EPA-600/2-74-009c
MARCH 1975
Environmental Protection Technology Series
State-of-The-Art For
The Inorganic Chemicals Industry;
Industrial Inorganic Gases
S7,,
^o
O
\
w,
C3
PRO'
Office of Research and DeveGopment
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
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 sou'rces to meet environmental
quality standards.
This report has been reviewed by the Office of Research and
Development. Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trades names or commercial
products constitute endorsement or recommendation for use.
-------
EPA-600/2-74-009c
March 1975
STATE-OF-THE-ART FOR THE INORGANIC CHEMICALS
INDUSTRY: INDUSTRIAL INORGANIC GASES
BY
James W. Patterson, Ph.D.
and
Roger A. Minear, Ph.D.
Project R-800857
Program Element 1BB036
ROAP 21 AZQ Task 029
Project Officers
Mr. Elwood E. Martin
Office of Water and Hazardous Materials Programs
Washington, D. C. 20460
and
Mr. Richard B. Tabakin
Industrial Pollution Control Branch
Industrial Waste Treatment Research Laboratory
Edison, New Jersey 08817
for the
Office of Research and Development
United States Environmental Protection Agency
Washington, D. C. 20460
-------
ABSTRACT
A literature and field study of the inorganic gas industry revealed
that the industry is dominated by (1) air separation plants producing
argon, nitrogen and/or oxygen, (2) hydrogen plants and (3) carbon
dioxide plants. The major effluent of the industry is cooling water,
which may be contaminated with raw product condensates, oil and grease,
and water supply and cooling water treatment chemicals. Spent scrubber
solutions from product purification may also constitute a significant
waste, although newer production technology eliminates this aspect, as
well as oil and grease.
iii
-------
CONTENTS
Section Page
I. Conclusions 1
II. Recommendations 3
III. Introduction 5
IV. Study Methodology 9
V. Industry Categorization 11
VI. Waste Characterization 19
VII. Control and Treatment Technology 47
VIII. Acknowledgements 49
IX. References 50
X. Appendix 53
v
-------
FIGURES
1. Schematic of Air Separation Process 12
2. Comparison of Production to Water Use 23
3. Distribution of Plant Production 25
4. Distribution of Water Use 26
5. Distribution of Air Separation Plant Production 27
6. Distribution of Air Separation Plant Water Use 28
7. Distribution of Waste Volumes 34
vi
-------
TABLES
Page
1. Summary of Gas Production Sites 7
2. Summary of Major Gas Producers 8
3. Summary of Production Lines Visited 10
4. Source of Raw Carbon Dioxide Gas 17
5. Summary of Water Use 21
6. Waste Volume per Ton of Product 32
7. Pollutant Concentration Factors 36
8. BOD and COD Concentrations 38
9. Mass Discharge of BOD and COD 39
10. Discharge of Chromium 41
11. Discharge of Zinc 42
12. Discharge of Oil and Grease 43
13. Summary of Temperature Data 45
vii
-------
I. CONCLUSIONS
The industrial inorganic gas industry consists of a large number
of facilities primarily involved in the production of argon, carbon
dioxide, hydrogen, nitrogen and oxygen. The industry may be classified
into three segments; air separation plants which produce argon, nitrogen
and/or oxygen by separation of air into the product components, hydrogen
plants which utilize various raw materials and methods of hydrogen gen-
eration, and carbon dioxide plants utilizing combustion, fermentation,
carbonate calcining or ammonia manufacture by-products as raw material.
Plants of the gas industry are typically located adjacent to either
their source of supply of raw material, or for customer convenience.
Wastewaters of the industry primarily result from cooling water
use, and contain in addition pollutants associated with processing and
product purification. The range of cooling water use is wide, and the
volume dependent upon the extent of cooling water recycle. The nature
of the wastewater is dictated by additives to recycled cooling water,
the use of lubricated versus non-lubricated compressors, the source of
the raw materials, and the product purification techniques used. Typi-
cal wastewater constituents include chromium and zinc from cooling water
treatment, BOD from process gas organic condensates, oil and grease from
lubricated compressors, and exhausted gas scrubber solutions employed
for product purification. In general, newer technology eliminates dis-
charges of oil and grease, as well as spent scrubber solution. Substi-
tution of other cooling water treatment chemicals for chromium and zinc
compounds is increasingly employed. Use of these and similar in-plant
modifications can significantly reduce or eliminate many of the more
critical pollutants of the industry.
Remaining problems are primarily associated with thermal discharge
and total dissolved solids relating to blowdown from cooling systems,
as well as process condensate from carbon dioxide plants. In these
facilities the raw material may contain significant quantities of con-
densible organics which, upon product purification, constitute a major
waste discharge.
-------
II. RECOMMENDATIONS
Current wastewater problems in the Inorganic gas industry may be
categorized as:
1. Those associated with use of older technology,
including use of scrubber solutions, and oil and
grease leakage from lubricated compressor.
2. Pollutants resulting from compounds added to recycle
cooling water systems, particularly chromium and zinc
compounds.
3. Raw material condensates generated in the process
of product purification.
Of these, the first category can be eliminated by use of newer technology,
which includes non-lubricated compressors and thermally regenerated puri-
fication units. 'However, there are many plants which incorporate the
older processes, and only the most rudimentary treatment techniques are
employed in those plants which do attempt to treat their wastewaters.
These plants require small scale, economical and reliable treatment
processes for their wastes.
Insofar as the second category, additives to cooling water systems,
chromium and zinc compounds should be substituted by other commercial
corrosion control and biocide chemicals. This has, in fact, taken place
in a significant number of gas plants, thereby eliminating the problem
of heavy metal discharge.
The problem of raw material condensates appears to be particularly
critical in the carbon dioxide segment of the industry, dependent upon
the type of raw material used. One carbon dioxide plant receiving raw
gas from an adjacent petrochemical plant, reported one pound of conden-
sate waste per pound of C0» product. Such wastes are complex and rich
in organics, and represent a significant pollution potential from carbon
dioxide facilities. Limited information was available from this study
to substantiate this concern, and expansion of the data base for carbon
dioxide facilities is recommended. Additionally, many of the C0_ plants
-------
received by-product from ammonia production facilities, and the poten-
tial for condensate pollution from this source is also high.
This report does not discuss the molecular-sieve-type processes
developed and used by the Linde Division of Union Carbide for the pro-
duction and/or purification of oxygen and nitrogen. These resin tech-
nologies are gaining increased attention in some oxygen and nitrogen
consuming industries. It is recommended that a study be performed to
compare the waste generation and energy requirements for the cryogenic
versus molecular-sieve processes.
-------
III. INTRODUCTION
The specific industries examined as the basis for this report
were those producing the inorganic industrial gases (SIC 2813), Argon,
Carbon Dioxide, Hydrogen, Nitrogen and Oxygen. Neon in crude form is
occasionally a by-product of the air separation process which yields
argon, nitrogen and oxygen, and is mentioned in that respect only.
Helium, which is a noncondensible gas in normal air separation pro-
cesses, was not included in the study.
General Description of the Industry
Although a given industrial plant may engage in more than one pro-
duction process, the industry has been divided into 3 specific cate-
gories, each of which has varying technology. These categories are
(1) Air Separation or Rectification, which encompasses production of
liquid or gaseous argon, nitrogen and oxygen (A/N/0), and occasionally
crude neon product, (2) hydrogen gas or liquid production and, (3)
carbon dioxide liquid or solid production. Specific variations in the
generation of final product within each of these categories is described
in Chapter V.
The physical form of the product and its purity is variable, and
dependent upon the particular market requirement. Many plants in the
air separation category serve specific industrial sites, and tailor
their process to the needs of that industry. Notable in this respect
are air separation plants designed to produce only gaseous oxygen for
pipeline delivery to adjacent steel mills or iron ore processing opera-
tions (taconite ore enrichment). When plant location is not remote,
only a portion of the product is sold via pipeline, and production is
geared to providing bulk liquid and gaseous argon and nitrogen, in
addition to oxygen, for direct bulk sale in tank trucks, or individual
cylinders. Many combinations of marketing procedures exist, and for a
given plant production for a given category may vary with time, with
attendant variation in plant operation procedures.
-------
Many plants, notably air separation plants and those producing
hydrogen, are captive industries within a larger manufacturing complex.
Frequently, this results in water use and wastewater treatment data
being undifferentiable from the total plant flow and waste character
data.
A summary of production sites in the ten regions of the Environ-
mental Protection Agency is given in Table 1. Of the total 472 plants,
308 or 65.2% can be attributed to 8 major producers. These are sum-
marized in Table 2 by manufacturer and EPA region. Production capaci-
ties range from a few tons/day of combined product gases and liquids to
ten thousand or more tons/day. Production data were available for less
than half of the total reported plants.
Distribution of the inorganic industrial gas manufacturing plants
is coincident with the major industrial regions of the country, as
indicated by the high density of production sites in EPA regions III,
IV, V, VI and XI. Principally, the major steel, petroleum and ferti-
lizer manufacturing areas either use large quantities of the various
gases, or produce by-product raw material which is sold to the industry
for gas manufacture.
-------
Table 1. Summary of Industrial Gases (Inorganic) Production Sites*
EPA Region
Product
Carbon Dioxide
Hydrogen
Argon/Nitrogen/
Oxygen
Totals
9
% of Total
I
2
4
4
10
2.1
II
12
16
11
39
8,3
III
14
20
35
69
14.6
IV
18
23
20
61
12.9
V
26
36
38
100
21,2
VI
15
28
27
70
14.8
VII
13
8
3
24
5.1
VIII
4
3
6
13
2.8
IX
13
29
28
70
14.8
X
7
4
5
16
3.4
Total
124
171
177
472
100.0
* Data from The Directory of Chemical Producers (Stanford Research Institute, 1972)
-------
Table 2. Summary of Major Producers* of Industrial Gases (Inorganic)
00
1.
2.
3.
4.
5.
6.
7.
8.
Producer
Airco, Inc.
Air Products
and Chem. , Inc .
Big Three
Industries
Burdett Oxygen
Company
Chemetron Corp.
and Cardox Div.
Houston Nat. Gas
Process Plants
Union Carbide
Totals
Major
Products
A/N/0
C02, H2
A/N/0
H2
A/N/0
H2
A/N/0
H2
A/N/0
H2, C02
A/N/0
C02, H2
co2
A/N/0
C02, H2
Region
I
3
0
0
0
0
3
0
5
11
II
4
6
0
1
3
9
0
7
30
III
6
13
0
3
8
6
2
20
58
IV
4
6
3
0
3
12
1
6
35
V
8
11
0
3
10
16
1
23
72
VI
1
7
10
0
5
10
3
11
47
VII
2
1
0
0
1
5
0
5
14
VIII
0
0
0
0
3
1
0
4
8
IX
6
4
0
0
1
4
0
12
27
X
1
0
1
0
1
2
0
1
6
Total
35
48
14
7
35
68
7
94
308
* Producers with more than five production sites.
Data from The Directory of Chemical Producers (Stanford Research Institute , 1972)
-------
IV. STUDY METHODOLOGY
Characterization of the subject industry was accomplished by (1)
use of published data describing the extent and distribution of
industrial sites and the principal manufacturing processes employed,
(2) selected site visits to a small but representative cross section of
manufacturing locations and (3) review of existing RAPP permit appli-
cations on file within the EPA regional offices. Frequently, this
latter source provided insufficient information, particularly regarding
product identification and/or production yields, to allow utilization
of all permit applications in summarizing general waste character.
Little specific information regarding waste treatment facilities
and their efficiency, cost, and operational problems was available
either from plants visited or from permit application files. Because
of the nature of the basic manufacturing processes and the individual
variations, however, the industry waste problems can be identified and
treatment solutions can be projected directly to known unit operations
and/or process modifications. Existing literature, and experience
demonstrated through plant visitations, has been utilized to accomplish
these, basic goals of the study.
A total of ten separate plants were visited, representing 22
separate production lines (exclusive of crude argon rectification and
deoxygenation). Table 3 delineates these production lines.
All data accumulated from these plant visits were provided by
plant or corporation personnel from their files. No independent
sampling and analysis was undertaken.
-------
Table 3. Summary of Separate Production Lines Existing at Plants
Visited
Process Line Number
Air Separation
A/N/0 10
N/0 2
02 only 2
N2 only 2
Hydrogen 5
10
-------
V. INDUSTRY CATEGORIZATION
This section presents a description of the basic production tech-
nologies employed in each of the 3 major subgroupings of the industry;
air separation, hydrogen production, and carbon dioxide production.
In general, this division is logical as each of these product lines
is distinct with a few exceptions, which will be pointed out where
appropriate.
Within each of these subgroups there are several variations relat-
ing to the actual manufacturing process used, manner in which a parti-
cular physical or chemical operation is effected, design of equipment
accomplishing a specific operation, and entry level of the product raw
material into the process. In some instances, these variations are the
result of advancing technology within the industry, and impact strongly
on the quantity and character of process waste water. The air separation
process best exemplifies this change in production technology.
Air Separation
The air separation process fractionates atmospheric air into its
component gases, at varying levels of purity depending upon need, by
fractional distillation and condensation. Normally, the products are
oxygen and nitrogen in both liquid and gaseous states. Crude argon and
neon may be separate by-products of the basic separation process, which
are then individually subjected to further purification. Argon recti-
fication and deoxygenation are generally closely connected with the
basic air separation system, frequently within the same cold box system.
Neon is noncondensible and recovered as neon enriched nitrogen, which
is subjected to separate purification. Of 16 air separation lines visited,
only one collected crude neon. This plant shipped it off-site for puri-
fication.
Conceptually, the sequence of operations in all air separation
plants is identical. These operations are represented schematically in
Figure 1. The filtration step is generally a simple dry inlet filter
to remove particulate matter, prior to the compression step. Depending
11
-------
Air Intake
Air Filtration
Air Compression
Water Removal
CO
„ and Organic Gas Removal
Air Liquifaction
Rectification
Produc t
Liquid - NZ,
Gaseous - N,
, A
, A
Figure 1. Sequence of operations in the Air Separation Process
12
-------
upon the overall process design, the air may be compressed from as
little as 60-80 psig to as high as 2700 psig. Intermediate pressures
of 80 to 100 psig were more common in those plants visited. The com-
pression is accomplished in stages with cooling after each stage,
during which moisture in the air condenses and is discharged. Generally
cooling is provided by jacketed noncontact systems, although a few
plants employed direct contact cooling after the final compression
stage„
Depending upon the nature of the compression unit, the condensates
may be essentially pure water or a significant contaminant stream.
Most recently constructed plants employ axial flow turbine or centri-
fugal compressors, which are nonlubricated. Although some nonlubricated
piston compressors were encountered, most reciprocating compressors were
lubricated and the condensate discharge constitutes an oily waste. Com-
pressors were driven by a variety of systems; electric motors powered
by commercial electricity, electric motors powered by steam driven
generators on-site, steam driven turbines, and gas fired turbines with
steam assist turbines utilizing the high temperature exit gas to fire
waste heat boilers. Occasionally, two multistage compressors were
used, with a cleanup step between the compressors.
Residual water, carbon dioxide and contaminant organic gases are
removed from the pressurized air stream prior to expansion and liqui-
faction. Older technology effected C0~ removal by direct contact with
strong caustic solution (10% NaOH). Only two plants were encountered
which used this process and it is generally acknowledge to be out of
date. Modern plants accomplish cleanup by passing the gas stream
through some combination of molecular sieves, alumina columns, silica
gel traps or reversing heat exchangers. In the latter case, C02, water
and organic gases are removed as solids or liquids in the heat exchanger
in counter current flow with product nitrogen and oxygen. The latter
are at very low temperatures. All of these cleanup units are thermally
regenerated off stream, and vent the contaminants to the atmosphere.
At this point, the compressed air is liquified and cooled by heat
exchanger systems and various expansion techniques. Cold product gases
13
-------
generally supply the cooling, although occasionally auxiliary nitrogen
and freon systems are used. Expansion cooling is commonly achieved by
turbine expanders although piston expansion engines are employed to a
limited extent. Depending upon the exact process some combination of
expansion processes may be used in conjunction with partial by-pass of
the expansion system.
The liquified or partially liquified air passes into separation
towers, generally at 70-90 psig. Depending upon the desired product,
the rectification or separation section (called the cold box) may con-
sist of only a high pressure column, or both a high pressure and a low
pressure column. For argon purification a third column is employed.
Gas enters into the low pressure column at a pressure of 5-9 psig.
Product take off points and exact entry and return points are dictated
by product purity, quantity and balance between liquid and gaseous
forms of the individual products desired. There is no wastewater
effluent associated with the cold box system.
Product is fed back through the heat exchanger system for cooling
incoming air, as may be waste gases prior to atmospheric venting.
Liquid product is sent to bulk storage. Gaseous product is usually at
low pressure and subsequently undergoes compression prior to bulk
storage, bottling or direct pipeline distribution to neighboring indus-
trial operations. Product gas at roughly 4 psig is compressed to 200
to 3000 psig, using nonlubricated centrifugal or piston multistage
compressors with cooling between each of the stages. Generally closed
water jacket cooling is used, although some operations employ freon
cooling units.
Argon product emanating from the separate argon rectification
tower contains roughly 2% oxygen and must undergo additional processing.
Deoxygenation is accomplished in a catalytic bed with ammonia and/or
hydrogen. Resultant water and dissolved oxides of nitrogen are condensed
and discharged. This may yield a wastewater high in nitrate and nitrite.
Argon is liquified by liquid nitrogen cooling, and residual nitrogen gas
impurity is vented prior to pumping to bulk storage.
14
-------
Hydrogen
Hydrogen production is accomplished primarily by steam-hydrocarbon
reforming (natural gas or propane). Catalytic cracking of ammonia, or
high purity hydrogen production by electrolysis of aqueous alkaline or
brine solutions are lesser used methods. The latter process also pro-
duces oxygen. Each of these three processes was encountered during
site visits. Other processes have been mentioned in the literature,
eg. steam - iron process, thermal dissociation of natural gas to produce
carbon black, fermentation and others in which hydrogen is a by-product
of other manufacturing processes. These processes are most commonly
employed by industries other then the gas industry. The by-product
hydrogen may be used on site or piped to adjacent independent facilities
for further processing for marketing.
The steam-hydrocarbon reforming process involves passing propane
(desulfurized if necessary) or natural gas (CH, ) and steam at roughly
100 psig into a heated (1500 F) reformer furnace. The reformer is
usually a gas fired catalytic bed of nickel, contained in alloy tubes
within a furnace. Products from this unit are H2, CO, CO- and water.
Carbon monoxide is converted to COp by a second catalytic converter of
cupric oxide, possibly with the addition of more steam and production
of additional hydrogen gas. The hot gases are cooled either by a water
jacket system or heat exchanger employing downstream product, prior to
CO- removal. Condensate is discharged from the cooling systems. A
second stage CO converter may be present, particularly in a propane
system.
Carbon dioxide removal is accomplished by scrubbing with a mono-
ethanolamine (MEA) solution. Purified hydrogen is further cooled
(water jacket) and dried with an alumina column. The product may be
stored and shipped in bulk, or piped directly to near-by customers.
Monoethanolamine is regenerated in a heated stripping tower, venting
C0« to the atmosphere and condensing MEA, which is cooled by water
jacket cooling and returned to the scrubber. Water condensate is dis-
charge from the unit.
Other hydrocarbon processes are used, but were not encountered in
15
-------
this study. These employ partial oxidation by mixing 95% oxygen with
preheated hydrocarbons (up to fuel oil) in a catalytic bed. A mixture
of hydrogen (34%), carbon monoxide (16%), carbon dioxide (2%), water
vapor (2%), methane (1%) and nitrogen (45%) is produced. Hydrogen pro-
duction by this method ranks second only to the reforming process.
Ammonia cracking involves vaporizing anhydrous liquid ammonia,
heating to 1600-1750 F, and passing over an active catalytic bed.
Nitrogen and hydrogen gases are produced in the volume ratio of 1 to 3.
Separation of the two gases is then accomplished. One plant visited
compressed the product gases and passed them through molecular sieves.
Nitrogen is retained and vented upon regeneration. Product hydrogen
contains less than 25 ppm nitrogen contaminant. Both the cracking
generator and gas compressors are water cooled. Another plant separated
nitrogen and hydrogen by differential liquifaction.
Electrolytic hydrogen generation is employed to a limited extent
in the industry, and was utilized by only one of the five hydrogen lines
visited in this study. The process is relatively simple, with elec-
trolysis of water in electrolytic cells to produce gaseous hydrogen and
oxygen. The gas mixture is separated into its two components by differ-
ential liquification. High purity products are obtained.
Carbon Dioxide
Product carbon dioxide may be in gas, liquid or solid form. Forma-
tion of the solid form incorporates processing the other two physical
forms. The starting material is gaseous C02 which may be generated on
site by burning of carbonaceous material, or received as a by-product
of other manufacturing processes either on-site or from nearby manu-
facturing plants. Flue gases, fermentation and carbonate calcining are
the major sources of commercial C02 production (1). However, of the 126
industrial producers (exclusive of brewers employing recovery and reuse
of C0?) listed in the Directory of Chemical Producers (Stanford Research
Institute, 1972) only 44 reported their raw C02 source. The distribu-
tion of these sources, shown in Table 4, indicates that ammonia by-
product is the major C0~ source. Representatives of the industry
indicate that generation of C02 by fuel combustion is declining in the
16
-------
Table 4. Distribution of Reported Carbon Dioxide Sources
Number of Percent of
Source Sources Sources
Natural Gas 7 16
Ammonia By-Product 22 50
Fuel Combustion 4 9
Hydrogen By-Product 1 2
Methanol By-Product 1 2
Coke Breeze 1 2
Natural Well 3 7
Sodium Phosphate By-Product 1 2
Flue or Stack Gas 2 5
Lime Kiln By-Product 2 5
44 100
17
-------
industry, being replaced by plants utilizing C02 by-product, principally
from fertilizer plants producing ammonia.
In-coming gas may contain up to 99% C02 from fermentation processes
and as little as 10% C02 from fuel burning or lime kiln sources. Other
sources are intermediate in C02 concentration. The one plant visited
received crude, hot C02 gas from an adjacent petrochemical manufacturing
operation. The gas yielded 1 lb. condensate (water and organics) per
Ib. of product CO-o
Low concentration C0~ gases employ reversible absorption concen-
tration systems such as ethanolamine or concentrated potassium carbonate
solutions. Desorption of concentrated C02 is achieved by temperature
elevation. Depending upon temperature and moisture content, processing
of the CO^-rich stream may involve passing it through a. series of cool-
ing stages to remove water and other condensable substances and possibly
an oxidation stage using permanganate solution, dichrornate solution or
catalytic oxygen oxidizers.
Purified and cooled gas is compressed, cooled, dried and possibly
polished by carbon beds and sand filters prior to cooling and liqui-
faction with liquid ammonia or other cooling systems. Liquid at 290
psig can be collected to storage by bleeding out at slightly reduced
pressure (200 psig). Noncondensable gases remaining at this point are
vented to the atmosphere. Liquid C02 is usually shipped in bulk to
local distribution centers for small container filling.
Solid C0_ is produced by dropping below the triple point pressure
by expansion in extruders and rotary expansion devices to produce
pellets, or large expansion chambers to produce solid CCL snow which is
then compressed hydraulically into blocks. Vapor remaining the chamber
is returned to the compression stream.
18
-------
VI. WASTE CHARACTERIZATION
This chapter presents water use data, and describes wastewater
sources, volumes and character. Information has been taken from two
sources:
(1) Data collected during visits to specific manufacturing
sites and
(2) RAPP permit applications on file with EPA.
Specific Water Uses
The primary water use in all three types of facilities is cooling
water. For air separation plants, this use is primarily associated
with air or product compressor cooling. Most compression operations
are multistage. For inlet air, 4 to 5 stages are typical; for product
piping or storage, 3 to 6 stages are common. The most frequent prac-
tice is to use jacketed cooling systems. Occasionally, direct contact
water cooling after the final compression state is employed. A variant
on the direct contact configuration, predominantly used in older plants,
is to remove carbon dioxide in the compressed gas stream by scrubbing
in strong caustic solution. Product compression cooling is always
closed jacket, to prevent product contamination.
In hydrogen manufacture, both hydrocarbon reforming and ammonia
cracking employ cooling of the hot gases exiting from the high tempera-
ture catalytic beds. For hydrocarbon reforming, additional cooling
follows the carbon monoxide conversion bed and is also used in the
regeneration of monoethanolamine scrubber systems. Product compression
is also accompanied by cooling, after the compression stages.
Carbon dioxide, when generated by combustion of hydrocarbons or
other low yield processes, must be cooled prior to enrichment in scrub-
ber systems since these systems are reversed (ie. heated) to yield high
concentration C02 gas. By-product C02 may be received at high tempera-
ture in admixture with condensible contaminants, in which case the
cooling process also serves as a purification step. Purification and
19
-------
liquifaction of the C02 involves compression, and requires cooling at
various stages of processing. Most of the cooling steps utilize water
cooling. Air and liquid ammonia or freon cooling systems are also
employed.
Boiler systems can represent significant water use, but the magni-
tude of this use is highly variable. The most significant use of boiler
systems in air separation plants is when compressors are steam driven
or when electric power is generated on-site with a steam powered genera-
tor. One instance was encountered where the primary compressor power
source was a gas fired turbine. High temperature exit gases in turn
fired waste heat boilers, thereby generating steam to drive assist
turbine compressors. Generally, boiler systems are present primarily
for plant heating and as stand-by units to vaporize backup liquid
product storage reserves for pipeline customers. Units are also avail-
able to provide rapid thawing of the process units upon plant shut down.
For hydrogen production by hydrocarbon reforming, high pressure steam
is one of the starting raw materials. This steam is produced on-site.
Water use in boiler systems depends upon the steam requirement, and
whether condensate return systems are employed.
Miscellaneous other water uses are generally small in comparison
to cooling and boiler systems. Among those encountered are;
Scrubber systems
General plant wash down
Spray systems used during bottle or tank filling
Sanitary systems
Water Use Volume
Total water use would be expected to reflect the size or production
level of the individual manufacturing facility, and the nature of its
products. This would be reflected in a rather narrow range of values
upon nornalizing water use to a "volume per unit of product" basis.
That this is not manifested is demonstrated in Table 5 and Figure 2,
where data are presented for 40 plants located throughout the United
States. Extremely high water use rates are directly attributable to
20
-------
Table 5. Water Use Summary for Industrial Inorganic Gas Facilities
Product
A/N/0
°2
N/0
N/0
N/0
N/0
N/0
N2
N/0
A/N/0
N2
N/0
N2
A/N/0
N/0
A/N/0
N/0
N/0
A/N/0
A/N/0
°2
A/N/0
N/0
°2
N/0
A/N/0
A/N/0
A/N/0
A/N/0
N2
Production, T/day
27
65
175
180
184
184
270 (02)
280
280
284
294
340
392
400 (02)
400 (02)
446
450 (02)
459
491
589
680
852
891
1,050
1,100
1,114
1,200
2,550
3,065
5,840
Water Use, MGD
.0235
.00095
.0436 to .0570
.21
.0801
.047
.09
.080
.090
0.40
.200
.06
.050
.36
.290
.440
9.65
.402
.0731
30.26
.130
.110
.940
14.40
.37
.290
0.100
2.003
.500 £,,•
.018
Water Use,
gal/ton Product
870
15
247-329
1,167
435
255
333
286
321
1,408
680
176
128
900
725
986
21,444
876
149
51,380 *
191
129
1,055
13,714 *
336
260
83
785
163
3
21
-------
Table 5 (Continued)
Product Production, T/day
N2, H2
A/N/0, H2
N/0, H2
A/N/0, H2
Hg (liq.)
H2, CO,
co2
CO,
co2
co2
179
250
984
5,870
16
289
85
112
200
200
Water Use, MGD
.05
.089
.0376
100
.19
.820
1.42
.0590
.020
.0082
Water Use,
gal/ton Product
279
356
38
17,035 *
11,875 *
2,837
16,705 *
524
100
41
* Once through cooling systems
22
-------
10,000
1,000
CO
g
4J
A
g
•rl
4J
U
I
100
20
0.001
00
O
OQ
O
O
0.01
1.0
0.1
Water Usage, MGD
Figure 2. Comparison of Plant Production to Water Use
10
-------
use of once through cooling water, by virture of plant location on a
lake, river, estuary or similar site. These plants are indicated in
Table 5, and have been omitted from Figure 2. Of the six plants not
using recycle cooling, four are air separation, one hydrogen and one
COg facilities. Water use variation likely reflects several factors,
among which are;
extent and nature of steam use
cost of water at source
quality of water from source
use of scrubber systems
method of reporting plant production data
With respect to this last point, some plant production data were given
only in terms of oxygen produced, while nitrogen production could have
been significantly high as well.
In one plant visited, use of steam generated electricity clearly
contributed to a high water use per ton of product (1055 gal/ton). Even
when cooling water represented essentially the total plant water use,
rates fluctuate considerably from plant to plant.
Evaluation of the air separation data of Table 5 reveals no corre-
lation between plant production capacity and cooling water utilization
per ton of product produced. The limited data for hydrogen and carbon
dioxide production facilities does indicate generally lower water use
per ton of product, for the larger plants.
Among those 40 plants for which data are reported in Table 5 the
median production capacity is 400 tons/day, and median water usage is
300 gal/ton of product. As shown in Figures 3 and 4, there was no
difference in distribution between the air separation plants as a group,
and all plants for which Table 5 lists data. However, within the air
separation plants (Figures 5 and 6), there is a significant difference
between those which produce only nitrogen and "oxygen (N/0), and those
which produce argon as well as the former products (A/N/0). Median
production capacity is 200 versus 650 tons/day for N/0 versus A/N/0
plants, while water usage is 350 versus 200 gal/ton product. Figures 5
and 6 exclude air separation plants not practicing cooling water recycle,
24
-------
3,000
I T
1,000
CO
g
4-1
i-l
4J
O
f
a
100
8**
O*
00
0°
10
I I
I
O ^ir Separation Plants
£ All Inorganic Gas Plants
I
I
I
_L
I
J_
10 15 20 30 40 50 60
Percentage £ Value Shown
70
80
90
Figure 3. Distribution of Production for Gas Plants
25
-------
3,000
I I
1,000
4J
O
T3
O
VJ
D,
03
60
0)
CO
0)
4J
100
10
Air Separation Plants
All Gas Plants
I I
I
I
I
I
1
I
10 20 30 40 50 60
Percentage < Value Shown
70
80
90
Figure 4. Distribution of Water Use for Gas Plants
26
-------
1,000
to
g
4-1
g
•H
4J
O
I
100
10
O A/N/O Plants
• N/0 Plants
I
I
I
I
I
I
_L
10 20 30 40 50 60
Percentage < Value Shown
70
80
90
Figure 5. Distribution of Production for Air Separation Plants
27
-------
1,000
4J
u
-o
o
M
CU
(0
00
(U
CO
J-i
(U
100
10
O A/N/0 Plants
A N/0 Plants
I
I
I
I
I
I
10 20 30 40 50 60 70
Percentage < Value Shown
80
90
Figure 6. Water Use in Air Separation Plants
95
28
-------
as well as plants which also produce either hydrogen or carbon dioxide,
or for which only oxygen production data was reported.
It is unlikely that the larger production of the A/N/0 plants, as
compared to the N/0 plants, is due solely to the production of argon.
More likely, it is characteristic that an air separation plant designed
to produce larger quantities of N/0 would also incorporate an argon
system as an adjunct. This reasoning, however, is not applicable to
the water use data. N/0 plants use more water per ton of product then
A/N/0 plants. This perhaps reflects less efficient cooling system
design for the smaller (ie. N/0 only) segment of the air separation
industry.
Average water usage, based upon the data of Table 5 excluding
once-through cooling systems are, for all plants, A/N/0 and N/0 respec-
tively 491, 537 and 546 gal/ton product. Average values are thus more
comparable then median values. The averages represent 9 A/N/0, 9 N/0
and 31 total plants from Table 5. The remaining 9 plants either used
once-through cooling or did not report total production tonnage.
Sources of Wastes
Sources of waste in industrial gas manufacturing plants are rela-
tively few, and easy to identify. These are:
process condensate
cooling water discharge
boiler blowdown
boiler condensate
boiler (and occasionally cooling) water treatment discharge
plant wash down
scrubber solution dump
paint stripping solutions
In the air separation process, in-let gas upon compression and
cooling produces a condensate waste. This water may be relatively
clean in plants using the newer nonlubricated compressors. Older tech-
nology employs lubricated units, which contribute oil to the condensate.
A few air separation plants use alkaline scrubbers for CO* removal after
29
-------
air compression and the spent caustic constitutes a waste. Beyond these
aspects, aqueous waste is not produced, as downstream cooling is con-
densate free and accomplished by closed systems. When C09 scrubbers
are not used, all moisture, CO- and trace atmospheric organic contami-
nants are removed by thermally regenerated system (gel traps, alumina
or reversing cryogenic heat exchangers). The deoxygenation process for
argon purification may produce water containg nitrogen oxides, where
ammonia is used as the reductant.
Cooling water character may be relatively unchanged for once
through systems, except for temperature increase and possible chlorine
addition at the intake. When cooling towers are employed, evaporative
losses are often significant, and overall constituent concentrations
are magnified by the factor employed to regulate the particular unit's
blowdown rate (ie. number of recycles). Chemical additions to control
corrosion and biological activity constitute contaminants in the blow-
down, and this plus evaporative concentration dictate the nature of
this waste stream.
Boiler operation can contribute to the waste stream in several
respects. Steam condensate may be discharged relatively clean, or may
contain volatile additives such as antiscaling amines. The boiler is
blown down either periodically or continuously, and blowdown generally
contains several additives. Boiler water pretreatment contributes
solids where coagulation or softening is practiced. Ion exchange
softening (generally with synthetic zeolites) produces waste brine
regenerant solutions, which require disposal.
Plant wash down will reflect general plant housekeeping practices
and could conceivably represent an oily waste depending upon the overall
nature of equipment in use. Specific wastewaters also result from
cleaning and stripping of paint from product gas cylinders. High lead
wastes can be associated with such operations.
The only significant difference between air separation and carbon
dioxide production lies in the nature of the raw C(>2 source. In one
plant visited, raw C09 received from an adjacent petrochemical manu-
facturing operation contained roughly 1 Ib. of condensate per Ib. of
30
-------
product C02« In addition to water, this condensate contained organic
compounds, including glycols. When organics removal is not accomplished
by condensation and bleed off or adsorptive processes, an oxidizing
scrubber such as permanganate or dichromate may be used. These solutions
may represent a disposal problem. Otherwise, compressor condensates,
boiler blowdown and related wastes, and cooling tower blowdown considera-
tions are similar to those of air separation units. Scrubber systems
for C0~ enrichment are normally regenerated, and typically represent a
waste problem only if periodic dumping is required.
The major hydrogen production process, hydrocarbon reforming, has
a water by-product condensate which should be essentially uncontaminatedo
The monoethanolamine scrubbing system is regenerated continuously and
some water condensate is generated, supposedly uncontaminated although
no substantive data were available. Other discharge sources are similar
to those previously discussed for air separation and carbon dioxide
facilities.
Hydrogen production by catalytic cracking of ammonia uses cooling
water after the cracking process, and for cooling after compression of
product. No specific wastewater information was available on hydrogen
production by the partial oxidation of hydrocarbons with 95% oxygen.
Since a mixture of gases is obtained, including CO, C02 and H20,
scrubbing and condensation would be expected, with unit operations
common to the reforming and ammonia cracking processes.
Quantity of Wastes
Wastewater volume data are tabulated in Table 6, on the basis of
gallons per ton of product. These data were restricted to plants for
which total discharge could be determined, and which did not use once
through cooling water. Where possible total production data were used
for computation. However, some production data were presented for
oxygen only, as indicated in the Table. Another uncertainty in these
data lies in whether or not the reported production is design capacity
of the plant or actual average production. The former is suspected,
on the basis of experience with those plants visited.
31
-------
Table 6. Waste Volume per Ton of Product
Product
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
N/0
N/0
N/0
N/0
N/0
N/0
N/0
N/0
N/0, H2
N2
N,
°2
H , N
H2, C02
CO-
Production Rate
284
284
321
400 (02)
446
852
2550
3065
270 (02)
280
284
284
340 (02)
400
459
891
1,100 (02)
280
294
392
5840
640
16
179
289
200
Discharge, Gal/T Product
35
357
62
300
538
48
196
82
70
54
56
35
21
35
544
169
78
43
75
51
1
74
2875
112
1211
12
32
-------
For 8 A/N/0 plants, the average discharge was 202 gal/ton product,
with values ranging from 35 to 538 gal/ton. Those 8 plants producing
N/0 only, had an average discharge of 123 gal/ton and a range of 21 to
544. A/N/0 and N/0 plants are thus comparable on this basis. Other
gas plants generally fell within the same range, except for two hydrogen
plants for which the water use was much higher. Figure 7 presents the
cumulative distribution of discharges (gal/ton product) for A/N/0, N/0,
and all plants. However, plants reporting oxygen data only were ex-
cluded. Median discharge for A/N/0 and all plants was equal, at approxi-
mately 650 gal/ton. Discharge for the N/0 plants was somewhat lower, at
450-550 gal/ton product. Based upon the limited amount of data for N/0
facilities, this difference may not be real. Figure 6 indicated that
the median water utilization for N/0 plants was 350 gal/ton product,
as compared to a median discharge of 450-550.
The average for all plants, except those reporting only partial
production data, was 303 gal/ton product. For comparison, the average
water intake volumes for all plants, A/N/0 and N/0, respectively, were
491, 537 and 546 gal/ton. These values are tabulated below:
All Plants A/N/0 Plants N/0 Plants
Average Inflow 491 537 546
Average Discharge 303 202 123
Percent Reduction 38.3 62.4 77.5
General analysis of water use data indicates that the major quantity of
waste water results from cooling tower blowdown. Water balances reported
suggest that evaporative losses are high, and that condensate volume in
air separation plants is small relative to the cooling tower blowdown
volume.
Character of Wastes
The nature of the waste stream is dictated by several factors.
Essentially these are:
1. Specific in-plant treatment processes and/or actual
33
-------
1,000
O
M
O.
00
Ci
-------
manufacturing technology in use.
2. Direct addition of process wastes and condensates
3. Discharge of in-plant boiler water treatment wastes
4. Evaporative concentration of intake water constituents
5. Nature and quantity of cooling and boiler water additives
6. Auxilliary activities such as plant washdown and gas
bottle paint stripping operations
These factors impact upon the chemical character of the waste discharge.
In addition, since the bulk of discharge water results from cooling
systems, elevated temperatures are also experienced.
Characterization of effluent nature by examination of permit appli-
cations proved to be difficult for several reasons. First, much reported
data is based upon a single grab or single composite sample. Secondly,
not all constituents are reported on all permit applications. In many
cases, plant effluent data are given but no corresponding data are
supplied for the intake water. In other cases, conversion of effluent
concentrations to Ibs/day and Ibs/unit product based on values given
for waste flow and production do not agree with those values reported
in the permit application. Complicating this situation further is the
fact that analytical data reported on influent and effluent concentra-
tions of what would be expected to be conservative constituents (those
not added or involved in chemical reactions leading to removal) do not
show agreement among themselves, nor correlate with calculated concen-
tration factors resulting from influent:effluent water ratios. Most
likely several factors are involved, among which are poor estimates of
flow, poor analytical precision, collection of influent and effluent
samples under significantly different conditions, and insufficient
numbers of samples to provide representative averages. Total mass
balances were not obtained in most instances.
Based on influent:effluent water ratios for 30 plants, the mean
discharge concentration factor was 4.5, with a range of 1.6 to 9.1. A
comparison of this value with effluent:influent ratios for selected
constituents is given for 19 plants in Table 7. Although there is a
great deal of scatter in the concentration data, a general trend of
35
-------
Table 7. Ratios of Effluent to Influent Concentrations
Product
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
N/0
N/0
N/0
N/0
N/0
N/0, Hg
N2
N2
N2
N2, H.J
°2
=2
V C02
oo2
Production
ton/day
3065
1114
852
400 (02)
321
981
340 (02)
280
270 (02)
1100 (02)
5840
392
294
179
640
16
289
200
Water ,
in: out
2.0
7.0
3.4
3.0
3.9
6.2
8.6
4.5
4.5
5.1
5.3
3.0
4.0
9.1
2.5
4.0
4.9
2.5
8.3
Total
Dissolved
Solids
1.4
0.69
2.9
1.8
5.8
6.9
2.9
1.9
4.8
2.3
*•
2.7
3.9
2.5
3.4
5.7
5.1
1.9
Sulfate
-
-
4.0
1.3
8.0
••
2.8
2.7
6.2
20.0
1.3
22.5
1.1
8.9
-
2.3
12.4
1.0
6.3
Chloride
3.6
-
-
2.8
0.95
-
2.0
2.9
2.6
2.3
1.4
3.3
1.0
4.3
-
2.2
5.2
13
1,7
Calcium
4.0
-
-
2.5
-
2.7
••
.74
1.4
4.4
-
2.2
5.2
1.1
36.5
Sodium
3.9
-
-
2.1
-
2.9
w»
3.6
3.1
3.8
4.4
3.5
5.9
-
M
5.2
2.0
Average
4.8
3.3
6.7
3.3
6.1
3.7
-------
increasing concentration with increasing water (influent:effluent)
ratios is observed. A comparison of the averages of Table 7 shows com-
parable values for dissolved solids, chloride and sodium, all of which
are close to the water discharge ratio. Except for the one large con-
centration factor of 36.5 reported for calcium at one carbon dioxide
plant, the calcium ratio (2.7) is also comparable. The sulfate
concentration factor is high, as are many of the individual values
reported. This likely results from sulfuric acid addition to recycled
cooling waters, for pH control. Sodium values would be expected to
reflect a similar increase to chloride and this is evident from the
data. In general, the concentration factors are of the same order of
magnitude, indicating that the primary contribution to increased
effluent concentrations is evaporation of cooling water, and where
appropriate, consumptive use of steam.
Data for nitrogen and phosphorus forms usually did not show similar
concentration factors. Usually, little or no change was evident, im-
plying that removal of these elements might be occurring in cooling
systems through biological action.
BOD and COD discharges are presented in Table 8 for 15 plants.
Significant increases in BOD and COD values are apparent for several of
these plants. There is no correlation between the water use (influent:
effluent) ratio and either BOD or COD concentration factors, for in-
dividual plants. However, the average concentration factors for BOD
and COD, at 6.9 and 7.0 respectively, are very close. Both are higher
then the average water use ratio of 4.8, indicating addition of organic
waste in the gas processing. This likely reflects organic condensate
blowdown, and oil and grease leakage from compressors, or organic
additives in the cooling water system. Table 9 presents BOD and COD
data on the basis of production. BOD discharge for the air separation
plants, with the exception of the one plant also producing carbon
dioxide, falls within the range of 0.0037 to 0.0156 Ibs. per ton pro-
duct. The average BOD for all plants for which sufficient data was
available is 0.0426 Ibs. per ton product. Average COD discharge is
much higher at 0.2481 Ibs/ton, although if the extremely high discharge
37
-------
Table 8. BOD and COD Concentrations and Concentration Ratios
u>
oo
Product
A/N/0
A/N/0
A/N/0
A/N/0
N/0
N/0
N/0
N/0
N/0, H2
N2
N2
N2, CO,
N2, Hg
co2
H2
Average
Production,
ton/day
3065
1114
400 (02)
321
891
280
270 (02)
-
1100 (02)
392
294
289
179
200
16
Water
in: out
2.0
7.0
3.0
3.9
6.2
4.5
4.5
5.1
0
4.0
9.1
2.3
2.5
8.3
4.9
4.8
Influent*
0
0
1.4
Ml
2
5
1
2
0
1.4
2
0.8
2
-
0
1.4
BOD
Effluent*
10
5
1.6
-
3
35
1
1
7
24
12
10
4
-
8
9.4
Ratio
-
-
1.1
-
1.5
7.0
1.0
0.5
-
17.1
6.0
12.5
2.0
-
-
6.9
Influent*
17
31
32.3
7
2
7
1
-
8
0
5
12
2
.4
0
8.9
COD, mg/1
Effluent*
60
20
62.2
214
40
49
0
-
18
79
190
40
16
16
70
62.4
Ratio
3.5
0.6
1.9
30.6
20.0
7.0
-
-
2.3
-
38.0
3.3
8.0
40.0
—
7.0
* Expressed in mg/1.
-------
t
Table 9. Mass Discharge of BOD and COD per Ton Product
u>
Product
A/N/0
A/N/0
N/0
N/0
N2
N2
N2, H2
N2, C02
co2
H2
Average, Ibs /ton
Production
tons /day
3065
321
891
280
392
294
179
289
200
16
BOD,
3
Ibs x 10 /ton product
6.8
-
4.2
15.6
10.2
7.5
3.7
101.0
-
191.8
42.6 x 10"3
COD,
3
Ibs x 10 /ton
40.8
111.2
56.2
21.9
33.6
118.6
14.9
404.0
1.3
1,678.4
248,1
product
x 10"3
-------
of the hydrogen plant is omitted this average drops to 0.0892 Ibs/ton.
Contaminants of significance, based on consideration of additives
to the water passing through the plant rather than increases due to
evaporation, seem to be sulfate from sulfuric acid addition, chloride
from brine discharge, and chromium, zinc and occasionally iron and
copper from corrosion inhibitor and biocide treatment of cooling water.
Nonmetallic organic biocide additives contribute to COD analysis values.
Oil and grease are present in some effluents and reflect plant washdown
practices or use of lubricated compressors. Data regarding these para-
meters are summarized in Tables 10 through 12.
In Table 10, chromium data are reported only for those plants
implying use of chromium by their permit applications. Some plants
showed no increase in chromium values between influent and effluent.
Others reported no values for chromium. In some cases information indi-
cated discontinuation of chrornate use. While effluent values range from
less then 1 to 23 mg/1, discharge per ton of product values vary over 3
orders of magnitude, reflecting the wider range of water discharge per
ton of product (Table 5).
Similar results are demonstrated for effluent zinc values in
Table 11. Concentrations vary over an order of magnitude while dis-
charge per ton of product varies over 3 orders of magnitude. Relatively
little data were available for oil and grease discharges, as shown in
Table 12.
A consideration of the purpose of chromium or zinc addition to
cooling waters, which is for corrosion and biological control, suggests
that from an economic point addition would be based upon maintaining an
effective concentration in the cooling water, and that effluent concen-
tration should be more consistent then mass discharge per ton of product,
Therefore, for these two parameters there appears to be little merit in
considering their discharge on a mass per ton of product basis, since
effluent volume shows no such correlation.
Tables 10, 11 and 12 indicate that one hydrogen plant (16 ton/day)
was consistently highest in pollutant discharge per ton of product, and
much higher then any other plant for which data were available. The
40
-------
Table 10. Selected Values of Chromium Discharge
Product
A/N/0
A/N/0
A/N/0
N/0
N/0
N2
N2
°2
H2
co2
co2
Average
Production,
tons /day
3065
2550
446
340 (02)
270 (02)
392
294
640
16
200
200
Water
Cone. Factor
2.0
4.0
1.8
8.6
4.5
4.0
9.1
4.0
4.9
1.0
8.3
4.7
Cone . in ,
mg/1
.01
-
.2
0.05
0.05
0
0.05
0.05
0
.2
.2
0.07
Cone . out ,
mg/1
.65
4.0
11.5
2.7
11.0
22.6
4.0
1.90
8.1
3.4
3.0
6.6
Chromium
Cone. Ratio
65
-
58
-
-
-
8
38
-
17
15
94.3
Discharge ,
Ibs x 103/ton prod.
0.4
6.5
51.6
0.5
6.5
4.8
2.5
1.2
194.2
1.1
0.3
24.5 x 10"3
-------
NJ
Table 11.
Product
A/N/0
A/N/0
A/N/0
N/0
N/0
N/0
N/0
N/0
N2
N2
N2
N2
H2
H2, C02
co2
co2
Average
Selected Values of Zinc Discharge
Production
3065
446
400 (02)
340 (02)
284
270 (02)
-
M
5840
392
294
280
16
289
200
200
Water
Cone. Factor
2.0
1.8
3.0
8.6
8.0
4.5
10.0
5.1
3.0
4.0
9.1
6.0
4.9
2.3
1.0
8.3
4.5
Cone . in ,
mg/1
.01
0.1
.067
.050
-
.050
.03
.05
.050
.004
.03
.05
0
.05
.02
.3
0.06
Cone, out,
mg/1
.51
3.4
4.77
.35
1.0
10.0
5.3
2.0
1.20
12.5
1.9
2.9
11.8
3.5
1.3
2.1
4.03
Zinc
Cone. Ratio
51
34
71
7/
-
200
177
40
24
3125
63
58
M
70
65
7
67.2
Discharge
Ibs x 103/ton prod.
0.3
15.2
11.9
0.06
0.3
5.9
m
-
0.01
2.7
1.2
1.1
282.9
35.4
4.4
0.2
25.8 x 10"3
-------
Table 12.
Product
A/N/0
A/N/0
A/N/0
N/0
N/0
N/0
N2
N2
«2
Average
Summary of Oil and
Production,
tons /day
3065
2550
289
284
284
280
392
280
16
Grease Discharge Data
Cone . in ,
tng/1
0
-
0
-
••
-
0
0
0
-
Cone . out ,
mg/1
11
15
5.0
30
30
20
15.8
19
10
17.3
Discharge
Ibs x 103/ton prod.
7.5
24.5
50.5
14.1
8.8
8.9
3.4
7.4
240.0
40.6 x 10"3
-------
values for chromium, zinc, and oil and grease for this plant were 0.1942,
0.2829, and 0.240 Ibs/ton while the respective averages for all other
plants (excluding that hydrogen plant) were 0.0075, 0.0060 and 0.0157
Ibs/ton. Thus, this single plant greatly influenced the averages pre-
sented in Tables 10 through 12.
Effluent pH values were always in the range of 6.0 to 9.0. In part,
this reflects the fact that most discharge water results from cooling
water systems which are controlled to maintain proper pH. Batch dumping
of alkaline scrubber water would significantly alter effluent pH. How-
ever, where used in plants visited, the concentrated caustic solutions
were contracted out separately for disposal.
Table 13 summarizes the reported temperature changes between
influent and effluent waters in terms of average summer and winter values,
These data are for both once through and cooling tower systems. For 31
plants, the mean temperature changes were 14.0 F for the summer and
14.5°F for the winter. Ranges of temperature change were -5 to 41 F for
summer and -5 to 49 F for the winter.
44
-------
Table 13. Summary of Temperature Information
Product
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/N/0
A/M/0
N/0
N/0
N/0
N/0
N/0
N/0
N/0
N/0
N/0
Production,
tons /day
5870
3065
2550
1114
852
589
446
400 (02)
321
287
284
1050 (02)
871
459
400
340 (02)
284
284
280
270 <02)
Inflow
Summer
65
80
78
65
75
80
64
65
70
85
84
87
70
85
70
73
85
74
85
60
Temp., °F
Winter
40
80
75
44
55
40
55
55
40
70
75
77
40
70
65
45
70
71
70
40
Outflow
Summer
82
87
90
106
75
82
86
85
80
90
90
98
80
90
65
83
90
90
90
85
Temp . , °F
Winter
50
78
85
93
55
42
74
75
60
80
82
83
45
80
60
63
81
84
81
70
-------
Table 13.(Continued)
Product
°2
°2
N2
N2
N2
N2
N/0, H2
Nu
n ) tin
H2
H2, C02
co2
Production,
tons/day
640
450
5840
392
294
280
1100 (02)
179
16
289
85
Inflow
Summer
75
85
75
70
70
84
64
70
77
77
63
Temp . , F
Winter
40
55
60
50
50
70
40
50
77
77
52
Outflow
Summer
85
95
85
90
90
90
85
106
85
102
94
o
Temp., F
Winter
50
66
70
80
80
80
70
90
82
94
86
-------
VII. CONTROL AND TREATMENT TECHNOLOGY
There are two approaches to handling waste problems in the inor-
ganic gas industry. In-plant control to eliminate the waste character-
istics of concern, and treatment of the waste stream to reduce the
contaminant to acceptable levels. Within the air separation industry,
in-plant controls are capable of complete elimination of the discharge
of oil and grease and caustic wastes, as well as chromium and zinc
associated with cooling tower blowdown. Current manufacturing technology
and cooling tower treatment practice have demonstrated that air separa-
tion plants can be pollution free, exclusive of their concentrating of
inlet water constituents.
Modern plants use nonlubricated compressors, yielding pure water
condensate. Carbon dioxide removal is accomplished by reversing heat
exchangers, silica traps or molecular sieves, all of which are thermally
regenerated and exhaust original air constituents back to the atmosphere.
Chrornate and zinc additives to cooling water systems can be and in many
plants have been replaced by other acceptable treatment chemicals. Any
problem beyond this aspect is not specific to this industry as such, but
a problem of cooling water blowdown in general. Primarily the solution
relates to use of acceptable additives for corrosion inhibition and
biological control, and discharge water of acceptable temperature dif-
ferential relative to the receiving system.
Where steam systems are employed, water pretreatment will result
in periodic brine or sludge discharge. Again this is a problem not
unique to the inorganic gas industry. One solution is to use electric
power from an off-site source.. Under these circumstances, air separation
plants require boiler systems only for backup purposes.
Existing plants utilizing older technology must apply oil removal
techniques to the compressor condensates. Many plants have already
taken this measure. Treatment ranges from simple catch tanks with sur-
face skimmers and submerged outlets, to systems with emulsion breakers,
skimmers and activated carbon clean-up. The technology and costs of
47
-------
these systems are well established (2). Caustic scrubber systems are
clearly outmoded and should be replaced by more modern CO- removal
systems. The alternative is to dispose of the waste caustic by contract,
as has been implemented by several plants.
Exclusive of those plants processing crude hydrogen or C02 from
adjacent petrochemical facilities, where condensible organics represent
a significant organic waste, hydrogen and CO, plants are analogous to
the air separation plants in their waste generation and control consider-
ations. Cooling tower blowdown, boiler blowdown and boiler water treat-
ment wastes are the principal waste streams. Carbon dioxide removal
and recovery systems, when properly designed and operated, have no liquid
wastes as they are closed cycle regenerative systems exclusive of cooling
water and condensate. Lubricated compressor systems are subject to the
same considerations as air separation units, relative to oil and grease.
For plants processing raw gas streams, location adjacent to large
petrochemical plants generally affords access to treatment units designed.
to handle the complex petrochemical waste. Contracting of condensate
wastes to these adjacent facilities is in practice at several plants.
One plant currently receiving bulk Ik for bottling will in the future
receive a raw EL stream. Clean-up is designed for the gas stream by
activated carbon with thermal regeneration and atmospheric venting.
48
-------
VIII. ACKNOWLEDGEMENTS
The cooperation of E. Martin and D. Becker of the Effluent Guide-
lines Division, and personnel of the regional offices of the Environmental
Protection Agency, greatly assisted the performance of this study.
49
-------
IX. REFERENCES
1. Shreve, R. N., Chemical Process Industries, 3rd. Ed., McGraw-Hill
Book Co., N.Y.
2. Patterson, J. W. and R0 A. Minear, "Wastewater Treatment Technology,"
2nd Ed., State of Illinois Institute for Environmental Quality
Report IIEQ 73-1, February, 1973.
50
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
A ccession No.
4. Title
State-of-the-Art For The Inorganic Chemicals Industry:
Industrial Inorganic Gases
7. Authoi(s) James W. Patterson, Ph.D.
Roger A. Minear, Ph.D.
ReportDat* fan. 1974
9. Organization
Department of Environmental Engineering
Illinois Institute of Technology
Chicago, 111. 60616
ft
10. Project tfo.
PE 1BB036 R/T 21 AZQ29
Contract/Grant No.
R-800857
13.
d Coveted
IS. Supplementary Notes
Environmental Protection Agency report number, EPA-600/2-7^-009c, March 1975
16. Abstract
A literature and field study of the inorganic gas industry revealed that the
industry is dominated by (1) air separation plants producing argon, nitrogen
and/or oxygen, (2) hydrogen plants and (3) carbon dioxide plants. The major
effluent of the industry is cooling water, which may be contaminated with raw
product condensates, oil and grease, and water supply and cooling water treatment
chemicals. Spent scrubber solutions from product purification may also constitute
a significant waste, although newer production technology eliminates this aspect,
as well as oil and grease.
17a. Descriptors
17b. Identifiers
17c. COWRR Field & Group
18. Availability
A bstractor
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON. O. C. 2O24O
Institution
WRSIC lOa (REV. JUINF I 9711
------- |