOD, for the biological process
effluent, with influent of about 40 mg/1 (5c). These preliminary carbon
experiments also indicated rapid breakthrough.
c. The biological ,treatability results cited above suggest
that treatment of the primary distillation wastewater by the proposed
aerated lagoon may be unsuccessful. Due to the relatively small quan-
tity of wastewater from primary distillation, serious consideration
should be given to separate treatment of this flow- Among processes
which should be considered are chemical coagulation and ozonation, per-
haps followed by carbon adsorption or biological treatment.
d. Initial plans for wastewater treatment in Area A called for
segregation of cooling water from process effluent, to be treated by neu-
tralization and an aerated lagoon which was to incorporate biodenitrifi-
cation (3h, ld(4)). However, serious questions have been raised concern-
ing the suitability of aerated lagoon treatment (4a). These questions
relate to the level of treatment achievable by the aerated lagoon, the
reliability of the process, and the fact that bioassay studies on Area A
wastes indicate a high degree of toxicity (4a). The presence of toxic
materials would seriously interfere with the efficiency of the proposed
biological treatment. Since the wastes are not well characterized, it
is not known which process effluent or effluents contribute to the tox-
icity effects. If biological treatment is to be implemented, however,
toxic waste streams must be segregated for pretreatment to reduce tox-
ixity, or alternately totally separate treatment by physical-chemical
means.
e. A recent architect/engineer treatment concept report to
Holston AAP suggests a sequence of activated sludge, followed by deni-
trification, mixed media filtration, and tertiary treatment if required
to remove toxic substances. The tertiary treatment process has not been
selected (5b). An activated sludge-denitrification system is to be pilot
tested at Holston AAP (15zi). Providing toxic components do not inter-
fere, this system should achieve acceptable treatment for BOD, nitrogen
and suspended solids. It will not achieve satisfactory phosphate re-
moval, phosphate being particularly high in the anhydride refining ef-
fluent (Table 9), as is sulfate and kjeldahl-nitrogen. Although the
wastes of Area A are not well characterized due to the high percentage
of dilution (cooling) water, such process stream constituents indicate
that a more reasonable and effective approach to selection of treatment
process would involve improved characterization of each process efflu-
ent, and consideration of individual treatment, such as phosphate pre-
cipitation, carbon adsorption, and chemical oxidation of those process
effluents which show unusual composition.
f. Sludges resulting from acetic acid concentration and an-
hydride manufacture and refining should be separately handled and dis-
posed. Recent information is that serious consideration is now being
given to incineration of these sludges in a coal gasification plant at
Holston AAP (15zi). Should this not prove feasible, the most probable
alternative is to dewater and remove these sludges to approved landfill.
-42-
-------
20. Impact of Air Pollution Control
Several side reaction products result from the manufacture of HMX
and RDX, and are returned with the used acetic acid to the acetic acid
recovery and purification process. These side products include methyl
nitrate, methyl acetate, propyl formate and propyl acetate. At present,
these contaminants are released as air pollutants in the vent gas from
the acetic acid azeotropic distillation system. Because of the toxic
and sensitive nature of methyl nitrate, these materials will ultimately
be contained within the system by condensation, and handled as liquid
wastes (3h). Preliminary biodegradation studies by the Army on decanter
"sweet water" from the acetic acid recovery process indicate better than
97% methyl nitrate reduction, with other contaminants reduced by at
least 90% (3i). Results are reported in Table 10. Based upon these
biological treatability studies, high effluent levels of organic con-
stituents are to be expected, and a tertiary process such as carbon ad-
sorption will likely be required.
21. Summary
Despite the lack of reliable waste characterization data, which
results from the high degree of process effluent dilution by cooling
water, and indications that at least some process effluents are toxic,
current plans call for biological treatment of acetic acid and anhydride
wastes. Serious questions have been raised concerning the suitability
of biological treatment, which will in any case be ineffective for phos-
phate control for the anhydride refining effluent. Final plans for
waste treatment should be preceded by waste characterization studies ,
and both biological plus physical-chemical pilot treatability studies
on individual as well as combined process effluents.
-43-
-------
Table 10. Biodegradability of Acetic Acid Recovery
Decanter Sweetwater (3i)
Component
Methyl Nitrate
Methyl Formate
Acetone
Methyl Acetate
Ethyl Acetate
Propyl Formate
Nitromethane
Propanol
Propyl Acetate
Concentration, mg/1
Initial
616
12
18
1,266
49
12,548
1,232
4,451
"High"
Final
19
-
-
25
-
502
16
13
999
Percent
Reduction
96.9
-
-
98.0
-
96.0
93.7
99.7
-
-44-
-------
SECTION III-B - NITRIC ACID-MANUFACTURE AND CONCENTRATION
22. Introduction
a. Facilities to manufacture weak nitric acid, by the ammonia
oxidation process (AOP), are available at Badger, Holston, Indiana,
Joliet, Newport, Radford, Sunflower and Volunteer AAP's. All except
Holston AAP also have facilities to produce concentrated nitric acid by
the sulfuric acid dehydration technique. Holston AAP uses a hot magnes-
ium nitrate dehydration process for nitric acid concentration (3f). Sim-
ilar processes are used to produce weak and concentrated nitric acid in
the commercial explosives industry, and that industry's waste character-
istics and pollution abatement technology have been described in a recent
report (4d).
b. Weak nitric acid is used to produce concentrated nitric
acid, for nitration of explosives intermediates. Spent nitration acid
is normally processed to purify and recover nitric and/or sulfuric acid.
Recovery is incorporated into the nitric acid concentration (NAC) pro-
cess, with spent nitration acid being bled into the dehydration step
along with weak nitric plus concentrated sulfuric acid. Since nitration
acid is a mixture of concentrated sulfuric and nitric acids, separation
of sulfuric and recovery of nitric acid occurs, where practiced, in the
NAC step, and recovery of spent nitration acid will be discussed as part
of the NAC process. At Holston AAP only, concentrated nitric acid is
mixed with liquid anhydrous ammonia to produce nitric acid/ammonium ni-
trate, used in HMX and RDX manufacture (3f). Unlike the commercial ex-
plosives industries, where ammonium nitrate is a commonly manufactured
and used explosives intermediate, the Holston AAP manufacturing process
for HMX and RDX represents the only current military production of
ammonium nitrate.
23. Waste Sources
a. In the AOP process, a mixture of ammonia gas and air is
combusted in the presence of a catalyst to produce nitric oxide plus
water. The gaseous products are cooled and excess air added to convert
nitric oxide to nitrogen dioxide. The water formed in the ammonia ox-
idation step reacts with nitrogen dioxide, producing nitric acid and
nitric oxide. Cooling and condensing results in the formation of dilute
(50-60%) nitric acid, which is sent to nitric acid concentration. The
nitric oxide produced in the final reaction is again oxidized by excess
air to nitrogen dioxide, which is converted to dilute nitric acid by
aqueous scrubbing in countercurrent adsorption towers.
b. Wastewaters from AOP facilities include cooling water and
leakage, and water used for clean-up of spills, floors and equipment.
Because AOP process effluents are normally mixed with large volumes of
cooling water, and frequently combined with wastes from NAC and sulfuric
acid processes, it is difficult to specify the pollutant discharge solely
associated with AOP operations. In general the AOP effluents are low in
pH and high in nitrate content, with low to moderate ammonia levels (8f).
-45-
-------
Flow data from the Holston AOP facility are summarized in Table 11.
Approximtely four percent of the total flow is process effluent. Based
upon the data of Table 11, 508 Ibs/day of ammonia are discharged. This
is equal to about 40 mg/1 in the combined flow, assuming no contribution
from floor washdown. Approximately 250 Ibs/day nitrate discharge is
reported for the Joliet AAP ammonia oxidation process (8c). This is
equivalent to about 4000 mg/1 nitrate for the AOP process effluent alone,
at a process flow of 12.3 gpm (17,760 gpd). Beyond the expected acidic
pH, no other wastewater characterization data specifically related to
AOP operation are available. However, the above cited data indicate
that AOP process effluent will contain appreciable quantities of nitrate
and ammonia nitrogen.
c. The standard nitric acid concentration (NAC) operation is
a continuous process in which dilute nitric acid is distilled in the
presence of sulfuric acid. The sulfuric acid acts as a dehydration
agent. Concentrated (98%) nitric acid is condensed from the distilla-
tion vapor, and diluted sulfuric acid is removed, normally for recon-
centration in on-plant sulfuric acid concentration (SAC) facilities.
Holston AAP uses a modification of this process, wherein 61% nitric acid
is concentrated by mixing with a hot solution of 72% magnesium nitrate
(made by dissolving magnesium carbonate in nitric acid), which acts as
the dehydrating agent. The concentrated nitric acid is removed as vapor,
condensed, cooled and stored. The diluted magnesium nitrate is recover-
ed, reconcentrated and reused. Wastewaters from NAC processes orginate
from acid spills, floor washings and cooling water (8f).
d. Except for Holston AAP, there are essentially no process-
specific wastewater characterization data available for the NAC process.
The applicability of the Holston data, which is based upon magnesium
nitrate dehydration rather than the more common sulfuric acid dehydra-
tion process, is questionable for extrapolation to NAC wastes at AAP's
operating sulfuric acid-based NAC facilities. Data from Holston AAP
indicate a flow of about 9000 gpm of contaminated process water, with
a cooling water and steam condensate flow of 1.4 MGD at a production
rate of 51 TPD* nitric acid (ld(2)). Waste characteristics of the com-
bined flow are summarized in Table 12. For comparison with the Holston
process flow, Newport AAP is reported, when in operation, to have a
process wastewater flow of only 10 gpm from its nitric acid concentra-
tion and recovery units (3f). Waste characteristics are not reported.
However, for Newport and other AAP's utilizing sulfuric acid in their
NAC facilities, sulfate plus contaminants present for Holston AAP
(Table 12) would be anticipated.
e. Because of the lack of available data on process effluent
from either AOP or NAC facilities, it is difficult to define the levels
of contaminants'. There are data available, however, on combined flows
from AOP plus NAC operations. Combined process wastewater data are
*tons per day
-46-
-------
Table 11. Holston AAP Wastewater Volumes (ld(2))
Source
1.
2.
3.
4.
5.
6.
7.
Ammonia Purge Pot
Slowdown
Ammonia Vaporizer
Slowdown
Adsorption Column
Slowdown
Adsorption Column
Demister Slowdown
Floor Washdown
Steam Condensate
Cooling Water
Total, gpd
Cooling Water, %
Process Waste, %
Volume , gpd
46,000
not measured
not measured
5-7
1,440
7,554
1,439,000
> 1 ,^9^ ,000
<96.3
>3.7
Comments
Approximately 88 Ibs/day NR^.
Approximately 420 Ibs/day NH-.
Approximately 112 Ibs/day HNO .
pH 0.1, TOC 28 mg/1, COD 95 mg/1,
NH3-N 37.2 mg/1, N03 + N02~N 100,000 mg/1.
Note: Production rate of 200 TPD equivalent 100% nitric acid.
-------
Table 12. Wastewater Characteristics of Holston AAP
Magnesium Nitrate-Based NAG Process (ld(2))
Parameter
Flow, MGD
pH
Total Solids, mg/1
Susp. Solids, mg/1
COD, mg/1
TOC, mg/1
NH3-N, mg/1
N02 + N03-N, mg/1
Average
1.8
6.3
603
7.1
36
6.8
0.5
29
Range
0.9-2.8
3.0-8.2
480-876
4.0-12.7
16-51
6-8
0.2-1.6
12-60
-48-
-------
presented in Table 13. Similar (averaged) data for commercial explosives
plants are also included in Table 13, for comparison. There is a great
deal of variability in the data of Table 13, due to variation in extent
of dilution of process effluents with cooling water, among other factors.
The data suggest however that the most significant wastewater character-
istics are low pH, with high ammonia and nitrate nitrogen, and high
sulfate levels.
f. At Joliet AAP, a Direct Strong Nitric Acid (DSNA) process
has been constructed and is undergoing start-up. This new process will
replace the existing AOP/NAC facility at Joliet. The DSNA process will
produce both weak and strong nitric acid, in a single operational se-
quence, without the use of sulfuric acid (8f). Although there is no
wastewater information on the new process, it is claimed to be a much
more pollution-free facility than the present AOP and NAC facilities at
Joliet AAP. Major wastewater flows will be cooling tower and boiler
blow-down, clean-up water and regenerant brine from water demineraliza-
tion units (8c, 8f).
24. Effects of Water Management
a. Two areas of water management are under consideration for
AOP/NAC facilities. These are: separation of cooling water from pro-
cess effluent; and use of process effluent from NAC facilities as water
source for both nitric acid production and NOo abatement. Based upon
the data of Table 11, approximately 96 percent of the total AOP dis-
charge at Holston AAP is cooling water. Separation of this flow would
reduce the process contaminated effluent from 1028 to 38 gpm. Similar-
ly, whereas the Holston NAC facility has a combined discharge of 1250
gpm (Table 12), the modern Newport NAC with flow segregation has a pro-
cess effluent of 10 gpm (3f). Thus, for existing AOP/NAC facilities,
flow segregation will result in a major reduction in the volume of
wastewater to be handled. Use of the new DSNA process at Joliet AAP
is expected to eliminate process effluent (8f).
b. The use of process effluent for the NAC facility at
Holston AAP, as a water source for nitric acid production and NOo abate-
ment, has been suggested (lr(l)). Advantages for this reuse concept in-
clude a 59 percent reduction in nitric acid discharge, and economic
benefits through acid recovery and reduced waste treatment costs. Pos-
sible obstacles to implementation include lack of demonstrated applica-
tion, and requirements imposed by physical-chemical manufacturing sys-
tems presently in use, safety, and product quality requirements (lr(l)).
The potential benefits that would result from effluent use however in-
dicate that serious consideration should be given to the application of
this concept.
25. Current Treatment Practice
a. Table 14 summarizes treatment now employed at AAP's for
AOP/NAC facility wastes. At all plants except Newport, the wastewater
-49-
-------
Table 13. Comparison of Characteristics of Combined Wastewaters from AOP plus NAC Facilities
Parameter
Flow, MGD
PH
BOD, mg/1
COD, mg/1
NH3~N, mg/1
N03 + N02~N, mg/1
Sulfate, mg/1
Joliet AAP
Areas 1 & 2 (3f)
21.2
7.2
na
na
na
26
373
Area 3 (8f)
11.2
2.7
<10
40
na
6.8
320
Radford AAP (3f)
na
na
13
46
1.2
163
na
Hols ton AAP (5 a)
3.9
na
7.2
19.4
0
94
na
Commercial
Explosives
Plants (4d)
0.49
2.3-3.1
13.5
208
191
206
312
Ul
o
Numbers in parenthesis are reference sources.
-------
Table 14. Current Treatment Practices at AOP/NAC Facilities (3f)
Location
Current Treatment
Holston AAP
Radford AAP
Joliet AAP
Badger AAP
Newport AAP
Volunteer AAP
No treatment.
Equalization, neutralization with soda ash for
spent TNT nitration acids. Neutralization with
lime for acid plant wastes, and sludge sedimen-
tation.
Neutralization.
Neutralization with lime, aeration.
Neutralization with lime. Calcium sulfate sludge
sedimentation.
Neutralization with lime. Calcium sulfate sludge
s e dimen t at ion.
-51-
-------
is predominately cooling water, contaminated with a much smaller flow of
process effluent. Current practice ranges from no treatment (Holston),
to neutralization, and at three plants calcium sulfate sludge removal.
Where treatment is applied, it is thus restricted to pH adjustment plus
partial sulfate removal by precipitation of calcium sulfate. The pre-
cipitation of calcium sulfate is limited to those process wastewaters
high in sulfate. Since AOP/NAC wastes have low to moderate sulfate
levels (Table 13), calcium sulfate precipitation would be minor except
in the presence of wastewaters representing a combination of AOP/NAC ef-
fluents and high sulfate (e.g. SAC) wastes. In that instance, calcium
sulfate precipitation would be most effective on the high sulfate stream
alone, rather than a combined waste stream.
b. Where treatment is employed, acceptable effluent levels
appear to result directly from process effluent dilution, rather than
application of good treatment technology. For example, neutralization
at Radford AAP is reported to be inadequate, with wide fluctuations in
the effluent pH and poor solids removal in the settling lagoons (4c) .
However, the (diluted) discharge is reported to have an average pH of
8.6 and average suspended solids of 1.2 mg/1 (3f). Joliet AAP reports
an effluent pH of 7.2 from Acid Areas 1 and 2, with a total combined flow
of 20 MGD (3f). Acid Area 3 pH ranges from 2.6 to 9.7, with a flow of
11.2 MGD (3f). Combined flow from Badger AAP ranges in pH from 1.8 to
11.5, with an average flow of 29 MGD. Nitrate and sulfate discharge for
this plant-wide combined discharge were 16-105 (72 average) and 225-600
(426 average) mg/1, respectively (3f).
c. Volunteer AAP uses a feed-forward/feed-back pH control
system, through three sequential treatment ponds, for neutralization
and suspended solids removal. The system appears effective for pH con-
trol, with acid area wastewater (AOP, NAG, SAC and Oleum manufacture)
being adjusted from pH 3 to 6.9 by addition of lime through the system.
Neutralization plant effluent ammonia, nitrate plus nitrate nitrogen,
and sulfate, are reported to be 1.9, 12.5 and 350 mg/1 (lr(l)). Efflu-
ent calcium is 177 mg/1, reflecting the high lime dosages used in neu-
tralization, although total dissolved solids are only 767 mg/1. Based
upon the performance of the Volunteer AAP neutralization system, effec-
tive pH control is possible. Other AAP's however provide less effective
control, and consequent inadequate neutralization. In no instance is
nitrogen or rigorous sulfate pollution abatement employed.
26. Modifications to Current Treatment
a. Treatment modifications currently under consideration for
five AAP's are listed in Table 15. In all except two instances (Joliet
and Volunteer AAP's), no provisions for nitrogen or sulfate control are
scheduled. At Joliet AAP however, essentially complete deionization is
under consideration (8c). Proposed modifications thus represent a full
spectrum from continuation of neutralization but no control of other
constituents, to removal of all suspended and dissolved pollutants.
-52-
-------
Table 15. Proposed Modifications to Current Treatment
Location
Modification
Reference
Holston AAP
Radford AAP
Joliet AAP
Badger AAP
Newport AAP
Volunteer AAP
Aerated treatment ponds.
Improved neutralization, and removal
of calcium sulfate sludge.
Sulfate removal by calcium and barium
precipitation; nitrate and ammonia
removal by ion exchange.
Reduction in process effluent by sub-
stitution of DSNA for AOP/NAC
facilities.
Improved neutralization.
Neutralization, and calcium sulfate
sludge removal.
Lime neutralization, clarification
and filtration. Filter cake to be
landfilled. Residual nitrate and
sulfate to be treated by ion
exchange.
3f
3f
8c
15 zb
3f
3J
15zp
-53-
-------
b. Holston AAP combined wastes, including nitric acid process
effluents, are scheduled to be treated in aerated biological waste
treatment ponds (3f, 4a). Flows from the nitric acid area contain low
values of BOD and COD. Other constituents such as nitrate and sulfate
will not be affected by pond treatment. Ammonia nitrogen might possibly
be biologically oxidized to nitrite or nitrate, depending upon the lagoon
design and mode of operation. However, aside from the small amount of
BOD, and possible ammonia nitrification, the major pollutants (nitrate
and sulfate) will not be reduced or eliminated in aerated pond treatment.
Because of this limitation of biological treatment, a recent EPA review
of Holston AAP recommended: (1) steam stripping of ammonia-rich waste
stream prior to pond treatment; (2) second stage polishing treatment of
the pond effluent by activated carbon or some similar physical or chem-
ical means; and (3) biological denitrification of the second stage ef-
fluent for nitrate control (4a). Sulfate removal was not discussed.
Alternatives to the EPA proposal are discussed and evaluated in para-
graph 27.
c. Radford AAP proposes to modify the present neutralization
facility, which now employs soda-ash, by: providing equalization tanks;
additional neutralization tanks for longer retention time and better mix-
ing; switching to lime neutralization; and providing calcium sulfate
sludge removal facilities (6a). The treatment system modifications should
result in improved pH control, but little sulfate and no nitrate control.
Implementation of more rigorous treatment was apparently discouraged as
a result of the regional U. S. EPA office indicating satisfaction with
neutralization only at Radford AAP, through 1977 (15y).
d. A new neutralization and ion exchange plant has been con-
structed at Volunteer AAP, to handle wastes from the acid manufacturing
facilities (3f). At Badger AAP, the new acid waste treatment facility
is limited to neutralization, and aeration to increase dissolved oxygen
levels (3f). Treatment at Newport AAP will be limited to neutralization,
with calcium sulfate sludge separation (3j). There is no provision for
eliminating nitrate, nor reducing sulfate below the solubility of calcium
sulfate, at either of these two AAP's. At Newport, neither the present
system, nor any proposed MCA* projects, will provide sufficient treat-
ment to satisfy the plant's present discharge permit (3j).
e. Pollution abatement plans for the plants cited above are
in distinct contrast to the proposed treatment system for Joliet AAP.
At Joliet, a $2.5 million waste treatment facility has been proposed,
which incorporates two-stage lime neutralization with calcium sulfate
sludge removal, barium carbonate treatment to precipitate residual sul-
fate as the barium salt and calcium as the carbonate salt, followed by
cation/anion exchange treatment to produce demineralized water for re-
use. Anion regenerant will be concentrated by evaporation, and sold
as an agricultural fertilizer material. Cation regenerant (dilute
sulfuric acid) will be treated by the sulfate removal process described
^"Military Const ruction-Army
-54-
-------
above (8c). Nitrate removal will occur in the anion exchange process.
Waste ammonia streams will not be handled in this proposed system, al-
though the cation exchanger should remove any trace ammonia present.
The consultant engineering firm which developed the concept design for
the Joliet system has recommended pilot testing the ion exchange system,
as well as an alternative biological denitrification process (8c).
27. Alternatives to Current Practice
a. The four parameters of wastewaters from nitric acid manu-
facture and concentration which are likely to require control are pH,
ammonia-nitrogen, nitrate-nitrogen, and sulfate. Current practice, par-
ticularly at Volunteer AAP, demonstrates that proper treatment facility
design and operation can achieve effective neutralization. Treatment
techniques and costs for pH control are equivalent to those well estab-
lished for other industry wastewaters (4d). The control of sulfate, re-
sulting from the use of sulfuric acid in NAG facilities, will be dis-
cussed in the following (sulfuric acid) section of this chapter. The
discussion in this section will therefore be restricted to control of
ammonia and nitrate.
b. Biological treatment for both ammonia and nitrate appears
to be an effective and acceptable type of treatment for munitions wastes,
providing that the wastewaters do not contain toxic materials at biolog-
ically harmful levels. Although most research on bionitrification of
ammonia has been directed toward wastewaters containing less than 60 mg/1
NHo-N, limited work on more concentrated wastewater has proven bionitri-
fication to be effective. Recent data on bionitrification processes are
summarized in Table 16. These data suggest that while the activated
sludge process is an effective nitrification process at moderate ammonia
levels, it is much less effective than either the oxidation ditch or
rotating disc for high ammonia levels. The feasibility of ion exchange
or other physical-chemical methods of ammonia control has also been
proven (16ze, 16zf), although the high nitrate levels in military muni-
tions wastewaters indicate that a combination of biological nitrification-
denitrification is likely to be the treatment method of choice.
c. A great deal more effort has been directed toward assess-
ment of nitrate control technology than for ammonia. Table 17 summarizes
the results of preliminary feasibility studies performed by the Army on
nitrate control technology, and indicates that on the basis of proven
treatment technology plus costs, biodenitrification is likely to be the
method of choice. However, recent developments in ion exchange and re-
verse osmosis technology, together with the prospect of valuable product
recovery, indicate that serious consideration should be given to these
alternatives. In the commercial explosives industry, evaporative ponds
and spray irrigation have found wide acceptance (4d). However, restric-
tions imposed by climatological factors, land availability, nature of
terrain, and volumes of wastewater indicate that these latter two al-
ternatives may not be feasible for the military munitions industry.
-55-
-------
Table 16. Ammonia Bionitrification Processes and Performance Data
i
Ul
Process
Activated Sludge
Activated Sludge
Oxidation Ditch
Rotating Disc
Detention
Time
10.7 Hr.
11.5 Hr.
4.5-6.5 Days
6.5-8.5 Days
2.0 Days
1.5 Days
2.0 Days
3.0 Days
Ammonia - N, mg/1
Influent
24.5
92.0
600
600
577
785
880
790
Effluent
1.0
1.0
138
72
5.0
2.3
1.5
1.7
Percent
Nitrification
96+
99+
77
88
99+
99+
99+
99+
Reference
I6za
16 zb
16zc
16zd
-------
Table 17. Summary of Nitrate Treatment Methods (3d)
Method
Biodenitrification
Algae Harvesting
Ion Exchange
Electrodialysis
Chemical Reduction
Reverse Osmosis
Distillation
Land Application
Removal
Efficienty, %
70-95
50-90
80-99
30-50
33-90
50-96
90-98
5-15
Approx. Cost,
$/MG
3.45-30
20-35
170-300
100-250
-
100-600
400-1000
na
-57-
-------
d. Of the three remaining nitrate treatment processes, bio-
logical denitrification has been most intensively studied. Most of those
investigations have been directed toward removing nitrate from municipal
wastewaters (4e). The U. S. Army has carried out pilot scale suspended
growth biodenitrification studies at two of its contractor operated mu-
mitions plants (3d). Initial studies at Sunflower AAP indicated that de-
tention times exceeding, ten days were required for 90 percent or great-
er nitrate reduction, at suspended solids levels of 100-200 mg/1. A mod-
ified 5000 gpd pilot unit tested at Radford AAP, and operated at 1000-
3000 mg/1 suspended solids, is reported to have achieved 93 percent den-
itrification with an 18 hour detention time. Average influent N03~N of
558 mg/1 was reduced to 47 mg/1 effluent concentration. At influent
NC>3-N concentrations below 400 mg/1 and suspended solids below 1400 mg/1,
the process yielded effluent nitrate levels of 1 mg/1, however (3i). The
primary disadvantages of the biodenitrification process are the need for
close pH and temperature control, and the large quantities of methanol
required. At a nitrate level of 500 mg/1, about 5 tons of methanol is
required per million gallons of waste treated (3f).
e. The recent development of anaerobic packed and suspended
growth bed columnar type biodenitrification processes is promising, in-
sofar as reduction in treatment time required. Pilot studies with a 26
gpm pilot suspended growth system have resulted in nitrate reduction
from 21.5 to 0.2 mg/1, representing 99 percent denitrification at a de-
tention time of only 6.5 minutes (16zg). Studies on a similar system
using a munitions wastewater containing approximately 100 mg/1 nitrate
nitrogen and with a 152 minutes detention time resulted in 98.9 percent
nitrate removal (16zh). Preliminary results such as these for high rate
biodenitrification indicate that this process, by improving process re-
liability and reducing required treatment time, may replace activated
sludge type denitrification as the preferred method. Pilot biodenitri-
fication studies in an upflow column are programmed to be performed at
Badger AAP (15y).
f. Ion exchange treatment for nitrate removal has been largely
limited to the abatement of low nitrate concentrations (16zi-16zk). One
historical problem with the use of ion exchange for nitrate treatment has
been the lack of resin specificity for nitrate. However, nitrate selec-
tive resins have recently become available (4f, 16zl). The only reported
ion exchange system in full-scale use for the treatment of high nitrate
concentrations is the Chemical Separation Corporation Continuous Counter-
Current (Chem-Seps) Ion Exchange System (3d). This system has been in-
stalled at the Farmers' Chemical Association, Inc. plant (Tyner, Tennessee),
a manufacturer of ammonium nitrate. The system is reported to reduce ni-
trate levels of 1230 mg/1 at 0.9 MGD, to effluent concentrations of 7-11
mg/1, in the presence of other ions. The regenerants used are nitric
acid and ammonium hydroxide, with ammonium nitrate (concentration about
19%) being recovered in the resin regeneration step.
g. Radford AAP has assessed the selectivity of other resins,
including liquid amine exchangers, for nitrate (3d). Of ten resins
studied, the liquid exchangers LA-1 and LA-2 (Rohm and Haas Co.) were
-58-
-------
found to be most selective for nitrate removal, without concurrent
sulfate removal. Nitrate removal without concurrent sulfate removal was
desired, in order to obtain a purer regenerant. However, the kerosene
solvent used with the liquid exchangers was felt to represent a possible
contaminant of the regenerant by-product which, if reused could result in
a potentially hazardous explosive mixture.
h. The application of the Chem-Seps process for munitions
facilities appears feasible, particularly at Holston AAP, which uses
ammonium nitrate as an intermediate in HMX/RDX manufacture. The Chem-
Seps process, in addition to nitrate removal, is reported to reduce
NHo-N levels from 340 to 2-3 mg/1. The sulfate level of the waste being
treated was only 72 mg/1 (16zm), and it is possible that the higher sul-
fate levels associated with explosives wastewaters may interfere with
effective nitrogen removal. However, preliminary studies on the Chem-
Seps System by the Army indicated reduction of nitrate from 1650 to 20
mg/1, even in the presence of high sulfate concentration (3i). This
process, using DOWEX Resin MWA-1, also reduced sulfate from 4320 to 200
mg/1, The process, for a 3 MDG flow, is estimated to require a capital
investment of $1,650,000 (3i).
i. Although there is some apparent reluctance to using a
nitrate treatment process which yields a fertilizer material, and which
would require the military to market the regenerant as a fertilizer, ion
exchange undoubtedly represents the best and most reliable nitrate treat-
ment process which has been proven in full-scale use. Serious consider-
ation is now being given by the Army to use of systems such as the Chem-
Seps process, particularly if improved control of manufacturing opera-
tions can reduce sulfate levels in the nitrate-rich wastewater (15y).
Indeed, the concept design of the acid waste treatment facility at
Joliet AAP incorporates such a system (8c).
j. If the acid waste can be concentrated to about 15 percent,
reverse osmosis is a technique by which recovery would be feasible for
acid reuse in the plant (3i). At the present time, application of re-
verse osmosis is largely limited to desalination and production of drink-
ing water, and recovery of products in the food processing and electro-
plating industries (16zn). Investigations of reverse osmosis as a meth-
od of concentrating nitrate wastes are few, and those reported in the
literature indicate variable results (16zm, 16zo). Advantages of re-
verse osmosis include its low energy requirement and corrosion free
operation, in contrast to evaporative systems. However, reverse osmosis
membranes are affected by pressure, chemical change and hydrolysis,
temperature and surface coating (16zp).
k. The Army has investigated on a pilot scale reverse osmosis
treatment of nitrate wastewaters. At pH 1.5, no nitrate removal occur-
red although 99 percent sulfate removal was achieved. Nitrate removal
efficiency increased with increasing pH, to 90 percent at pH 7.0. Sul-
fate removal was not found to be pH dependent, with reverse osmosis
achieving 99 percent sulfate separation at all pH values tested. How-
ever, it appears that the lower limit of osmosis treatment is near 20 mg/1
NOg-N, and that achieving lower concentrations would not be economically
-59-
-------
feasible. Reverse osmosis application thus appears to be best suited
for nitrate recovery, with subsequent treatment of the dilute permeate
by other means such as biodenitrification or ion exchange. The Army is
continuing to investigate reverse osmosis, utilizing low-pH resistant
sulfonated polyphenylene oxide (SPPO) membranes, with evaporative concen-
tration of the process brine to recover mixed nitric and sulfuric acid
for reuse. This process, if proven feasible, will both eliminate the
need for neutralization and provide product recovery.
1. While reverse osmosis must still be considered as being
in the developmental stage, two other proven technologies are available
which can provide effective nitrate control. On the basis of economics
only, biological denitrification would be the method of choice, while
on the basis of process reliability and prospect of by-product recovery
ion exchange is most promising. With proper process design and good
control, either technology should achieve effective nitrate removal.
28. Impact of Air Pollution Control
Significant levels of nitrogen oxide are emitted from AOP and nitric
acid plants (3f). At some plants these emissions are not controlled,
although EPA emission standards have been promulgated for nitric acid
plants. Thus, control technology must be implemented at the military
nitric acid facilities. Where controls are in use, such as for NAG
emissions from Holston AAP, they are typically rather ineffective scrub-
ber systems (3f). Newport and Volunteer AAP AOP's use catalytic corn-
busters, designed to reduce nitrogen oxides to below 200 ppm (3f) .
Nitrogen oxide control procedures under consideration by the military
include: new AOP and DSNA production facilities which would reduce
emissions; incineration; sulfuric or nitric acid scrubbing with recovery;
catalytic reduction; and dry (molecular sieve) absorption (16zq). None
of these control techniques generate water pollution, and there is thus
no expected impact from air pollution control on water pollution from
nitric acid manufacture or concentration.
29. Summary
a. Despite the lack of reliable wastewater characterization
data for process effluents of AOP and NAG facilities, there is sufficient
evidence to establish that the effluents are of low pH, with high ammonia
and nitrate levels for both the AOP and NAG processes. The latter process
effluent also contains moderate to high sulfate levels, due to the use of
concentrated sulfuric acid as a dehydrating agent.
b. Technology for control of pH, ammonia and nitrate is well
established. However, among military munition and propellant plants
there is much inconsistency in the extent of waste treatment utilized or
planned. Volunteer and Joliet AAP's propose a full spectrum of treat-
ment, including neutralization, sulfate re'moval by precipitation or ion
exchange, and nitrogen removal by ion exchange. Other plants, with
equivalent wastewaters, propose only neutralization, perhaps accompanied
by some minor degree of sulfate control through lime treatment, with
calcium sulfate precipitation. The discrepancies in levels of treatment
-60-
-------
to be implemented indicate that no consistent effluent standards have
been established by the Military. In the absence of definition of
such standards by either the military or the U.S. Environmental
Protection Agency, the paradoxical spectrum of treatment levels programmed
will likely continue.*
*Note added in press. Interim/Final guidelines for Explosives Point
Source Category were issued in the Federal Register, Volume 41, No.
47, pp 10180, March 9, 1976.
-61-
-------
SECTION III-C - SULFURIC ACID - MANUFACTURE, CONCENTRATION AND RECOVERY
30. Introduction
a. Sulfuric acid is manufactured, and concentrated to oleum,
for use as a dehydration agent in the nitric acid-based nitration of
products such as TNT and nitroglycerin. In most instances, the spent
nitrating acid mixture is processed for recovery of both nitric and sul-
furic acids. In some instances the recovered sulfuric acid is concen-
trated and sold commercially, as at Radford AAP, while at other plants
the recovered acid is regenerated and reused. Military munitions plants
which recover sulfuric from spent nitrating acids are Radford, Joliet,
Badger, Newport and Indiana AAP's. Recovered sulfuric acid is a by-
product of the NAC process described in the preceding section of this
chapter.
b. In the sulfuric acid concentration (SAC) process, dilute
sulfuric acid yielded from the NAC operation is concentrated to 93% by
evaporation. Plants operating SAC processes include Radford, Joliet,
Badger, Volunteer, Newport and Indiana AAP's. This concentrated acid
is used to produce oleum at the first four named AAP's above. In oleum
manufacture, spent sulfuric acid is decomposed and elemental sulfur is
oxidized to sulfur dioxide, then catalytically oxidized to produce sul-
fur trioxide gas. The gas is absorbed with concentrated sulfuric acid
in absorption towers, to yield 30-40% oleum. This is the new sulfuric
acid regeneration (SAR) process. The SAR process produces oleum directly,
plus sulfuric acid of different concentrations, 98% sulfuric acid being
the weakest.
31. Waste Sources
a. Process waste sources, in addition to the NAC-based sul-
furic acid recovery operation described previously, are the SAC and
oleum facilities.
b. Wastewaters from the SAC process originate from occasional
acid spills, floor washings and water used for cooling (8f). Wastewater
generated in oleum production consists mainly of cooling water, contam-
inated by acid spills. At Joliet, an additional waste source is acid
tank car drainage and washout (8f). The major wastewater from the SAR
process is a weak (2.5%) sulfuric acid solution, which results from
cooling and scrubbing of combustion products (3f, 8f). At Newport AAP
this flow is reported to total 180 gpm (3f). Waste characterization
data for SAC, oleum and SAR facilities are limited. Table 18 summarizes
available data. The oleum facility at Joliet AAP is no longer in use.
Oleum is currently purchased from an outside source, pending start-up
of the new SAR facility.
c. The wastewaters from all except tank car clean-up are
predominantely cooling waters, which result in high dilution of process
contaminated effluents. The major characteristics expected for the
process effluents from SAC, oleum and SAR facilities would be low pH
and high sulfate levels.
-62-
-------
Table 18. Waste Characteristics of Sulfuric Acid Manufacture and Concentration
Parameter*
PH
COD
BOD5
Nitrate-N
Ammonia-N
Kjeldahl-N
Sulfate
Suspended Solids
Tot. Dissolved Solids
Oleum Production
Joliet (8f)
6.8
30
10
7.7
-
-
153
44
484
Joliet (lg(3))
7.8
-
-
-
-
-
133.4
-
-
Joliet (3f)
2.3-9.8
-
-
31
-
-
153
-
-
Radford (3f)
-
51
12
2.4
0
0.3
-
4
274
Tank Car
Clean-up (lg(3))
7.6
-
-
-
-
364
5875
73.5
6401
I
cr\
OJ
I
*A11 parameters mg/1, except pH.
-------
32. Current Treatment Practice
a. Treatment of wastes associated with sulfuric acid manufac-
ture and concentration is, at present, restricted to pH control by neu-
tralization with lime or soda-ash. The waste is frequently combined with
wastewaters from AOP/NAC facilities, prior to neutralization. At Radford
AAP , SAC wastewater is treated together with wastes from TNT production
facilities and a red water incineration unit, by soda-ash addition. Oleum
wastewater is treated at a separate neutralization facility, also with
soda-ash (Ac). The combined TNT/SAC effluent from the first neutraliza-
tion facility is reported to be adequately neutralized (4c). For a com-
bined wastewater flow of 0.16 MGD from the neutralization pond, the
effluent averages 2167 mg/1 sulfate, with a maximum of 3128 mg/1 (3f).
The oleum wastewater treatment facility (4.2 MGD) has been cited as per-
forming inadequately insofar as pH control, and has an effluent sulfate
concentration of 47 mg/1 (4c).
b. At Joliet AAP, treatment is restricted to neutralization
of acid spills (3f). Badger AAP discharges oleum wastewater to an
evaporation/percolation pond, without pretreatment. Wastewaters from
the SAC facilities are treated with lime, for neutralization, as is the
SAR wastewater at Newport AAP and SAC/oleum effluents at Volunteer AAP.
Like Radford, Volunteer AAP has been cited for poor pH control (lr(l)).
c. There is no question that proper technology is available
and can be implemented for pH control (4d). There is at present, no
effort directed toward sulfate control, although sulfate may in fact be
reduced to some extent in those plants utilizing lime for neutralization,
as a result of calcium sulfate precipitation. Calcium sulfate is a
relatively soluble salt however, as shown in Figure 8, which relates
effluent calcium to effluent sulfate concentration, on the basis of a
theoretical molar solubility product of 2.4 x 10~5 (16zr). This rela-
tionship predicts effluent sulfate levels exceeding 1000 mg/1, for
effluent calcium concentrations below 90 mg/1.
33. Modifications to Current Treatment
a. Modifications which have been proposed for existing
treatment facilities, to improve sulfate control, include calcium and/or
barium sulfate precipitation at Volunteer and Joliet AAP's, and multiple-
effect evaporation for acid waste concentration and recovery at Badger,
Radford and Sunflower AAP's (3i). At Newport, and apparently other
plants, no specific sulfate removal technology is under consideration.
Newport AAP will use lime neutralization, which provides partial sulfate
precipitation. However, the proposed treatment will not satisfy the ef-
fluent sulfate requirement of Newport's present NPDES permit (3j).
b. The most comprehensive treatment system is that proposed
for Joliet AAP. It is to consist of calcium and barium sulfate pre-
cipitation, followed by ion exchange treatment (8c). The solubility of
barium sulfate is much less than calcium sulfate, as shown in Figure 8.
Barium is a toxic metal, and ion exchange will serve both to remove barium
-64-
-------
0,001
10,000pr-
1000
D)
E
LLJ
oo
UJ
RESIDUAL BARIUM, mg/l
0.01 1.0
10 100
RESIDUAL CALCIUM, mg/ I
10
1000
Figure 8 - Solubility of Calcium Sulfate, K = 2.4 x 10 5 (16zr)
sp
Solubility of Barium Sulfate, K = 1.1 x 10~10 (16zs)
sp
-65-
-------
from the precipitation effluent, and to remove other trace constituents.
The specific steps of the proposed Joliet treatment system are: two-
stage neutralization with lime, resulting in formation of calcium sul-
fate sludge; addition of barium carbonate to form mixed barium sulfate
and calcium carbonate sludges; and ion exchange for effluent from the
precipitation processes. Sludges will be stored in lined lagoons. The
first step in ion exchange treatment is to be the cation unit, which will
be regenerated with dilute nitric or sulfuric acid. This regenerant will
be treated with other acid wastes, as described above. Effluent from the
cation unit will be treated by anion exchange, on the hydroxide cycle.
Regeneration will be ammonium hydroxide, with regenerant partially evap-
orated to yield fertilizer feedstock containing 45 percent ammonium ni-
trate. The deionized effluent will be used for boiler feed and other
plant water needs (8c).
34. Alternatives to Current Practice
a. Methods proven feasible for sulfate treatment, including
those described above, include:
(1) Reverse Osmosis - Separation of sulfate from the
wastewater by use of suitable membranes, to produce water for plant re-
use and concentrated brine for reclamation processes.
(2) Ion Exchange - Removal of sulfate by use of sulfate
specific resins, producing reclaimed water and a useful by-product re-
sulting from the regeneration of the ion exchange bed.
(3) Evaporation - Combined use of reverse osmosis or
ion exchange and evaporation techniques to recover sulfuric acid.
(4) Precipitation - Reduction in sulfate by use of lime
and/or barium to produce calcium and barium sulfate precipitate.
(5) Calcination - Produce calcium sludge by precipita-
tion with lime, followed by high temperature calcination to evolve sulfur
dioxide for sulfuric acid production, plus lime recovery.
b. Reverse osmosis has been investigated at pilot scale, in
combination with nitrate removal. High sulfate removal efficiencies
(99+%) are reported even at acidic pH. Permeate sulfate levels of below
100 mg/1 can be achieved, with brine sulfate levels above 4000 mg/1 (3i).
However, membrane hydrolysis at low pH greatly decreases useful membrane
life. In the absence of more resistant membranes, neutralization would
likely be required for the reverse osmosis feed stream. This may result
in precipitation, and fouling of the membrane by solids. Limited experi-
ments with acid resistant sulfonated polyphenylene oxide (SPPO) reverse
osmosis membranes have indicated good performance however, and may en-
hance the feasibility of reverse osmosis treatment.
c. The most technically feasible method of sulfate treatment
appears to be precipitation. However, the solubility of calcium sulfate
-66-
-------
is high (Figure 8), and lime treatment may not be feasible for more
stringent effluent requirements. The use of barium to precipitate sul-
fate has been proposed (8c), but cost and the possibility of exceeding
effluent barium levels appear to be major disadvantages.
d. The economic and technical difficulties associated with
treatment for pollutants such as nitrate and sulfate has led to several
applications of waste disposal by land irrigation, in the commercial ex-
plosives industry (4d). The use of land application appears practical
where suitable soil and climatic conditions exist. The soil has the in-
herent ability to be able to remove materials from the wastewater by both
microbial and crop utilization, and by physical adsorption. However, only
the biological processes affect the nitrates since the soils, which are
generally negatively charged, allow the nitrate to pass unimpeded, de-
pending upon the hydraulic loading rate (3d). Limitations include the
large land area required, and the possibility of contamination of ground
water.
35. Impact of Air Pollution Control
Both SOX and sulfuric acid mist emissions result from sulfuric acid
and oleum plants and SAC facilities (3i). Control procedures under study
include demisters, molecular sieves, catalytic reduction and scrubbing.
All processes incorporate recovery, with the catalytic reduction process
yielding elemental sulfur. One scrubber process uses glycol as an absorp-
tion medium, from which S0~ is recovered with glycol being recycled. The
molecular sieve process demists stack gases and adsorbs the contained S0«.
The sieve bed is cyclically regenerated, with the effluent recycled for
conversion to sulfuric acid (3f). There is no apparent water pollution
impact from proposed air pollution control procedures now under consider-
ation. Volunteer AAP utilizes Mahon fog filters to remove acid mist from
sulfuric acid recovery of TNT nitrating acids (3f) . This wastewater,
which is acidic and high in TNT, nitrate and sulfate, can be handled by
standard wastewater treatment procedures described in this chapter.
36. Summary
a. As in the case for nitric acid production, there is inade-
quate information available to accurately characterize process effluents
from sulfuric acid facilities. Noteworthy aspects of the effluents in-
clude low pH and high sulfate levels. Several acid plants have been cited
for improper pH control. The technology of pH control is well established,
and there seems to be no technical reason why these plants cannot achieve
effective control, other than from lack of commitment of resources to the
problem.
b. Similarly, methods of control of sulfate have been suffic-
iently well established to allow at least two plants (Volunteer and Joliet
AAP's) to reach the design stage for sulfate abatement. Other plants,
however, anticipate at best only minor sulfate control, through precipita-
tion resulting from lime addition for pH control. While effective,
available control methods are either costly or yield by-product sulfate
-67-
-------
brine or sludge. Additional research on methods of sulfate control is
warranted. The impetus for sulfate control must originate from a na-
tional requirement, if all sulfuric acid facilities are to meet the
levels of performance projected for Joliet and Volunteer AAP's.
-68-
-------
SECTION IV - NITROCELLULOSE
37. Introduction
a. Nitrocellulose is currently manufactured at Radford and
Badger AAP's by a batch process. A continuous process has been developed
and is being installed at Radford AAP. Indiana and Sunflower AAP's also
have capability for nitrocellulose production, but are not currently
manufacturing this product (2b). Between 30 and 40 Ib of dry, fluffed
cellulose (cotton linters or wood pulp) is the normal batch size. This
is submerged in about 1600 Ibs of mixed (nitric plus sulfuric) nitrating
acid. The composition of the mixed acid is adjusted to the grade (ni-
trogen content) of nitrocellulose required. The nitration reaction is
exothermic, and the charge temperature is kept below 37-40 C by cooling.
b. After about 25 minutes in the nitrator, the charge is trans-
ferred to a centrifuge, where spent acid is removed, and the nitrated
cellulose is washed in a large excess of water. The spent acid is pumped
to a tank, where a portion is fortified for reuse and the remainder is
sent to acid recovery. The crude nitrocellulose is usually pumped as a
water slurry to the purification area, where it goes through an elaborate
series of water washes, boiling treatments, neutralizations and heating
steps to stabilize the nitrocellulose (NC).
c. First, the acid content of the crude NC is reduced to a low
(0.25-0.50 percent) level by water washing. Then, the NC receives sev-
eral boiling treatments to destroy unstable sulfate esters and nitrates
of partially oxidized cellulose by acid hydrolysis. Next, the product
is beaten in a Jordan beater to reduce the NC fiber length and remove
traces of occluded acid. Finally, the NC is boiled in dilute sodium
carbonate solution, then washed with water until free of alkali. After
purification, the NC is centrifuged to approximately 30 percent moisture,
and then processed in accordance with the specific end-use requirements
of the batch. Nitrocellulose is a principal ingredient of most propel-
lants (single-base; combined with nitroglycerin in double-base propel-
lants; combined with nitroglycerin and nitroguanidine in triple-base
propellents). NC is used in smokeless powder, rocket grains, ball powder
and mortar increments, as well as some explosives (2b).
d. In the continuous process, linters are continually nitrated
and dumped to a centrifuge to recover acid. The crude NC is washed,
recentrifuged, slurried in water and boiled. The remaining steps are as
described above for the batch NC process.
38. Waste Sources
a. In the manufacture of nitrocellulose, repeated washing
yields wastes with varying acid content, from highly concentrated acidic
effluents to alkaline ones, some having an elevated temperature. Volumes
of wastewater generated, per Ib of NC produced, have been variously re-
ported as a lowest value of 16 gal/lb (2b) up to a high value of 100 gal/
Ib (16zt). Figure 9 is a schematic of current water use at Radford AAP.
-69-
-------
o
1
H
i
620,
rt s
- 33
Be
NC
Wa
Pun
Wa
Pun
Tot
2O to Acid Sewer Recovered Wat
Nitrator 620,000 I
620, 000 " C
784,300
000 r '241 '40° BY-p"«
^TA 1 200,000 Used
iling Tub House Bee
3f> Slurry
: Wash 730,000 563,000 Jordan Gl an
sh Lines 43,300 ' Agitator Glc
ip Gland 30,500 pump Gland
ip Tub 179,800 Line Wash Jub Add
al 984,300 43'JUU Line Wash
Total
\$ Total Acid
90-110ppm
Fines
998, 000 Used
( ,241 ,400 By- pass)
Boiling Tub Pits
2,239,400
1
To Neutralization Plant
er Filtered Water
2,244,600
*
109,800 455,000 226,000
iters Poacher Blender
6% Slurry 5% Slurry
d 26,200 274000 NC Wash 343,000 52 500 Vac Filter 184,000
nd 17500 ' Washout 48,000 ' Line Wash 29,500
4,900 Tub Wash 8,000 A^OQQ Cleanup 11,500
i-r'nnn — _ c,\nnrl 11 OOfl ' ... -_ rif»nn Tnh AW
2,900 Line Wash Line Wash 45,000 Tail Water Pump Gland 600
41,300 41,30U To|fl| 455i000 329,800 Tofa| 226<000
* A-n «nn 8* jl""X
400,800 343,000 2UU,000
Line Wash
29,500
__ Wringer
"* 77^ 800 Wash Lines 12,000
Wash Load 5,900
Poacher Pits Total Fines Wash Tubs 400
1.441.400 200-400 ppm Linc H 000 ^9'500 .
(Cleaning 11,500
Glands 5,700
To Recovered Water Tank Totd 49(500
IMC at 30# H2O
To Dehy Press
5,200
Figure 9 - Current NC Water Use (gpd) at Radford AAP - One Line
NC Capacity; 144,000 Ib/day (Pulp), or 120,000 Ib/day
(Linters) (3u)
-------
Wastewater volumes for that facility are summarized in Table 19. Of the
total, 29.1 percent is cooling water used only during the summer months.
The boiling tub operations use 200,000 gpd of recycled wastewater from
the combined discharge of the last four sources shown in Table 19. Ex-
cluding cooling water, the process water use for the flows presented in
Table 19 is 16.9-20.3 gal/lb NC. Approximately 45,000 gpd of this water
is used to transport the NC from process to process, as a 3-10 percent
solids slurry. The wastewater discharge from total NC production at Rad-
ford AAP is reported to be about 10 MGD (15zp).
b. Modernization of the NC process at Radford AAP will result
in a continuous cellulose nitration process. As presently designed (Fig-
ure 10), the continuous process will use 432,000 gpd recycle cooling
water (summer only) versus 1 MGD for the batch process, and 441,880 gpd
recycle wastewater for the boiling tubs versus 998,000 gpd with the pres-
ent batch system. The boiling tub water will be recycled to an acid
content which will permit economical acid recovery (15zp). Water needed
for the remaining operations of the continuous process will be identical
to the present system, but will be used counter current to process flow,
with ultimate disposal to the NAG acid recovery process (15zp). In
addition, the finished NC will be transported by slurry (150,000-160,000
gpd) to modernized continuous automated single- and double-base propel-
lant lines (3u).
c. As shown on Figure 9, the batch process wastewaters contain
considerable amounts of acid and suspended solids. A discharge of 0.7-
1.0 Ib sulfuric acid/lb NC, and 0.3-0.4 Ib nitric acid/lb NC, has been
reported (2b). The suspended solids are essentially fine particles of
NC. A typical size distribution of these NC fines is presented in Fig-
ure 11. Fifty percent of the particles are less than 2 microns in size.
There is no NC in solution, since NC is essentially insoluble in water
(2b). In addition to NC fines and nitrating acids, present wastewaters
contain alkali from neutralization steps, and soluble organic compounds
such as esters of oxy-cellulose (16zt).
d. Good wastewater characterization data are available from
Radford AAP (In3). Selected parameters of the waste discharge for each
NC processing step are summarized in Table 20. For each source described
in Table 20, there are normally several sequential processing steps, such
as fill and drain rinses. The wastewater generally becomes more dilute
with each sequential step, so that the range of data reported in Table 20
spans the first, most concentrated rinse to the last, most dilute rinse.
For example, sequential samples taken at Boiling Tub House #1019 showed
pH values of 1.4, 1.6, 3.1 and 3.2. On the basis of waste character-
istics, the Boiling Tub House effluent is much different than other
sources, being very acidic and high in nitrate, but low in suspended
solids. The other process effluents are more nearly neutral and low in
nitrate, but high in suspended solids as a result of the NC fines
discharged.
e. The Boiling Tub wastes at Badger AAP are also character-
ized by low pH (range 0.4-3.3, average 1.4), high nitrates (range 100-
-71-
-------
Table 19- Wastewater Volumes from NC Production at Radford AAP (3u)
Source
Nitration Cooling
Boiling Tubs
Beaters
Poachers
Blenders
Wringer
TOTAL
Volume*, gpd
1,000,000
998,000
400,800
343,000
423,000
273,800
3,438,600
Percent Use
29 . 1
29.0
11.7
10.0
12.1
8.0
100.0
*Flow per NC line. NC capacity per line is 120,000-144,000
Ibs/day.
-72-
-------
Acid
Acid Acid
i
Cool ing Water
bummer Unly 432,000gpd
432,000gpd
441 ,i
Reuse or Treat "1 ,880 gpd ( 1*TA)
and U.scharge p|us 378,000 gpd
t
Nitrator
Centrifuge 72° 9Pd Recovered Acid
-I 1
9880 qpd 1 .
Oo/3 'A ^^^__^_^^^«
i
SlurryTub Convey Fines
ir "~ ~" ~" """ — • ----- — i
j nrtA nno ,,^\
606 ,300 gpd 1
Figure 10 - Water Balance for Continuous NC Line (3u)
-------
'.71—
o; to
u <
84
£ — 50
I— UJ
< N
z S
u n 16
oi ^^
2.3
X
P-
MEASUREMENT METHOD:
COULTER COUNTER
I 1 1 1 1 i 1
2 3 4 5 6 8 10
PARTICLE SIZE, MICRONS
Figure 11 - Boiling Tub Pit Water Nitrocellulose Fines (2b)
-74-
-------
Table 20. Wastewater Characteristics of Radford AAP NC Processes (ln(3))
Source
Boiling Tub Houses
#1019
#2019
Jordan Beater Houses
#1022
#2022
Poacher-Blender Houses
#1024 (Poach)
#2024 (Poach)
#1024 (Blend)
Wringer House
pH
1.4-3.2
1.1-3.9
7.2
7.6-9.1
6.6-9.8
5.5-9.0
6.0
7.4-8.2
Suspended Solids ,
mg/1
Range
1-22.0
1-62.5
-
290-1054
52.0-632.0
23.0-526.0
463-495
343-828
Average
8.30
10.0
140
580
278.9
214.1
479.0
518
Filtered COD,
mg/1
Range
32-270
105-152
-
-
-
72-685
—
-
Average
103.5
136.0
-
31.0
-
414
—
135
Nitrate+Nitrite-N ,
mg/1
Range
7.1-1000
9.0-1000
-
0.6-4.0
1.0-60.0
100-70.0
30.0-34.0
-
Average
406.8
277.3
-
2.2
21.1
26.9
32.0
-
-------
1350 mg/1, average 700 mg/1) and sulfate (range 75-5100 mg/1, average
2600 mg/1). The waste is thus similar to Radford AAP, and although sul-
fate data for Radford are not available, they are likely equivalent to the
Badger wastewater. Composite wastewater suspended solids concentration
for the Badger NC line is 43 mg/1 (2b).
39. Effects of Water Management
a. There is significant opportunity for reduction in waste-
water volume, and mass pollutant discharge, by water management. Use of
a cooling tower and recycle system will reduce water discharged at Rad-
ford AAP during the summer by 1 MGD (3h). The two basic pollutants re-
sulting from NC manufacture are nitrating acids and NC fines. Figure 12
shows a proposed recycle water management program, which would reduce
process wastewater discharge by 90-95% (I6j). The program involves re-
cycle of boiling tub water through the nitrator, up to a total acid con-
tent of 13 percent. With a 13 percent maximum acid level, this flume
line water can be cycled up to six times, and then discharged to the NAC
acid recovery plant for recovery of nitric and sulfuric acids (3u). Im-
plementation of recycle and acid recovery will save about $10 million in
capital investment for denitrification facilities at Radford AAP (15zp).
b. For the more neutral, but suspended solids-laden wastewater
from the remaining batch processes, centrifugation treatment allows both
recovery of NC fines and recycle of the water. Fines collected by cen-
trifugation will be blended into the NC process, where practical, with
the remainder of the fines being disposed as solid explosive wastes (3u).
One problem encountered in the reuse of water in the NC purification
process is the buildup of sodium nitrate and sodium sulfate, resulting
from neutralization of acids with sodium carbonate. The buildup of these
salts will be only about 27 mg/1 per cycle for the conventional batch
line and 9 mg/1 for the continuous line (3u) . This may necessitate blow-
down from the recycle system, and Radford AAP has proposed a blowdown
flow of 200,000 gpd (6a). If blowdown could be avoided, the NC process
could be totally closed, with complete wastewater recycle plus recovery
of NC and acids.
c. A similar recirculating system has been suggested for the
continuous NC process, with recycle of Boiling Tub water to achieve 15-
20 percent acid levels, and centrifugation for NC fines control (3u).
40. Current Treatment Practice
a. At both Badger and Radford AAP's, spent nitrating acid is
separated from the crude NC and processed for acid recovery. At Radford
AAP, acid wastewater generated from washing the NC after each process
step in the Boiling Tub House is diverted to settling pits, for partial
removal of NC fines. The more neutral wastes from the Jordan Beater
House, Poacher-Blender House and final wringer house go to a separate
set of settling pits. Average NC fines concentration for the Boiling
Tub discharge is less than 10 mg/1, and ranges from 104-580 mg/1 for the
-76-
-------
1,000,000 to
Cooling Tower
109,800
Cool, Water F/
Solvents Area
455,000
103,334
Recycled 6X
6$ Slurry
274,000
Line Wash
41,300
400,800
Boiling
Tub Pits
Flume Water
103,334
277,666
Make up
water from
acid boil
343,000
25$ Slurry
52,500
Line Wash
45,000
Tail Water
329,000
273,800
Poacher Pits
Fi nes removed
to < 15 ppm
226,000
Slurry
200,000
Line Wash
29,500
Wringer
49,500
To Acid Recovery
To Recovered
HO Tank
After Fines
Removal
NC at 30$ HO
to Dehy Press 5,200
Figure 12 - Proposed NC Water Balance (6a)
-------
other processes (Table 20). No information is available on the efficiency
of fines removal in the settling pits, but the small size of the fines
(Figure 11) suggests that significant quantities of fines overflow the
settling pits.
b. The pit effluent is currently treated, along with acid area
wastewater, with lime in a neutralization plant. The neutralization fa-
cility consists of two baffled lagoons operated in parallel, preceded by
a slaked lime slurry feeder. The influent is heavily diluted with cool-
ing water (4c). The plant is reported to perform inadequately, with
fluctuations in pH, and high effluent levels of solids, nitrate, sulfate
and COD (4c). Sulfate control would be enhanced through calcium sulfate
precipitation, if process wastewater were treated prior to dilution with
cooling water. Influent and effluent characteristics are presented in
Table 21. Effluent suspended solids are 535 mg/1 (average), representing
a removal efficiency of only 44 percent.
c. Badger AAP also uses settling pits for roughing treatment
of NC wastewaters. Overflow for the Boiling Tub Houses settling pits
flows to a waste acid neutralization facility, where lime slurry is added.
Control of pH is automated, with a feedback pH controller (3f). The sys-
tem does not function adequately, with effluent pH values of 1.7-12.2
reported for influent pH of 0.4-3.2 (la(3)). The effluent also contains
high levels of nitrate (434 mg/1), sulfate (1605 mg/1), and suspended
solids (287 mg/1). Influent suspended solids to the Badger AAP neutral-
ization facility averaged only 6.1 mg/1 (la(3)). The high effluent value
likely reflects discharge of precipitates yielded in the neutralization
reactions.
d. The neutralization plant at Badger AAP receives only Boil-
ing Tub House settling pit overflow. Wastewater from other NC processing
houses, after settling pit treatment, is either discharged without further
treatment or recycled for use in NC slurry transport lines (3f). This
wastewater is reported to have the following characteristics: pH, 1.5-
6.9; nitrate - N, 7.1-31.8 mg/1; sulfate, 63-195 mg/1; COD, 64-112 mg/1;
and suspended solids (NC fines) 63.5-138 mg/1 (la(3)).
e. Thus, on the basis of the above information, there is
essentially no effective treatment applied to wastewaters of NC manufac-
ture. Treatment is limited to partial removal of suspended solids by
settling pits, and some rather ineffective efforts to adjust pH, with
the exception that there is no pH treatment at all for Badger AAP waste-
waters other than from the Boiling Tub Houses.
41. Modifications to Current Treatment
a. Both Badger and Radford AAP's have approval for, and are
implementing changes in current waste treatment practice. Modifications
related to control of NC fines will be tested at Radford, with successful
technology transferred to Badger AAP. Both plants have on-going programs
to improve treatment of acidic wastewaters.
-78-
-------
Table 21. Performance Data on Radford AAP Facility
for NC and Acid Area Wastes (ln(2))
Parameter
pH
Total Solids
Volatile Solids
Suspended Solids
Settleable Solids
Nitrate-N
Sulfate
COD
Ave. Concentration, mg/1*
Influent
6.3
3563
396
956
0.24
271.9
1199
-
Effluent
6.4
3397
376
535
0.07
299.8
1208
80
Percent
Removal
-
4.7
5.0
44.0
70.8
+10.3
+ 0.7
-
*Except pH.
-79-
-------
b. At Badger AAP, a new industrial wastewater treatment fa-
cility is under construction. The facility will provide equalization,
mixing, aeration, and neutralization through chemical addition (la)2)).
This facility will treat all industrial wastewater concurrently dis-
charged through the Badger main plant outfall. This includes wastewaters
from the nitric acid plant, NC production and solvent recovery. The new
facility should be effective in pH control, but does not incorporate pro-
cesses for BOD, nitrate, sulfate or effective suspended solids removal.
A second treatment plant is thus under construction to treat the effluent
from the waste acid neutralization facility (3f). This second facility
is intended to provide reduction of BOD and solids. Depending upon the
solids separation technology utilized, this second plant should achieve
effective BOD and suspended solids control, although nitrate and sulfate
will be unabated.
c. At Radford AAP, two modifications to current treatment are
undergoing implementation. One modification is for NC fines and the
second for acidic wastewaters. Installation of centrifuge units is plan-
ned for NC fines control (6a). Centrifugation would follow the settling
pits. In prototype centrifuge tests at Radford, a DeLaval unit operating
at 5,500-9,000 gph reduced NC fines in Boiling Tub pit effluent (188 mg/1),
poacher pit (477 mg/1) and tail water (230 mg/1), to below 25 mg/1) (6a).
Average flow was 8,000 gph. Centrifugation is claimed to have the fol-
lowing advantages (6a):
(1) no additives are required (which would contaminate
the recovered NC)
(2) direct return of NC fines for reuse
(3) reuse of the clarified effluent.
Current plans are to install a 10,000 gph centrifuge, on a trial basis,
at Radford AAP by 1976 (4c).
d. In addition, improved neutralization facilities, similar
to those to be installed at Badger, are planned for Radford AAP. Modi-
fications to the present neutralization lagoons will include equalization
tanks, additional tanks for improved mixing and longer retention times
for pH control, settling tanks with flocculation capability for improved
removal of the settleable solids, and mechanical sludge removal and sludge
dewatering (6a). Since the major shortcoming of present neutralization
facilities at both Radford and Badger is lack of adequate reaction time,
compounded by feedback automatic lime addition, the use of flow equal-
ization, together with better mixing and longer reaction tank retention
times, should provide improved neutralization. As at Badger AAP however,
the proposed Radford treatment plant has no provision for control of BOD,
nitrate or sulfate.
42. Alternatives to Current Practice
a. The single unique constituent of wastewaters from nitro-
cellulose manufacture is NC fines. Other wastewater parameters include
-80-
-------
acidic to near-neutral pH, sulfate, nitrate and moderate to high COD.
These latter waste constituents are typical of acid manufacturing wastes,
and pH, nitrate and sulfate may be controlled by the technology discussed
in Section III of this chapter.
b. The biodegradability of the COD constituent of the waste-
waters is unknown. Measured soluble COD's ranged up to 414 mg/1 at
Radford AAP (Table 20). This likely represents soluble organic com-
pounds such as esters of oxycellulose (16zt). NC itself is essentially
insoluble, and not readily biodegraded (2b). Thus, the soluble organics
of NC wastewater presumably consist of incomplete reaction products,
plus products of alkaline neutralization and other NC cleaning processes.
In biodegradation studies on caustic-digested NC fines, an activated
sludge process (bench-scale) reduced BOD and COD from 240 and 724.5 mg/1
to 27.3 and 324.2 mg/1, respectively (3h). Thus, although good BOD re-
moval was shown (88.6%), a large fraction of the COD and TOC (about 45%)
was not removed by biological treatment. It is probable that the same
situation would hold for the NC manufacture wastes, with effective re-
moval of the BOD fraction by biological treatment, but high effluent
COD levels.
c. The insolubility of NC, and its presence in wastewater as
fine suspended solids, indicates that any of several solids removal
technologies, such as centrifugation, coagulation and/or filtration,
could be employed. Centrifugation has been pilot tested at Radford AAP,
achieving suspended solids reduction to less than 25 mg/1 (6a). Other
solids removal processes include dissolved air flotation, coagulation-
flocculation, granular filtration and resonating filters. Dissolved
air flotation was unsuccessful when tested at Radford AAP (6a). Un-
doubtedly coagulation and flocculation would be effective on NC fines,
and is a proven process for similar applications of suspended solids
removal. However, the process is expensive when compared to other al-
ternatives and additionally results in a sludge from which NC could not
be recovered. In granular filtration tests at Radford, Boiling Tub pit
effluent was reduced to below 15 mg/1 suspended solids, but granular
filtration of poacher pit effluent resulted in suspended solids levels
exceeding 40 mg/1 (6a). By comparison with centrifugation, granular
filtration is thus less effective.
d. A newly developed micropore filtration process has proven
successful in pilot treatability studies on pulp and paper mill waste-
waters. The patented "Hydroperm" system (HYDRONAUTICS, Inc.) has
achieved 88 to 100 percent removal of suspended solids at a filration
pressure of 15 psi (16zv). Particle characteristics of the pulp and
paper mill wastes should be similar to NC fines. However, approximately
30 percent of NC fines are less than one micron in size (Figure 11), and
no information is available on the minimum size of effective particle
removal by the micropore filtration process.
e. Tests at Radford in NC fines removal with a resonance
filter were inconclusive. Difficulties included a too fragile filter
-81-
-------
medium, with too large pore size to remove the finer fraction of sus-
pended solids (6a). Thus, among the available proved processes, co-
agulation and centrifugation are most effective, with centrifugation
having the advantage of NC recovery. In summary, the current Army
program for pH control and NC fines abatement appears to be drawn from
best available technology. When coupled with available water management
techniques, and sulfate and nitrate abatement procedures, the wastewater
resulting from NC manufacture can be totally controlled.
43. Summary
Although wastewaters from NC manufacture are not currently receiv-
ing adequate treatment, beyond partial pH adjustment, at either Badger
or Radford AAP's, major efforts have been initiated at both plants for
improved pollution abatement. The program at Radford AAP in water con-
servation and reuse plus recovery of NC and acids, is particularly
notable. Current plans call for essentially complete recycle and re-
covery, with the exception of some blow-down from the acid neutraliza-
tion wash. Although Badger AAP also will incorporate some new treatment
procedures, they are essentially limited to pH, solids and BOD control.
A more rigorous abatement effort is indicated for Badger AAP, encompass-
ing both recycle and recovery, plus nitrate and sulfate control.
-82-
-------
SECTION V - RDX AND HMX
44. Introduction
a. RDX (cyclotrimethylenetrinitramine) is a major military
explosive, used as a component in many explosive and propellant formu-
lations. HMX (cyclotetramethylenetetranitramine) is a by-product of RDX
manufacture, and is also used as a component of explosive formulations
as well as solid missile propellants. Mixtures of RDX and wax, called
"Composition-A" explosives, are used in artillery shells. Compositions
A-3 and A-5 are also used as booster charges in many Army munitions, in
lieu of tetryl. Mixtures of RDX or HMX with special plasticizers and
solvents give rise to numerous plastic explosives and demolition charges
such as "Composition-C," or "PBX" explosives. RDX and HMX find wide
application in bomb and artillery shells. For this purpose they are
mixed with TNT to form mixtures called "Composition-B," "cyclotols" or
"octols." RDX and HMX, and the various explosive formulations derived
from them, are produced only at Holston AAP. However the composition
explosives which incorporate RDX and HMX are loaded at several Army and
Navy LAP plants, and LAP air scrubber and clean-up waters thus contain
these explosives as contaminants.
b. RDX and HMX are manufactured by the same processing stages,
using the same ingredients. The two different products result from
using different proportions of reactants. The reactants for RDX and
HMX, for a hypothetical 100 Ib reactant charge, are presented in Table
22.
c. In RDX manufacture, the ingredients are charged to a
reactor at 75°C. The initial crude product contains about 79% RDX, 6%
HMX and various intermediate products. The reaction mixture goes through
an aging and simmering process to convert these intermediates to RDX or
decompose them. HMX is not considered detrimental to KDX performance,
and no attempt is made to separate or recover it (2c).
d. The reaction mixture is cooled, and most of the crude RDX
crystallizes and precipitates. Most of the supernatant liquor is drawn
off by vacuum and transferred to a recovery building. The processing of
this supernatant was described in Section III of this chapter. After
vacuum filtering, the crude RDX is washed with water. The wash water is
withdrawn by vacuum and recycled to the simmering process (Figure 7,
Section III). The RDX is then slurried in additional water and trans-
ferred to a recrystallization process. There, cyclohexanone is added
to the slurry, the slurry heated, and the cyclohexanone distilled, with
RDX recrystallizing in particles of acceptable size. The RDX in water
slurry is then poured into special vacuum carts (nutsches) and most of
the water withdrawn. The resulting explosive contains about 10% resid-
ual moisture (2c). When RDX is heated and mixed with wax, TNT or other
ingredients of composition explosives, this residual water separates
and is decanted as a wastewater. The major incorporation product is
Composition B, which contains 60.5% RDX, 38.7% TNT and 0.8% wax. In-
corporation water will contain TNT in addition to RDX.
-83-
-------
Table 22. Reactants for RDX and HMX Manufacture (2c)
Reactant
RDX, Ibs
HMX, Ibs
Ammonium Nitrate
98% Nitric Acid
Hexamethylenetetramine
Acetic Acid
Acetic Anhydride
11.0
17.0
18.0
54.0
-84-
-------
e. For HMX manufacture, the same production steps are employed.
About 27% of the crude product is RDX. Most of the RDX is extracted in
the acid draw-off by vacuum and washing steps, and eventually is added to
RDX production. The HMX recrystallization process is usually accomplished
with acetone, rather than cyclohexanone. The major wastewater discharges
for HMX production occur during the dewatering steps, and during incor-
poration into composition explosives such as Octol.
f. RDX and HMX solubilities will influence their concentrations
in wastewater, which may also contain acetic anhydride, cyclohexanone (for
RDX) or acetone (for HMX). Selected solubilities are presented in Table
23. As is indicated by those data, both RDX and HMX are extremely solu-
ble both in warm water and organic solvents.
g. At many LAP facilities, shells and bombs are filled with
explosives which incorporate RDX or HMX. Typically, the composition ex-
plosive is ground or otherwise broken up, melted and poured into shells
and bombs. RDX and HMX-contaminated wastewaters from these LAP opera-
tions include fume and dust control scrubber water, and washwater from
spills and floor and equipment washdown. Similar wastewaters may also
result from explosive steamout from rejected shells (2c).
45. Waste Sources
a. The major wastewater sources from RDX and HMX manufacture
result from dewatering of the explosives in nutsches, decanting water
from hot composition explosive blending, and floor and equipment clean-
up. In addition, explosives contaminated wastewater results from dust
control by scrubbers at the Holston AAP explosives packaging buildings.
Most discharges flow through catch basins, to provide removal of settle-
able solids, before wastewater discharge.
b. Pollutant discharge data are presented in Table 24 for RDX
manufacture,, and Table 25 for HMX manufacture. The wastewater discharge
resulting from explosives packaging (Table 24, Building N-7) results from
air pollution scrubber operation. Greatest wastewater volumes from pro-
duction are associated with the nitration process and nutsch dewatering,
with an additional high flow from incorporation. Highest concentration
of explosives results from nutsch dewatering, the incorporation step and
scrubber operation in explosives packaging. These latter two sources
also contain significant TNT levels. All catch basin effluents discharge
to an industrial sewer and thence to the Holston River, without addition-
al treatment. It has been reported that Holston AAP discharges 920-1200
Ibs/day of RDX, and lesser amounts of HMX and TNT to the river (4b).
c. Table 26 presents flow data, from a 1971 study, of process-
contaminated waste volumes and non-contact discharge volumes. Non-
contact water includes heat exchange cooling, steam condensate and pump
seal water. Differences in flow volume between Table 26, and Tables 24
and ?5, likely reflect changes in daily explosives production rates. Con-
taminated process water for most of the manufacturing buildings represents
only a small fraction of the total discharge to the catch basins.
-85-
-------
Table 23. Selected Solubilities of RDX and HMX (2c)
Solvent
Water @ 25°C
Water @ 83°C
Acetone @ 30°C
Cyclohexanone @ 30°C
Acetic Anhydride @ 30°C
KDX Solubility
7.6 mg/1
1.3 g/1
69.0 g/1
84.0 g/1
49.0 g/1
HMX Solubility
-
0.14 g/1
22.0 g/1
53.0 g/1
13.0 g/1
-86-
-------
Table 24. Pollutant Discharge from KDX Manufacture at Holston AAP (5d, 5e)
i
CO
Production
Source
Nitration
Reactor (D-2*)
Filtration and
Washing (E-3*)
Recrystallization
(G-2*)
Nutsch Dewatering
(H-2*)
Incorporation-
Composition B
(J-2*)
Explosives
Packaging (N-7*)
Catch Basin
Influent (I)
or Effluent (E)
I
E
I
E
I
E
I
E
I
E
I
E
Wastewater
Volume , gpd
49,760
49,760
13,856
13,856
27,218
27,218
346,523
346,523
118,803
118,803
742
742
Concentration, mg/1
BOD
2359
2116
582
1163
896
339
2771
3229
5895
4500
3747
3151
RDX
5.06
2.31
4.15
3.53
0.0
0.0
49
33
13.12
11.10
13.0
6.0
HMX
2.17
0.96
1.87
0.00
0.0
0.0
1.5
0.7
0.10
0.06
14.0
8.0
TNT
-
-
-
-
24.2
12.1
237
129
*Indicates Holston AAP Building for which data reported.
-------
I
00
oo
Table 25. Pollutant Discharge from HMX Manufacture at Holston AAP (5e)
Production
Source
Nitration
Reactor (D-6*)
Filtration and
Washing (E-6*)
Re crystallization
(G-6*)
Nutsch Dewatering
(H-6*)
Catch Basin
Influent (I)
or Effluent (E)
I
E
I
E
I
E
I
E
Wastewater
Volume , gpd
18,800
18,800
-
135,807
135,807
-
Concentration, mg/1
BOD
2393
1840
5183
6098
331
322
394
375
RDX
0.0
0.0
70.1
6.1
5.2
5.2
21.5
20.9
HMX
0.0
0.0
45.7
12.2
10.6
4.4
34.5
9.2
*Indicates Holston AAP Building for which data reported.
-------
Table 26. Discharges to Catch Basins at Holston AAP (ld(2))
i
oo
Building
D-6
D-8
E-8
G-8
G-4
H-6
H-8
Product
HMX
RDX
RDX
RDX
RDX
HMX
RDX
Process
Wastewater, gpd
2,500
2,900
1,640
1,440
77,740
10,740
54,480
Non- Contact
Water, gpd
38,710
39,420
6,480
5,420
39,540
12,500
2,880
Total
Flow , gpd
41,210
42,320
8,120
6,860
117,280
23,240
57,360
Percent
Process Wastewater
6.1
6.8
20.2
20.9
66.1
46.2
94.8
D - Buildings are Nitration Reactors
E - Buildings are Filtration and Washing Operations
G - Buildings are Recrystallization Processes
H - Buildings contain Nutsch dewatering equipment
-------
d. Limited information is available on wastewater
characteristics associated with handling and loading of RDX and HMX
based explosives at LAP facilities. Available data are tabulated in
Table 27- Wastewaters from melt-pour operations result from condensed
steam and washwater, primarily. Bomb Plant A at NAD Crane (Table 27)
also includes a demilitarization facility. Bomb Plant B was not in
operation at the time of preparation of this report. Other plants also
handle RDX and HMX, and have associated wastewaters. For example,
NAVPRO Magna grinds HMX-based explosives, and uses catch basins for
partial wastewater treatment. No explosive concentration data are
available, however (15zj). Evidence of groundwater contamination by RDX
has been reported at NAD Crane (14y). Water samples taken from on-
station monitoring wells have been found to contain 0.0 to 7.9 mg/1 RDX,
with nearby streams containing RDX at 3.0-10.0 mg/1. Most wells sampled
were shallow, and there appears to be a strong correlation between stream
water and well water RDX content. TNT, monitored concurrently with RDX,
was much lower in both streams and wells (14y). In addition, RDX levels
showed rapid attenuation at greater distances away from streams. Over
a distance of 28 ft from one well, RDX was reduced from 4.1 to 0.24 mg/1
46. Effects of Water Management
a. The most notable effort in water management associated
with RDX and HMX manufacture is the return of the first wash water after
nitration to the nitration process. This wash water, which is acidic
and high in nitrate, ammonia, acetate, explosives products and reaction
intermediates, would otherwise represent a major pollutional source.
One of the five Composition B lines at Holston AAP is to undergo mod-
ernization. This modernized line will incorporate several water manage-
ment techniques to eliminate pollutant discharge. These will include:
elimination of the decantation water resulting from melting of RDX, re-
cycle of spray water used to cool molten Composition B, and recycle of
fume scrubber water, after carbon treatment. Wastewater from the re-
maining four Composition B lines, which will not undergo modernization,
will be handled by wastewater treatment processes described in para-
graph 48.
b. Based upon the data of Table 26, it appears that the
segregation of non-contact from process cooling water offers a signif-
icant opportunity for water management and reduction in contaminated
wastewater volume. Discharge for the "D" Buildings to the catch basins
contains less than 7 percent process effluent, while discharge from all
except Buildings G-4 and H-8 represent less than 50 percent process
effluent. Segregation of non-contaminated discharges would provide
greater detention time in the catch basins, and result in smaller waste-
water volumes for additional treatment.
c. Aside from the possibility of dry rather than wet
dust control systems, discussed in paragraph 50, other opportunities
for effective water management are limited.
-90-
-------
Table 27. RDX and HMX Wastewaters Resulting from LAP Activities
LAP Facility
Joliet AAP
Milan AAP
Iowa AAP
NAD Crane
Activity
Melt-Pour
LAP Laundry
Melt-Pour
Me It -Pour
Air Scrubber and
Melt-Pour
(Rockeye Plant)
Me It -Pour
(Bomb Plant A)
Melt-Pour
(Bomb Plant B)
Volume ,
gpd
6,200
9,600*
-
-
—
-
-
Concentration, mg/1
RDX
87-180
1.1
1.1-2.2
18-24
0.0
17.9
12.8-41.6
HMX
1.0-16.0
0.0-0.9
0.0-5.5
Treatment
Activated Carbon
None
Catch Basin &
Evap . Pond
-
Catch Basin
Catch Basin
Catch Basin
Reference
8f
8f
31
13e
14f
14f
14f
*20 gpm, based on 8-hour operation.
-------
47. Current Treatment Practice
a. With the exception of activated carbon treatment of
LAP wastewater at Joliet AAP, and some limited application of solar
evaporation ponds to wastewaters at other LAP facilities, treatment of
RDX and HMX wastewater from both manufacture and LAP is limited to catch
basins or baffled sumps. These devices are intended to remove only the
larger settleable solids and, if subjected to hydraulic overload, will
have very low efficiency of removal even for that constituent of the
wastewater. Based upon the data of Table 24, the efficiency of RDX
removal at Holston AAP ranges from 0.0 to 91.3 percent, with an average
removal efficiency of 33.2 percent. Effluent RDX concentrations range
from 0.0-33 mg/1, as shown in Tables 24 and 25. The efficiency of HMX
removal at Holston AAP ranges from 40-100 percent, averaging 62.1 per-
cent removal efficiency. This higher removal may reflect the lower
solubility of HMX.
b. Both Milan AAP and NAVPRO Magna use catch basins,
followed by solar evaporation ponds, for LAP wastewaters. In neither
case is there a surface discharge from the ponds. However, the use of
evaporation ponds depends upon proper climatological conditions, and
facilities located in geographical regions which do not provide these
conditions must utilize alternate treatment methods.
c. Joliet AAP treats one Composition B LAP wastewater
stream by a combination diatomaceous earth filtration and activated
carbon adsorption process. The process uses two carbon columns, con-
nected in series and piped so that the columns may be used in either
order. Because the carbon cannot be effectively regenerated, the carbon
in the No. 1 column is replaced monthly (lg(3)). Spent carbon and back-
washed solids from the filter are burned. Wastewater is pumped through
the system at a fairly constant rate of 12-15 gpm. Operating on an
eight-hour shift, the treated wastewater volume is about 6200 gpd.
Table 28 presents operating data for the system. Although removal ef-
ficiency is good for suspended solids and TNT, RDX removal is less ef-
ficient, with an average effluent RDX level of 19 mg/1, and maximum of
46 mg/1. Further, most removal occurs in the first carbon column, as
shown in Table 29. The diatomaceous earth filter provides essentially
no removal, and the second carbon column very little additional removal
after the first.
d. While activated carbon is effective in removing TNT
from wastewater, its efficiency is significantly less for RDX, as shown
in Figure 13. It has been reported that activated carbon adsorptivity
of RDX is only about 16-17 percent of carbon adsorptivity for TNT (15zk),
This indicates that in a mixed RDX plus TNT wastewater, only a fraction
of the TNT capacity will have been exhausted prior to RDX breakthrough.
Further, it has been reported that TNT will displace and leach adsorbed
RDX from a carbon column (14z). This does not imply that activated car-
bon cannot be used for RDX removal and, in fact Yorktown Naval Weapons
-92-
-------
Table 28. Performance of Joliet AAP Treatment System (lg(3))
Parameter*
PH
Total Solids
Suspended Solids
TNT
RDX
Influent
Range
6.8-8.4
903-1790
22-336
156-235
87-180
Average
7.9
1401
138
178
145
Effluent
Range
6.7-8.0
762-1497
0-7
0-25
0-46
Average
7.7
1070
1.2
3.7
19
Average Percent Removal
-
23.6
99.1
97.9
86.6
U)
I
Concentration in ing/I, except pH.
-------
Table 29. Performance of Individual Units of Joliet AAP Treatment System (lg(3))
Treatment
Unit
Raw Waste
Filter
1st Column
2nd Column
Suspended Solids
Concentration ,
mg/1
138.5
108.6
8.4
1.2
Cumulative
Percent Removal
-
21.6
93.9
99.1
TNT
Concentration ,
mg/1
178.2
175.7
14.7
3.7
Cumulative
Percent Removal
-
1.4
91.8
97.9
RDX
Concentration ,
mg/1
145.2
148.9
30.1
19.5
Cumulative
Percent Removal
-
(-2.5)
79.3
86.6
-------
1.8
a.
a.
O
o
z
z
LU
u
z
o
3.3MIN. RETENTION TIME
AVG. FEED CONCENTRATION
9.8 PPM TNT
4.5 PPM RDX
TOTAL COLUMN VOLUME : 2.0 F
3.5 GAL /MIN / SQ . FT
1000
2000 3000 4000
BED VOLUMES OF EFFLUENT
5000
Figure 13 - Competitive Adsorption of RDX and TNT Using
Filtrasorb 400 at Keyport Test Site
-------
Station currently has a 20 gpm system planned, designed to reduce RDX
below 1 mg/1 (15zh). The carbon process would be most effective for
wastewater devoid of TNT, such as the manufacturing wastewaters of
Hols ton AAP, but would also be effective for wastewater containing TNT,
although the treatment capacity for RDX would be much less than for
effective TNT treatment. Since the activated carbon is not normally
regenerated, effective RDX treatment would require excess carbon, and
increase the overall cost of treatment.
48. Modifications to Current Treatment
a. Holston AAP, the manufacturing facility for RDX and HMX,
has proposed to biologically treat these and other industrial process
effluents from their Area B activities by aerated ponds (3f). There
seem to be serious questions concerning the capability of biological
treatment for RDX and HMX, however. In assessing the potential effect-
iveness of the proposed biological treatment at Holston AAP, the U. S.
Environmental Protection Agency has pointed out that it is most unlikely
that biological processes will degrade complex organic materials such as
RDX and HMX (4a). A subsequent architect/engineer report to Holston AAP
recommended treatment of Area B wastes by biological treatment (denitri-
fication followed by trickling filtration), multi-media filtration,
break-point chlorination and physical adsorption (5c). This latter pro-
cess may be either activated carbon or synthetic resin.
b. Many studies have been made on the biological treatability
of RDX and HMX. Osman and Klausmeier have reported, for example, no
breakdown of RDX during studies on the microbial degradability of RDX,
ammonium picrate and TNT (16o). Trickling filtration experiments with
HMX at 2.2-2.5 mg/1 resulted in effluent levels of HMX of 2.0-2.5 mg/1 (14f).
Green, studying explosives degradation in a 70 liter pilot activated
sludge unit with 5 days hydraulic detention time, found only 14-42 per-
cent reduction in RDX and 53.4 percent removal of HMX (2c). Pardee, in
soil suspension tests, has reported partial breakdown of RDX. Repeated
dosing of RDX to 7 to 19 mg/1, followed by degradation periods of 12 to
26 days, resulted in residual RDX levels below 3 mg/1 (2c). Despite
these results, U. S. Navy proposes to install a demonstration oxidation
ditch at NAD McAlester for their Plant "B" wastes, the effluent of which
is expected to contain less than 0.5 mg/1 TNT and less than 0.5 mg/1 RDX
(14m). NAD McAlester has also planned a diatomaceous earth filter and
activated carbon treatment system for their Plant "A" explosives con-
taminated wastewaters (14m).
c. Several LAP facilities have programs to install activated
carbon treatment systems. These include Milan and Kansas AAP's, York-
town Naval Weapons Station and NAD McAlester. Milan will treat waste-
waters for their composition explosive melt-pour lines by activated
carbon (31). Yorktown Naval Weapons Station will use a diatomaceous
earth filter plus two carbon columns in series, and operate at 20 gpm
wastewater flow. The Yorktown system is expected to produce effluent
levels of 1 mg/1 each of TNT, nitrocellulose and RDX (15zh). NAD McAlester
-96-
-------
also will use diatomaceous earth plus carbon treatment systems. Two
complete systems are planned at McAlester, each with a treatment capac-
ity of 80 gpm. System performance is expected to achieve 0.5 mg/1 ef-
fluent TNT and RDX (14m).
d. Based upon the carbon performance and other data presented
in paragraph 47 above, it is extremely doubtful that a carbon system
treating TNT plus RDX can achieve equal effluent levels for both explo-
sives (e.g., 1.0 mg/1 for Yorktown or 0.5 mg/1 for NAD McAlester). The
preferential adsorption of and greater capacity for TNT displayed by
activated carbon indicates that such treatment systems must be designed
and operated on the basis of RDX treatment, and that this will perforce
result in more rigorous treatment for TNT.
e. Kansas AAP also plans to use diatomaceous earth filtration
plus carbon, for wastewater from melt-pour operations. However, waste-
water from a detonator manufacturing operation, containing lead azide,
lead styphnate and RDX, will be chemically treated. Sodium nitrite will
be used to break down lead azide, and caustic to deactivate lead styph-
nate and RDX (lh(3)). Caustic hydrolysis of RDX reportedly occurs in
1-1/2 to 2 hours, yielding breakdown products of ammonia, formaldehyde
and nitrite (2c).
f. Air Force Plant #78, which has wastewater associated with
HMX use in LAP activities, including HMX grinding and incorporation,
plans to dispose of wastewater from equipment cleanup and the HMX area
laundry by discharge to catch basins, and ultimate disposal to evapora-
tion/percolation ponds (15zk).
g. Beyond these activities, and the water management program
associated with modernization of one of Holston AAP's five Composition B
lines, no other types of treatment are programmed for implementation.
Research is proposed however, to evaluate alternate treatment technologies.
49. Alternatives to Current Practice
a. The Army plans to investigate several methods of RDX and
HMX control, in pilot scale studies at Holston AAP. Methods under con-
sideration include: reverse osmosis; activated carbon adsorption;
polymeric resin adsorption; and anaerobic biological degradation (3i).
There are no existing data on RDX or HMX treatment by reverse osmosis or
anaerobic degradation. The anaerobic digestion process is notable for
its capability to degrade complex organic compounds, but is also generally
recognized as a system extremely sensitive to toxicity. The toxicity of
RDX and HMX has not been investigated to any significant extent (2c).
The potential success of anaerobic degradation however will require low
toxicity, plus capability of the anaerobic microbial population to de-
grade RDX and HMX.
b. The use of regenerable polymeric resin for RDX and HMX
treatment would have a significant advantage over one-time use of carbon
-97-
-------
in terms of economics, if the resins are effective in removing the
explosives. A novel approach is under investigation by NSWC White Oak,
whereby RDX would first be concentrated and hydrolyzed on polymeric
resin (14z). This approach is based upon preliminary studies which in-
dicated that the relative adsorptivities of TNT and RDX were approxi-
mately the same on resin as on carbon, and further that the rate of
alkaline hydrolysis by the resin, when on the hydroxide cycle, could be
described by the equation,
k [RDX] [OH~]
dt
where k = 0.33 + 0.05 moles /min (14z). The hydrolysis reaction, based
upon experimental data, is reported to be (14z) :
RDX + 3 OH~ - 0.99 NH + 1.26 N0~ + 0.12 NZ + 1.13 N20 + xO^O
Others have reported that alkaline hydrolysis of RDX produces quantita-
tively two moles of nitrite (out of a possible three nitro groups) for
each mole of RDX reacted. Alkaline hydrolysis of HMX produces 3 moles
of nitrite out of a possible four nitro groups. Successful RDX treat-
ment by alkaline hydrolysis has been reported (2c, lh(3)). HMX has been
noted to be more resistant to alkaline decomposition than is RDX, how-
ever (2c). The success of the adsorption plus hydrolysis process de-
pends first upon the ability to concentrate RDX or HMX on the resin, and
second upon the extent of hydrolysis achievable. A commercial resin,
XAD^n bench scale tests, reduced RDX from an initial level of 45 mg/1,
to below 0.02 mg/1 (14z).
c. Results from recent pilot plant studies at NAD Hawthorne
indicate however that the XAD polymeric resin is a much less effective
adsorbent than activated carbon (14zb) . The wastewater was treated by
a sequence of air flotation, followed by a cooling tower. Effluent
from the cooling tower was split into stream 1, which went through mixed
media (sand and coal) pressure filtration followed by activated carbon
adsorption; and stream 2 which passed through diatomaceous earth filtra-
tion followed by Rohm and Haas Amberlite XAD-4 polymeric resin adsorption.
Table 30 summarizes the treatment results for TNT and RDX from several
composition explosive wastewaters. For TNT, both carbon and polymeric
resin provided effective treatment, consistently yielding effluent TNT
levels below 0.5 mg/1. Activated carbon was even more effective for
RDX, producing effluent levels of 0.02 mg/1 or less. However, with the
single exception of the Composition A-3 wastewater, the performance of
the polymeric resin system was much less effective. Some RDX removal
resulted in the first resin column, but no additional removal is seen
for the second of the two series units. Table 31 summarizes overall
treatment efficiency for Streams 1 and 2, based upon plant influent,
and final plant effuent values. The results of this study indicate
that activated carbon treatment is more reliable for RDX control, with
the capability of reducing RDX to below 0.02 mg/1 while simultaneously
treating for TNT.
-98-
-------
Table 30. Treatment for TNT and KDX by NAD Hawthorne Pilot Plant (14zb)
Explosive
Comp . A- 3
TNT
RDX
Comp . B
TNT
RDX
HEX - 1
TNT
RDX
Mixture
TNT
RDX
Influent
Concentration ,
tng/1
11.4
79.9
422
103
314
92.2
279
145.7
Flotation
Effluent*
81.9
82.8
389
141.9
323
144.5
256
302.2
Cooling
Tower*
620.0
72.3
348
147.9
330
158.0
283
205.1
Stream 1*
Sand
Filter
434.2
87.9
304
171
308
128.5
285.3
221.8
Activated
Carbon
< 0.5
0.02
< 0.5
< 0.01
< 0.5
0.02
< 0.5
0.01
Stream 2*
Diat.
Earth
422.5
-
328
192.3
281
155.4
272.3
170.5
Resin
Column-1
4.35
0.62
2.3
25.5
2.6
66.8
3.0
59.0
Resin
Column- 2
< 0.5
< 0.01
< 0.5
20.4
< 0.5
61.6
< 0.5
60.7
*Values given are for effluent from each treatment unit in mg/1 concentration.
-------
Table 31. Pilot Treatment Plant Efficiencies, NAD Hawthorne (14zb)
Wastewater
Comp . A- 3
Cotnp . B
HBX - 1
Mixture
Constituent
COD
Suspended Solids
TNT
RDX
COD
Suspended Solids
TNT
RDX
COD
Suspended Solids
TNT
RDX
COD
Suspended Solids
TNT
RDX
Percentage of Removal
Stream 1*
92.5
84
90
99.5
97
99.5
99.5
99.5
98
98.5
99.5
99.5
97
84
99.5
99.5
Stream 2**
32
53.5
13
99.5
91.5
99.5
99.5
78
86.5
97
99
30
83
76.5
99
59
*Sand Filter/Activated Carbon
**Diatomaceous Earth/Resin
-100-
-------
50. Impact of Air Pollution Control
a. Water scrubbers are widely used in melt-pour and other
composition explosive handling steps. These scrubber systems generate
wastewaters with characteristics similar to clean-up and other LAP
wastewaters. For example, the wastewater reported in Table 24 for ex-
plosives packaging at Holston AAP results from dust control by an air
scrubber system. At most facilities where air pollution control is not
now employed, current plans are to use scrubber systems. NAD Crane, for
example now uses scrubbers at their Rockeye Facility, and plans to in-
stall the same system at two bomb loading plants (14f). Scrubber over-
flow rates for these additional systems are estimated at 0.5 gpm per
scrubber, with four scrubbers at each bomb loading plant. The major im-
pact of scrubber installation is to increase total plant wastewater
volume. Treatment processes applicable for other waste sources should
apply equally well to the air pollution scrubber water.
b. As an alternative to wet-type scrubbers, Air Force Plant
#78 has installed dry "absolute" filters for HMX grinding process dust
control. This filter system, which will control particles down to 0.3
micron size, has a major advantage of not producing wastewater. The
units used at Plant #78 (American Air Filter Co., Hospital Filter Units)
are reported to be effective, and trouble free in operation (15zk).
51. Summary
a. RDX and HMX are manufactured at Holston AAP, and loaded
at several Army, Navy and Air Force LAP facilities. The wastewaters
associated with manufacture have been adequately characterized, although
much less information is available on process wastewaters from LAP oper-
ations. Despite the lack of complete characterization data, specific
treatment processes for these wastes have been selected at several plants,
b. Although Holston AAP does not currently treat wastewaters
from RDX and HMX manufacture, biological treatment by an aerated lagoon
is proposed. Prior attempts to biologically degrade RDX and HMX have
been largely unsuccessful, and serious reconsideration is due the pro-
posal of Holston AAP to treat the manufacturing wastewaters by an aerated
lagoon.
c. More successful treatment has been achieved with activated
carbon, such as for the Joliet AAP LAP wastewater. There is evidence to
indicate that carbon may preferentially remove TNT, and carbon treatment
processes designed for wastewaters containing both RDX or HMX, and TNT,
should be designed for effective RDX or HMX removal. This might repre-
sent an over-design of the system for TNT treatment.
d. Although early work indicated that polymeric resin adsorp-
tion for RDX and HMX was promising, more recent pilot studies at NAD
Hawthorne yielded inconsistent treatment results. Additional studies
appear warranted however, in light of the possibility of polymeric resin
regeneration or hydrolysis of RDX or HMX by the hydroxide form of anion
-101-
-------
exchange resins. With the exception of Air Force Plant #78, most LAP
wastewater treatment systems under design or construction are based
upon activated carbon treatment. Air Force Plant #78 will use evapor-
ative ponds, as are now in use. at both Milan AAP and NAVPRO Magna*. Such
ponds are effective in semi-arid regions.
*Also designated NIROP Magna.
-102-
-------
SECTION VI - NITROGLYCERIN
52. Introduction
a. Nitroglycerin is used primarily as an ingredient of double-
base and triple-base propellants. It is manufactured at Radford and
Badger AAP's, NOS Indianhead and NAVPRO Magna. Nitroglycerin (NG) is
manufactured by the batch process at Radford and Badger AAP's. Radford
also has the newer continuous (Biazzi) NG process, as do both of the
Navy facilities. The batch process at Radford is no longer in use, and
the batch process at Badger is scheduled for replacement by the continu-
ous process (3f). Thus, batch manufacture of NG will soon be phased out
totally in the military munitions industry, being replaced by the Biazzi
system.
b. The batch NG process employs three steps. In the first
step, glycerol is added to a mixture of concentrated sulfuric and nitric
acids. The nitration mixture is agitated, and cooled to hold the tem-
perature below 25°C. After the reaction is completed the reacted prod-
uct (a mixture of NG, dinitroglycerin (DNG), water, and spent sulfuric
and nitric acids) passes to a gravity separator where the second step
involves gravity separation of the NG from the nitration acids, and de-
cantation of the NG. Spent acid is normally sent to recovery. In the
final step, the NG is washed with aqueous sodium carbonate solution and
water to remove and neutralize residual acidity. Several sequential
water washes may take place. The NG is then taken to storage, usually
in rubber lined carts, until used in propellant manufacture.
c. The continuous NG process is an automated system in which
the nitroglycerin is produced by similar reactions to the batch process.
At Radford AAP, for example, the Biazzi system incorporates the follow-
ing steps (6e). The reaction components are first combined in a temper-
ature controlled reacter. After nitration, the NG and acid mixture flows
to an acid separator, where excess acids are removed. The NG then re-
ceives three 16-percent soda ash washes and two fresh water washes, to
remove residual acids and impurities. The NG flows to an emulsifier
where it is mixed with an equal volume of a three percent soda ash solu-
tion and conveyed to the NG storehouse. When the NG is retrieved from
the storehouse, the emulsifying solution is decanted to catch basins,
to remove non-soluble NG by gravity settling.
53. Waste Sources
a. Wastewater flows from NG manufacture result from the soda-
ash solution and water washes following the nitration step, and from NG
storage when dilute soda-ash solution is decanted before NG use. The
only batch NG process in operation is at Badger AAP. Table 32 presents
wastewater characteristics for this operation. For comparision, aver-
age wastewater characteristics for batch NG manufacture in the commercial
explosives industry are also included in Table 32 . No data on nitro-
glycerin content of the Badger AAP wastewater flow are available. For
-103-
-------
Table 32. Wastewater Characteristics of Batch
Nitroglycerin Production
Parameter*
Flow, MGD
PH
Temperature , °C
Kjeldahl-N
Nitrate-N
Sulfate
COD
Nitroglycerin
Sodium
Badger AAP (la(3))
Range**
0.06-0.17
1.7-9.5
10-19.4
1.1-5.1
0.5-200
62-415
18-340
-
-
Average
0.11
4.7
14.6
2.5
116.6
242.6
109.1
-
-
Commercial Explosives
Industry (4d)
0.0097
2.7-10.0
-
23.0
5564
3154
2260
315-12,700
13,323
*A11 parameter values in mg/1 except flow, pH and temperature.
**Maximum and minimum values taken from four-hour composites.
-104-
-------
commercial explosives plants, NG content of manufacturing wastewater
ranges from 315-12,700 mg/1. The nitroglycerin content of process waste-
water will represent both soluble NG and NG present above the solubility
limit. Nitroglycerin solubility in water at 15°C is 1,270 mg/1 (16zu).
b. The pH of wastewater from NG manufacture ranges from low
for the first (sour water) wash after nitration, to alkaline for sub-
sequent washes with sodium carbonate to neutralize residual acidity. In
addition, the wastewater is high in nitrate and sulfate from the nitra-
tion acid mixture, and COD from the organic reactants and products of
NG manufacture.
c. Waste characteristics for the continuous Biazzi NG process
at Radford AAP are presented in Tables 33 and 34. Other flows associated
with NG manufacture are a 15,000 gpd flow for the air compressor build-
ing and a 20,000 gpd flow from the glycerin soda solution refrigeration
house. These latter two flows are noncontact and contain no explosives
contaminants (6e). These two flows are currently combined with waste-
water from the nitration house and discharged to the receiving water.
The manufacturing wastewater (Table 33) is notably high in sulfate, ni-
trate, alkalinity and dissolved solids. Nitroglycerin averages 1,300
mg/1 in the wastewater, while dinitroglycerin averages 850 mg/1. The
COD is also high, reflecting in part the NG and DNG content. NG and
DNG have, respectively, theoretical COD's of 0.8 and 1.02 mg COD/mg NG
and DNG. While DNG gives 100 percent of the theoretical value in the
COD test, NG yields only 18.5 percent (3u). Thus, NG plus DNG account
for 86 percent of the COD value in Table 33, but only 19 percent of the
COD value in Table 34. The NG store house waste represents only 5000
gpd and is more alkaline than the manufacturing wastewater. In other
constituents, the storehouse waste is similar but more dilute.
d. No wastewater characterization data are available for NG
manufacture at NAVPRO Magna or NOS Indianhead. However, both facilities
use the continuous Biazzi process, and the wastewaters should be similar
to data presented in Table 33. At NOS Indianhead, an extensive and com-
prehensive environmental monitoring program, titled the Navy Environ-
mental Protection Data Base (NEPDB) Program has recently been initiated.
The program involves sampling at nine stations on the facility for a
variety of standard pollution parameters such as BOD, heavy metals and
oil and grease. A future monitoring program will extend to analysis for
explosives pollutants such as NG, DNG or nitrocellulose and to character-
ization of individual process effluent streams (15zl, 14zc, 14zr).
54. Effects of Water Management
a. A water management program proposed for the Radford AAP
continuous NG process would reduce wastewater discharge by 64 percent,
from 50,000 to 18,000 gpd (6a). The program incorporates the following
components:
-105-
-------
Table 33. Combined Wastewater Characteristics of Radford AAP
Continuous NG Nitration and Spent Acid Buildings
9463 and 9466 (6e)
Parameter*
Flow, MGD
pH
BOD
COD
Nitrate-N
Sulfate
Total Alkalinity (CaCO )
Suspended Solids
Dissolved Solids
Nitroglycerin
Dinitroglycerin
Range
-
8.4-9.2
1.5-6.5
1,000-1,400
7,500-20,000
534-3,550
9,000-16,400
3.0-63.3
68,000-98,950
800-1,800
520-1,180
Average
0.015
8.6
4.5
1,228
13,280
1,416
12,700
23.0
81,626
1,300
850
*A11 parameter values in mg/1 except flow and pH.
-106-
-------
Table 34. Wastewater Characteristics of Radford AAP
NG Store Houses 9471 and 9472 (6e)
Parameter*
Flow, MGD
pH
BOD
COD
Nitrate-N
Sulfate
Total Alkalinity (CaCO )
Suspended Solids
Dissolved Solids
Nitroglycerin
Dinitroglycerin
Range
-
10.2-11.3
2.4-4.1
460-1456
270-665
20-179
7,500-18,000
3.3-22.1
2,952-30,848
83-490
41-248
Average
0.005
10.5
3.2
912
477
130
11,400
11.3
13,905
266
130
*A11 parameter values in mg/1 except flow and pH.
-107-
-------
(1) Installation of a refrigeration unit at the glycerin
soda solution refrigeration house. Net water savings
amounts to 20,000 gpd.
(2) Use of 10,000 of the 15,000 gpd cooling water flow
from the air compressor house as washwater for spent
acid wash. Net water savings are 10,000 gpd.
(3) Use of the first washwater from NG manufacture (after
the soda-ash solution neutralization step) as emulsi-
fier water from NG storage. Net water savings are
1200 gpd.
(4) Reuse for NG storage water, after adjustment of the
soda ash content to 16 percent, as the first neutral-
ization wash. Net water savings amounts to 1200 gpd.
The wastewater discharges remaining, after implementation of this program,
would include the remaining 5000 gpd of air compressor cooling water flow
which contains some oil, a 10,000 gpd flow from spent acid washing, a
1200 gpd 16% soda- ash neutralization flow, 420 gpd clean-up water and
1575 gpd (winter only) in-line heating water (6a). The net change in
actual process contaminated water involves reduction from a present flow
of 14,000 gpd to 11,600 gpd, with the balance due to savings in non-
contact water use. The process water use reduction is thus 2,400 gpd,
due to sequential use of a single 1200 gpd flow as first NG washwater,
second NG emulsifier (storage) water and finally for soda-ash supple-
mented NG neutralization.
b. The water management program proposed for Radford AAP is
not unique to the manufacture of NG by the continuous Biazzi process ,
and would be equally effective for both the batch process at Badger AAP
and the Biazzi process at NOS Indianhead.
55. Current Treatment Practice
a. Process wastewater at all production facilities are routed
through catch basins, for removal of non-soluble nitroglycerin. At Rad-
ford AAP, these catch basins have been reported to contain very little
accumulation of solids (ln(3)). At both Radford AAP and NOS Indianhead,
the catch basin overflow is discharged without further treatment (3f ,
b. Badger AAP uses percolation/evaporation ponds for final
disposal of the NG process wastewater. At Badger, all wastewater from
the NG manufacturing area flows to two percolation ponds , where the
liquid leaches into the ground (la(3)). The combined wastewater flow to
the pond is reported to have an average pH of 4.7, average nitrate-N
concentration of 117 mg/1 and average sulfate concentration of 240 mg/1.
The ponds have sandy bottoms and the leaching system is reported to oper-
ate effectively (la(3)). At 25 percent capacity production of NG, the
ponds receive 120,000 gpd of wastewater (2b).
-108-
-------
c. NIROP Magna follows a similar program for wastewater dis-
posal. Nitroglycerin manufacture wastewater is discharged to earthen
sumps. There is some removal by percolation and substantial evaporation.
However, twice per year the sumps are allowed to dry up, and the sedi-
ments decontaminated for residual NG and DNG by placing explosives in
the bottom of the sumps and detonating the explosives. A similar pro-
cedure has been employed at one commercial explosives plant (4d). Many
commercial NG plants also treat their acidic wastewater for pH, by per-
colation of the waste through crushed limestone beds or other neutral-
ization techniques (4d) .
56. Modifications toCurrent Treatment
a. Among the four nitroglycerin-producing facilities, only
Radford AAP has a program for additional treatment. Current plans, in
addition to water management efforts, include installation of a gutter-
ing system to collect the wastewater at specific processing buildings
and direct it to primary clarifiers to remove settleable solids, grit
and oily material such as insoluble NG. The clarified effluent will be
treated with caustic and sodium sfclfide, to decompose NG and DNG (15zp).
The effluent will then be treated in a 250,000 gpd secondary treatment
process (6a). An activated sludge system is proposed as the secondary
treatment method (3f). The collection and treatment facilities at Rad-
ford have been funded (4c).
b. Some success has been achieved with biodegradation of ni-
troglycerin (3h). Preliminary laboratory work with wastewaters from NG
production indicates that some components are readily biodegraded, but
it has been reported that at least one component of the wastestream is
recalcitrant to direct biological attack (3h). Based upon oxygen util-
ization respirometry studies at Radford AAP, this recalcitrant constit-
uent appears to be NG. Standard substrate solutions to which DNG was
added showed higher oxygen utilization rates than for the standard sub-
strate alone, while rates were depressed in standard solutions to which
NG had been added (6a).
c. Further, nitroglycerin is reported toxic above 600 mg/1
(15y). In another study on nitroglycerin manufacturing effluent,
gravity clarification yielded NG levels of 900 to 2100 mg/1. This waste
was treated by activated sludge (16zt). At 16 hours aeration time, suc-
cessful treatment was achieved, with a decomposition product of nitrite.
The maximum concentration of NG which could be treated by the activated
sludge process was 400 to 500 mg/1 (16zt). The evidence thus indicates
that biological treatment of high NG wastewaters, such as described in
Tables 32 and 33, is not feasible without dilution or pretreatment for
NG removal. Further, the proposed biological treatment system would not
control the high nitrate, sulfate and alkalinity levels of the wastewater.
57. Alternatives to Current Practice
a. Various treatment methods have been investigated for con-
trol of NG and DNG. Physical processes studied include: reverse osmosis;
-109-
-------
adsorption on to polymeric resins; and activated carbon adsorption.
Chemicals investigated for treatment include: lime saponification;
oxidation with ozone or permanganate; and reduction with sulfide salts.
b. Preliminary results of research studies at Radford AAP in-
dicated that reverse osmosis is capable of removing NG and DNG (6a) .
These results have not yet been confirmed in pilot studies. The extreme
pH values associated with acid washing and soda ash neutralization dur-
ing NG manufacture would likely be detrimental to reverse osmosis cell-
ulose acetate membranes. Thus, pH adjustment prior to reverse osmosis
treatment would be necessary. In addition, the cellulose acetate mem-
brane adsorbs and becomes saturated with NG (3u). Reverse osmosis treat-
ment using currently available membranes thus does not appear feasible.
c. Studies at Radford indicated that activated carbon adsorp-
tion of NG was not feasible, but that adsorption on polymeric resins was
possible (6a). In reticulated resin adsorption column tests, using Rohm
and Haas resin XAD-4, NG removal was excellent, but DNG removal only
fair (6a). The resin requires regeneration with ethyl alcohol or other
organic solvent.
d. Lime and caustic have been reported successful as chemical
treatment techniques for NG. Nitroglycerin is hydrolyzed very slowly at
neutral pH. It also reacts rather slowly in concentrated caustic (1.5%
NaOH), with a reported decomposition rate of 0.24 per mole-second (2b).
Decomposition of NG with lime, at a lime dosage of 2 gm/1, is reported to
require up to three days, yielding decomposition products of calcium sul-
fate, calcium sulfite and calcium salts of low molecular weight organic
acids. The treated waste has a high residual pH (15zt). Saponification
of NG wastewater with lime-at pH 11 has been reported at Radford to yield
nitrate, plus glycerin which could be treated biologically (15x). In
addition to decomposition of NG, the use of lime, plus calcium chloride
or calcium sulfate, would have an additional advantage by reducing the
high alkalinity of the wastewater. The pertinent reactions to remove
alkalinity by precipitation are given below.
2 NaHC03 + Ca(OH)2 -» CaC03 + Na2C03 + 21^0 [I]
Na7CO, + CaSO, -» CaCO + Na SO [21
L 5 4 324
Calcium carbonate is precipitated in each of the reactions, thereby re-
ducing the carbonate alkalinity. In laboratory studies at Radford AAP
on this process, alkalinity was reduced from 16,400 to 70 mg/1. Addi-
tion of anionic polymer resulted in effective precipitate sedimentation
(6e). Based upon these results, approximately 70 Ib of lime and 140 Ib
of calcium sulfate. are required per 1000 gallons of wastewater to reduce
alkalinity to acceptable levels (6e). Calcium carbonate sludge sedimen-
tation can be accomplished within five minutes with addition of 5 mg/1
anionic polymer flocculant.
-110-
-------
e. By adding excess lime in the alkalinity precipitation ex-
periments, resulting in higher pH, complete NG and DNG decomposition
occurred within 30 minutes at pH above 11 (6e). The excess lime required
would be about 20 additional Ibs per 1000 gallons of wastewater. The
Radford study indicated that removal of alkalinity, plus decomposition
of NG and DNG in a single treatment step were possible (6e).
f. Both ozone and permanganate have been studied as oxidants
for decomposition of NG and DNG. Permanganate dosages up to 200 mg/1,
at reaction times of one hour, resulted in only about 40 percent decom-
position of NG and essentially no removal of DNG (6e). Although higher
permanganate dosages would likely improve NG treatment, the method does
not appear to be economically feasible at the high dosages necessary for
NG, and, in any case, is ineffective for DNG.
g. Laboratory-scale ozonation studies indicate effective de-
composition of both NG and DNG, at long reaction times and high ozone
dosage. NG nitration wastewater, after lime plus calcium sulfate treat-
ment for removal of alkalinity, was treated with ozone at an ozone ad-
dition rate of 4160 mg/hr ozone per liter of wastewater. During the
first ten minutes of treatment 100 percent of the ozone bubbled through
the treatment column was consumed, but during the remaining three hours
of treatment only 1000 to 1400 mg/hr/liter were consumed in the oxida-
tion reaction. Ozone treatment at pH values of 4.9, 7.2 and 8.6 indi-
cated that pH had little effect on ozone treatment. At six hours treat-
ment time and pH 4.9, NG at 324 mg/1 and DNG at 122 mg/1 initial con-
centrations were completely decomposed (6e). The ozone requirement was
23 mg ozone per mg NG and DNG.
h. Sodium sulfide solution has been used for many years to
decompose NG during equipment cleanup (4g). In studies at Radford AAP,
NG nitrator wastewater was treated with sodium sulfide dosages of 100
to 700 mg/1. Initial NG and DNG concentration of 351 and 122 mg/1,
respectively, were totally decomposed within 30 minutes at sodium sul-
fide dosage of 320 mg/1 (6e). Treatment pH was 8.6. These results in-
dicate that in terms of treatment, chemical dosage and reaction time
required, lime treatment and sodium sulfide treatment yield equivalent
results. Table 35 presents a comparison of treatment chemical costs for
lime, permanganate, ozone and sulfide treatment. On the basis of total
cost, sulfide treatment is most economical. However if removal of al-
kalinity is also required, single-step lime plus calcium sulfate removal
of alkalinity plus NG and DNG becomes most economical. The major dis-
advantages of this treatment technique are the volumes of calcium car-
bonate sludge generated which must be disposed of, and the added soluble
sulfate from calcium sulfate treatment chemical use.
i. There are certain potential problems associated with the
use of sulfide as a treatment chemical, including the possibility of
residual toxic sulfide ion in the effluent. Further, sulfide treatment
of wastewater which contains insoluble NG may represent an explosive
hazard, since the saponification and reduction reactions are exothermic.
-Ill-
-------
Table 35. Treatment Chemicals Cost for Nitroglycerin
Manufacturing Wastewater Treatment (6e)
Treatment
Cost per
1,000 gal
Lime-Calcium Sulfate Treatment
a) For Alkalinity Only
b) Increment for NG and DNG Decomposition
c) Total Cost
Polymeric Resin (Regeneration by ethanol)
Ozone Oxidation
Sodium Sulfide Decomposition
$ 2.38
$ 0.18
$ 2.56
$10.00
$38.50
$ 0.81
-112-
-------
With concentrated NG, where the heat of reaction is not dissipated,
danger of explosion due to thermal shock has been cited (4g). In
addition, the sulfide treatment method is reported to liberate sulfur
compounds with a very disagreeable pungent odor (4g). Despite these
factors, the sulfide decomposition process has been selected for treat-
ment of NG wastewaters at Radford AAP (15zp).
58. Summary
a. Wastewaters from nitroglycerin manufacture are common to
both the military and commercial explosives industries. The wastewaters
typically vary in pH from acidic to alkaline, and are high in NG, DNG,
nitrate and sulfate. Current treatment is limited to catch basins, with
final discharge to waterways or evaporative ponds. Although several
treatment processes have been proposed, most are ineffective for NG, DNG
or both. Only sulfide decomposition is proven, and Radford AAP plans to
use sodium sulfide treatment followed by activated sludge.
b. However, there are several undesirable side-effects of
sulfide treatment, including: the presence of residual toxic sulfide
ion in the process effluent; the liberation of disagreeable odors during
sulfide decomposition; and the possibility of explosive hazard due to
exothermic reaction. These aspects of sulfide treatment indicate that
new research and development efforts should be undertaken to assess al-
ternative processes for both decomposition and recovery of NG and DNG.
-113-
-------
SECTION VII - SELLITE
59. Introduction
Sellite (sodium sulfite) is manufactured at Volunteer and Joliet
AAP's. It is used in the purification of crude TNT to remove unwanted
isomers, leaving a-TNT. Sellite is manufactured by burning sulfur to
produce sulfur dioxide, S02- The product gas is countercurrent scrubbed
with water to remove sulfur trioxide and other impurities, followed by
absorption of S02 by a solution of sodium carbonate (soda ash) in a
countercurrent packed tower. The sodium sulfite effluent from the tower
is recirculated through the SC>2 absorption tower until the desired sel-
lite strength (16% Na2S03> is obtained. This solution is then used in
TNT purification. Figure 14 is a schematic of sellite manufacture.
60. Waste Sources
a. Wastewaters from sellite manufacture result from overflow
from the gas scrubber water tank, spills of soda ash and sellite solu-
tions, and floor washings and spill clean-ups (8f). The sellite efflu-
ent is characterized by a widely fluctuating pH, low dissolved oxygen,
and high concentrations of ash, sulfate, sulfite and total dissolved
solids. This waste flow discharges into a small lagoon where it is
aerated and neutralized, and is then discharged. Waste flow, with one
of the two sellite manufacturing units at Joliet AAP in operation, has
been reported as ranging from 0.062 to 0.614 MGD with an average flow
of 0.222 MGD. Flow during this period of measurement was strongly in-
fluenced by rainfall runoff into the effluent ditch (lg(3)). An aver-
age flow of 0.173 MGD results if flow measurements on days of heavy
rainfall are disregarded (lg(3)). A separate study of sellite wastewater
discharge indicated a range of 0.058 to 0.737 MGD, with an average flow
of 0.213 MGD, or 148 gpm (3b). The estimated actual process effluent
per sellite manufacturing unit has been estimated at 0.062-0.125 MGD,
exclusive of surface runoff (8d).
b. Limited data are available on the process effluent prior
to neutralization, lagooning and aeration. Table 36 presents available
data. The data indicate an anoxic discharge which at the average flow
rate of 148 gpm represents 227 Ibs/day COD, 1246 Ibs/day sulfate and
1694 Ibs/day sulfite. The anoxic condition of the wastewater results
from the reaction of sulfite with dissolved oxygen to produce sulfate.
The data of Table 36 were collected during summer months (June and July),
and in cooler months, the effluent contains trace amounts of dissolved
oxygen (3b).
61. Current Treatment Practice
Wastewater discharged from sellite manufacture is currently treated
by pH adjustment with soda ash, followed by oxygen addition by only one
mechanical surface aerator in a small flow-through lagoon. The lagoon
-114-
-------
SODIUM
CABONATE
SOLUTION
r
SODIUM
SULFITE
SOLUTION
STACK GAS
(SO , ACID MIST)
ABSORPTION
TOWER
(PACKED
SELLITE TO
TNT AREA
so2 + so3
SO,
SCRUBBER
TOWER
(PACKED)
FURNACE
EFFLUENT
SCRUBBER
WATER
WATER
MAKEUP
SCRUBBER
WATER TANK
SULFUR
Figure 14 - Sellite Manufacture at Joliet AAP (15zb)
-115-
-------
Table 36. Sellite Wastewater Characteristics Before Treatment (3b)
Parameter*
PH
COD
Dissolved Oxygen
Sulfate
Sulfite
Flow, gpm
Range
2.3-3.5 (minimums)
17-493
0.0
287-2410
450-2260
40-512
Average
2.3-5.6
128
0.0
701
953
148
*A11 in mg/1, except pH and flow.
-116-
-------
effluent is discharged without further treatment. Tables 37-39 summarize
effluent characteristics from the lagoon, as a result of three separate
monitoring studies. Table 39, presenting both lagoon influent and ef-
fluent data, indicates that the lagoon system is ineffective in producing
an effluent of acceptable pH and dissolved oxygen concentration, although
sulfite is 96 percent removed in the aeration process. Since sulfate
concentration does not increase by a stoichiometric equivalent to the re-
duction of sulfite, removal may result from air stripping of sulfur oxides,
as well as oxidation of sulfite. The dissolved oxygen data of Table 39
are somewhat atypical, since results of additional monitoring efforts on
lagoon effluent revealed effluent dissolved oxygen levels of 0.4 to 10.6
mg/1 when the surface aerator was in operation, with oxygen levels below
3.0 mg/1 occurring only 27.7 percent of the time (3b). With the aerator
not operating, the lagoon effluent oxygen ranged from 0.0-3.0 and aver-
aged 0.6 mg/1. Thus, mechanical aeration does improve effluent oxygen
and sulfite levels, but it is likely that additional aeration capacity
is required to provide more consistent results.
62. Modifications to Current Treatment
a. In order to provide more effective and consistent treat-
ment for sulfite, longer aeration periods and additional mechanical aera-
tion capacity are required. The present surface aerator has a rated
capacity of 14 Ib/hr oxygen input, and is estimated to be operating near
rated capacity (3b). To achieve a consistent effluent dissolved oxygen
level of 3 mg/1, added capacity of 8 Ibs/hr oxygen input is reported
necessary (3b). The performance of the lagoon system would also be en-
hanced by flow equalization, as indicated by the almost 13-fold varia-
tion in flow rate shown in Table 36.
b. Poor pH control is evidenced by all available data (Tables
37-39), but inadequate treatment has been ascribed to equipment malfunc-
tion rather than poor process design (3b). The pH system is set up for
proportional control, which yields imprecise control capability. Im-
proved performance would result from installation of a feed-forward con-
trol system, and the effluent should be monitored by an alarm, to warn
of equipment malfunction.
c. No attempt is made to control effluent sulfate, which has
been reported to reach 1750 mg/1 (Table 38). Current pH control is with
soda ash, which yields soluble reaction products. Use of lime would be
equally effective in neutralization, given the detention time of the
lagoon system, and might have the additional benefit of partially re-
moving sulfate by precipitation of calcium sulfate.
63. Summary
In summary, the type of technology applied to sellite wastewater
is appropriate, with the possible substitution of lime for soda ash in
neutralization. Treatment deficiencies result from inadequate treatment
system aeration capacity and detention time, plus poor neutralization
performance, rather than other factors.
-117-
-------
Table 37. Sellite Wastewater Treatment System
Effluent Characteristics (lg(3))
Parameter*
Temperature , °F
PH
Acidity
Alkalinity
Total Solids
Suspended Solids
Dissolved Solids
TOG
Sulfate
Sulfite
Range
65.0-79-0
2.1-7.8
9.0-720.0
0.0-382.0
824-2030
1.0-40.0
784-2015
6.0-10.2
530-1400
10.0-1600
Average
72.0
5.3
209.7
125.0
1453
17.1
1438
8.5
928.6
330
*A11 in mg/1,.except pH and temperature,
-118-
-------
Table 38. Sellite Lagoon Effluent Characteristics (8d)
Parameter*
Temperature, °F
pH
Color
Turbidity
Hardness
Dissolved Oxygen
BOD5
COD
Sulfate
Chloride
Sodium
Suspended Solids
Dissolved Solids
Range
63-73
3.3-9.8
5-50
3.3-150
190-449
0.0-10.3
0.0-94.7
8.0-199
398-1750
39-301
137-1350
2-37
338-11,806
Average
71.3
7.1
36.0
24.7
342.1
4.3
12.7
44.4
752.8
105.6
471.3
19.7
2408
*A11 in mg/1, except pH and temperature.
Table 39. Comparison of Lagoon Influent and Effluent
Sellite Wastewater Characteristics (3b)
Parameter*
pH
Dissolved Oxygen
COD
Sulfate
Sulfite
Influent
2.3-5.6
0.0
128
701
953
Effluent
4.3-10.0
0.0
38
1008
35
-119-
-------
SECTION VIII - PROPELLANTS
64. Introduction
a. Propellents may be divided into three classifications:
solvent, solventless and composite. There are many variations in for-
mulations of propellants within each classification. Solvent propellants
include single and multi-base compositions with NC as the major ingred-
ient plus stabilizers and catalysts. Double base propellants incorpor-
ate nitroglycerin, while triple base propellants include both nitro-
glycerin and nitroguanidine. The basis for the solvent classification
is the use of one or more NC solvents in the formulation process. Sol-
ventless propellants also contain NC plus typical ingredients of solvent
propellant, but are blended in a x^ater-wet state without addition of
organic solvent. Composite propellants contain, in addition to typical
double-base ingredients, significant amounts of oxidizers such as ammoni-
um perchlorate or HMX, plus powdered aluminum. Additionally, composite
propellants are based on a variety of organic binders, substituted for
the double-base matrix of NC-NG. The latter types of composite propel-
lants are discussed in Section IX.
b. Solvent propellant manufacture produces smokeless powder,
powder which may be extruded into solid rocket motors, and casting pow-
der for^composite propellants. Solventless propellant is produced as a
"carpet roll," which may be extruded into solid rocket motors, or cut
into mortar increments. Casting of composite propellant is more proper-
ly a loading then manufacturing operation and, as such will be discussed
in Section IX of this chapter.
c. The wastewaters of propellant manufacture reflect both the
constituents of the products, and the blending process (e.g., solvent vs.
solventless). Wastewater treatment technology for propellants will be
discussed in separate sections below, according to the two-propellant
classification of solvent and solventless.
SECTION VIII-A - SOLVENT PROPELLANTS
65. Introduction
For purposes of waste characterization and evaluation of treatment
technology, solvent propellants will include single and multi-base pro-
pellants, and ball powder. Ball powder production is a special fora of
the solvent process. Badger AAP is the only military plant which manu-
factures ball powder (3a).
66. WasteSources
a. The process steps are essentially the same in the pro-
duction of solvent-type single, double and triple base propellants.
-120-
-------
Major differences are in the specific chemicals and explosive ingredients
added. In batch single base propellant manufacture, nitrocellulose is
shredded and then mixed with ethyl alcohol, to dehydrate the NC and form
a colloidal dough which is pressed into blocks. Pressing yields a weak
alcohol solution. The blocked dough is then blended thoroughly with
ethyl ether, DNT and other ingredients, depending upon the exact formu-
lation desired. The dough is blocked once again and then extruded
through dies, resulting in strands having a fixed number of perforations.
The strands are cut to a specified length and transferred to a solvent
recovery area.
b. In the solvent recovery area, the cut powder is washed with
water previously heated to 60°C. This step partially removes residual
solvents and DNT from the perforations within the pellets. After the
water wash is completed the pellets are purged with methane gas, to re-
move final traces of solvents. Solvents are recovered from the methane
(16zw). The powder is then air dried. The manufacture of double base
solvent propellant is similar to the single base, with the addition of
several processing steps for adding nitroglycerin and other chemicals
for the formulations desired. Nitroglycerin composite double-base pro-
pellants require additional processing steps for the blending of ammonium
perchlorate, HMX and powdered aluminum.
c. Triple base propellant production is a somewhat more com-
plicated operation. About 400 Ib of propellant is made per batch (2c).
After shredding and dehydration, nitrocellulose is mixed with nitro-
glycerin and acetone to form a pre-mix slurry. About 200 Ib of nitro-
guanidine blended with about 4 Ib of ethylcentralite, a stabilizer and
waterproofing agent, is slowly added to the pre-mix and blended. Inert
gas is bubbled through the mix to remove volatile solvents, until the
mixture reaches the desired viscosity. The dough is then blocked and
extruded through dies, and the strands cut. Twenty pound batches of
propellant are placed in trays and forced air dried. The triple base
propellant grains are then glazed with graphite prior to packout (2c).
d. At NOS Indianhead, NC is purchased either water-wet or
alcohol-wet. If obtained water-wet, the water is pressed from the NC,
which is then rewet with ether. Subsequent processing steps, to manu-
facture double-base propellant, are as described previously (15zl).
e. Alcohol solution from the dehydration and first blocking
steps is recovered by a rectification (distillation) process (ln(3)).
The weak alcohol solution is passed through a steam purged still, with
alcohol vapor recovered at the top and condensed. Steam condensate and
residual still liquid are discharged as wastewater from the bottom of
the still (ln(3)).
f. Volumes of wastewater generated in single and multi-base
propellant manufacture are relatively small, and consist almost entirely
of cleanup water from shift and weekly equipment cleanup and floor wash-
down operations. Screens, dies and often metal objects which have come
-121-
-------
into contact with propellant are cleaned by chemical digestion, using
concentrated sodium hydroxide. In this operation the screens, dies and
metal parts are hand packed into a wire basket. The cleaning tank is
charged with 500 gallons of a twenty percent sodium hydroxide solution
and heated to 217°F. The basket is then lowered into the cleaning tank.
After completion of the digestion process, the basket is removed, drained
and lowered into a rinse tank where the parts are cooled and rinsed in
500 gallons of rinse water. When the cleaning operation is completed,
the water from both tanks is discharged (6a). Currently, at Radford AAP,
the spent caustic is blended into the TNT red water (6a) . This caustic
solution is reported to total 1,200 gpd (3f). A similar discharge at
NOS Indianhead is reported to range from 20-300 gpd (15zl).
g. Limited data are available on waste volumes associated
with single and multi-base propellant manufacture and clean-up opera-
tions. Table 40 summarizes data from Radford AAP. The data, taken from
three sources, do not show agreement in waste volumes. For example,
average daily volumes of 500 and 86,000 gpd have been reported for single
base propellant manufacture at Radford AAP (ln(3), 6e). The larger vol-
ume incorporates an unknown portion of cooling water, however. Tables
41 and 42 present wastewater characteristics, for single and multi-base
propellant manufacture, for the high and low discharge volumes given in
Table 40. Despite the differences in reported flow volumes for the two
data sources, there is little difference in the concentrations of in-
dividual constituents. As shown in Table 41, the larger reported flow
contained higher levels of total organic carbon and COD, while in Table
42, the reverse situation holds. Beyond these two parameters, plus ni-
trate for single base, there is little to differentiate between the
wastes.
h. The heated water used to partially remove residual solvent
from the extruded and cut powder, referred to as "water-dry waste," is
diluted with cooling water and discharged at Radford AAP (16zw). This
waste is reported to contain significant quantities of ethyl alcohol,
diethylether and DNT. Table 43 presents characteristics of this waste,
prior to dilution with cooling water. Although the data are limited,
the high organics level does indicate the presence of solvents at sig-
nificant concentrations.
i. A further wastewater source is propellant conveyance water
in the propellant sorting operation. Table 44 summarizes waste char-
acteristics for the Radford AAP propellant sorting process.
j. One further source of wastewater from solvent propellant
manufacture is the solvent recovery process, for recovery of alcohol
from the dehydration water. Recovery of alcohol by distillation results
in a residual wastewater consisting of still bottoms. These still bot-
toms, from two rectification units at Radford AAP, are reported as 21 100
and 8,640 gpd respectively. In addition, there is a cooling water dis-
charge for each alcohol rectification unit of about 430,000 gpd (ln(3)).
-122-
-------
Table 40. Wastewater Volumes for Solvent Propellant
Manufacture at Radford AAP
Propellant
Single Base
Multi-Base
Combined
Wastewater Volume, gal
Daily Average
500
86,000*
660
74,000*
5,000
Periodic Washdown
3,700
4,960
Reference
ln(3)
6e
In (3)
6e
4c
*Includes some cooling with process water.
-123-
-------
Table 41. Wastewater Characteristics of Single Base
Propellant Manufacture at Radford AAP
Parameter
Flow, gpd
Temperature , °F
pH
Total Solids
Suspended Solids
Total Organic Carbon
COD
N02 + N03 Nitrogen
Ethyl+Diethyl Ether #
Concentration (In (3))*
Average
500
65.9
8.7
2233.6
8.1
29.9
43.9
159.4
4.1
Range
-
56-80
7.6-10.0
10-25,358
1-161
5-420
10-195
0.1-1920
4.0-7.0
Concentration (6e)*
Average
86,000
69.2
7.6
262
6.9
83.7
627.4
5.4
-
Range
-
60-77
6.7-8.7
10-1068
1-21
5-777
10-8755
0.4-21.6
-
*A11 units in mg/1 except flow, temperature and pH.
//As reported by U.S. AEHA.
-124-
-------
Table 42. Wastewater Characteristics of Multi-Base Propellant
Manufacture at Radford AAP
N3
Un
I
Parameter
Flow, gpd
Temperature, °F
pH
Total Solids
Suspended Solids
Total Organic Carbon
COD
N02 + N03 Nitrogen
Concentration (ln(3))*
Average
660
-
7.95
472.4
31.6
1012.0
194.5
4.4
Range
-
-
7.3-8.7
57-1677
12.5-67
7-5300
17-510
0.25-17.3
Concentration (6e)*
Average
74,000
78
7.6
637
16
10
107
11.6
Range
-
53-90
3.2-8.9
70-1805
0-45
2-56
5-1195
4-73
*A11 units in mg/1 except flow, temperature and pH.
-------
Table 43. Characteristics of Water-Dry Waste
from Radford AAP (6e)
Parameter
Flow, gpd
Temperature, °F
PH
COD, mg/1
TOC, mg/1
NO, -Nitrogen, mg/1
Range
-
155-165
6.55-7.75
593-28,600
96-3525
5-45
Average
15,000
150
7.26
7,295
1,011
19
Table 44. Characteristics of Propellant
Conveyance Wastewater at Radford AAP (6e)
Parameter
Flow, gpd
Temperature , °F
PH
Total Solids, mg/1
COD, mg/1
TOC, mg/1
N0,-Nitrogen, mg/1
Range
-
60-79
7.4-8.6
91-226
85-478
14-71
3.5-13.0
Average
210,000
69
7.7
130
174
30
8.0
-126-
-------
Still bottom waste characteristics for Radford and Badger AAP's are
presented in Table 45. The pH of the still bottoms is maintained near
neutrality by pumping an 8 percent caustic solution into the lower sec-
tion of the still (ln(3)). The addition of caustic serves to decompose
NC fines in the still bottoms (la(3)). The wastewater at Radford AAP
is discharged, after dilution with cooling water, to the Radford C-Line
Acid Neutralization Facility (ln(3)). Cooling water accounts for 97
percent of the total flow for the alcohol rectification area and the com-
bined flow, as a result of the high dilution of still bottoms by cooling
water, is relatively uncontaminated.
k. A comparison of the two Radford discharges described in
Table 45 suggests that the discharge of Bldg. 1502 was diluted prior to
the point of sample collection. This conclusion is substantiated by the
fact that the second discharge is up to three hundred times more con-
centrated in solids than the first. Notably high, in addition to solids,
is organic matter. The wastewater would also be expected to contain re-
sidual NC and NC decompositon products. With the exception of higher
average pH and nitrogen, the Badger AAP discharge is intermediate to the
two Radford discharges.
1. The manufacture of ball powder, a double base solvent
propellant, at Badger AAP involves processing steps and compounds not
used in other solvent propellants. Nitrocellulose is ground and washed
in benzene plus ethylacetate to extract impurities. The benzene is re-
moved by decantation, followed by vacuum distillation. The purified NC
in ethylacetate is then shaped into uniform balls in a second still.
Sodium sulfate and collagen protein is added to the solution to prevent
ball agglomeration during the distillation step. The solvent is boiled
off, and the ball powder is washed, size separated by wet screening, im-
pregnated with NG, washed, rolled moisture dry, coated with graphite,
screened, blended and packed out. Wastewaters consist of the two washes,
plus cooling water from solvent distillation. The wash and screening
waters pass through separate clarifiers, where settleable solids are
removed. The clarifiers overflow to an industrial waste sewer, which
discharges a variety of plant wastes into a series of three shallow
lagoons (la(3)). All cooling water is recycled through a cooling tower
(3a). Approximately half of the process water use is in the wet screen-
ing operation (3a). Process water use is summarized in Table 46. Wet
screen sizing accounts for over half of the total process water use,
with wash water representing an additional major fraction. The wash
water is heavily contaminated with collagen, benzene and sodium sulfate
(3a). Detailed waste characterization data are not available on the ball
powder process effluents (3a). Solvent levels in selected effluents are
given in Table 47. The ball still effluent, due to its high solvent con-
centration, and the wash water, due to its large volume, account for most
of the solvent discharge for the ball powder processes. In addition to
solvents the presence of collagen presents a major problem due to its
high BOD, and foaming characteristics (3h).
-127-
-------
Table 45. Characteristics of Solvent Recovery Still Bottoms at
Radford, (ln(3)) and Badger (la(3)) AAP's
K)
oo
Parameter*
Temperature , °F
PH
Total Solids
Suspended Solids
TOC
COD
BOD5
N02 + NOv-Nitrogen
Radford-Bldg 1502
Range
-
6.7-7.4
25-98
3-12
15-40
32-144
24-78
-
Average
-
7.1
49.7
7.2
29.7
104.0
44.7
3.0
Radford-Bldg 1503
Range
-
6.5-9.0
4060-5997
2264-2545
500-1240
3400-3640
1850-3500
-
Average
-
7.4
5146.7
2451.3
824.3
3520.0
295.0
—
Badger
Range
57-7120
6.0-11.5
612-4639
40.9-427
31-1160
48-1560
-
2-425
Average
110
9.0
1885.0
251.0
516.8
347.2
-
143.7
*A11 parameters except temperature and pH in mg/1.
-------
Table 46. Process Water Use in Ball Powder
Manufacture (3a)
Process
NC Storage Pits
NC Grinding
Extraction
Weighing
Ball Still
Washing
Wet Screening
Coating
Roll and Dewater
Miscellaneous
Total (D
Volume , gpd
101,856
164,475
170,163
108,398
109,626
591,752
1,728,000
87,290
9,094
34,272
3,104,926
Percent of
Total
3.3
5.3
5.5
3.5
3.5
19.0
55.7
2.8
0.3
1.1
100.0
(1) Represents approximately 30 percent of full-scale
production of ball powder.
-129-
-------
Table 47. Solvent Levels in Selected Ball Powder
Process Effluents (31)
Process
Grinding
Extraction
Ball Still
Washing
Coating
Ethyl
Acetate, mg/1
-
12.0-36.2
140.5-1942.0
35.5-93.4
0.1-18.1
Ether,
mg/1
31.5-69.6
2.4-5.6
-
-
-
Benzene ,
mg/1
-
5.1-16.7
0.1-12.1
<1.0
0.04-0.4
-130-
-------
67. Effects of Water Management
a. Several opportunities exist for effective water management
in the manufacture of solvent propellants. Some, such as the complete
recycle of cooling water for ball powder manufacture at Badger AAP, have
already been implemented. At Radford AAP, dry sweeping of propellant
manufacturing buildings is employed prior to washdown, to reduce propel-
lant loss through washdown (In(3)).
b. Recycle of cooling water from solvent rectification at
Radford AAP provides an opportunity for significant water savings. This
water, used for cooling of the solvent recovery stills, is uncontamina-
ted and could be reused in another area or recycled through a cooling
tox
-------
b Ball powder wash and screening water, after flowing
through settleable solids catch tanks, is combined with other P^nt
wastes and discharged into a series of three shallow lagoons. Screened
catch tanks are also used at Radford AAP, although they are ported to
be ineffective (4c, ln(3)). At NOS Indianhead press water from NC pur
chased water-wet is treated in an activated carbon column prior to dls
charge (15zl). No data are available on the carbon column performance.
c. One commercial explosives manufacturing plant, which pro-
duces smokeless (ball) powder plus NG and other intermediates except NC,
uses extended aeration activated sludge followed by lagooning to treat
the combined waste flow (4d).
d. The activated sludge unit operates with 24 hours aeration
at 8000-9000 mg/1 mixed liquor suspended solids. There is little excess
sludge wasting required for the plant. The lagoon has 5-10 days deten-
tion time, and is aerobic to facultative anaerobic, depending upon tem-
perature and waste loading to the lagoon. Table 48 presents operating
data on the treatment system.
e. Although only limited data are given for the activated
sludge effluent, this process results in BOD5 removal exceeding 95 per-
cent. Table 48 also summarizes the treatment efficiencies for the act-
ivated sludge plus lagoon treatment sequence. Effluent nitrogen levels
are below 0.3 mg/1. The effluent from the lagoon is spray irrigated on
a four acre field, at 3 inches hydraulic loading per month.
f. Dickerson (16zx) has also reported the use of biological
treatment for a smokeless powder explosives waste. Based upon pilot
studies, a two-stage trickling filter was employed to treat process ef-
fluents ranging up to 4500 mg/1 in BOD,.. Operation at hydraulic load-
ings of 20 MGAD* on both units, with a recycle ratio of 20:1, yielded
BOD removal of 97.5 percent and effluent BOD of 90 mg/1. Lower hydraulic
loading rates (8-16 MGAD) slightly improved treatment efficiency, yield-
ing a final effluent BOD of 55 mg/1. However, severe problems were ex-
perienced with clogging of the filters at the lower hydraulic loading
rates.
69. Modifications to Current Treatment
a. Although preliminary treatment feasibility studies are
underway on wastewaters of solvent propellant manufacture, only at Rad-
ford AAP are actual modifications in progress (3f). These modifications
consist of three projects as follows:
(1) provide open drain guttering and primary solids
separation facilities at individual process build-
ings to remove settleable solids,
*Million Gallons per Acre Per Day
-132-
-------
u>
u>
I
Table 48. Smokeless Powder Wastewater Treatment Plant Operation
and Average Waste Concentration (4d)
Parameter*
pH - Range
Total Solids
Suspended Solids
BOD5
COD
BOD5/COD
Phosphorus
Chlorides
Oil and Grease
Ammonia-N
Kjeldahl-N
Nitrate-N
Nitrite-N
Sulfate
Influent
6.5-8.8
1646
292
628.0
1175
0.53
0.28
97.0
35.0
-
-
-
-
351.4
Activated Sludge
Effluent
7.4-8.4
-
-
29.2
117
0.16
-
-
-
-
-
-
-
—
Lagoon Effluent
7.5-10.5
1137
47
16.8
114
0.15
0.15
107.1
8.0
0.25
0.17
0.04
0.1
76.6
Percent
Removal
_
30.9
83.9
97.3
90.3
-
46.4
-
77.1
-
-
-
-
78.2
*A11 concentrations in mg/1, except pH.
-------
(2) A sewer system consisting of existing and new sewer
lines to carry the wastewater to a proposed acti-
vated sludge secondary treatment facility. Separate
sewer systems will be provided for uncontaminated
cooling water and storm water runoff.
(3) the proposed secondary treatment facility, which will
have a capacity to treat 3 MGD of combined waste-
water.
b. The activated sludge process, based upon experience at the
commercial explosives plant described above (4d) , should provide effect-
ive removal of organic solvents and other biodegradable constituents.
There is some limited data however to indicate that the water-dry waste
is stimulatory to filamentous bacteria, which may develop in the acti-
vated sludge process (16zw). If a significant population of filamentous
organisms does develop, a condition called "bulking" occurs in which the
activated sludge will not settle in the secondary clarifier. The sludge
floe then carries over the clarifier weir, resulting in deterioration of
effluent quality, and inability to sustain an acceptable microbial popu-
lation in the activated sludge process. The effectiveness of the acti-
vated sludge process in removing soluble NC is unknown, although removal
of NC fines carried over from the proposed catch tanks should be effect-
ive. Removal of particulate explosive is likely to be by entrapment of
the particulate material in the activated sludge floe.
c. Radford AAP has proposed treatment for the spent caustic
solution used to clean screens, dies and other metal objects which have
come into contact with propellant (6a). The proposed treatment involves
installation of a neutralization task, where the spent caustic solution
will be treated with concentrated sulfuric acid. The neutralized waste
will then be hauled by tank truck to a proposed waste propellant incin-
erator. This treatment process is to be implemented only if the red
water, into which the caustic is now blended, can no longer be sold (6a) .
70. Alternatives to Current Practice
a. Major sources of pollutant discharge from solvent propel-
lant manufacture include the solvent recovery and water dry operations.
These wastes are high in both solvents and dissolved organic propellants.
Numerous methods have been proposed to treat the solvents contained in
the solvent recovery and water dry wastes. These methods include bio-
logical treatment, air stripping, activated carbon absorption,ozonation
and reverse osmosis. Reverse osmosis is normally ineffective'in remov-
ing low molecular weight organics such as solvents from wastewaters.
Table 49, which presents results of laboratory scale reverse osmosis
treatment of solvent rich waste streams from Radford AAP, confirms the
inability of the reverse osmosis process to effectively remove organic
solvents. Equally ineffective was ozone oxidation, as shown by the data
of Table 50. Ether removal by ozone treatment is attributable primarily
to stripping by the ozone gas stream, rather than chemical oxidation (6e)
-134-
-------
Table 49. Results of Reverse Osmosis Treatment of
Solvent Wastewater (3y)
Constituent
Percent Removal
Alcohol
Ether
Dinitrotoluene
Dipheny1amine
Dibutylphthalate
Ethyl Centralite
Dimethylphthalate
Ammonium Perchlorate
20
38
75.5
100
100
100
25
95
Table 50. Ozone Oxidation Treatment of Solvent
Wastewater (3y)
Constituent
Percent Removal
Alcohol
Ether
Dinitrotoluene
Diphenylamine
Dibutylphthalate
Chemical Oxygen Demand
Total Organic Carbon
33
100
38
100
100
20
11
-135-
-------
b. Laboratory scale air stripping studies on solvent wastes
resulted in removal of only 20 percent of the organic carbon within
three hours, and 40-45 percent removal during a total of 22 hours of
treatment (6e). With the exception of ether, for which air stripping
was completely effective within one-half hour, the process does not
appear to have applicability to solvent wastes.
c. Biological treatment is capable of removing solvents from
the wastewaters (3f, 6a). However, recent activated sludge process
treatability studies were only partially successful in treating a water
dry waste from Radford AAP (16zw). Treatment of influent BOD5 concen-
trations of 176 to 888 mg/1 yielded only 65-76 percent BOD removal. Ef-
fluent BOD values up to 316 mg/1 were observed. Further, the waste was
stimulatory to filamentous growth, which normally interfereswith effi-
cient activated sludge process operation. Successful activated sludge
treatment of a commercial solvent propellant waste has been reported
however, where the solvent propellant waste was mixed with and diluted
by other plant process waste streams (4d).
d. One possible beneficial use of solvent wastes may be as
organic substrate carbon donors for biodenitrification. Preliminary
studies indicate that the organic solvent wastes are effective carbon
donors in the biodenitrification process (15y).
e. Activated carbon adsorption appears to provide best sol-
vent removal from solvent recovery and water dry wastes. Complete re-
moval of ethyl alcohol, acetone and diethyl ether, and essentially com-
plete COD removal resulted from carbon treatment of Radford AAP waste-
waters (6e). Adsorption capacities on Witco Grade 718 activated carbon
for ethyl alcohol, acetone and diethyl ether respectively were 0.82,
1.06 and 1.11 Ibs solvent per cubic feet carbon (6e). The effective-
ness of carbon treatment suggests the possibility of steam regeneration
of the carbon, with solvent recovery and reuse.
f. Carbon is equally effective in removing soluble organic
propellant constituents from the wastewater (3y). Results are presented
in Table 51, along with results of reverse osmosis and ozone treatment.
g. Solvent rectification still bottoms, although much more
concentrated than other solvent wastes, have characteristics and con-
stituents similar to the solvent wastes (Table 45). Based upon the
success of activated carbon treatment for both solvent and soluble pro-
pellant ingredients of the solvent wastes, still bottoms would likely
be best handled by blending with the solvent wastes , and treating the
combined flow with activated carbon.
h. The spent caustic cleaning solution, used to clean metal
dies and screens, represents a small but highly concentrated waste
stream. In addition to the caustic itself, the wastewater contains deg-
radation products of propellant contaminants on the metal parts. The
-136-
-------
Table 51. Treatment Results for Dissolved Organic
Propellant Ingredients (6e)
Ingredient
Dinitrotoluene
Diphenylamine
Dibutylphthalate
Initial
Concentration ,
mg/1
316
0.9
0.17
Reverse
Osmosis
125
0.9
0.0
Ozonation'^-'
230
0.0
0.0
Activated
Carbon
2.1
0.0
0.08
(1) Seventy-five minute ozone treatment.
-137-
-------
waste contains essentially only soluble constituents, which are neither
amenable to physical nor biological treatment (3f, 6a). As discussed
in prior sections of this chapter, the propellant constituents are also
not effectively treated by chemical means. In the absence of availa-
bility of treatment methods, the problem of this waste seems best handled
by seeking an alternate metal cleaning technique, which either results in
no waste effluent, or yields one susceptible to treatment. One alternate
technique which appears promising is the substitution of the hot caustic
with an organic solvent. This cleaning system may permit recovery of both
the cleaning media and the propellants. Organic solvents which have
proven successful in cleaning the metal parts within less than one hour
contact and at ambient temperature include methanol, tetrahydrofuran,
acetone, benzene and ethyl acetate (15zp). In acetone, the most effect-
ive solvent, NC concentrations up to 2 percent do not interfere with
cleaning efficiency. Above this concentration, cleaning rate declines
rapidly (31). Substitution of solvent for caustic cleaning will elim-
inate discharge of spent caustic contaminated with propellant, and pos-
sibly allow recovery of propellant plus recycle of solvent.
i. Various physical and chemical techniques have been attempt-
ed to treat the collagen wastes associated with ball powder manufacture.
Coagulation, chlorine oxidation, aeration and steam stripping have all
been attempted with only limited success (16zx). Best and most consis-
tent results have been achieved by biological treatment, including both
activated sludge (4d, 16zx) and trickling filtration (3h, 15y, 16zx).
However, Badger AAP has tested the trickling filtration process for
treatment of collagen waste, and found that the present trickling filter
at the plant lacks adequate treatment capacity for the collagen waste-
water volumes associated with full mobilization. Installation of an
activated sludge treatment system has been recommended (9c).
71. Summary
a. Volumes of wastewater associated with solvent propellant
manufacture are relatively small compared to effluents of most other
explosives and propellants. The wastes, however, are high in organic
solvents and dissolved propellant constituents. Inadequate wastewater
characterization data are available on either waste volumes or waste-
water constituents, and an effort to better characterize the wastes is
warranted.
b. Wastes are typically not treated prior to discharge, with
the limited exception of some partial pH adjustment. There appears to
be a significant potential for application of water management tech-
niques leading to reduced water use, wastewater recycle and product re-
covery, although some residual waste volume would still remain to be
treated.
c. Some effort, largely unsuccessful, has been directed to-
ward treatment for solvents removal from wastewaters such as the "water
-138-
-------
dry" waste. Among the many methods assessed, only activated carbon and
biological treatment appear to be both effective and reliable. Biologi-
cal treatment is currently employed at one commercial explosives plant,
and has been proposed for both Radford and Badger AAP's. While probably
effective for solvent degradation, the possibility of biotransformation
of waste propellant constituents into even more undesirable compounds
cannot be ignored. The propensity of many plants to propose biological
treatment reflects only economic factors and, until the possibility of
biotransformations has been more thoroughly explored the use of biologi-
cal treatment should be viewed with caution.
-139-
-------
SECTION VIII-B - SOLVENTLESS PROPELLANTS
72. Introduction
Solventless propellants are used to make mortar increments and some
extruded rocket motors. Rocket motors are manufactured at NOS Indianhead
and Badger and Radford AAP's. The latter AAP also makes mortar increments.
73. Waste Sources
a. Solventless propellant (rolled powder) is prepared in a
batch process in which chopped NC, NG and various sensitizers and in-
hibitors are slurried in water, centrifuged to a wet cake, and dried to
a paste. After blending, the mixture is rolled into homogeneous plastic
sheets. These sheets are then slit to width, and made into "carpet rolls."
The carpet rolls are extruded in a hot press to form solid rocket motors.
The extrusions are cut to length, dried and cured, and then trimmed to
required dimension on a lathe. For mortar increments, the rolled sheets
are sewed together, die cut and punched.
b. The specific processing steps of Solventless propellant
manufacture are as follows (6e):
(1) NC, NG and other constituents including a lead salt
are first premixed. Water is used only for building
and equipment clean up.
(2) The premix is transferred to another operation where
water is added to form the slurry. This slurry is
centrifuged. The centrate is recycled back to the
slurry mix, for reuse. The water is discharged only
when the propellant formulation is changed.
(3) The centrifuged propellant is then tumbled to blend
and partially dried to a paste, which is fed to
rollers. Water use results from clean up, dust con-
trol processes and automatic sprinkler fire systems
upon accidental firing of propellant on the rollers.
(4) The rolled propellant is then extruded into solid
rocket motors or sewed and cut into mortar incre-
ments. Water use results from clean up, cooling,
and chip removal from machining and sawing operations.
Table 52 presents wastewater characterization data for the various pro-
cessing steps at Radford AAP. All wastewaters are notably high in NG,
with the slurry mix wastewater containing 1500 mg/1 NG plus 800 mg/1 DNG.
Lead is also high in this waste, up to 300 mg/1 as is organic material
(BOD, COD, TOC), nitrate and dissolved solids. The Premix Blender and
Machining clean-up waters contain more moderate levels of pollutants,
-140-
-------
Table 52. Wastewater Characteristics from Solventless Propellant
Manufacture at Radford AAP (6e)
Constituent
PH
BOD
COD
TOG
Nitrate
Suspended Solids
Dissolved Solids
Nitroglycerin
Dinitroglycerin
Lead
Flow, gpd
P remix
8.3
22
91
27
3.2
45
393
20
-
0.7
200
Slurry
Mix
6.4
680
3526
1081
177
13
1117
1500
800
0-300
1500
Blender
7.0
-
31.7
10
6.8
2.4
74
21
-
0.5
45 ,000
Rolling (2)
8.9
-
31
13
6.9
6.9
462
10
-
-
85 ,000
Rolling <2>
Cleanup
10.4
-
197
137
-
179
1582
7.3
-
0.3
7200
Machining
7.2
-
59
-
14.8
1.9
96
15
-
-
30,000
(1) All values except pH and flow in mg/1. Values reported are averages.
(2) Values reported for "Rolling" represent normal total flow. Values reported
for "Rolling Clean up" represent major clean up water only.
-------
while the Rolling clean up water is high in organics, nitrate, and
solids. The high organic and solids content of the Rolling clean-up
water is reported due to the use during clean up of asbestos plus clean-
ing compounds (ln(3)). Solventless propfcllant manufacture at Radford
AAP is reported to result in a total wastewater discharge of 3 MGD (3f).
Of this, a significant fraction is cooling water. Waste characteriza-
tion data are not available for Badger AAP or NOS Indianhead. At Badger
AAP, as a result of separation of process from cooling water, and imple-
mentation of various water management techniques described below, the
process flow is of small enough quantity to dispose to a evaporation/
percolation pond, from which there is no overflow (la(3)). Wastewater
volumes are presented in Table 53. Above 88% of the total water is re-
lated to cooling use.
74. Effects of Water Management
At Radford AAP, centrate slurry water is recycled back to the slurry
mix tank, significantly reducing the discharge of NG and other constitu-
ents shown in Table 52, The water is discharged to waste only upon a
change in propellant formulation (6e).
An extensive water management program has been implemented at Badger
AAP (3a). Procedural changes have resulted in a 50% reduction in water
consumption in the Paste Area and further reductions are possible, pri-
marily in cooling water reuse by recycle (3a). At Badger AAP, almost
all process water is recycled, with the exception of the dowel and
spiral wrap flush and clean up water (Table 53). This process flow
represents 988,520 gallons/month. Cooling water from the Roll and Press
Area totals 3.9 million gallons per month and from the Paste Area about
7.6 million gallons per month. This latter flow has been reduced
approximately 50% by elimination of cooling water use in winter to pre-
vent drain line freeze up, and by reducing flow rate during other peri-
ods of operation (3a).
75. Current Treatment Practice
a. At present, solventless propellant wastewater treatment
at Radford AAP is restricted to catch tanks, to remove readily settle-
able suspended solids. Catch tank effluents are discharged without fur-
ther treatment. From the data of Table 52 most pollutants of the waste-
water are in soluble form. Hence, the existing catch basins appear to
be of limited utility (ln(3)).
b. At Radford AAP, a new wastewater collection and treatment
system has been programmed (3f). This will include combined collection
of process wastewater with cooling and storm water, primary settleable
solids removal, and secondary treatment by processes which have not yet
been specified (3f). Pilot studies on biological treatment are currently
underway (15zp).
-142-
-------
Table 53. Badger AAP Rocket Area Water Usage (3a)
Area
Use of Water
Gallons/Month
Paste
Total
Roll and Press
Total
Finishing
Total
TOTAL USAGE
Tank Wash
Wet Floors
Cooling (Hydraulic System)
Wash Down
Cooling (Vacuum System)
Cooling (Hydraulic System)
Dowel and Spiral Wrap Flush
25,700
216,720
3,696,700
3,939,120
167,000
604,800
3,316,320
4,088,120
604,800
604,800
8,632,080
-143-
-------
c. At Badger AAP, process wastewaters are screened to re-
move propellant chips and flakes. The screened effluent, which contains
smaller particles of solid propellant, discharges through baffled ditches
to a closed pond (la(3)>. Baffling enhances the sedimentation of sus-
pended solids in the ditches, which are periodically cleaned out. The
wastewater in the pond is lost through evaporation and leaching. The
pond is reported to overflow only during periods of intensive rainfall
(la(3)). Except during these periods, there is no discharge of process
wastewater for Badger AAP solventless propellant manufacture.
76. Alternatives to Current Practice
a. The primary constituents of concern in solventless propel-
lant process wastewater are organics including nitroglycerin and dinitro-
glycerin, nitrate, suspended solids including solid propellant, and lead.
With the exception of lead, treatment alternatives for the other constit-
uents have been assessed in preceding sections of this chapter.
b. Lead may be removed by a variety of processes, including
hydroxide or sulfide precipitation and ion exchange, the two most ef-
fective and widely used techniques (16zy). Lead levels in Radford AAP
wastewaters of up to 300 mg/1 have been reported (6e), although most
process effluents contain sub-mg/1 levels (Table 52). At lead concen-
trations below 5 mg/1, and particularly for large volume flows, hydrox-
ide precipitation treatment is relatively ineffective and uneconomical
(16zy). Therefore, ion exchange treatment appears to be most directly
applicable to lead bearing wastewaters from solventless propellant man-
ufacture. Ion exchange or sulfide precipitation has the further advan-
tage of possible recovery of the lead upon regeneration of spent resin.
c. Preliminary studies on ion exchange treatment of solvent-
less propellant slurry mix water have indicated that while several res-
ins have limited efficiency, Duolite ES 63 resin is capable of reducing
lead to about 0.01 mg/1, at flow rates up to 2.7 gal/min/cubic feet of
resin. At higher flow rates, treatment efficiency drops off rapidly.
Although saturation capacity of this resin for lead exceeds 3 Ib./cu.
ft., significant lead leakage occurs above a loading of 2 Ib. lead/cu.
ft. of resin (6a).
77. Impact of Air Pollution Control
Currently, a portion of the Radford AAP wastewater from the sol-
ventless propellant blending area is air pollution dust scrubber water
from rotoclones (ln(3)). It is proposed to partially recycle this
scrubber water (3u), with overflow discharges to the proposed main
plant treatment system (6a).
78. Summary
a. Wastewaters from solventless propellant production result
primarily from building clean-up and equipment washdown. According to
-144-
-------
characterization data for Radford AAP, the wastewaters are high in BOD,
nitrate, suspended solids, NG and DNG, and lead. With the exception of
Badger AAP, these wastes are discharged without treatment. Except for
lead all constituents are, however, treatable as described in previous
sections of this chapter. Through implementation of several water man-
agement techniques, Badger AAP has sufficiently reduced process waste
volume to allow utilization of percolation/evaporation ponds. Although
there are possible questions concerning the possible contamination of
soil and ground water from leaching of propellant wastewater, the water
management program implemented by Badger AAP has been effective in it-
self, and appropriate elements of this program should also be implemen-
ted at Radford AAP.
b. Although pilot biological treatability studies are to be
undertaken at Radford AAP, caution is indicated in the face of current
lack of knowledge of possible biotransformation of propellant wastewater
constituents into environmentally unacceptable compounds. In any event,
biological treatment is ineffective for lead control, although other
treatment technologies are available. Among these processes, ion ex-
change appears to offer most promise.
-145-
-------
SECTION IX - MISCELLANEOUS LAP ACTIVITIES
79. Introduction
The three major load and pack operations at military munition
plants are: the casting of melt-pour explosives; the casting of plastic-
bonded explosives; and the casting of composite propellents into solid
rocket motors. All operations involve processing steps for intermedi-
ates.
a. Major wastewater constituents associated with LAP casting
of melt-pour explosives are discussed in Section II (Trinitrotoluene)
and Section V (RDX and HMX) of this chapter.
b. This section is devoted primarily to: wastewaters from
casting of composite propellants; wastewaters associated with process-
ing of secondary ingredients of pressed explosives ; and casting of
plastic bonded explosives.
SECTION IX-A - CAST PROPELLANTS
80. Introduction
a. Cast propellants are so identified because slurry-
propellant formulation is either poured directly in the rocket case
(composite propellant) or formulation blsnded "in-place," (composite
double-base propellant) followed by curing to form a solid rocket motor.
Most large rocket motors are cast. Smaller motors are typically extrud-
ed, as described in Section VIII. There are exceptions, such as the
small TOW motor, which is cast in plastic forms at Radford AAP (6a).
b. Rocket motors may be cast from either fine NC-base cast-
ing powder with NG added to form a double base, or from aluminum powder
oxidizer, and other ingredients stabilized in a polymer binder such as
polybutadiene or polybutene. Both types of casting formulations include
trace amounts of catalysts and stabilizers in addition to typical oxi-
dizers such as ammonium perchlorate or HMX oxidizer.
c. For the polymer based motor, ingredients including ammoni-
um perchlorate are pre-mixed and blended in large kettles to form a
slurry. A polymer cross-linking agent is added just prior to completion
of mixing, and propellant slurry is poured into the rocket case. Upon
curing, the slurry solidifies into a rocket motor. In casting the double
base propellant, casting powder is loaded into the rocket case, and
acetone plus NG added to dissolve the powder. Nitrogen or dry air is
bubbled through the mixture to remove the acetone, and yield a dense gel.
The loaded rocket is then cured in an oven, during which the gel
solidifies.
-146-
-------
81. Waste Sources
a. Wastewaters are associated with dust abatement procedures,
and equipment and building clean-up. Typically, water use is minimized
in the actual vicinity of propellant casting operations, as moisture is
detrimental to propellant quality (15zm). Wastewaters from dust abate-
ment result from grinding of ammonium perchlorate and HMX. No character-
ization data are available on either dust abatement or clean-up waste-
waters. Double-base propellants are cast in closed systems and are thus
not generally subject to polluting wastewaters. Because of the small
volumes of wastewater involved, and the widespread use of evaporation
ponds for final disposal, there is little activity or need in water
management and recycle.
b. Size reduction is required for both ammonium perchlorate
and HMX where used in composite and double-base propellants. Prior to
blending these oxidizers into the propellant, size reduction is accom-
plished in hammer mills, with size control by screening. Sized oxidizer
is then sent to the blending operation by either forced air conveyers,
or preweighed and transported in containers. At Radford AAP, dust from
the grinding and weighing operations is removed by a duct system to a
wet scrubber. The flow from the scrubber is discharged without treat-
ment. A dry-type dust collection system has been proposed, with the
collected dust incinerated (6a).
c. Both Air Force Plant #78 and NAVPRO Magna grind HMX. At
Magna, dust is vented to a wet scrubber system (15zj). Overflow, plus
HMX equipment and process building washwater is discharged through
catch tanks to earthen sumps. The sumps are allowed to dry up by evap-
oration (annual evaporative rate of 56 inches/yr.), and the dried resi-
due burned. At some process points, catch basins are preceded by nylon
filter bags to enhance solid HMX removal (15zj). AF #78 uses air pre-
fliters plus a hospital-type "Absolute Filter" (American Air Filter Co.)
to control HMX grinding dust (15zk). This filter system removes par-
ticles down to 0.3 microns in size. Building and equipment washwater
is discharged into a collection pit, and transferred to an evaporation
pond. The pond residue is burned (15zk).
d. Many LAP plants grind ammonium perchlorate. Both AF #78
and NAVPRO Magna have dry dust control systems for the grinding oper-
ations (15zj , 15zk). The Magna grinding equipment is totally enclosed,
and there is no escape of dust. AF #78 uses a baghouse control system.
Both plants, however, use water clean-up of the grinding equipment and
buildings. At AF #78, washwater is collected in concrete lined sumps.
When the sumps are filled, the contents are transferred to evaporation
pits, where the residue is burned (15zk). Magna uses a similar system
of catch tanks and evaporative ponds.
e. Other LAP plants grinding ammonium perchlorate typically
use wet scrubber systems for dust control. At Longhorn AAP, vacuum
-147-
-------
ducts exhaust dust laden air to a wet scrubber. The scrubber bath is
stagnant, with bath contents dumped once per shift to a collection sump.
Sump contents are transported by truck to a large evaporative pond (15zm).
Clean-up water is discharged by surface ditches to a large lake located
on the plant property. Ammonium perchlorate transport containers are
washed, with washwater stored in a 35,000 gallon tank and bled to surface
drainage when surface flow is sufficient to provide dilution (3f).
f. Redstone Arsenal has a wet scrubber system for ammonium
perchlorate grinding dust control (15zo). Scrubber water is shipped in
drums to the scrap propellant burning ground for disposal. Building and
equipment washdown water discharges through sumps to creeks. Due to the
high water solubility of ammonium perchlorate, sumps would be ineffective
in controlling discharge of the oxidizer. At the Aerojet Sacramento
plant, a. similar system is used for dust control. Ammonium perchlorate
grinding dust is removed by wet scrubbing. Scrubber water is retained
in a holding pond, and recirculated. The ammonium perchlorate grinding
and conveyance systems are designed to prevent dust escape and thereby
avoid building and equipment washdown, and as a result there is no clean-
up water associated with this process (15zm). Both Redstone and Aerojet
transport ground ammonium perchlorate by forced air conveyance systems
(15zm, 15zo). Conveyer air is recirculated at Aerojet (15zm), while
Redstone exhausts the air through baghouse collectors (15zo).
g. Clean-up of propellant handling equipment such as slurry
kettles, and propellant loading buildings, is normally by a routine of
dry hand cleaning, followed by solvent cleaning and final wash with
soapy water. At NAVPRO Magna, any floor spills are scraped and hand-
wiped by freon-damp rags. This is followed by an acetone wipe, and a
soap and water wash (15zj). Floors are also cleaned in this manner at
the end of each shift. Soapy water discharges through catch tanks into
evaporation pits. The presence of any solid propellant in the catch
tank results in a special investigation by plant management. Typically
the wastewater is soapy wash water, with traces of acetone (15zj).
Equipment is cleaned by scraping, freon wiping and acetone wash. Water
is normally not used.
h. AF #78 first washes buildings and equipment with a high
pressure water jet, followed by hand wiping with methyl chloroform.
Water is collected in concrete lined sumps, and transported by tank
trucks to evaporative ponds (15zk). The water is evaporated, and resi-
due burned. Aerojet (Sacramento) uses no water at all in building
clean-up. Cleaning is by dry and solvent processes only (15zm). Cast-
ing bowls, mixers and other equipment are cleaned by wiping with tri-
ch lore thane followed by a water wash. Solvent is collected in concrete
simps and disposed by burning. Water is discharged to evaporation ponds
(15zm), Redstone Arsenal uses a combination of solvents plus detergents
and water for equipment and building clean-up. Wash waters discharge
through catch sumps to surface drainage (15zo). Equipment clean-up
water is partially recirculated. All of the above approaches appear to
be effective.
-148-
-------
SECTION IX-B - PRESSED EXPLOSIVES
82. Introduction
Secondary ingredients of pressed explosives which impact on process
water quality at LAP facilities are primarily powdered aluminum, and
ammonium picrate (Explosive "D"). Powdered aluminum is used in a variety
of pressed explosives, including HBX-1 (NAD Hawthorne), H-6 (NAD Crane,
NWS Yorktown), Tritonal (NAD Crane, NWS Yorktown, NAD McAlester, NAD
Hawthorne), Minol (NAD McAlester), and other composition explosives.
Ammonium picrate is used in Picratol formulation, a blend of TNT and
Explosive "D". Other significant explosives, used in detonators, are
lead azide and lead styphnate. Lead azide is no longer manufactured.
The Army is investigating methods to destroy lead azide (3h).
83. Waste Sources
a. Aluminum is received at LAP plants in powdered form, and
processing is limited to screening, weighing, transporting and blending
with the composite explosive mixture. Wastewaters associated with pow-
dered aluminum result from dust abatement and building and equipment
clean-up. No characterization data are available on these wastewaters.
Many plants exhaust aluminum dust directly to the atmosphere, without
dust control. Others use wet scrubbers. Plants without dust control
plan to incorporate scrubber systems, and handle the scrubber water in
a variety of ways. NAD McAlester will use an overflow lagoon (15za).
Since aluminum is relatively insoluble, most scrubbed aluminum would
be expected to remain in particulate form, and unless solids removal
occurs in the lagoon, aluminum will be present in the lagoon discharge.
b. NWS Yorktown currently uses a wet scrubber system for
powdered aluminum dust control (15zh) . Wet scrubbing was selected in
preference to dry collection and recovery, due to the possibility of
explosive hazard (14r). Scrubber overflow discharges to baffled sumps,
and thence to surface drainage (15zh). Sumps are mucked out weekly,
with sludge burned. In contrast to the Yorktown system, NAD Hawthorne
successfully uses a dry cyclone type dust separator for their aluminum
sieving operation (15zc).
c. Ammonium picrate is used at NWS Yorktown, and NAD Crane
and Hawthorne, to formulate Picratol, which contains approximately half
TNT and half ammonium picrate (14zb). Process wastewaters from Picratol
manufacture are currently discharged through catch basins to surface
drainage (15ze, 15zh, 14zb). At NWS Yorktown, however, Picratol dust is
vented to a dry collection system (14r).
84. Treatment Practice
a. Various treatment methods have been assessed for ammonium
picrate wastewater. Biological degradation is not effective (15ze, 16o).
-149-
-------
A comprehensive pilot study of physical-chemical treatment processes
for Picratol wastewaters, carried out at NAD Hawthorne, indicated that
activated carbon adsorption is most consistently effective (14zb).
Treatment results are presented in Table 54.
b. There is effectively no information on the wastewater
characteristics or treatability of wastewaters associated with lead
styphnate use. Wastewaters originate from building and equipment wash-
down, and dust scrubber systems. At Kansas AAP, the wastewater is re-
tained in holding pits, where it is treated with caustic to "kill" the
lead styphnate (lh(3)). The wastewater is then discharged to holding
ponds. The pH of the caustic-treated wastewater is unknown. Although
some lead hydroxide precipitate may be removed in the treatment process,
most effective treatment by lead precipitation occurs near pH 10, with
lead being significantly more soluble at higher or lower pH (16zz).
-150-
-------
Table 54. Pilot Plant Treatment of Ammonium Picrate (14zb)
Treatment
Process
Air Flotation
Sand Filter
Activated Carbon
Polymeric Resin
Concentration, mg/1
Influent
86
419
86
419
86
419
86
419
Effluent
52
413
57
414
<1
<1
< 1-17
27-270
Percent
Removal
39.5
1.5
33.7
1.1
98.8+
99.8+
80.2-99.8+
35.6-93.6
-151-
-------
SECTION IX-C - PLASTIC BONDED EXPLOSIVE
85. Introduction
The production of plastic bonded explosives (PBX) is a relatively
small experimental operation, compared to production of other military
explosives. Currently PBX is loaded at NWS Yorktown (14z). As with
cast propellants, solvents are used in the formulation and loading of
PBX. At Yorktown, these solvents are collected separately and disposed
of by incineration. Solvent- and explosive-contaminated wastewaters
are discharged to surface drainage. No characterization data are avail-
able on these wastewaters. No treatment system has been proposed for
these wastes at present (14z), although the U. S. Navy is conducting
research programs to develop water-soluble and other binders more
amenable to pollution-free disposal (15zr).
86. Summary
There is sparse information available on wastewater volumes or
characteristics of the minor LAP activities discussed in this section.
In general, such wastewaters originate from air pollution abatement,
and building and equipment clean-up procedures, and presumably reflect
the constituents of the loaded explosives and propellants. Despite
widespread LAP activities, little effort seems to be currently directed
toward either waste characterization or treatment of the wastes de-
scribed in this section. A waste characterization program is indicated,
in order that treatment needs for these minor LAP facilities can be de-
fined, and effluent criteria established.
-152-
-------
APPENDIX 1 - REFERENCES
1. U. S. ARMY ENVIRONMENTAL HYGIENE AGENCY
a. Badger Army Ammunition Plant
(1) WQESS1 No. 24-004-72, 29 Apr-30 Oct. 1971
(2) WQMC No. 24-038-73/74, 23-27 Apr. 1973
(3) SESIWSS3 No. 24-038-70/71, 13-20 May 1970
b. Burlington Army Ammunition Plant
(1) SSIWS4 No. 24-002-70, 18-29 Aug. 1969
c. Cornhusker Army Ammunition Plant
(1) SES5 No. 24-010-71, 7-11 Dec. 1970
d. Holston Army Ammunition Plant
(1) SES No. 24-026-70, 12-16 Jan. 1970
(2) WQESS No. 24-021-71/72, 19 Mar-28 Jun. 1971
(3) WQBS3 No. 24-010-73, 31 July-10 Aug. 1972
(4) WQMC No. 24-005-73, 7-11 Aug. 1972
e. Iowa Army Ammunition Plant
(1) WQES No. 24-003-72, 13-17 Sept. 1971
(2) WQBS No. 24-009-73, 10-19 July 1972
(3) Carbon Column System Removal Efficiency Study
No. 24-033-73/74, 2 May-22 Jun. 73
f. Indiana Army Ammunition Plant
(1) SES No. 24-002-71, 13-17 July 1970
g. Joliet Army Ammunition Plant
(1) SES No. 24-001-71, 5-10 July 1970
(2) WQMC No. 24-030-72/73, 5-16 Jun. 1972
(3) WQESS No. 24-024-72/73, 5-19 Jun. 1972
(4) WQESS No. 24-032-72/73, 5-20 Jun. 1972
T,Water Quality Engineering Special Study
2. Water Quality Monitoring Consultation
3. Sanitary Engineering Survey & Industrial Waste Special Study
4. Special Study of Industrial Waste Survey
5. Sanitary Engineering Survey
6. Water Quality Biological Study
-153-
-------
h. Kansas Army Ammunition Plant
(1) SES No. 24-009-70, 22-25 Sept. 1969
(2) WQES No. 24-004-73, 28 Aug.-l Sept. 1972
(3) WQMC No. 24-010-74, 27-31 Aug. 1973
i. Lake City Army Ammunition Plant
(1) SES No. 24-029-70/71, 12-15 Oct. 1970
(2) WQESS No. 24-030-71/72, 16-31 Aug. 1971
j. Longhorn Army Ammunition Plant
(1) SES No. 24-023-70, 5-8 Jan. 1970
(2) WQESS No. 24-014-72/73, 3-19 Apr. 1972
(3) WQMC No. 24-041-73/74, 4-8 Jun. 1973
k. Lone Star Army Ammunition Plant
(1) WQES No. 24-029-72/73, 22-26 May 1972
(2) WQMC No. 24-015-74, 24-28 Sept. 1973
1. Louisiana Army Ammunition Plant
(1) SES No. 24-024-70, 8-14 Jan. 1970
(2) SESS No. 24-005-71, 1 May-15 Aug. 1970
(3) WQESS No. 24-006-72, 2-14 Dec. 1971
(4) WQMC No. 24-025-73, 8-12 Jan. 1973
m. Milan Army Ammunition Plant
(1) WQES No. 24-018-73/74, 5-9 Mar. 1973
n. Radford Army Ammunition Plant
(1) WQMC No. 24-007-73, 11-22 Sept. 1972
(2) SSIWS No. 24-011-69/70, 13-21 Jun. 1969
(3) WQESS No. 24-001-72, 4-18 Oct. 1971
o. Redstone Arsenal
(1) Water Pollution Evaluation Visit No. 24-009-69 24-28
Feb. 1969
(2) SESIWSS No. 24-030-70/71, 20 Apr.-l May 1970
(3) WQES No. 24-002-72, 13-17 Sept. 1971
p. SESS No. 24-007-70/71, 1 Jan 1970-31 Oct. 1970
"Evaluation of Toxicity of Selected TNT Wastes on Fish"
q. Twin Cities Army Ammunition Plant
(1) WQESS No. 24-034-72/74, 2-19 Oct. 1972
-154-
-------
r. Volunteer Army Ammunition Plant
(1) SES No. 24-003-71, 10-14 Aug. 1970
s. SESS No. 24-017-70/71, "Special Study of the Effect of Alpha TNT
on Microbiological Systems and the Determination of the Bio-
degradability of Alpha TNT." Jan.-Aug. 1970.
U. S. ARMY MEDICAL BIOENGINEERING R&D LABORATORY
a. Technical Report 7403, "Mammalian Toxicity and Toxicity to
Aquatic Organisms of Four Important Types of Waterborne
Munitions Pollutants," March 1974.
b. U. S. AMEERU Report No. 73-07, on "Munitions Production Products
of Potential Concern as Waterborne Pollutants," June 1973.
c. Technical Report 7404, "Munitions Production Products of
Potential Concern as Waterborne Pollutants," April 1974.
d. Burrows, D. 1973. Literature Review on the Toxicity of RDX
and HMX.
U. S. ARMY PICATINNY ARSENAL
a. Technical Report 4533, "Preliminary Water Management Study for
Ball Powder & Rocket Manufacturing Area Badger AAP," July 1973.
b. Technical Report 4552, "Sellite Effluent Pilot Aeration and
Neutralization, Joliet AAP," April 1973.
c. Technical Report 4554, "A Laboratory Study of Carbon Adsorption
for Elimination of Nitrobody Waste from AAP's," Jun. 1973.
d. Technical Report 4568, "Abatement of High Nitrate Concentrations
at Munitions Plants," Aug. 1973.
e. "Initial Water Utilization Survey for Processes Employed by
GOCO AAP's," July 1972.
f. Report No. 96020.007, "Army Munitions Plants Modernization
Program-Pollution Abatement Review - Final Report. TRW.
August 1973.
g. Report No. 96020.009, "Water Quality Standards and Regulations
Applicable to Army Ammunition Plants - Volume II," Jan. 1973.
TRW.
h. "Pollution Abatement Engineering Program for Munition Plant
Modernization - 3rd Briefing for Senior Scientist Steering
Group," Feb. 1974.
-155-
-------
i. "Pollution Abatement Engineering Program for Munition Plant
Modernization - 4th Briefing for Senior Scientist Steering
Group," Feb. 1974.
j. "Pollution Status Memorandum, Newport AAP," Apr. 1974.
k. "Pollution Status Memorandum, Longhorn AAP," Jun. 1974.
1. "Pollution Status Memorandum, Milan AAP," May 1974.
m. "Report on an in-depth Pollution Survey at Joliet AAP," TRW
Aug. 1974.
n. "Pollution Status Study, Louisiana AAP," Jun. 1974.
o. Letter, SARIO-CE, Iowa Army Ammunition Plant, IA, 16 Oct. 1973,
subject: "Request for Technical Information."
p. Griffin, D., "Joliet Army Ammunition Plant Pollution Discussion
and Abatement Plans," Technical Report 4368, Manufacturing
Technology Directorate, Picatinny Arsenal, NJ, June 1972.
q. "Process Design Criteria Memorandum - Nitroguanidine Facilities,"
AMC Project No. 5742632, Hercules Incorporated, Wilmington, DE,
February 1972.
r. Letter, AMSMU-PP-PGB, HQ, MUCOM, Joliet, IL, (undated) May 1973,
subject: "Request for Information on TNT Waste Flows from
Cornhusker AAP."
s. Letter, SARIO-CE, Iowa Army Ammunition Plant, IA, 16 October 1973,
subject: "Request for Technical Information."
t. Letter, SMUMO-S, Milan Army Ammunition Plant, TN, 10 May 1973,
subject: "Request for Information on TNT Waste Flows from
Milan Army Ammunition Plant (MAAP)."
u. Evans, J. L., "Project Status Report, Project No. BP 4932 5724114,"
USAMUCOM, Dover, N.J., 1 January 1973.
v. "Technical Evaluation Study FY 73 MCS-AMC TNT Industrial Waste
Treatment," A.M. Kinney, Inc., Consulting Engineers, June 1972.
w. Eskelund, G. R., "TNT and Related Facility Modernization,"
Picatinny Arsenal, May 3, 1973.
x. "Water Management" A Tool for Water Pollution Abatement at
Army Munitions Plants," internal document, Picatinny Arsenal,
undated.
y. "Project Status Report," RCS-AMCRD-127, USAMUCOM, Dover, N.J. ,
1 June 1973.
-156-
-------
z. "Pollution Abatement Engineering Program for Munitions Plant
Modernization," Technical Memorandum 2170, Picatinny Arsenal,
February 1975.
za. Maybury, D. H. and J. L. Evans, "Propellant Plant Pollution
Abatement Improvement of Water Utilization at Radford Army
Ammunition Plant," Technical Report 4562, Picatinny Arsenal,
December 1973.
U. S. ENVIRONMENTAL PROTECTION AGENCY
a. "Report on Holston Army Ammunition Plant," March 1973, NFIC.
b. "Results of Organic Analyses of Samples from Kingsport, Tenn.,
Apr. 1973," 20 Jun. 1973, NFIC.
c. "Report on Waste Disposal Practices, Radford Army Ammunition
Plant, Radford, Virginia," U. S. Environmental Protection
Agency, Region III, May 16, 1973.
d. Patterson, J. W. and Minear, R. A. , "State-of-the-Art for The
Inorganic Chemicals Industry: Commercial Explosives," U. S.
EPA Report EPA-600/2-74-009b, March 1975-
e. "Nitrification and Denitrification Facilities Wastewater
Treatment, U. S. Environmental Protection Agency, Technology
Transfer, August 1973.
f. Grinstead, R. and K. C. Jonos, "Nitrate Removal from Waste-
water by Ion Exchange," U. S. Environmental Protection Agency
Report No. 17010 FSJ, 1971.
g. Ottinger, R. S., J. L. Blumenthal, D. F. Dal Porto, G. I. Gruber,
M. J. Santy and C. C. Shih, "Recommended Methods of Reduction,
Neutralization, Recovery or Disposal of Hazardous Waste, Volume
VII," U. S. Environmental Protection Agency Report EPA-670/2-73-
053-g, August 1973.
HOLSTON DEFENSE CORPORATION, KINGSPORT, TENN. (HOLSTON AAP)
a. "Development Report No. ASQ-8A, Stream Monitor at HOC," Oct. 1973.
b. "Transmittal of Findings on Nitrogen Removal for Wastewaters
Emanating from Area "A", Holston AAP," DA CERL, 6 Dec. 1973.
c. "HAAP Industrial Waste Treatment Facilities Meeting - Summary
of Treatment Concept," 1 November 1973.
d. "Material Balance and Waste Characterization of Explosives
Finishing and Analytical Laboratories," Development Report No.
ASQ-6, Holston Defense Corp., Kingsport, Tenn., undated.
-157-
-------
e. "Material Balance and Waste Characterization - Explosives
Manufacturing Development Report No. ASQ-4," HoIston Defense
Corporation, September 1972.
6. RADFORD ASMY AMMUNITION PLANT - THIOKOL CORP.
a. "ADPA Meeting at Radford AAP - Water Pollution Abatement and
Control," 29 Jan. 1974.
b. Small, M., "Report of Visit to Radford Army Ammunition Plant,"
Radford, VA, March 1973.
c. Smith, L., Internal Engineering Report - Production Engineering
Project PE-290, Hercules Incorporated, Radford Army Ammunition
Plant, Radford, VA, Oct. 1971.
d. "Clean Stream Program Monitoring Analyses," Radford AAP,
Sept. 6, 1973 - Feb. 28, 1974.
e. Smith, L. L. and R. L. Dickenson, "Final Engineering Report on
Production Engineering Project PE-249 (Phase II) Propellant
Plant Pollution Abatement," Radford Army Ammunition Plant,
Radford, Virginia, May 1974.
7. LONGHORN ARMY AMMUNITION PLANT - THIOKOL CORP.
a. Brochure, July 1973.
b. Water Sample Logs., Dec. 1973 - May 1974.
8. JOLIET ARMY AMMUNITION PLANT
a. Consoer, Townsend and Associates, 1972, Tetryl Industrial Waste
Treatment - WPC: Concept Design, JAAP, Joliet, Illinois.
b. Memorandum for Record, SGRD-UBG, Environmental Quality Division,
USAMBRDL, 10 September 1973, subject: "Studies of Red Water
Condensate and Tetryl Refiner Ditch Soil and Water Samples" -
Joliet Army Ammunition Plant (JAAP), Joliet, IL.
c. "Concept Design Waste Treatment Facilities for FY-70 Acid
Plants, Joliet Army Ammunition Plant," A. M. Kinney, Inc.,
Consulting Engineering, Nov. 1971.
d. "Phase I'Report Process Water Management Study," Project AMC
5724114-00, Joliet AAP, 1973.
e. "Final Design Analysis Waste Treatment Facilities for FY-70
Acid Plants," Joliet AAP, A. M. Kinney, Inc., 5 May 1971.
-158-
-------
f. Ghassemi, M. , "Report on an In-depth Pollution Survey at Joliet
Army Ammunition Plant," TRW, Inc., August 1974.
9. BADGER ARMY AMMUNITION PLANT
a. Rosenblatt, D., "Report of Visit to Badger Army Ammunition
Plant," Baraboo, WI, March 1973.
b. Roth, E. S. , "Cleaning of Screens and Dies Used in the Manu-
facture of Smokeless Propellant," Report TD-123, Olin Corp.,
Badger Army Ammunition Plant, Baraboo, Wise., 28 Nov. 1973.
c. Augsburger, R. L., "Disposal of Collagen Waste from the
Manufacture of Ball Powder Propellant," Technical Report PI-05,
Olin Corp., Badger Army Ammunition Plant, Baraboo, Wise.,
6 June 1975.
10. EDGEWOOD ARSENAL
a. Rosenblatt, D. H. , G. E. Lauterbach and G. T. Davis, "Water
Pollution Problems Arising from TNT Manufacture," EASP 100-94,
Edgewood Arsenal, MD, May 1971.
11. FRANKFORD ARSENAL
a. Letter, Schwarzkopf Microanalytieal Laboratory, Woodside, NY,
to Frankford Arsenal, PA, 6 January 1972, subject: "Separation
and Identification of Impurities Present in a Sample of
Trinitroresorcinol (our B 13650)."
12. U. S. ARMY
a. Department of the Army, "Military Explosives," Department of
the Army Technical Manual 9-1300-214, Washington, DC, 1967.
13. U. S. ATOMIC ENERGY COMMISSION
a. Bullerdiek, W. A., "Removal of TNT from Waste Water Using a
Solvent Extraction Process," Technical Report No. 89, Burlington
AEC Plant, July 6, 1964.
b. "Comparative Efficiency of Four Activated Carbons for TNT
Removal from Water," Technical Report 101, Burlington AEC Plant,
12 Jan. 1967.
c. "Air and Water Pollution Control," Technical Report No. 136,
Burlington AEC Plant, 6 Feb. 1968.
d Meek L "Removal of TNT from TNT Contaminated Water Using Fly
Ash," Burlington AEC Plant Tech. Report No. 212, 10 February 1972.
-159-
-------
e. "Internal Development Progress Report, 2nd Quarter FY 1974,"
Burlington AEC Plant, Burlington, Iowa, March 8, 1974.
f. Brinck, C. R., "Solvent Regeneration of TNT Laden Carbon,"
Iowa Army Ammunition Plant, Technical Report No. 228, 22 Nov.,
1974.
14. U. S. NAVY
a. "Modernization Study of Naval Ordnance Field Activities,"
A. T. Kearney Inc., March 1974:
(1) "Executive Summary"
(2) Volume I "Study Overview," Ch. I, II, V & VI
(3) Volume II "Appendix - Task III - provision of Data &
Information"
b. "Environmental Impact Assessment Covering Fuels for the
Torpedo Mk 46 and Mk 48 Program," PM 0402, 4 October 1972.
c. "Construction and Evaluation of the Pilot Oxidation Ditch
for the Biodegradation of TNT," Ord 0332.
d. "Waste Treatment Oxidation Ditch of NAD McAlester," P-046.
Extract.
e. "Bench Scale Disposal Studies to TNT in the Soil," Progress
Report in Task 27, NAD Crane, 18 July 1974.
f. "Concept Engineering Report - TNT Waste Control Program for
US NAD Crane," Catalytic, Inc., Oct. 1972.
g. Brochure, "NAD Crane Laboratories & Facilities," 1972.
h. "The Environmental Protection Data Base Program at NAD Crane."
i. "The Microbial Degradation of Explosives," J. L. Osmon & R. E.
Klausmeier, NAD Crane, 1973.
j. "The Effect of TNT on Microorganisms," R. E. Klausmeier, J. L.
Osmon and D. R. Walls, NAD Crane.
k. "Explosives and the Environment," NAD Crane, Report No. OEEL/C
73-217, 6 Jun. 1973.
1. Brochures, NOS Indianhead.
m. "Environmental Impact Statement Concerning Modernization of
Plant "A"," MCON Project P-006 NAD McAlester, Feb. 1974.
-160-
-------
n. "Biological Degradation of TNT," NOL White Oak, 31 Mar. 1973.
o. "Ground Water Sampling Program for Possible Explosive Contamina-
tion," NAD McAlester, 17 Apr. 1974.
p. "Biological Degradation of TNT," E. P. Pardee and J. D. Hodge,
NAD McAlester, Nov. 1967.
q. "Further Studies of TNT Pollution," NAD McAlester, May 1970.
r. "Air and Water Pollution Control Study - NWS Yorktown," Pope,
Evans and Robbins, 30 Dec. 1970.
s. "Warhead Melt Cast Loading Plant Modernization," NWS Yorktown
P-347 and P-346, Mar. 1974.
t. Klausmeier, R. E. , "Progress Report on Task 27 to Picatinny
Arsenal," Naval Ammunition Depot, Crane, Indiana, undated.
u. Hoffsommer, J. C. , "Report on TNT Analysis of Water Samples,"
USNOL, White Oak, April 26, 1971.
v. "The Environmental Protection Data Base Program at NAD Crane,"
Public Works Department, NAD Crane, Indiana.
w. Memorandum, "Semi-Annual Analysis of Waste Water Streams,"
10 July 1973, NAD McAlester, Okla.
x. Pollution Chart, Drawing F-1671, 22 Jun. 1974, NAD McAlester,
Okla.
y. Kent, R. C. , G. Reid and J. M. Stone, "Naval Ammunition Depot,
Crane, Indiana Pilot Test - Final Report," Naval Civil Engineer-
ing Laboratory, Port Hueneme, Calif., May 1973.
z. "Pollution Abatement of Explosives," Naval Surface Weapons
Center, White Oak, Silver Spring, Maryland, undated memorandum.
za. "Pollution Potential of Explosives in Water," Navy Research
and Technology Work Unit Summary, Agency Accession No. DN 485105,
1 November 1973.
zb. "Report on Pilot Plant Investigation for Treatment of Explosive
Contaminated Wastewater," Naval Ammunition Depot, Hawthorne,
Nevada," November 1974.
zc.
Collins, J. M. and D. E. Morris, "General Plan for Environmental
Monitoring at Naval Ordnance Station, Indianhead, Maryland,"
NCEL, Port Hueneme, Calif., December 1972.
-161-
-------
15. PERSONAL COMMUNICATIONS
a. Personal communication. HAAP personnel.
b. Burlinson, N., U. S. Naval Ordnance Laboratory, White Oak, MD,
5 Apr. 1973. "
c. Chalk, R. , U. S. Army Natick Laboratories, Natick, MA, May 1973.
d. Kite, D., HQ, MUCOM, Picatinny Arsenal, NJ, November 1972.
e. Russo, S., Picatinny Arsenal, NJ, Aug. 1973.
f. Eskelund, G., Manufacturing Technology Directorate, Picatinny
Arsenal, NJ, May 1972.
g. Hebert, P., U. S. Army Environmental Hygiene Agency, Aberdeen
Proving Ground, MD, November 1972.
h. Letter, Sanghavi, A., Department of Health, State of West
Virginia, Charleston, WV, 26 Mar. 1973.
i. Sanghavi, A., Department of Health, State of West Virginia,
Charleston, WV, March 1973.
j. Bearae, U. S. Corps of Engineers, Huntington, WV, March 1973.
k. Karshina, G., Picatinny Arsenal, NJ, Nov. 1973.
1. Simmons, C. B. , Radford Army Ammunition Plant, VA, Nov. 1973.
m. Matsaguma, H., Picatinny Arsenal, NJ, Nov. 1973.
n. Broudy, P., Frankford Arsenal, PA, Nov. 1972.
o. Broudy, P., Frankford Arsenal, Philadelphia, PA, Nov. 1973.
p. Belser, D. , Remington Arms Company, Inc., Lake City Army
Ammunition Plant, MO, November 1973.
q. Reasons, K., Remington Arms Company, Inc., Lake City Army
Ammunition Plant, MO, November 1973.
r. Broudy, P., Frankford Arsenal, PA, Dec. 1973.
s. Reasons, K. , Remington Arms Company, Inc., Lake City Army
Ammunition Plant, MO, November 1972.
t. Wright, T., Twin Cities Army Ammunition Plant, MN, January 1974.
u. Mathes, G. , Iowa Army Ammunition Plant, Burlington, IA, 10 Apr.
1973.
-162-
-------
v. Griffin, W. R., Lone Star Army Ammunition Plant, Texarkana
TX, 4 Dec. 1972.
w. Allen, T., Milan Army Ammunition Plant, Milan, TN, 10 Apr. 1973.
x. Personal Communications, Radford Army Ammunition Plant Personnel
24 Jan. 1974.
y. Picatinny Arsenal personal communication, 23 September 1974.
z. John A. Brown, 18 September 1974.
za. NAD McAlester Personal Communication, July 16, 1974.
zb. Personal Communication, Joliet Army Ammunition Plant Personnel,
18 March 1974.
zc. Personal Communication, Hawthorne NAD, 18 July 1974.
zd. Personal Communication, Holston AAP personnel, 30 January 1974.
ze. Personal Communication, Crane NAD personnel, 16 August 1974.
zf. Personal Communication, NAD Hawthorne, 18 July 1974.
zg. Personal Communication, NAD McAlester, 16 July 1974.
zh. Personal Communication, NWS Yorktown, 12 September 1974.
zi. Personal Communication, Gerald R. Eskelund, 16 December 1974.
zj. Personal Communication, NAD Magna Personnel, July 22, 1974.
zk. Personal Communication, Air Force Ammunition Plant #78
Personnel, July 27, 1974.
zl. Personal Communication, NOS Indianhead Personnel, May 23, 1974.
zm. Personal Communication, Aerojet General Corp., Sacramento,
Calif., 19 July 1974.
zn.
Personal Communication, Longhorn AAP Personnel, 15 July 1974.
zo. Personal Communication, Redstone Arsenal Personnel, 13 Sept.
1974.
zp. Personal Communications, Picatinny Arsenal Personnel, Aug. 18-
22, 1975.
zq. Personal Communication, NSFC White Oak, 12 October 1975.
-163-
-------
zr. Personal Communication, NAVSEA Personnel, 11 October 1975.
16. GENERAL PUBLICATIONS
a. Walsh, J. T. , R. C. Chalk, and C. Merritt, Jr., "Application of
Liquid Chromatography to Pollution Abatement," Anal Chem, 45(7):
In press, 1973.
b. Quarterly Progress Report No. 1, "Research and Development Study
for Identification of Colored Pollutants of TNT Wastewaters,"
Contract No. DAAG17-73-C-0073, Tufts University, MS, 31 March
1973.
c. Long, L. H., ed., "The 1972 World Almanac and Book of Facts,"
Newspaper Enterprise Association, Inc., New York, NY, 1971.
d. Green, J. M., 1972, Biodegradation of Selected Nitramines and
Related Pollutants, M.S. Thesis, East Tennessee State University.
e. Verschragen, P., "Comparative Investigation of Some Methods for
the Determination of Nitrogen Content of Nitrocellulose," Anal
Chem Acta, 12:227-230, 1955.
f. "Water Resources Data for Iowa 1971," Geological Survey, De-
partment of the Interior, Iowa City, IA, 1972.
g. Rosenblatt, David H., "Investigations Related to Prevention and
Control of Water Pollution in the U. S. TNT Industry," in
Pollution, Engineering and Scientific Solutions, E. S.
Barrekette, Editor, Plenum Press, N. Y. , 1973.
h. Schulte, G. R. , Hoehn, R. C. and Randall, C. W. , "The Treata-
bility of a Munitions-Manufacturing Waste with Activated Carbon,"
Proc. 28th Purdue Industrial Waste Conference, Purdue University,
W. Lafayette, Indiana, May 2-4, 1973.
i. Schott, S., Ruchhoft, C. C. and Megregian, S. , "TNT Wastes,"
Indust. 2nd Engr. Chemistry, 35:10:1122, Oct. 1943.
j. Hales, R. A., Almy, E. G. , Young, A. A. and Pratt, C. D. ,
"Disposal of Nitrotoluene Waste Liquors," U. S. Patent 2,362,066,
1945.
k. Kozlorowski, B. and Kucharski, J., Industrial Waste Disposal,
Pergamon Press, New York, 1972.
1. Edwards, G. and Ingram, W. T. , "The Removal of Color from TNT
Wastes," Jour. San. Engr. Div., Americ. Soc. Civil Engrs., 81,
Separate No. 645, 1955.
-164-
-------
m. Prengle, H. W., Jr., "Report on Advanced Ozone Oxidation System
for Complexed Cyanides," Houston Research, Inc., Houston, Texas,
1974.
n. Rudolfs, W., Industrial Wastes Their Disposal and Treatment
Reinhold Pub. Co., New York, 1953.~
o. Osman, J. L. and Klausmeier, R. E. , "The Microbial Degradation
of Explosives," Developments in Industrial Microbiology 14-247-
252, 1973. ~ *"•'
p. Nay, M. W. , Jr., Randall, C. W. and King, P. H. , "Biological
Treatability of Trinitrotoluene Manufacturing Wastewater,"
Jour. Water Poll. Control Fed., 46:3:485-497, 1974.
q. Ruchhoft, C. C. , LeBosquet, M. , Jr., and Meckler, W. G., "TNT
Wastes from Shell-Loading Plants," Ind. Eng. Chem. 37:937, 1945.
r. Nay, M. W. , Jr., Randall, C. W. and King, P. H. , "Factors
Affecting Color Development During the Treatment of TNT Waste,"
presented at 27th Purdue Industrial Waste Conference, Purdue
Univ., May 2-4, 1972.
s. Spano, C. A., Chulk, R. A., Walsh, T. T. , and D. Pietro, C. ,
"Abatement of Nitrobodies in Aqueous Effluents from TNT Pro-
duction and Finishing Plants," in Pollution Engineering and
Scientific Solutions, E. S. Barrekette, ed., Plenum Press,
New York, 1973.
t. Eliassen, R. , "Wartime Operating Problems in Municipal and Army
Sewage Treatment Plants," Sewage Works Journ. 16:363, 1944.
u. Solin, V. and Kustka, M. , "The Treatment of Waste Waters Con-
taining TNT by Sprinkling on Ashes," Sci. Pap. Inst. Chem. T
Technol. , Prague, Fac. Technol. Fuel Wat., 2:1:247, 1958.
v. Solin- V. and Burianek, K., "Removal of TNT from Industrial
Waste," Jour. Wat. Poll. Control Fed., 32:-:110, 1960.
w. Southgate, B. A., Treatment and Disposal of Industrial Waste-
Water, His Majesty's Stationery Office, London, 1948.
x. Madera, V., Solin, V. and Vucka, V., "The Biochemical Reduction
of Trinitrotoluene," Sc., Pap. Inst. Chem. Technol., Prague,
Fac. Technol, Fuel Wat., 3:1:129, 1959.
y Allen L A "The Effect of Nitro-Compounds and Some Other
Substances in Production of Hydrogen Sulphides in Sulphate
Reducing Bacterial in Sewage," Proc. Soc. Appl. Bact., 2:26,
1949.
-165-
-------
z.
za.
Wilkinson, R. J., "Treatment and Disposal of Sewage and Waste
Waters from Shell-Filling Factories," Jour, and Proc. Inst.
Sewage Purif., 145, 1945.
Poduska, R. A. and J. F. Andrews, "Dynamics of Nitrification
in the Activated Sludge Process," presented at 29th Industrial
Waste Conference, Purdue University, May 1974.
zb. Mutton, W. C. and S. A. LaRocca, "Design for Biological Treat-
ment of Concentrated Ammonia Wastewaters," presented at 46th
Annual Conference, Water Poll. Control Fed., Cleveland, Ohio,
October 1973.
zc. Prakasam, T.B.S., Y. D. Joo, E. G. Srinath and R. C. Loehr,
"Nitrogen Removal from a Concentrated Waste by Nitrification
and Denitrification," presented at 29th Indust. Waste Conf.,
May 1974.
zd. Lue-Hing, C. , A. W. Obayashi, D. R. Zenz, B. Washington and
B. M. Sawyer, "Nitrification of a High Ammonia Content Sludge
Supernatant by Use of Rotating Discs," presented at 29th
Indust. Waste Conf., Purdue Univ., May 1974.
ze. Adams, C. E. , D. A. Krenkel and E. C. Bingham, "Investigations
into the Reduction of High Nitrogen Concentrations," Proc. 5th
Internat. Water Poll. Research Conf., 1970.
zf. Bingham, E. C., "Fertilizer Maker Stops Nitrogen," Water and
Waste Engr., F-4, Nov., 1972.
zg. Jeris, J. S. and R. W. Owens, "Pilot Scale High Rate Bio-
logical Denitrification at Nassau County, N.Y. ," presented
at N.Y. Water Poll. Control Assoc. Winter Meeting, January
1974.
zh. Tucker, D. 0., C. W. Randall and P. H. King, "Columnar De-
nitrification of a Munitions Manufacturing Wastewater,"
presented at 29th Indust. Waste Conf., Purdue Univ., May 1974.
zi. Roblich, G. A. , "Methods for the Removal of Phosphorus and
Nitrogen from Sewage Plant Effluent," Internat. Jour. Air
Water Poll., 7:427, 1963.
zj. Elissen, R., B. M. Wyckoff and C. D. Tonkin, "Progress Report-
Reclamation of Re-Usable Water from Sewage," Tech. Report 49,
Dept. of Civil Engr., Stanford Univ., 1965.
zk. Eliassen, R., "Ion Exchange for Reclamation of Reusable Water
Supplies," presented at 1965 Conf. Amer. Water Works Assoc.,
Portland, Ore., 1965.
-166-
-------
zl. "Amberlite LA-1," Tech. Note No. IE-41058, Rohm and Haas Co.,
1969.
zm. Bingham, E. C. and 1. C. Chopra, "Unique Closed Cycle Water
System for an Ammonium Nitrate Producer Using Chem-Seps
Continuous Countercurrent Ion Exchange," presented at Internat.
Water Conf. , Engr. Soc. of West. Penn. , 32nd Ann. Meet.
Pittsburgh, Penn., 1971.
zn. Kaup, E. C., "Design Factors in Reverse Osmosis," Chem. Ener.
80:8:47, 1973.
zo. Eliassen, R., and G. Tschobanoglous, "Removal of Nitrogen and
Phosphorus from Wastewaters," Envir. Sci. Tech., 3:6:538, 1969.
zp. Spatz, D. D., "Industrial Waste Processing with Reverse Osmosis,"
Osmonics, Inc., Minneapolis, Minn., 1971.
zq. Rigo, H. G. , W. T. Mikuchi and M. C. Davis, "Control of Nitro-
gen Oxide Emissions for Nitric Acid Plants," in Pollution
Engineering and Scientific Solutions, E. Bar ekette, editor,
Plenum Press, New York, 1973.
zr. Freiser, H. and Q. Fernando, Ionic Equilibria in Analytical
Chemistry, John Wiley and Sons, Inc., New York, 1966.
zs. Butler, J. N. , Ionic Equilibrium - A Mathematical Approach,
Addison-Wesley Pub. Co., Inc., Reading, Mass., 1964.
zt. Koziorowski, B. and Kucharski, J., Industrial Waste Disposal,
Pergamon Press, New York, 1972.
zu. Stephen, A. and T. Stephen, Solubilities of Inorganic and
Organic Compounds, MacMillan Pub. Co., New York, 1963.
zv. "A Unique Filtration System for Treating Industrial Waste
Water Effluents," Summary Report, Hydronautics, Incorporated,
Laurel, Md., October, 1974.
zw. Albert, R. C. , Hoehn, R. C. and Randall, C. W. , "Treatment of
a Munitions-Manufacturing Waste by the Fixed Activated Sludge
Process," presented at the 27th Annual Purdue Industrial Waste
Conference, Purdue University, 1972.
zx. Dickerson, B. W., "Treatment of Powder Plant Wastes," Proc.
6th Purdue Indust. Waste Conference, West Lafayette, Ind., 1951.
zy. Patterson, James W., Wastewater Treatment Technology, Ann
Arbor Science Publishers, Inc., Ann Arbor, Mich., 1975.
-167-
-------
zz. Patterson, J. W. , Scala, J. J. and Allen, H. E. , "Heavy Metal
Treatment by Carbonate Precipitation," presented at 30th Ann.
Purdue Indust. Waste Conference, May 1975.
zzz. Andren, R. K. , R. McDonnell, J. M. Nystron and Bruce Stevens,
"Removal of Explosives from Wastewater," Presented at 30th
Annual Purdue Industrial Waste Conference, 6-8 May 1975.
-168-
-------
TECHNICAL REPORT DATA
(Please read instructions on the reverse before completing}
REPORT NO.
EPA-600/2-76-213C
2.
!. RECIPIENT'S ACCESSIO!*NO.
TITLE AND SUBTITLE
State-of-the-Art: Military Explosives
and Propellants Production Industry (3 vols)
Vol. Ill - Wastewater Treatment
5. REPORT DATE
October 1976 (Issuing Date")
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
Patterson, James; J. Brown; W. Duckert;
J. Poison; and N. I. Shapira
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORG'XNIZATION NAME AND ADDRESS
American Defense Preparedness Association
Union Trust Building
15th and H Street, N. W.
Washington, D. C. 20005
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
R 802872
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
- Cin., OH
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/600/12
15. SUPPLEMENTARY NOTES
Vol. I - The Military Explosives & Propellants Industry
Vol. II - Wastewater Characterization
16. ABSTRACT
This study has surveyed the military explosives and propellant
manufacturing industry, covering both "GOGO" and "GOCO" facilities. Sources of
wastewater, volumes, and pollutant constituents have been reported where such
data existed.
Treatment technology currently in use at the various installations has been
described, including effectiveness of pollutant removal and secondary (air and
solid) waste generation. Systems under development at these military
installations have also been examined and evaluated in light of available
information.
The report consists of three volumes. Volume I presents general conclusions
and recommendations and describes the industry's manufacturing operations.
Volume II presents the bulk of the data concerning the wastewaters and the
treatment systems now in place. Volume III reviews and summarizes data from the
first two volumes and describes and evaluates the new treatment processes under
development at this time.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
Wastewater
Industrial wastes
Explosives
Waste treatment
Propellants
Water pollution control
Chemical wastes
Military
Manufacturing
13B
15B
3. DISTRIBUTION STATEMENT
Public Distribution
19, SECURITY CLASS (This Report}
Unclassified
21. NO. OF PAGES
181
20. SECURITY CLASS (This page/
Unclassified
EPA Form 2220-1 (9-73)
-169-
------- |