EPA-600/2-77-200
September 1977
Environmental Protection Technology Series
TREATMENT OF AMMONIA PLANT
PROCESS CONDENSATE EFFLUENT
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Off ice of Research and Development, U.S. Environmental Protec-
tion Agency, have been grouped into nine series. These nine broad categories were
established to facilitate further development and application of environmental tech-
nology. Elimination of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY
series. This series describes research performed to develop and demonstrate instrumen-
tation, equipment, and methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the new or improved tech-
nology required for the control and treatment of pollution sources to meet environmental
quality standards.
REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved for
publication. Approval does not signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
-------
TECHNICAL REPORT DATA
rfsd IntifUftioHt OH rtr rcirnr twitter
RtGl^iiNT'S ACCI*SI0*« WO,
, if Tki Treatnaent of Ammonia Plant Process
Condensate Effluent
DATE
September 1977
CODC
7 *UT«0«»Si c.J. Romero, F.Yocum, J.H.Mayes, and
D. A, Brown (Gulf South Research Imtltute)
• Pf nrOflMiNG ORGANIZATION nCPOHT NO.
9 PERFORMING a«e«NI£ATi0». NAMf AND ADDRESS
Louisiana Chemical Association
251 Florida Street
300 Taylor Building
Baton Rouge. Louisiana 70801
10
NO.
1BB6JO
! 1 CONTHA
11 CONTHAeT/fiRANT NO
Grant S 802 90 8
12 SPO**SO«INC AGfcNCv
AMD AOOSfSS
EPA, Office of Research and Development
Industrial Environmental Reseai ch Laboratory
Research Triangle Park, NC 27711
ia. TYP* or mt.*Qmr AMD remioo covtmo
Final: 7/74-8/77
14. SPONSOftfNQ AGENCY CODE
EPA/800/13
Mail Drop 62, 919/541-2547.
project officer for this report is Ronald A. Venesia,
The report gives results of an examination of contaminant content and
selected treatment techniques for process condensate from seven different ammonia
plants. Field tests were performed and data collected on an in-plant steam stripping
column with vapor injection into the reformer furnace stack. Bench scale steam strip-
ping was studied on several different plant process condensates for comparative
purposes. Data for design of a commercial steam stripper were obtained on the bench
scale unit. Design conditions for the commercial unit were given. Four different
methods of treating the stripper overhead were compared. The results indicate that
stripping the process condensate and injecting the vapor into the reformer stack
offers a viable control technology for reducing the amount of ammonia and methanol
discharged to the environment.
KEY WORDS AND DOCUMENT ANALYSIS
OISCHIPTORS
Pollution
Ammonia
Industrial Processes
Condensates
Treatment
Steam
Stripping (Distil-
lation)
Carblnols
b IDiNTIFIERS/Off N ENDED TERMS
Pollution Control
Stationary Sources
Steam Stripping
Reformer Furnace
Methanols
F»sM/Gf«mp
1SB
07B
13H
07D
07A
07C
1 JISTHIBUTION STATEMENT
Unlimited
19 SECURITY CLASS (1 tut Krpottj
Unclassified
21 NO Of FACES
*$*
20 secu*it v ci ASS
Unclassified
33
IZKM it-7J)
-------
EPA-600/2-77-200
September 1977
TREATMENT
OF AMMONIA PLANT
PROCESS CONDENSATE EFFLUENT
by
C.J. Romero, F. Yocum,
J.H. Mayes and D.A. Brown
(Gulf South Research Institute)
Louisiana Chemical Association
251 Florida Street
300 Taylor Building
Baton Rouger Louisiana 70801
Grant No. S802908
Program Element No. 1BB610
EPA Project Officer; Ronald A. Venezia
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
CONTESTS
Figures. , v
Tables vii
1. Introduction,......,,,.,.,...,,..,.,.».,....»»...«.,,„...«....,,,.1
The nitrogen cycle related to surface water.... ,...l
2. Sunreary and Coneiuslona .. 5
Characterizations of ammonia plant process condensate........5
Bench-scale data. ,.,<•>
C0t*«rcia.l unit........................ ......... — .... .10
Cone lus ions ............14
3, Background......,..»,.,...... ..,.,...,,........,...15
U.S. aaoonia production. .*........ 15
Amentia plant waatewaters 15
Developing new technology for renewal of asraonia
front process condensate with subsequent recycle ......19
4. State-of-the-Art 21
Mlerobial nitrification and denitrification, ,.21
Selected ion exchange ,.....,. .23
Chlorlnatlon-dechlorination .24
Anmonia stripping. ,24
Reverse osmosis 25
Effect of various treatment processes on removal of
nltrogsn compounds ......,.,.,,.....,,..,....,,,...25
j. Characterization of Anmonla Plant Process Condensate. .26
Contaminant identification., 26
Trace metal analyses of the process effluent. ..27
6. Development of Stripping Data fron Laboratory and Bench-Scale
Data , 29
Introduction. 29
Bench-scale stream stripping of process condensate,.........29
?. Design of Comae re la i Anasonia-Methanol Steals Stripper 42
8. Disposition of the Stripper Tower Overhead......................,50
Introduction. 50
Direct discharge to the atmosphere.,,..... 50
Reinjection into the primary furnace Inlet 50
Injection of stripper overhead vapor into the furnace stack.56
Precipitation of the awnonia with magnesium phosphate and
biotreatnent of the nethanol contaminated wash water...... 56
lii
-------
Contents (continued)
9. Evaluation of Commercial Steam Stripper with Overhead
Injection into the Furance Stack 59
Introduction 59
Commercial stripper processing conditions 59
10. Economic Comparisons of Selected Treatment Schemes for Removal
of Ammonia from Process Condensate 69
Introduction 69
Process characterization schemes for economic evaluation....70
Summary of economic evaluation . .82
References 84
iv
-------
FIGURES
Number Page
1 The nitrogen cycle. ,, 2
2 Mass balance equations with reflux 8
3 Ammonia plant locations ......16
4 General process flow diagram of a typical ammonia plant............17
5 Organization chart 20
6 Pilot steam stripper located at GSEI, New Orleans 30
7 Model of pilot stream stripper (GSRI, New Orleans)..... 31
8 Diagram of apparatus, to gather vapor—liquid data. 39
9 Equilibrium curve for ammonia/methanol wastewater system...........41
10 Steam consumption vs. water content in overhead and percent feed
taken as overhead . .44
11 Pressure drop vs. tower diameter ......45
12 McCabe-Thiele method for theoretical stages 47
13 Packing height required for overhead water content 49
14 Stripper overhead to primary reformer 52
15 Percent increase in heat required to maintain reformer temperature
vs. water content in stripper overhead 53
16 Steam temperature vs. water content in stripper overhead 54
17 Gibb's free energy for ammonia and methanol reactions at furnace
stack temperatures .58
18 Plant schematic showing location of five sample points for
test runs ..60
19. Ammonia/methanol sample train for stripper overhead analysis 61
-------
Figures (continued)
Number Page
20 Ammonia/methanol sampling train for stack analysis 63
21 Atmospheric steam stripper discharge via primary reformer stack....71
22 Reinjection of steam stripped process condensate into primary
reformer via steam injection 73
23 Vanadium pentoxide catalyst absorption 76
24 Magnesium ammonia phosphate process 80
vi
-------
TABLES
Number Page
1 Methanol, Ammonia, and Carbon Dioxide Concentrations 6
2 Metal analyses 7
3 Economic Evaluation of Treatment Schemes 12
4 Component and Material Balances for Commercial Unit 112
5 Furnace Stack Analysis .'....... 13
6 Ammonia and Methanol Removal via Furnace Stack Injection .....13
" V . '
7 Contaminants in the Process Condensate from a 907 m. Ton/Day
Amonia Plant , 18
8 Plant Treatment of Process Condensrate. — 21
9 Effect of Various Treatment Processes on Nitrogen Removal 25
10 Methanol, Ammonia, and Carbon Dioxide Concentrations 26
11 Ratio of Ammonia-to-methanol-to-Carbon Dioxide 27
12 Metal Analyses 27
13 Preliminary Results of Heavy Metal Analyses on Grab Samples 28
14 Analyses for Methanol-Ammonia-Carbon Dioxide Acquired by Steam
Stripping in Bench Scale Unit 33
15 Mass Balances around Pilot Steam Stripper 37
16 Vapor and Liquid Equilibrium Data (Process Condensate from
Company 200) 40
17 Vapor and Liquid Equilibrium Data (Process Condensate from
Company 500 and Company 600) 40
18 Process Condensate Assumed for Column Design 42
19 Gibb' s Free Energy Calculations 57
vii
-------
Tables (continued)
Number Page
20 Average Chemical Analysis for All Runs on the Steam Stripper 62
21 Field Data on Process Condensate Stripper and Stack Analysis »64
22 Field Data 65
23 Production Unit Average Process Conditions 66
24 Source of Contaminants in Furnace Exhaust Stacks 66
25 Total Component Analysis of the Furnace Fuel Gas 66
26 Stack GAs Analysis Prior to Stripper Overhead Injection 67
27 Theoretical Conversion of Ammonia in Stripper Overhead to NOx 67
28 Average Stack Emission Values with Stripper Overhead Injection....68
29 Economic Evaluation of Various Processes ...82
30 Process Cost Ratios and Cost per Liter of Influent 83
viit
-------
SECTION 1
INTRODUCTION
Compounds containing the element nitrogen are becoming increasingly
important in wastewater management programs because of the many effects that
nitrogenous materials iLn wastewater effluent can have on the environment.
In its various forms, nitrogen can deplete dissolved oxygen levels in receiv-
ing waters, stimulate aquatic growth, exhibit toxicity toward aquatic life,
affect chlorine disinfection efficiency, present a public health hazard, and
affect the suitability of wastewater for reuse. Biological and chemical proc-
esses which occur in wastewater treatment plants and in the natural environment
can change the chemical form in'which nitrogen exists. Such changes may elimi-
nate one deleterious effect of nitrogen while producing, or leaving unchanged,
another effect. It is important, therefore, to review the chemistry of
nitrogen and the effects that the various resulting compounds can have on the
environment prior to the detailed discussion of the results of this grant
program.
The relationship among the various nitrogen compounds and the transforma-
tions which can occur may be presented schematically in a diagram known as the
nitrogen cycle, which is illustrated in Figure 1. The atmosphere serves as
the ultimate reservoir of nitrogen gas. From this reservoir, nitrogen is
removed naturally by electrical discharge and artificially by chemical manu-
facturing. The nitrogen gas is returned to the atmosphere by the action of
denitrifying organisms. In the fixed state, nitrogen can undergo the various
reactions shown in the nitrogen cycle diagram. The aspects of particular
importance to the grant program and its effect on surface waters are discussed
in detail later in this report.
THE NITROGEN CYCLE RELATED TO SURFACE WATER
Since the presence of nitrogen is essential to aquatic and marine life in
certain regulated amounts, there is a balanced cycle of its presence within
surface waters. Nitrogen may"be added to this system any of several ways:
Natural Sources
1. Atmospheric nitrogen fixation solution and dispersion by rain.
2. Atmospheric nitrogen fixation and dispersion through contact and
subsequent fallout.
3. Atmospheric nitrogen fixation by algae and bacterial species.
4. Presence in subsurface ground water, surf ace entrance and subsequent
runoff.
-------
JT.C.OfUy
MATTER
ORGANIC
N
ATMOSPHERIC
NITROGEN
ANIMAL
PROTEIN
ORGANIC
N
HJBjrjW -t»MMUi j,« vrthiM. c
WT^ (FIXATION)
PLANT
PROTEIN
ORGANIC
N
figure 1. The nitrogen cycle.
-------
Man-Caused Sources
1- Industrial wastewaters contaminated with various compounds of
nitrogen.
2. Agricultural runoff from land containing previously applied
nitrogenous fertilizers.
Natural Sources
Ammonification, nitrification, assimilation, and dentrification can occur
within the aquatic environment. Ammonifications of organic matter are carried
out by microorganisms. The ammonium thus formed can be assimilated by algae
and aquatic plants, and the resultant growths can create water quality problems.
Nitrification of ammonium can occur with a resulting depletion of the
dissolved oxygen content of the water. To oxidze 1.0 mg/1 of ammonia-nitrogen,
4.6 mg/1 oxygen is required. Denitrification produces nitrogen gas which
may escape to the atmosphere. Because anoxic conditions are required, the
oxygen-deficient hypolimnion (or lower layer) of lakes and the sediment layer
of streams and lakes are important zones of denitrification action.
Man-Caused Sources of Nitrogen in Waters
The activities of man have increased naturally-occurring quantities of
nitrogenous compounds in the aquatic environment. These sources have been
principally (1) fertilization of agricultural land, (2) combustion of fossil
fuels, (3) wastewater from fertilizer manufacturing facilities, (4) wastewaters
from other organic-based production facilities, and (5) other sources such
as livestock feed lots, poultry and egg production. These manmade sources
can affect the environment through biostimulation. of surface water, toxic
contributions to surface waters, and contamination of drinking water.
Biostimulation of Surface Waters—
A major problem in the field of water pollution is eutrophication,
excessive plant growth and/or algae "blooms" resulting from over-fertilization
of rivers, lakes, and estuaries. Results of eutrophication include deterio-
ration in the appearance of waters, odor problems from decomposing algae, and
lower dissolved oxygen levels which can adversely affect fish life. Eutrophi-
cation is of most concern in lakes because nutrients which enter tend to be
recycled within the lake and build up over a period of time.
Toxic Contributions to Surface Waters—
The principal toxicity problem is from ammonia in the molecular form
(NH ) which can adversely affect fish life in receiving waters. A slight
increase in pH may cause a great increase in toxicity as the ammonium ion
(NH.) is transformed to ammonia in accordance with the following equation.
Nfll" + OH~ > NH, + H00
4 < 3 2
Factors which may increase ammonia toxicity at a given pH are: greater
concentrations of dissolved oxygen and carbon dioxide; elevated temperatures;
-------
and bicarbonate alkalinity. Reported levels at which acute toxicity is
detectable have ranged from 0.01 mg/1 to over 2.0 mg/1 of molecular ammonia-
nitrogen.
Public Health Considerations—
When chlorine, in the form of chlorine gas or hypochlorite salt, is added
to wastewater containing ammonium, chloramines, which are less effective
disinfectants, are formed. The major reactions are as follows:
NHff + HOC1 »• NH2C1 (monochloramine)-H20 + H+
NH2C1 + HOC1 >• NHC12 (dichloramine) + H20
NHC12 + HOC1 >• NC13 (nitrogen trichloride)-H20
Only after the addition of large quantities of chlorine does free avail-
able chlorine exist. If the effluent ammonia-nitrogen concentration were
20 mg/1, about 200 mg/1 of chlroine would be required to complete the reactions
with ammonium and organic compounds. Only rarely is this level of chlorine
addition ("breakpoint" chlorination) used in wastewater treatment. Obviously,
ammonia would have to be present in large quantities in any industrial effluent
to cause any serious disinfectant problem through chlorine losses.
-------
SECTION 2
SUMMARY AND CONCLUSIONS
Based on a survey of the fertilizer industry, the guidelines
division of the Environmental Protection Agency promulgated discharge standards
for ammonia producers on April 8, 1974. These effluent limitation guidelines
set the amount of ammonia which could be discharged from a plant process
condensate as 50 kg/day. This represented an average 50 mg/1 concentration in
the process effluent from a 907 m. ton/day ammonia plant discharging an aver-
age 757 1/min from the process area.
Recognizing that the problem of meeting this limitation existed
in plants whose only product was ammonia, Louisiana ammonia producers sought
to develop the necessary technology to meet the guidelines. Since the only
product for the majority of the ammonia producers in Louisiana is anhydrous
ammonia, a wide base for program development was established. Through their
industrial membership in the Louisiana Chemical Association (LCA), a joint
research grant involving the EPA and LCA was established. The participants
were as follows:
1. Louisiana State Science Foundation
2. Environmental Protection Agency Industrial and Environmental
Research Laboratory
3. Participating companies through the Louisiana Chemical Association
a. Air Products and Chemicals, Inc.
b. American Cyanamide Company
c. Borden Chemical Company
d. C.F. Industries, Inc.
e. W.R. Grace & Company
f. I.M.C. Corporation
g. Monsanto Company
h. Olin Corporation
CHARACTERIZATIONS OF AMMONIA PLANT PROCESS CONDENSATE
Contaminant Identification
Seven different plant process condensate sources were represented by the
production ;operations of the eight participating industries. Each representative
stream was analyzed for the ammonia, methanol, and carbon dioxide components.
The analytical results of all samples for each stream were averaged and are
presented in Table 1. For each process involved, the catalyst age and sever-
-------
TABLE 1. METHANOL, AMMONIA, AND CARBON DIOXIDE CONCENTRATIONS
Stream
Number
100
200
300
400
500
600
Average
Ammonia
(mg/D
800
1,041
858
1,015
825
700
873
Methanol
(mg/1)
459
362
618
972
559
172
524
Carbon Dioxide
(mg/1)
1,137
2,470
2,559
2,789
1,258
642
1,809
ity of operation (amount of condensate or excess steam) was noted.
Metal analysis was performed on samples for each stream to determine cop-
per, nickel, iron, zinc and chromium concentrations. Such metal contaminants
could affect the process for one of the proposed treatment methods calling for
reinjection of the stripper overhead in the reformer feed stream. Samples
were obtained in propylene containers for laboratory purposes. Analyses of
the samples indicated that no trace metal contaminant problem existed if
reinjection of the stripped condensate back into the process was contemplated.
Metal analyses results are reported in Table 2.
BENCH-SCALE DATA
A review of stream characterization data indicated that stream 700 was
too low in ammonia, methanol, and carbon dioxide concentration to be classified
as a representative sample stream. Therefore, values for stream 700 are
deleted. Table 1 shows concentrations of the three contaminants from repre-
sentative streams, the average values being ammonia, 873 mg/1; methanol, 524
mg/1; and carbon dioxide, 1809 mg/1. Four representative streams were selected
for bench-scale steam stripping tests; 100, 200, 300 and 400. A total of 61
runs was made utilizing the process condensates from these four production
sources.
Process Effluent from Company IOC)
Initially the pilot steam stripper contained a packed bed depth of 2.2 m
using 6 mm (1/4 inch) Rasching rings. Results of runs 1-7 indicated that
steam stripping the process condensate was feasible but did not produce desired
ammonia and methanol concentrations in the effluent bottoms. Bottoms from
runs 5, 6, and 7 of pilot steam stripper were collected and stored to deter-
mine if additional packing height would be required. These collected bottoms
were reprocessed through the pilot steam stripper as feed (for runs 8, 9, and
10) to determine if further separation of ammonia, methanol and carbon dioxide
would take place. Similarly, run 10 overheads were collected and used as feed
for run 11. Results indicated that additional packing height would reduce the
ammonia, methanol, and carbon dixoide concentration for a single once-through
run in the pilot steam stripper.
-------
TABLE 2. METAL ANALYSES
Sample I.D.
Cu
mg/1
Ni
mg/1
Fe
mg/1
Zn
mg/1
Cr
mg/1
Company No. 100
Feed <0.5 <0.5 <0.3 1.5 <0.5
Stripper Overhead <0.5 <0.5 <0.3 1.86 <0.5
Company No. 200
Feed " <0.5 <0.5 <0.3 <0.1 <0.5
Stripper Overhead <0.5 <0.5 <0.3 <0.1 <0.5
Stripper Bottoms <0.5 <0.5 <0.3 <0.1 <0.5
Company No. 400
Feed ^<0.02 <0.2 <0.1 <0.02 <0.2
Company No. 500
Feed <0.02
-------
Feed
Overhead
pm
pa
Key:
P = Overhead
F = Feed
R = Reflux
B = Bottoms
feed
overhead
bottoms
methanol
ammonia
Figure 2. Mass balance equations with reflux.
-------
Runs 15-18 were made on the bench stripper without the addition of reflux.
Contaminants in the stripper bottoms from these tests were from 1 to 85 mg/1
methanol and from 50 to 160 mg/1 ammonia. The reduction of these two contam-
inants in the stripper bottoms was dependent on the percent overhead-to-feed
ratio. For example, if the feed rate of process condensate to the steam
stripper was 100 kg/mln, then 10 kg of condensed overhead would represent 10%,
and 5 kg/min condensed overhead would represent a 5% overhead rate.
Test runs 19-28 were conducted with a portion of the overhead refluxed.
These tests did not give satisfactory results because the reflux pump did not
function properly.
Runs 29 and 31 were without reflux, while runs 30 and 32 were with reflux.
In these runs, ammonia in the stripper bottoms was reduced to very low limits
(<5 mg/1). The percent overhead-to-feed ratio used to achieve these results
was quite high (>10%). For example, a typical 907 m. ton/day plant generally
produces about 45,000 kg/hr. If this 45,000 kg/hr is fed to a steam stripper,
enough steam must be added to produce the desired overhead-to-feed ratio. If
a 10% overhead rate is needed for this separation, then enough steam is added
to vaporize 4,500 kg/hr (total) of water, ammonia, methanol and carbon dioxide.
If these stripped overheads are reinjected into the primary reformer furnance,
the amount of overhead generated (for reinjection) from the steam stripper
affects the amount of energy required for condensation, pressurizing, re-
evaporation, and Injection into the primary reformer furnace for reclaiming of
ammonia and methanol. In order to decrease this overueau rate and still
achieve the desired separation (<20 mg/1 ammonia and methanol) in the strip-
ped bottoms, either packing height or refluxing rate has to be increased.
Process Effluent from Company 300
Test runs 33, 34, 36, 38, and 39 were without reflux while runs 35 and 37
were with reflux. Runs 33-39 were performed with the overhead less than 5% of
the feed rate. For test runs without reflux, the concentrations of the contam-
inants in the stripper bottoms ranged from 10 to 35 mg/1 for methanol and 25
to 53 mg/1 for ammonia. For test runs with re*.lux, the concentrations of the
contaminants in the stripper bottoms ranged from 1 to 7 mg/1 for methanol and
29 to 39 mg/1 for ammonia.
Process Effluent from Company 400
Test runs 40, 42, 44, and 46 were without reflux, and runs 41, 43, 45,
and 47 with reflux. The amount of methanol in the stripper bottoms was influ-
enced significantly by the addition of reflux. Methanol in the bottoms was
119-129 mg/1 and ammonia, 73-145 mg/1, for runs made without any reflux. For
runs with reflux, the methanol varied between 30 and 32 mg/1 and the ammonia
between 62 and 100 mg/1. With the addition of reflux, the reduction in con-
centrations averaged 25.7% for ammonia and 74.9% for methanol.
Process Effluent Condensate from Company 200
An effort was made in the final runs to perfect the stripping technique
for operating the pilot steam stripping column. Test runs 48, 49, 51, 53, 55,
9
-------
57, 59 and 60 were performed without reflux, while runs 50, 52, 54, 56, 58,
and 61 were performed with the addition of reflux. An effort was made to keep
the reflux ratio (R/P as shown in Figure 2) at approximately 1:1. Residual
methanol concentration in the stripper bottom varied between 4 and 100 mg/1
without reflux and between 1 and 53 mg/1 with reflux. Residual ammonia concen-
tration in the stripper bottom varied between 14 and 81 mg/1 without reflux
and between 12 and 55 mg/1 with reflux.
It was determined that a 5 to 6% overhead rate was optimum for the pilot
scale equipment and process conditions. Under these conditions, the methanol
and ammonia concentrations in the stripper bottoms would be <15 mg/1 and <20
mg/1, respectively.
Pilot (Bench) Steam Stripper Mass Balance
Mass balances were determined from the data collected during the operation
of the bench steam stripper in order to validate the steam stripping data.
These mass balances were determined for methanol, ammonia, and total mass flow
rates. Standard deviation (sum of squares) was calculated for this data
revealing + 9.9% for methanol, + 12.21 for ammonia and + 7.0% for a total mass
balance around the stripper column.
VaporLiquid Data
Vapor-liquid equilibrium data were determined for three of the sources of
process condensate for use in comparative design calculation. The source
stream and its vapor-liquid equilibrium can be represented by the following
equations:
Stream 200; y = 147(x) 110
Stream 500; y = 123.5(x) 456
Stream 600; y - 232(x) 100
where y is the mole fraction NH~ in the vapor and x is the mole fraction NIL
in the liquid. •* J
COMMERCIAL UNIT
Design of Commerical Ammonia-Methanol Steam Stripper
Design calculations were made for a steam stripper column using pilot
plant vapor-liquid equilibrium data. Vapor-liquid data would help to estab-
lish the necessary depth of packing to reduce the ammonia and methanol content
to the specified level.
In the design calculations, both stripper columns using reflux and steam
stripping without reflux were considered. Economic considerations would
determine use and extent of reflux, which in turn influence packed bed
depth and steam load to the stripper column. As stated eariler, refluxing
theoretically increases the height of packing, depending on the reflux ratio
used. However, refluxing uses more steam input because the portion of con-
densed overhead sent back to the column must be reheated and vaporized. Since
10
-------
reflux requires more steam input, and since the diameter of the stripping
column is determined by the liquid (feed)-vapor (steam) load, a high reflux
could increase stripper column diameter. The following design conditions were
used as the basis for a commercial size steam stripper:
1. 907 m. ton/day ammonia production.
2. 757 1/min process condensate wastewater stream with the following
concentration.
a. Ammonia concentration of 1000 mg/1.
b. Methanol concentration of 750 mg/1.
3. Removals of 98% of the ammonia and 99% of the methanol showing the
following concentration in the bottoms:
a. Maximum ammonia concentration of 20 mg/1.
b. Maximum methanol concentration of 5 mg/1.
Stream Stripper Overhead Disposal
There were four potential options investigated for disposal of the steam
stripper overheads:
1. Reinjection
2. Precipitation with magnesium phosphate
3. Adsorption by vanadium pentoxide
4. Injection of the stripper overhead into the reformer stack
Several large ammonia producers have installed process condensate steam
strippers which are discharging to the atmosphere. Analysis of the stripper
bottoms indicates that this operation does reduce the ammonia in the stripper
bottoms to the desired level. The net result, however, is that the contami-
nants have been removed from the water and redistributed into the surrounding
atmosphere.
Economic Comparison of Treatment Schemes
Four processes were economically evaluated (January 1977 figures) for
their cost effectiveness in reducing the ammonia and methanol present in the
process condensate (Table 3). Ten-year straight-line depreciation with 8%
interest charges were utilized for comparison. The cost-benefit ratio of
stack injection outweighs other systems. The ammonia and methanol contained
in the overheads from the stripper are reduced by 59.3 and 74.7 percent,
respectively, with an increase of NOx in the final stack emission of 95.3 kg/hr.
11
-------
TABLE 3. ECONOMIC EVALUATION OF TREATMENT SCHEMES
Atmospheric
Steam Stripping
With With
Stack Reformer
Inj ec t ion Inj ec t ion
Vandium Magnesium
Pentoxide Phosphate
Adsorption Precipitation
Variable Cost
Fixed Cost
Total
Recovered Credit
Total Annual Cost
Cost per Liter
Cost/m. ton NH
$368,000
62,000
430,000
None
$430,000
$0.0012
$1.49
$814,500
134,400
984,900
61,000
$887,900
$0.0026
$3.20
$ 890,500
348,400
1,238,900
61,000
$1,177,900
$0.003
$3.94
$1,380,000
325,400
1,706,300
288,000
$1,418,300
$0.004
$4.41
Evaluation of Commercial Steam
Stripper with
Overhead Injection
into the Reformer
Furnace Stack
The commercial column was designed with 9.1 meters of stripping section
packed with Pall rings. The stripper's overhead line was injected into the
furnace stack approximately 8 meters above ground level. Total stack height
measured 32 meters. Component and material balances of the 65 tests are
listed in Table 4.
TABLE 4. COMPONENT AND MATERIAL BALANCES FOR COMMERCIAL UNIT
Feed
Overhead
Bottoms
Steam
Total
Percent Reduction
Ammonia
mg/1 kg/hr
487 39.2
4,750 37.9
7 1.3
— —
86.4
96.8
Methanol
mg/1 kg/hr
262 21.1
2,610 20.8
3.4 0.3
_ _
46.5
98.8
Flow
kg/hr
30,500
7,980
81,200
8,680
Furnace Stack Analysis—
According to the Gibb's free energy calculation, the decomposition of
ammonia to nitrous oxide in the furnace stack is highly probable in the
presence of oxygen. If 100% of the ammonia (37.9 kg/hr NH_) out of the steam
stripper were converted to nitrogen dioxide (N0_) in the primary reformer
furnance stack by the following equation:
12
-------
then the 37.9 kg/hr or NH. would be converted into 102.6 kg/hr (260.7 ppm) of
NO^ (Table 5). However, 15.5 kg/hr of ammonia was detected at the primary
reformer stack discharge. A reduction of 22,4 kg/hr or a decomposition of
some 59.2% of ammonia was observed. Also found in the primary furnance stack
discharge outlet was 5.3 kg/hr of methanol, a reduction of 15.6 kg/hr or a
decomposition of 74.7% of methanol. However, an increase from 67.6 kg/hr (172
ppm) to 95.3 kg/hr (242 ppm) of nitrogen oxide was observed. This increase of
27.7 kg/hr (70 mg/1) or 40.9% in nitrous oxide can be related to the ammonia
reduction observed.
If the 22.4 kg/hr decomposition of ammonia were converted into nitrous
oxide, this would represent an increase of 60.7 kg/hr of NO™. Since only 27.7
kg/hr increase of N0? was found, some of the ammonia must have decomposed into
N and H
TABLE 5. FURNACE STACK ANALYSIS
Component
Ammonia
Methanol
NOx
Furnace
Outlet
ppm kg/hr
0.0
0.0
172
0.0
0.0
67.2
Stripper
Overhead
ppm kg/hr
4750
2610
0.0
37.9
22.4
0.0
Primary Reformer
Discharge
ppm kg/hr
39.3
13.4
242
15.5
5.3 ,
95.3
Effectiveness of Ammonia and Methanol Removal Via Furnace Stack Injection—
Measurements of the concentration of methanol and ammonia exiting the
furnace stack were compared to theoretical calculated values for the stripper
overhead being discharged directly. Ground level concentrations for these two
cases were also calculated. Results are shown iii Table 6.
TABLE 6. AMMONIA AND METHANOL REMOVAL VIA FURNACE STACK INJECTION
Emissions
Ib/hr g/sec
Maximum Downwind
Ground Level-Concentrations
)
Actual Measurements
Ammonia
Methanol
34.1
11.6
4.28
1.46
12.8
4.4
If No Decomposition
Ammonia
Methanol
83.6
45.9
10.56
5.78
31.6
17.3
13
-------
CONCLUSIONS
Several conclusions can be drawn from the laboratory and plant evaluations
of the removal of ammonia from process condensate via steam stripping:
1. Steam stripping is a viable process for the reduction of ammonia
and methanol in ammonia process condensate streams and will achieve
established EPA guidelines.
2. Injecting the overheads from the steam stripper into the reformer
furnace stack can effectively reduce the amount of ammonia and
methanol discharge to the atmosphere.
3. In the commercial unit evaluated, it is possible to strip the con-
densate and recycle the bottoms to the boiler feedwater system.
The bottoms could be used for cooling tower make-up without further
treatment, depending on final ammonia and methanol concentrations.
4. Pilot plant data on steam stripping of ammonia plant process con-
densate compared favorably with data from a full-scale commercial
unit tested in an ammonia plant.
5. Trace metal levels in the condensate will not present a problem
in the recycle of stripped bottoms to the boiler feed water treat-
ment system. Trace metals would not present any problem If the
overheads from the steam stripper were recycled to the primary reformer
furnace in the ammonia process.
6. The concentration of ammonia in the process condensate varies with
the age of the primary reformer catalyst and severity of process
conditions.
7. Reinjection of process steam stripper overheads into the primary
reformer furnace would require a stripper bed with reflux for
concentrating the overhead. A preheater would be required prior to
injection in the primary furnace and/or a large heat increase in
the primary furnace itself.
8. Comparison of alternate treatment schemes for the atmospheric
reduction of ammonia and methanol showed that venting the steam
stripper overheads via the reformer furnace stack was the least
costly.
14
-------
SECTION 3
BACKGROUND
With the creation of the Environmental Protection Agency and its mandate
to set effluent guidelines for various industries, many ammonia producers in
Louisiana began to investigate means of reducing ammonia and organics from
the process condensate they were discharging. As late as 1973, efforts had
been made by the Guidelines Division of the Environmental Protection Agency
to set permissible levels of ammonia which could be discharged into receiving
waters.
After making an industry survey, the Guidelines Division initially set
the amount of ammonia which could be discharged from a plant process conden-
sate as the equivalent of 50 kg/day, which represented an average concentra-
tion of 50 ppm in the process effluent from a 907 m. ton/day ammonia plant
discharging an average of 757 1/min from the process area. Data for setting
this maximum ammonia concentration were based on assuming removals over this
maximum in adjoining process equipment.
U.S. AMMONIA PRODUCTION
As of 1975, approximately 16-million m, tons of synthetic ammonia were
produced annually in the United States. This product was manufactured in 30
states in some 88 plants. Figure 3 shows the location of these plants. By
1977, an additional 2-million plus m. tons wi- T->e added to the manufacturing
capability of the industry.
In general, ammonia production units are located in areas where there
is abundant natural gas. This material forms the basic raw material for
cracking and furnishes fuel for the manufacturing process. Because of the
availability of natural gas in Louisiana and Texas, these two states produce
over 50 percent of the anhydrous ammonia in the United States. Louisiana
alone produces nearly 30 percent.
AMMONIA PLANT WASTEWATERS
The technology of cracking methane for hydrogen production and combining
with atmospheric nitrogen to manufacture ammoniahas advanced significantly
in the past 10 years. Over 98 percent of the ammonia production in the
United States is done by the catalytic steam reforming of natural gas (see
schematic, Figure 4). Some wastewater is an unavoidable product of the
manufacturing process of ammonia via natural gas cracking. There are several
sources of the effluent from an ammonia production facility:
15 •-''•-.. "
-------
Legend
O Closed or idle plant
0 Operating plant (capacity under 200,000 TPY)
Operating plant (capacity over 200,000 TPY)
/\ New plant (capacity over 200,000 TPY)
Figure 3. Ammonia plant locations.
-------
Natural
Gas
Feedstock
1 Desulfurization
Fuel
1
Primary
Reformer
Product of
Air
Overhead
Process
Condensate
Steam
Stripper
n
Steam
Bottoms
I
Secondary
Reformer
High Temp. Shift
Low Temp. Shift
CO 2
Absorber
Methanation
Ammonia
Synthesis
I
NIL
CO,
I
co2
Stripper
Steam
Purge Gas
Figure 4. General process flow diagram of a typical ammonia plant.
17
-------
1. Process condensate as a result of the cracking process.
2. Pump gland and sealant water. ;
3. Process area washdowns and leaks.
4. Cooling tower blowdown, where applicable. ;
5. Boiler blowdown, where applicable. *
6. Raw water clarifier underflow, where applicable.
The wastewater most highly burdened with ammonia contaminant is the process
condensate. Typical analyses from a 907 m. ton-per-day production facility
are shown in Table 7.
TABLE 7. CONTAMINANTS IN THE PROCESS CONDENSATE FROM A
907 M. TON/DAY AMMONIA PLANT
Component
Concentration
(mg/1)
Output
NH
Organics
CO
COD
Process
, mainly methanol
Condensate
600
200
200
200
- 1,000
- 1,000
- 2,800
- 1,200 -
653
218
218
600
- 1,088 kg/day
- 1,088 kg/day
- 3,039 kg/day
- 757 1/min
The major contaminants in the process condensate are methanol, ammonia
However, with respect to the effluents and emis-
is a process
contaminant only if the wastewater is to be reclaimed an3 used in further
process areas.
and carbon dioxide (CO-).
sions, the only pollutants are methanol and ammonia. CO-
The amount of process condensate is approximately 1150 liters/m. ton
of ammonia produced. Total ammonia production in the United States will be
approximately 18,144,000 m. ton/yr in 1980, corresponding to approximately
21.2 billion liters/yr of process condensate. Based on an average ammonia
concentration of 800 mg/1, this represents the equivalent of about 16,950 m.
tons per year. Within the State of Louisiana, there will be approximately
5,440,000 m. tons of anhydrous ammonia produced in 1978. This represents
about 6.4-billion liters of process condensate containing approximately
5,400 m. tons of ammonia.
The possibility of land disposal instead of further treatment has been
considered for this ammonia process condensate. Shipping and handling a
product which is 95% water and 4% ammonia, methanol and other contaminants
over long distance would be uneconomical. The possibility of using this
condensate for crop irrigation is remote for Louisiana. However, if an
ammonia plant were located in a^ area where this process water could be used
for irrigation.and if 11.21 g/m of ammonia per acre per season were needed,
116.9 liters/m of process wastewater would be required. This would corre-
spond to a square meter of land flooded to a depth of 12 cm with process
water.
18
-------
DEVELOPING NEW TECHNOLOGY FOR REMOVAL OF AMMONIA FROM PROCESS CONDENSATE WITH
SUBSEQUENT RECYCLE
The only product from the majority of ammonia producers in Louisiana is
anhydrous ammonia. Recognizing the need to develop technology which would
be compatible with all EPA regulations, the ammonia producers, through their
industrial membership in the Louisiana Chemical Association, participated in
an EPA grant which was jointly supported by the following entities (see
organization chart, Figure 5):
1. Louisiana State Science Foundation
2. U.S. Environmental Protection Agency, Industrial and Environmental
Research Laboratory
3. Participating Companies through the Louisiana Chemical Association
a. Air Products and Chemicals, Inc.
b. American Cyanamid Company
c. Borden Chemical Company
d. C.F. Industries, Inc.
e. W.R. Grace & Company
f. IMC Corporation
g. Monsanto Company
h. Olin Corporation
The principal objective of the program was to establish technology to
remove the environmental pollutants ammonia and methanol and the process
contaminant carbon dioxide. Anticipating future effluent requirements, this
program was designed to establish the technology to lower,discharge of
ammonia to the environment well below the EPA guideline of 50 kg/day and to
minimize the discharge of methanol. The following program outline was
developed.
1. Review previously developed information to evaluate possible
technology transfer.
2. Evaluate steam stripping as a viable process.
3. Evaluate reflux of stripper overhead to concentrate ammonia.
4. Investigate disposal of ammonia concentrate.
a. Consider reinjection of concentrated stripper overhead into
cracking furnace feed and the effect of this reinjection
on increased furnace heat requirements.
b. Investigate injection of concentrated stripper overhead into
furnace exhaust stack and the effect of stack temperatures
on the decomposition of ammonia and methanol.
c. Study economics of, adsorbing ammonia on vanadium pentoxide
and subsequent recovery oxides.
5. Evaluate stripper bottoms as feed to:
a. Recycle to demineralize system.
b. Use directly in low pressure boiler.
cl Discharge into receiving waters.
d. Discharge in cooling tower make-up.
e. Recycle to water demanding process if available.
19
-------
Environmental Protection
Agency
Louisiana Chemical Association (LCA)
Grantee
Gulf South Research Institute (GSRI)
Subcontractor
LCA Ammonia
Producers/Advisory
Technical
Committee
Project Direction
LCA Henry A. Landrum
GSRI James H. Mayes
Laboratory Development
GSRI Project Engineer
Analytical
Support
Technician
Demonstration Unit
GSRI Project Engineer
Analytical Support
Engineering Support
Metallurgical Engineer
Support
LCA Joint Analytical and
Technical Support
Figure 5. Organization chart.
20
-------
SECTION 4
STATE-OF-THE-ART
Until several years ago, concentrations of nitrogen compounds in surface
waters had not presented any serious problems due to excessive biostimulation
(eutrophication). The microbial nitrification and denitrification of munici-
pal wastewaters in conventional treatment plants was the primary means of
keeping the nitrogen cycle in balance. Increased awareness of the impact of
nitrogen entering the environment has led to investigations of other means
of removal, particularly from more concentrated industrial sources, such
as the process condensate from large ammonia plants. Some methods investi-
gated were:
1. Microbial nitrification and denitrification.
2. Ion exchange.
3. Chlorination - dechlorination.
4. Ammonia plant process condensate steam stripping with air venting
of stripper overhead.
5. Reverse osmosis. *
None of these treatment systems offers an industry-wide solution to the
reduction of the contaminant level in the ammonia plant process condensate
effluent. Each will be discussed in turn to point out the limitations of
industrial plant applications.
MICROBIAL NITRIFICATION AND DENITRIFICATION
Several modern ammonia plants treat their process condensate effluent
with an aeration lagoon. Data from one such operation are shown in Table 8.
Biological treatment under these conditions gives excellent reduction of
methanol, but the ammonia is only partially reduced.
TABLE 8. PLANT TREATMENT OF PROCESS CONDENSATE
Process Condensate Bio-Pond
Bio-Pond Influent Effluent
Component (mg/1) (rag/1)
Ammonia 800-1100 100-650
COD 2200-2800 100-400
BOD ' 1600-2800 • 150-250
pH ' 8-9 8-8.5
21
-------
It has been the practice of many treatment plants to convert nitrogen,
in the form of ammonia, to nitrates through the process of nitrification(1).
This oxidation of ammonia is performed by a specific group of microorganisms
which have growth rates that are highly temperature-sensitive. In a study
by Bingham, et^ al.(2), a pilot unit trickling filter was evaluated under
laboratory conditions for the conversion of ammonia to nitrate. The results
of this study indicated that ammonia removal was dependent on hydraulic flow
rate, temperature, and inorganic carbon concentrations. Ammonia removal
through the trickling filter amounted to 20 to 40 percent even at the best
hydraulic loadings. Thus, even with good carbon conversion, the nitrifica-
tion of ammonia would take considerable residence time or a number of trickling
filters in series.
Under anaerobic conditions, microorganisms utilize chemically bound
oxygen for the final hydrogen acceptors. Therefore, in an anaerobic environ-
ment, the nitrate from the nitrification treatment system may be reduced or
denitrified to gaseous forms of nitrogen. This principle has been employed
in previous investigations(2) for the removal of nitrate nitrogen from
agricultural runoff. For proper denitrification, an organic carbon source
must be present so that the microbes can perform their normal metabolic
activities. Past investigations have indicated that methanol may be the
most economical source of supplemental carbon(1). Thus, the presence of
methanol in the ammonia plant process condensate effluent was initially
regarded as encouraging for the biological approach to effluent treatment.
Unfortunately, the concentration levels of the ammonia in the effluent were
so high that satisfactory treatment could not be achieved in a reasonable
length of time. Also, the addition of phosphates is necessary for the
reactions to occur.
Considerable work was done by Bingham, et al.(2) on biological denitri-
fication of effluent from an ammonia and ammonium nitrate plant at Farmers
Chemical Association in Harrison, ..Tennessee. The ammonia and nitric acid
plant effluent amounted to 2.7x10 liters per day and contained 100 mg/1
of ammonia nitrogen and 120 mg/1 of nitrate nitrogen. The ammonium nitrate
plant effluent was 3.79x10 liters per day and contained 2500 mg/1 ammonia
nitrogen and 10,000 mg/1 nitrate nitrogen. Based on the study by Bingham
for denitrification of this stream, the optimum COD-to-nitrate ratio was
3.2:1 (about 6:1 methanol:nitrogen). The necessary retention period was 25
to 30 days.
Eckenfelder(3) discusses the relationsip between residence time and
temperature for the nitrification reaction. If the reaction temperature
drops from 15° to 6°C, the residence time for the same level of nitrification
will approximately double. (Estimates are based on nitrogen reduction in
domestic sewage.)
According to data presented by Johnson(4), activated sludge plants show
removals of organic nitrogen ranging from 50 to 85 percent and a total
nitrogen removal of 16 to 75 percent. Johnson further states that the
removal of nitrogen is a function of the BOD-to-nitrogen removal ratio and
that an increase in the nitrogen content of an effluent would reduce nitrogen
removal.
22
-------
The often-used rule of thumb for domestic sewage is 15 mg/1 organic
nitrogen and 10 mg/1 ammonia nitrogen. Obviously, in most activated sludge
plants the organic nitrogen is biodegraded while the ammonia nitrogen remains
essentially unchanged. Microbial nitrification-denitrification systems for
ammonia contaminated process condensate effluents are limited by two additional
drawbacks: (1) For high concentrations of ammonia-nitrogen, the retention
times necessary to achieve realistic reductions are too great; the impounding
of areas is a waste of valuable land. (2) During winter operation (lower
temperatures), treatability levels would fall below acceptable standards.
SELECTED ION EXCHANGE
In other research, Bingham, et^ al. (5) concluded that ion exchange
offered the best solution to the ammonia and nitrate removal from the effluent
from the Harrison, Tennessee, plant. Supporting this conclusion was an
extensive investigation of various methods of reducing nitrogenous compounds
in the Farmers Chemical Association plant. The research and development
program was supported and financed in part by the United States Environmental
Protection Agency. The following processes were investigated:
1. Biological nitrification.
2. Biological denitrification.
3. Stripping of ammonia.
4. Precipitation of ammonia as magnesium ammonium phosphate.
5. Recovery of ammonium nitrate by reverse osmosis.
6. Recovery of ammonium nitrate by continuous ion exchange.
Bingham, et al.(5), summarized their conclusions regarding the above
treatment processes as follows:
1. Microbial nitrification of ammonia nitrogen in plant effluents
over laboratory and plant scale trickling filters was ineffi-?
cient (indicating inadequate residence time) and temperature
sensitive.
2. Biological denitrification of nitrate nitrogen in plant effluents
under laboratory and plant scale anaerobic conditions in stabili-
zation ponds also proved ineffective.
3. Air stripping of ammonia nitrogen under laboratory and plant-
scale conditions showed promise. (Stripped by-product was vented
to atmosphere.)
4. Precipitation of ammonia nitrogen as magnesium ammonia phosphate
showed promise if the treatment process could be integrated into
existing operations.
5. Ion exchange apparently would provide effluent water of adequate
quality for reuse or discharge.
To achieve success with the ion exchange system, the recovered by-
product, ammonium nitrate, would have to be concentrated and added to the
finished product of the plant. Potential features of the treatment system
included:
23
-------
1. Use of both nitric acid and ammonium hydroxide as regenerants on
cation and anion units since both these compounds are in-plant
products and thus are preferred regenerants.
2. Use of strong acid cation exchange resins and weak base macroreticu-
lar anion exchange resins.
3. High total dissolved solids as calcium carbonate.
Farmers Chemical Association had a plant producing ammonia .and nitric
acid which were subsequently converted to ammonium nitrate by evaporation
and drying. The ion exchange regeneration products could be charged to the
product evaporation units. The regenerant stream from the ion exchange beds
contained 85% water. A considerable capital investment was necessary to
concentrate this stream. For operation of the ion exchange unit, it was
necessary to utilize about twice the amount of regenerant chemicals (nitric
acid and ammonia) present in the effluent. Thus, the process is not economic-
ally attractive.
CHLORINATION-DECHLORINATION
In a study by Atkins and Schegner(6) sponsored by the EPA, the feasibility
of using chlorination followed by dechlorination with granular activated
carbon for the removal of ammonia nitrogen from effluent water was demonstrated.
This study was conducted on a domestic sewage effluent with an average
concentration of 300 mg/1 ammonia nitrogen. Several findings of this study
were:
1. The ammonia removal process tends to depress the pH in nonbuffered
systems and might necessitate adjustment of the final effluent.
2. The chloride content of the wastewater was increased from 193 mg/1
to 293 mg/1.
3. Dissolved oxygen levels of the final effluent were between 1 and 2
mg/1, necessitating re-aeration.
4. Complete removal of ammonia nitrogen from wastewater required a
chlorine-to-ammonia feed rate of 9-to-l.
5. If pretreatment is inadequate, considerable chlorine will be
consumed by other impurities in the water, increasing both chlorine
and activated carbon cost.
6. High ammonia nitrogen contaminated effluents would create excessive
chloride concentrations in the final effluent.
The chlorination-dechlorination process offers excellent treatment
possibilities for waste streams which have ammonia nitrogen concentrations
within the order of magnitude of domestic sewage (300 mg/1 average). Streams
contaminated with higher ammonia nitrogen concentrations (i.e., 1000 mg/1)
present economic and chloride contamination problems.
AMMONIA STRIPPING
Several investigators have studied the aeration of aqueous effluent for
the removal of ammonia(7-9 ). The important criteria appear to be pH, air-
24
-------
to-water ratio, and contact time. Stripping towers for the treatment of
process condensate have proven effective at several plant installations;
however, all of these systems are vented to the atmosphere. The best removal
in aeration columns, reported by Rohlich(7), was 92% at pH of 11, an air-to-
water ratio of 500, and a packed bed depth of 2.1 m. Less bed depth
led to serious reduction in efficiency and aeration rate; at a lower pH,
efficiency should be less due to the formation of ammonium ions. Gulp and
Selechta(8) have reported removals of ammonia up to 80% at pH of 9.3 and 98%
at pH of 10.8 at air-to-water ratios of 8QQ and a contact time of 0.5
min. Other than the steam-stripping proposed by Kellogg (10), ho research
has been reported on the stripping of an ammonia-methanol aqueous mix-
ture .
REVERSE OSMOSIS
It would be possible to develop a membrane to remove ammonia from
water. However, several factors would have to be evaluated: (1) the cost
effectiveness of the process, (2) the disposal of the resulting ammonia solu-
tion if only anhydrous ammonia is produced, and (3) the relationship of the
methanol present in the condensate to the reverse osmosis action. ±0 date,
no development program has been initiated to investigate the commercial
potential of removal of ammonia by membrane action. . ,.-
EFFECT OF VARIOUS TREATMENT PROCESSES ON REMOVAL OF NITROGEN COMPOUNDS
The effect of removal of nitrogen compounds by the previously discussed
process method is shown in Table 9. These values are averages from the
literature.
TABLE 9. EFFECT OF VARIOUS TREATMENT PROCESSES ON NITROGEN REMOVAL
Removal of
Total Nitrogen
Treatment Process (%)
Biotreatment 10 to 20
Reverse Osmosis 50 to 90
Dialysis 30 to 60
Breakthrouch Chlorination 80 to 90
Ion Exchange 80 to 95
Ammonia Stripping 80 to 90
25
-------
SECTION 5
CHARACTERIZATIONS OF AMMONIA PLANT PROCESS CONDENSATE
CONTAMINANT IDENTIFICATION
The participating companies represented seven different plant process
condensate sources. Each stream was analyzed for ammonia, methanol and
carbon dioxide. Results of several samples of each stream were averaged and
are presented in Table 10.
TABLE 10. METHANOL. AMMONIA, AND CARBON DIOXIDE CONCENTRATIONS
Stream Ammonia Methanol Carbon Dioxide
Number (mg/1) (mg/1) (mg/1)
100 ,800 459 1137
200 1041 362 2470
300 858 618 2559
400 1015 972 2789
500 825 559 1258
600 700 172 642
Average 873 524 1809
A review of stream characterization data indicated that stream 700 was
too low in ammonia, mehtanol, and carbon dioxide concentration to be classified
as a representative sample stream. Therefore, values for stream 700 are
deleted. Table 10 shows concentration of the three contaminants from repre-
sentative streams, the average values being ammonia, 873 mg/1; methanol, 524
mg/1; and carbon dioxide, 1809 mg/1. Age of the catalyst and severity of
operations affect the amount of each of these contaminants. However, the
ratio of the three remains fairly constant. The ratios of methanol and vi
carbon dioxide to ammonia are shown in Table 11.
26
-------
TABLE 11. RATIO OF AMMONIA-TO-METHANQL--TO-CARBON DIOXIDE
Stream
Number
100
200
300
400
500
600
Average
Ammonia
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Methanol
0.57
0.35
0.72
0.96
0.68
0.25
0.58
Carbon Dioxide
1.42
2.37
2.98
2.75
1.52
0.92
1.99
TRACE METAL ANALYSES OF THE PROCESS EFFLUENT
There are two areas of concern regarding trace metals: recycling of
stripper overheads to the process and recycling of stripper bottoms to the
boiler water system. Trace metals in the feed stream could poison the
catalyst system on the process side and might cause problems on the steam
generation side. Samples of condensate from each of the participating
companies were analyzed, with the results shown in Table 12. As a further
check on where the metals in the process condensate wculd go in the stripping
operation, runs from the bench-scale operation were sampled. These results
are shown in Table 13. Levels of trace metals in the process condensate
detected (or below the detection limit of the instruments used) would not
present problems with regard to further processing or recycle.
TABLE 12. METAL ANALYSES
Sample I.D.
Company No. 100
Feed
Stripper Overhead
Company No. 200
Feed
Stripper Overhead
Stripper Bottoms
Company No. 400
Feed
Company No. 500
Feed
Company No. 600
Feed
Company No. 700
Feed
Cu
(mg/1)
<0.5
<0.5
<0.5
<0.5
<0.5
<0.02
<0.02
<0.02
0.045
Ni
(mg/1)
<0.5
<0.5
<0.5
<0.5
<0.5
<0.2
<0.2
<0.2
<0.2
Fe
(mg/1)
<0.3
<0.3
<0.3
<0.3
<0.3
<0.1
<0.1
<0.1
<0.1
Zn
(mg/1)
1.5
1.86
<0.1
<0.1
<0.1
<0.02
<0.02
<0.02
<0.02
Cr
(mg/1)
<0.5
<0.5
<0.5
<0.5
<0.5
<0.2
<0.2
<0.2
<0.2
27
-------
TABLE 13. PRELIMINARY RESULTS OF HEAVY METAL ANALYSES ON:GRAB SAMPLES
Metal
Cu
Zn
Ni
Cr
Fe
R-27, 28
Feed
(mg/1)
<0.5
<0.1
<0.5
<0.5
<0.3
R-29
Overhead
(mg/D
<0.5
<0.1
<0.5
<0.5
<0.3
R-30
Overhead
(mg/1)
<0.5
<0.1
<0.5
<0.5
<0.3
R-23
Bottom
(mg/1)
<0.5
<0.1
<0.5
<0.5
<0.3
R-30
Bottom
(mg/1)
<0.5
<0.1
<0.5
<0.5
<0.3
Several samples of process condensate were analyzed for the presence of
methylamines, and none were detected. In order to verify that the gas
chromatograph (GC) was capable of detecting methylamines, a gas sample of
methylamine was received and tested. It was concluded that methylamine was
not present in the process condensate. A mass spectrometric analysis con-
firmed that methylamines were not present in the process condensate.
28
-------
SECTION 6
DEfllOPMENT OF STRIPPING DATA FROM LABORATORY AND BENCH-SCALE DATA
INTRODUCTION
As indicated in the stream characterization section, the average values
for the three contaminants were: ammonia 873 mg/1, methanol 524 mg/1 and
carbon dioxide 1809 mg/1 when the lowest stream analysis was omitted.
Streams 100, 200, 300, and 400 were thought to be more representative of the
level of contaminants to be found in the ammonia process condensate. It was
decided to perform bench-scale tests on these streams.
A technology review indicated that steam stripping o£ the process
condensate would reduce the ammonia and methanol contaminants. However,
there were several questions concerning the stripper operations if the
overhead were condensed for reinjection into primary reformer process.
• If the overhead were condensed, would the methanol and ammonia form a
second phase?
• Would the carbon dioxide form a carbonate product with the ammonia?
0 Would the stripper remove the desired amount of contaminants?
BENCH-SCALE STREAM STRIPPING OF PROCESS CONDENSATE
A bench-scale unit was constructed to clarify operating conditions for "
the steam stripper with condensation of the overhead and subsequent reinjection
into the primary reformer. Drum samples of process condensate from several
different plants were transported to the laboratory unit for test runs. The
laboratory staff operated the small column, obtaining analytical data for
interpertation and operational changes.
The pilot stream-stripper system is depicted schematically in Figures 6
and 7. The column was made from a 7 cm diameter piece of Derakane, with an
overall length of 4.3 m. Packing height was 3.4m with 1.6 cm (5/8
inch) polypropylene Pall rings as the fill material. Feed was measured to
the system through a calibrated rotameter, and a peristaltic pump was used
for flow continuity. Steam for stripping was preset by a needle valve and
measured by a calibrated orifice. To minimize all process losses, the
bottoms and overhead were condensed in a refrigerated bath.
Analytical Teehinques
Analyses of the process condensate from the participating industries
and from the pilot steam-stripper for ammonia, methanol, methylatnines, and
29
-------
Overhead Line to Condenser
Feed Line to Stripper
ometer
for Overhead,
JI Bottom, and
Steam
Temperatures
Stripper
Column
Overhead
Line to
Condenser
Sight
Glass for
is
valve to Control
Overhead
and Bottoms
Condenser
Figure 6. PilQt steam stripper located
at GSRI, New Orleans.
30
-------
Condenser
Feed
Pump
Thermocouple
0
Steam
In
Rotameter
Feed
Overhead
M
Pressure
Gauge
t^j
V
N
! \ _„,
' v \ i
A n\ \ A
Orifice
Plate
Produc
(To be sent to primary reformer)
Boiler
Feed Make-up
Bottom Out
Manometer^
Figure 7. Model of pilot steam stripper (Gulf South Research Institute - New Orleans).
-------
and carbon dioxide presented some difficulties, primarily in the analysis of
the ammonia and the methylamines. Ammonia analysis was initially done on a
gas chromatograph, along with the analyses of methanol, methylamines, and
C0_. After trying several column packing materials and varying conditions
of the instrument, including thermal conductivity and flame ionization detec-
tors, only the analysis of the methanol was considered to be reliable. The
retention time of the ammonia was very close to that of the water, and there-
fore could not be separated with reproducible accuracy. The GC did not
indicate the presence of methylamines or carbon dioxide.
The data presented in this report were obtained using a specific anion
ammonia electrode. The presence of methylamine could result in erroneous
answers for ammonia concentrations, thus it was important to establish
whether or not methylamine was present in the process effluent. The presence
of methylamine was determined by two separate means. Samples of the process
condensate were collected and analyzed by a mass spectrometer. The results of
these tests for methylamine were received, and known dilute samples were made
up. These spiked samples of methylamine were then analyzed by gas chromato-
graphy. The GC analysis did show the presence of methylamine in the amount
added to the samples. Thus, if methylamine were present in the process
effluent, the GC analysis would identify it.
The analysis of carbon dioxide, shown in Table 14, was done using the
inorganic carbon (1C) side of a Beckman Model 915 total organic carbon (TOC)
analyzer. All reported carbon dioxide values were in the form of carbonates
or bicarbonates, depending on the pH of the sample. The total carbon (TC)
side of the Beckman 915 was used in conjunction with the 1C side to find total
organic carbon by subtracting the 1C from the TC. This TOC value represents
the concentration of all organic carbons.
Process Effluent from Company 100
Initial runs were made on the pilot steam stripper having a packed bed
depth of only 2.2 m and 0.6 cm (1/4 inch). Rasching rings. Results of runs 1-7
indicated that steam stripping of the process condensate was feasible, but
that desired ammonia and methanol concentrations were not attainable in the
effluent bottom. The first 7 runs (see Table 14) were made with the overhead
rate varying between 2.5 and 20 percent. Analysis of the bottoms from these
runs indicated substantial amounts of ammonia and methanol remaining in the
effluent bottoms.
Effluent bottoms from runs 5, 6, and 7 of the pilot steam stripper were
collected and stored to determine if additional packing height would be re-
quired. These collected bottoms were re-processed through the pilot steam
stripper as feed for runs 8, 9, and 10 to determine if further separation of
ammonia, methanol, and carbon dioxide would take place and if the bottoms of
this rerun would be essentially free of methanol and ammonia. These projec-
tions were confirmed, and the results indicated the need for the following
equipment changes:
1. Addition to the column height (increase packed bed depth)
2. Change to a more efficient packing
32
-------
TABLE 14. ANALYSES FOR METHANOL-AMMONIA-CARBON DIOXIDE ACQUIRED BY STEAM STRIPPING IN BENCH SCALE UNIT
co
CO
Run
So.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Overhead Bottom
Rate Rate
mi/mm ml/min
20.0
19.0
8.0
4.5
34.0
18.0
28.5
28.5
19.0
11.0
5.0
28.0
20.0
6.0
18
5.2
22
19
8
40
26
30
14
26
53
26.8
47
64
30
45.5
34
58
250
185
185
185
200
197
200
230
230
220
340
250
250
245
395
390
450
450
500
410
420
475
495
400
420
460
355
385
250
265
240
272
Feed
Rate
ml/min
220
170
170
170
165
165
165
195
195
195
300
220
220
220
260
360
392
390
390
320
320
360
360
360
375
375
260
260
260
260
260
260
Reflux
Rate
ml/min
_
_
_
-
-
-
-
-
-
-
-
—
-
-
-
_
-
-
58
68
68
68
68
-
-
35
108
108
-
25
-
42
Steam
Rate
kg/hr
3.22
3.83
2.61
2.17
4.17
2.86
2.17
3.22
2.38
1.88
3.36
3.47
3.22
2.17
4.76
3.63
5.22
4.99
4.99
5.44
4.85
5.44
4.85
4.85
6.03
6.03
5.89
6.35
3.04
4.76
3.36
5.21
Feed mg/1
Meth NH3 CO,
Runs
1174
1040
1040
1040
1526
1526
1526
89
89
89
1422
Runs
581
581
581
257
257
280
280
280
280
280
258
258
258
260
260
300
300
317
317
317
317
Feed Overhead Bottom
Overhead mg/1 Bottoms mg/1 TOO TOC TOC
Meth NH3 O>2 Meth NH3 GO mg/1 mg/1 mg/1
1-11 on Company No. 100
620
580
580
580
2700
2700
2700
65
65
65
570
1450
1600
1600
1600
1548
1548
1548
342
342
342
2097
12-32 qn
900
900
900
1200
1200
1100
1100
1100
1000
1000
1150
1150
1150
1000
1000
1200
1200
1100
1100
1100
1100
2816
2816
2816
3399
3399
3381
3381
3381
3263
3263
3227
3227
3227
3157
3157
3212
3212
3131
3131
3131
3131
17880
7096
17740
16450
3333
10443
16443
613
774
1422
57000
Company
5345
6451
10160
8063
23200
7653
9358
25112
6519
16965
10456
5729
23122
3143
5788
10158
9454
2889
3699
2461
3076
6600 18000
5200 9595
6500 17000
9700 25824
2000 1595
2700 8452
5600 12162
580 1369
600 1672
570 2097
10400 22218
No. 200
4000 8569
3300 14501
19500 40326
22000 32838
15000 111228
10000 28174
22000 34393
60noO 126309
20000 14571
49000 55146
29000 42555
49500 45819
19000 27199
9500 13827
16500 16500
37500 39757
28500 31405
8000 15396
10000 16966
13000 13200
9000 75438
193
43
171
171
20
64
173
7
14
32
444
34
92
383
16
85
-
-
69
-
-
-
35
-
-
-
-
-
-
-
-
-
180
60
98
90
25
54
110
7
8
9
240
80
140
160
40
160
50
62
140
19
54
45
100
29
11
15
9
6
4
5
1
4
186
136
155
166
53
98
172
15
21
14
640
158
264
667
87
272
66
105
270
39
97
76
161
52
21
28
21
14
11
11
10
7
1174 17987 381
1028 10336 64
1028 13763 254
1028 14976 232
1235 4312 69
1235 10400 152
1235 16330 384
397 1436 57
397 1685 71
397 2360 81
1229 38120 813
834 7077 61
834 7890 107
834 13205 304
680 11168 32
680 25912 173
579 10416 29
579 12405 40
579 32704 123
664 8707 29
664 16256 38
661 12389 37
661 59744 73
661 8090 28
629 4083 23
629 7536 22
59 12509 27
59 11291 22
648 4920 171
648 4547 19
648 4539 19
648 4445 19
(continued)
-------
TABLE 14 (continued)
OJ
Run
No.
Overhead Bottom
Rate Rate
ml/min ml/min
Feed
Rate
ml/min
Reflux Steam
Rate Rate
ml/min kg/hr
Meth
Runs 33-39 on
33
34
35
36
37
38
39
16.8
13.4
46.5
12
39
15
27
380
390
430
380
420
350
345
350
360
360
340
340
310
310
—
-
34
-
29
-
-
3.90
3.90
5.44
3.90
5.44
3.90
4.35
615
615
615
621
621
621
621
Runs 40-47 on
40
41
42
43
44
45
46
47
11
44
9
40
12
32
8.6
28
395
430
430
460
360
390
330
350
350
350
395
395
345
345
320
320
28
30
25
19.
3.90
5.44
3.90
6.12
3.36
5.13
2,68
2 3.90
1143
1143
1134
1134
805
805
805
805
Runs 48-61 on
48
49
50
51
52
53
54
55
56
57
58
59
60
61
5.5
11.8
30
11.6
23.3
15.5
32.5
17.3
33
7
i.4
14.3
9.9
19.6
350
350
395
320
323
365
385
385
400
300
310
290
370
380
330
380
380
260
260
310
310
350
350
255
255
210
320
320
-
-
24.
-
11.
-
16
-
17.
-
7.
-
-
11.
3.36
3.63
,6 3.95
2.89
8 3.63
3.63
3.95
3.95
3 5.44
2.68
6 3.63
3.63
3.63
0 4.67
448
448
448
397
397
397
397
448
448
454
454
454
429
429
Feed fflg/1
NH,
Company
825
825
825
890
890
890
890
Company
1000
1000
1060
1060
1000
1000
1000
1000
Company
1000
1000
1000
930
930
930
930
380
880
1100
1100
1100
968
968
co2
No.
2763
2763
2763
2354
2354
2354
2354
No.
3260
3260
3260
3260
2390
2390
2244
2244
No.
1888
1888
1888
1521
1521
1521
1521
1532
1532
1595
1595
1595
1525
1525
Overhead mg/1
Meth NH3 C02
300
14646
17498
24693
20914
18908
14771
7320
400
38205
35518
46433
47035
24400
35138
22578
27745
200
17626
11602
16101
3400
9162
7897
7582
7523
7825
13917
14356
7291
11450
13216
18000
21500
27750
25000
31250
27750
26250
21250
23000
23500
30000
26250
37000
31000
32750
40000
30500
38750
11000
19000
19200
18900
16675
21500
32250
28750
17750
28750
29375
28549
39358
34943
47271
35926
39953
17739
61028
45305
80681
60045
29381
36186
24427
31104
23870
14710
21120
9190
15240
15858
14180
14900
11970
14861
12000
15300
19375
Feed Overhead Bottom
Bottoms mg/1 IOC TOC TOG
Meth NH, C02 mg/1 mg/1 mg/1
10.6
23.4
1.0
35.5
1.0
7.3
1.0
108
3.3
122
6
119
30
128
32
100
64
11
23
53
16
4
4
1
52
25
4
30
29
25.2
53.0
29.0
33.0
39.0
43.0
35.5
142
96.0
145
100
73
74
80
62
67
81
55
38
22
27
15
14
12
42
41
19
32
27
40
37
18
37
18
18
11
48
26
59
29
286
280
117
158
213
125
137
92
95
91
110
48
143
205
136
132
147
765
765
765
739
739
739
739
1754
1754
1696
1696
652
652
612
612
515
515
515
415
415
415
415
418
418
415
415
435
416
416
13109 64
16272 59
21461 24
19232 72
18485 53
15961 56
8251 21
38624 139
35925 45
40074 131
46688 40
8013 78
9868
6662 77
8483 32
6510 43
4012 50
5761 34
2506 37
4157 25
4325 26
3868 25
4065 30
3265 13
39
4061 56
3271 37
4173 36
5284 40
-------
3. Addition of a reflux system
With the modifications completed on the pilot stream stripper, a series
of test runs was made on the unit. Test were conducted at various feed, over-
head, and reflux rates. Run 11 was performed on overheads collected from
run 10. The data for these tests are presented in Table 14.
Process Effluents from Company 200
An additional 1.2 m of packing was added to the pilot steam stripper,
making a total packed depth of 3.4 m. The packing material was changed from
Rasching rings (6 mm) to Pall rings (16 mm). This type of packing has been
proven to give better vapor-liquid contact with larger loading capabilities.
It also enables for a more uniform vapor-liquid loading during the stripping
operation. (The term vapor refers to the steam to the stripper column, and
the term liquid refers to the process condensate feed to the pilot steam
stripper.)
Results from first runs (12-14) made on the modified pilot steam stripper
indicated that residual methanol in the bottoms of the steam stripper ranged
from 34 to 383 mg/1 and the ammonia ranged from 80 to 160 mg/1. The amount of
overhead (2.5%) stripped from the feed in this test corresponded to the high
bottom concentrations. The 14% overhead rate corresponded to the low concen-
tration of the effluent bottom from the pilot steam stripper. Results of
these tests indicated that once through 3.4 m of packing was not enough to
effect the desired removals (less than 20 mg/1) of methanol and ammonia.
This column height deficiency could be made up by the addition of reflux.
With reflux, a portion of the condensed overhead is returned to the stripping
column, as shown in Figure 2. Reflux gives the column greater capability and
flexibility and theoretically adds packing height to the column. However, the
addition of reflux requires additional load on the overhead condenser and to
the top of the column. Addition stripping steam would be required which would
in turn increase the vapor-liquid load to the column. However, with reflux,
the net result would be a decrease in the methanol and ammonia concentrations
in the stripper bottom effluent.
Runs 15-18 were made on the bench stripper without reflux. Contaminants
in the stripper bottoms from these tests were from 1 to 85 mg/1 methanol and
from 50 to 160 mg/1 ammonia. The reduction of these two contaminants in the
stripper bottoms was dependent on the percent overhead-to-feed ratios. For
example, if 100 kg/min of process condensate was the feed rate to the steam
stripper, then 10 kg/min of condensate overhead would represent a 10% overhead
rate, and 5 kg/min of condensed overhead would represent a 5% overhead rate.
Test runs 19-28 were conducted with a portion of the overhead refluxed.
This test did not give satisfactory results because a reflux pump did not
function properly.
Test runs 29 and 31 were performed without reflux while runs 30 and 32
were with reflux. In these runs, ammonia in the stripper bottoms was reduced
to very low limits (<5 mg/1). The percent overhead-to-feed ratios used to
achieve these results was high (>10%). A typical 907 m. ton/day ammonia plant
35
-------
generally produces about 45,400 kg/hr (757 1/min) of process condensate waste
stream. If this 45,400 kg/hr is fed to a steam stripper, enough steam will be
added to produce the desired overhead-to-feed ratio. For example, if a 10%
overhead rate is needed for this separation, then enough steam is added to the
column to vaporize the ammonia, methanol, carbon dioxide, and water to make up
a total of 4540 kg/hr, which must be condensed prior to further in-house
processing. The greater the overhead rate, the more steam is required, and the
larger the vapor-liquid load to the column. Vapor-liquid loading to a column
primarily controls the diameter sizing of a column. Therefore, if the overhead
rate can be minimized, the economics for further handling of the overheads
will be improved, especially in the case of reinjection of the stripped over-
heads into the primary reformer furnace process stream. The amount of over-
head generated (for reinjection) from the stripper has a significant influence
on the amount of energy required for condensation, pressurizing, re-evaporation,
and injection of these overheads into the primary reformer furnance process.
In order to decrease this overhead rate and still achieve the desired levels
(<20 mg/1 ammonia and methanol) in the stripped bottoms, the packing height
and/or the refluxing rate has to be increased in the steam stripping column.
Process Effluent from Company 300
Test runs 33, 34, 36, 33, and 39 were without reflux, while runs 35 and
37 were with reflux. All runs were performed with the overhead rate less than
5% of the feed rate. The concentrations of the contaminants in the stripper
bottom ranged from 10 to 35 mg/1 for methanol and 25 to 53 mg/1 for ammonia
for test runs without reflux. The concentration of the contaminants in the
stripper bottoms ranged from 1 to 7 mg/1 for methanol and 29 to 39 mg/1 for
ammonia for test runs with reflux.
Process Effluent from Company 400
Test runs 40, 42, 44, and 46 were performed without reflux, and runs 41,
43, 45, and 47 were performed with reflux. The addition of reflux had a
significant effect on the amount of methanol in the stripper bottoms. Meth-
anol in the bottoms varied between 199 and 128 mg/1, and the ammonia between
73 and 145 mg/1 for runs made without reflux. For runs with reflux, the
methanol varied between 30 and 32 mg/1 and the ammonia between 62 and 100 mg/1
in the steam stripper effluent bottoms. With the addition of reflux, the
reduction in concentrations averaged 25.7% for ammonia and 74.9% for methanol.
Process Condensate from Company 200
An effort was made in the final runs to perfect the stripping techniques
for operating the pilot steam stripping column. Test runs 48, 49, 51, 53, 55,
57, 59, and 60 were performed without reflux, while runs 50, 52, 54, 56, 58,
and 61 were performed with the addition of reflux. An effort was made to keep
the relux ratio at approximately 1:1. Residual methanol concentrations in the
stripper bottom varied between 4 and 10 mg/1 without reflux and between 1 and
53 mg/1 with reflux. Residual ammonia concentration in the stripper bottom
varied between 14 and 81 mg/1 without reflux and 12 and 55 mg/1 with reflux.
36
-------
It was determined that from 5 to 6% overhead-to-feed rate was the optimum
for the bench equipment and process conditions. Under these conditions, the
niethanol and ammonia concentrations in the stripper bottoms would be <15 mg/1
and <20 mg/1, respectively.
Bench Stream StripperMass Balance
Mass balances were determined from the data collected during the operation
of the tench steam stripper to validate the steam stripper data. These mass
balances were determined for the methanol and ammonia content, as well asJthe
volumetric throughputs. The volumetric balances were made by measuring the
feed with a calibration peristaltic pump and comparing with an in-line rota-
meter. After cooling, the overhead bottoms were measured with a graduated
cylinder. The deviation of the feed, with the sum of the overhead and bottoms,
is shown in Table 15. Considering that flow rates varied from 10 to 250
ml/min, the overall volumetric balances for the runs evaluated were fairly
consistent.
Vapor-Liquid Efflilibrium Data
Prior to initiating the actual stripping runs on the bench unit, experi-
mental vapor-liquid equilibrium measurements of ammonia in the process con-
densate were made. These data could be roughly correlated with the height of
equivalent packing in the stripping column and would be needed in any subsequent
design of a full-scale commercial steam stripping unit. Equilibrium data for
the ammonia-water and methanol-water systems can ke found in the literature.
However, a literature search revealed no data for the quaternary system of
ammonia-methane1-earbon dioxide-water.
TABLE 15. MASS BALANCES AROUND PILOT STEAM STRIPPER
Run
No
R-29
1-30
1-31
1-32
R-33
1-34
R-35
1-36
1-37
R-38
1-39
1-40
1-41
1-42
1-43
R-44
Methanol Ammonia
% Error % Error
5.1
-3.8
1.5
-15.7
16.2
10.0
8.2
25.3
-2.9
16.4
2.7
15.7
12.2
5.0
1.4
-21.2
-15.7
-14.9
54.6
21.2
8.1
3.9
5.0
3.3
2.2
56.3
160.0
-6.9
5.9
34.6
-5.5
1.1
% Error Run Methanol
By Volumetric No. /= Error
-9.9
-14.8
-13.3
-15.2
-4.4
-5.1
-1.7
-3.2
-1.0
-2.7
-2.8
-2.2
-l'.'l
-4.6
-5.2
-7.2
1-45
1-46
1-47
R-48
1-49
R-50
1-51
R-52
R-53
R-54
1-55
R-56
fe-57
R-58
R-59
R-60
1-61
1.7
8.2
0.3
10.8
6.4
13.9
54.7
-9.1
-4.2
-1.4
16.0
11.5
2.4
6.6
10.7
9.3
4,5
Ammonia
% Error
4.5
8.4
1.1
26.2
-2.2
11.2
42.2
-3.5
-6.6
-4.9
4.6
-5.0
15.8
17.2
-11.2
4.3
7.4
% Error
By Volumetric
1.9
7.2
1.9
7.9
17.9
4.7
-6.9
-8.0
-2.7
-11.1
2.0
5.5'
-2.4
-2.7
-12.5
0.2
2.3
37
-------
Standard deviation was made on the data presented in Table 15. The formula
used was as follows:
c2 - nEx2 - (Zx)2
b n(n-l)
Vapor-liquid analyses were conducted on raw process condensate received
from streams 200, 500, and 600. A diagram of the testing apparatus is pre-
sented in Figure 8. The following procedure was used to collect samples for
analysis of the ammonia in the recovered vapor and liquid.
1. A total volume, of 600 ml was initially charged to vessel A.
2. Ice bath water was circulated by pump B to condenser C to
cold finger D and back to ice bath reservoir E.
3. Heater F was then turned on.
4. Valves 1,' 2, and 4 were closed.
5. Valve 3 was opened and vented to atmosphere.
6. After still reached 100°C read by thermometer G, valve 3 was closed.
7. Approximately 20 min af.ter systems reached equilibrium main-
taining 100°C temperature,valve 1 was opened, and a 5 ml vapor
sample was collected at H.
8. Syringe J was attached to needle I, valve 4 was opened, and approxi-
mately 15 ml was drawn off at the same time that the vapor sample
was being collected.
9. After samples were collected, valves 1 and 4 were closed, and a 20-
min period was designated between withdrawing samples.
10. Vapor and liquid samples were diluted and analyses for ammonia were
made and recorded.
The first process effluent tested was from stream 200. The results are
shown in Table 16. To check the variability of the runs, a least squares-fit
of the data for the vapor (y) versus the liquid (x) equilibrium values was
determined. The equation was as follows:
Stream 200: y = (147)(x) - 100
Where y = mole fraction of ammonia in vapor
x = mole fraction of ammonia in liquid
Subsequent runs on stream 500 and 600 produced the results shown in Table
17. The least-squares vapor-liquid equations for the two streams were as
follows:
Stream 500: y = (123.5)(x) - 456
Stream 60.0: y = (232) (x) - 100
The vapor-liquid equilibrium data as represented by the preceding equations are
plotted in Figure 9.
38
-------
Boiling flask
Circulating pump
Condenser
Cold fiuger
Reservoir
Heater
Thermometer
Collection point
Needle
Syringe
Valves
Figure 8. Diagram of apparatus to gather vapor-liquid data.
-------
TABLE 16. VAPOR AND LIQUID EQUILIBRIUM DATA
(PROCESS COMPENSATE FROM COMPANY 200)
Vapor
mg/1 NH
5400
2700
1820
1650
1100
930
725
700
690
515
930
475
250
220
700
Run 1
Liquid
3 mg/1 NH3
335
275
218
185
129
113
83
75
60
54
96
45
28
27
74
Run
Vapor
mg/1 NH3
8500
7700
4950
2550
1900
1255
870
1060
1030
1050
965
2
Liquid
mg/1 NH3
470
365 '
250
187.5
162.5
106.5
78.5
100.0
84
95
80
Run
Vapor
mg/1 NH3
8875
4440
2500
1750
1650
3
Liquid
mg/1 NH3
330
220
180
155
135
TABLE 17. VAPOR AND LIQUID EQUILIBRIUM DATA
(PROCESS COMPENSATE FROM COMPANY 500 AND COMPANY 600)
Company Number 500
Vapor Liquid
(mg/1 NH,) (mg/1 NH )
(y) (x)
Company Number 600
Vapor Liquid
(mg/1 NH.,) (mg/1 NH,
(y) (x) ^
16440
12240
11425
7440
5100
3096
2495
1850
684
456
380
240
228
148
144
79
72
82
54
41
24
17
11
8
7
6
5
A
24600
18720
10080
6600
3840
3120
1980
1140
708
336
276
222
171
150
40
-------
12.5 -
CM
O
!-*
X
CO
PC
s
10.0 -
O
CX
c
•H
C
O
•H
4J
O
ctf
s-l
0 2.5 5.0 7.5 10.0
Mole Fraction NH3 in Liquid; Nf^xlO4
-Figure 9. Equilibrium curve for ammonia/methanol
wastewater system.
41
-------
SECTION 7
DESIGN OF COMMERCIAL AMMONIA-METHANOL STEAM STRIPPER
!
Using pilot plant and vapor-liquid equilibrium data, design calcula-
tions were made for a steam stripping column. The pilot stripper achieved
98 percent removal of ammonia and 99 percent removal of the methanol. The
vapor-liquid 'data would-help establish the necessary depth of packing to
reduce the ammonia and methanol content to the specified level.
The bench-stripper data corresponded to a concentration of 30 mg/1
ammonia and 5 mg/1 methanol in the bottoms from the column. Vapor-liquid
equilibrium data indicated that a straight line existed in the dilute ammonia
concentrations in the process condensate. The column was designed for a 907
m. fon/day ammonia production unit with 757 1/min process condensate. The
design conditions and component analysis of the process condensate are shown
in Table 18. ,
TABLE 18. PROCESS CONDENSATE ASSUMED FOR COLUMN DESIGN
Component mg/1 pph flow
Ammonia
Methanol
Carbon dioxide
Water (%)
1000
750
1500
99.675
. 100
75
150
99.675
t.
Plant capacity 907 m. tons/day ;
Process condensate 757 1/min
Stripper Effluent
Ammonia 20 mg/1
Methanol 5 mg/1
To develop otpimum design conditions, mass and energy equations were
developed with the stripper overhead as the variable. These data were
correlated and plotted, and are shown in Figure 9 as kg/hr of steam consumption
vs. water content in overhead; kg water/kg pure overhead; and/or percent feed as
overhead. (Pure overhead = combined methanol, ammonia and C0» content in the
overhead.) Theoretical process conditions used to develop the design were as
follows:
1. 45,400 kg/hr feed rate to the steam stripper with the composition
as shown in Table 16.
2. Low pressure steam (3.16 kg/cm gauge).
3. Vapor density (0.769 g/l).3
4. Liquid density (0.862 g/cm ).
42
-------
5. Liquid viscosity (0.29 centipoise).
6. 3.8 cm (1.5 inch) Pall rings for packing material with a factor of
3.9.
7. Feed preheated prior to stripper column entry.
8. 98 percent removal of ammonia and 99 percent remova.' of methanol
and carbon dioxide.
With the above conditions, approximately 150 kg/hr total of ammonia,
methanol, and carbon dioxide are fed into the steam stripper along with the
process wastewater. If there were an infinitely large column, it would be
possible to use only enough steam to bring the column and its contents up
to temperature for removal of the contaminant. In this case, there would be
145.5 kg/hr of pure overhead (44.5 kg/hr of ammonia, 33.7 kg/hr of methanol,
and $7.3 kg/hr of carbon dioxide), without any water carryover into the
overhead. At this time, such an operation is not economically feasible.
However, if enough water was taken with this 145.5 kg/hr of contaminant, cal-
culations could be made to determine how much steam would have to be added
to the stripper. For example, on the abscissa (Fig.10), 1 kg water/kg
pure overhead would represent a 50% water and 50% ammonia, methanol, and
carbon dioxide. If the 50% ammonia, methanol, and carbon dioxide represents
145.5 kg/hr, then the 50% water represents 145.5 kg/hr. Therefore, 145.5
kg/hr of water divided by 145.5 kg/hr of pure overhead represents the 1
kg/hr in the abscissa. The 10 kg/hr would represent 1455 kg/hr of water and
145.5 kg/hr of ammonia, methanol, and carbon dioxide. Also shown on the
abscissa is a scale showing the percent feed as overhead. For example, the
1 kg of water/kg pure overhead represents 145.5 kg/hr of water and 145-5
kg/hr of ammonia, methanol, and carbon dioxide, for a total of 291.0 kg/hr
of overhead. If the feed rate is 45,400 kg/hr, there will be 0.6% feed in
the overhead (291.0 kg/hr divided by 45,400 kg/hr multiplied by 100). The
10 kg of water/kg of pure overhead has a total of 1455 kg/hr of water with
145.5 kg/hr of ammonia, methanol, and carbon dioxide, representing a total
overhead of 1600.5 kg/hr. In this case, there will be 3.53% feed as overhead.
The concentration in the overhead can be determined using the data in Figure 10.
Figure 11 is a plot of the tower diameter vs. pressure drop for steam
rates from 454 to 9072 kg/hr. With this plot, the required column diameter
for the separation can be determined where the pressure drop and steam rate
are known.
The water content of the stripper overhead was minimized to reduce the
total volume for further handling. If an ammonia concentration of 6 percent
in the overhead could be achieved, the total overhead volume would be reduced
to some 19-38 1/min. The design was based on packed tower concepts. The
basic equation for packed tower design was:
AP = a(10BL) (--| )
5
where AP = Pressure drop
a,B = Constants for a particular packing
G = Vapor mass velocity
L = Liquid mass velocity
Pr = Vapor density
J 4'3
-------
13
12
11
10
60
o
o
o
'-' 7
c
o
01
C
O
4J
CO
Basis:
907 m. ton/SD Ammonia Plan:
98% Aramonia Removal
4.20 kg/cm
Feed Preheated
II
I
I
10 20 30
Water content in overhead, kg H90/kg.
40
Pure overhead.
05 10 15
Percent feed as overhead, %
Figure 10. Steam consumption vs. water content in overhead and percent
feed taken as overhead.
44
-------
20.d
60
c
10.
9.0
8.0
7.0
6.0
5.0
4.0
o
ea
14-1
o
S-i
-------
Vapor-liquid data from company 500 were used for design because of the
similarity of these data to process condensate composition. The McCabe-
Thiele method of stepping off theoretical trays between equilibrium and
operating lines (Fig. 12) was used. Another method to check calculations
was the use of Henry's law, which simplifies design equations to give
theoretical packing height. The equations are summarized below.
To Calculate Height of Packing
Number of Units —
From literature, NH_ stripping is gas-phase controlled. Therefore, the
gas-phase resistance equation was used :
.
NTOG " (y-y*)B-(y-y*)T ln (y-y*)T
Height of Unit —
MG
HTOG " HG + IT* (V
m
Where:
B
H - °G
" '
Packing Height —
Z = H N TOG
Where :
y = Mole fraction NH
M - Slope of equilibrium lime
2
G = Gas mass velocity (Ib moles /hr-ft )
m 2
L = Liquid mass velocity (Ib moles/hr-ft )
" G' = Gas rate (lb/hr-ft2)
L' = Liquid rate (lb/hr-ft2)
yG = Gas viscosity (Ib/ft-hr)
uL = Liquid viscosity (Ib/ft-hr)
?„ = Gas density (lb/ft3)
46
-------
12.5 .
io.o -
Hole Fraction NH,, in Liquid, NFL x 10
Figure 12,
2,5 5.0 7 .-5
McCabe-Thlele method for theoretical stages.
10.0
4?
-------
P, = Liquid viscosity (Ib/ft )
L 2
!>„ = Gas diffusivity (ft /hr)
2
DT = Liquid diffusivity (ft /hr)
Li
Z = Packing height (ft)
cj> = 0.01
J = 0.22 Constants for specific packing;
a = 7.0 based on Rasching rings
B = 0.39 (similar to 3.8 cm [1-1/2"! Pall rings)
Y = 0.58
Subscripts: T = Top of column
B = Bottom of column
Superscripts: * = Equilibrium value.
The procedure used for the column design was:
1. Select overhead composition; water content (with maximum of
20 ppm NH_ in bottoms).
2. Use Figure 10 to determine steam required for overhead composition.
3. Use Figure 10 to obtain tower diameter to give a pressure drop =
4.2 g/cm /meter of packing.
4. Calculate theoretical height by McCabe-Thiele.
5. Multiply by appropriate efficiency factor to obtain actual height.
The results are shown in Figure 12 as a plot of tower height required
to obtain specific overhead composition with and without refluxing. The
plot shows that an increase in packing height will reduce the water content
overhead, which in turn reduces steam consumption. A comparison between
pilot plant data and theoretical calculations showed the pilot stripper was
27 percent efficient, while refluxing increased the efficiency to 36 percent.
48
-------
98 percent NH3 removed max.
20 ppm ML maximum in bottoms
5 ppm methanol maximum in bottoms
3,8 kg/on2 steam
Feed Preheated
09? m. ton/day
f \
5 10
A: Water content in overhead
_L
15 20
(kg H90/kg of pure overhead)
I
B: Percent feed as overhead
Figure 13. Commercial column design.
Packing height vs. overhead water content,
-------
SECTION 8
DISPOSITION OF THE STRIPPER TOWER OVERHEAD
INTRODUCTION
There were several options available for treatment of the steam stripper
overhead.
1. Direct discharge to the atmosphere.
2. Reinjection into the primary reformer furnace inlet.
3. Injection into the base of the furnace stack.
4. Precipitation of the ammonia with magnesium phosphate and biotreat-
ment of the methanol residuals.
5. Adsorption of the ammonia utilizing a vanadium pentoxide packed
bed.
Options 4 and 5 were to be investigated from an economic standpoint to
give an indication of the total cost-benefit comparisons of the various
processes (see Section 10).
DIRECT DISCHARGE TO THE ATMOSPHERE
Several large ammonia producers have installed process condensate steam
strippers which are discharging to the atmosphere. Analysis of the stripper
bottoms indicates that this operation does reduce the ammonia in the stripper
bottoms to the desired level. The net result, however, is that the contami-
nants have been removed from the water and redistributed into the surrounding
atmosphere.
REINJECTION INTO THE PRIMARY FURNACE INLET
With the recycle of the stripper bottoms to the boiler feed water
makeup station, the reinjection of the stripper overhead into the primary
furnace inlet would produce total plant recycle. Two aspects of the reinjec-
tion process were investigated: (1) effects of trace metal contaminants on
the reformer catalyst and (2) effects of added energy requirements in sparg-
ing and vaporizing the stripper overhead.
Effects of Trace Metal Contaminants on the Reformer Catalyst
The metal analysis of the process condensate and samples of the overhead
and feed from the bench unit operations had indicated only trace amounts of
any metals which might interfere with or subsequently poison the process.
The resulting data did not indicate there would be any adverse affect from
these trace metals if the overhead were reinjected via primary reformer.
50
-------
Effects of Added Energy Requirements in Sparging and Vaporizing
the Stripper Overhead
The primary reformer feed is a proportioned mixture of steam and natural
gas at approximately 38 kg/cm and 315°C. It is preheated to this tempera-
ture prior to the furnace inlet. The stripper overhead would have to be
pressured for injection at these conditions. A schematic of the flow condi-
tions is shown in Figure 14.
Since there is no preheat source, the heat of vaporization and sensible
heat required to bring it up to process conditions would have to come from
the steam and methane. The net result is an overall reduction in the tempera-
ture of the feed to the primary furnace. The temperature of the primary
reforming operation is critical to the conversion of the methane and steam.
to carbon monoxide and hydrogen. If the temperature of the furnace inlet is
decreased, the only way of achieving the desired conversion-temperature is
to decrease the quantity of feed to the furnace, which results in an overall
reduction in plant production capacity. Energy and mass balance calculations
were performed around the point of reinjection to determine the net decrease
in the primary furnace inlet temperature due to the sparged; stripper overhead.
Three different process temperature conditions, 49°C, 60°C,,and 121°C, were
tried for the condensed stripper. <
To determine the effect of the amount of stripper overhead on the
production capacity, material and energy balances were determined on the
reinjection of varying amounts the overhead. The following basis was assumed
for these calculations.
907 m. ton/day ammonia plant
45,400 kg/hr of process condensate
Process condensate contaminants
NH 1,000 mg/1
Methanol 750 mg/1
CO 1,500 mg/1 2
Primary furance inlet at 38 kg/cm and 315°C
Stripper overhead at 49°C, 60°C and 121°C.
A plot of the inlet conditions of the primary reformer versus varying amounts
of stripper overhead is shown in Figure 16.
For the design conditions initially specified, the net decrease in the
primary reformer inlet temperature would be 21°C if a 6.73% overhead rate
from the steam stripper was injected into the process at the point shown in
Figure 14. This 6.73% overhead rate corresponds to a water content of 20
kg/hr of pure overhead (2909 kg/hr of water and 145.5 kg/hr of ammonia,
methanol, and carbon dioxide, for a total of 3054.5 kg/hr of condensed
overhead). Thus, the amount of water present in the stripper overhead is
critical to the amount of reduced temperature for the ammonia production.
Figure 15 was prepared to determine the amount of extra heat which
would have to be added for varying amounts of stripper overhead injected
into the furnace inlet. This plot represents the percent increase in furnace
51
-------
Overhead
Feed
:
Steam In
Bottom Out
Steam Stripper
315 °C
38 kg/cm Steam to Primary Reformer
Reinjection
of Overhead
Concentrate
Primary Reformer
Figure 14. Stripper overhead to primary reformer.
52
-------
5.0
i 1 • r
Basis; 907 tn. ton Ammonia Plant
4.0
•e
8
•H
3
cr
u 3.0
PL,
c
M
rt
01
tn
01
03
o
c
4-1
C
QJ
O
i r»
2.0
1.0
Stripper Overhead Temperature
49° C
Figure 15.
5 10 15 20
Water content (kg H_0/kg pure overhead)
Percent increase in heat required to maintain reformer
temperature vs. water content in stripper overhead.
53
-------
316
312
308
o
o
0)
H
3
4->
«J
304
0)
300
296
292
Basis: 907 m. ton Ammonia Plant
Steam at 38 kg/cm
316°C
Stripper Overhead Temperature
49CC
121°C
10
15
Water content (kg H_0/kg of pure overhead)
20
Figure 16. Steam temperature vs. water content in stripper overhead,
54
-------
heat needed to maintain am exit temperature of 825°C and 38 kg/cm2 in the
primary reformer. Again, three different temperatures were used for the
process condensate, and various amounts of water were included in the stripper
overhead. ^For example, the 20 kg HgO/kg of overhead is approximately 95%
H2^* at ^°c» a 3.1% increase in heat input is necessary to maintain the
same conditions prior to injection of the steam stripper condensate.
Approximately 440 m of natural gas per m. ton of ammonia is required
as the heating fuel in the primary furnace. If a 3.1 percent increase in
heat is necessary, an additional 12,300 m /day of natural gas would be
needed to maintain production capacity.
Minimum^_gtripper Overhead to Achieve gatisfactory Bottoms Concentrations
of Contaminants
If the removal of the ammonia and methanol contaminant vapor could be
accomplished with minimum stripping, the quantity of stripper overhead
condensate would he reduced considerably. The impact of reduced water
reinjection into the furnace inlet would be reflected in a smaller reduction
in the temperature, for example, if the overhead from the stripper were
reduced to 5 kg H20/kg of pure overhead components (1.92% of feed taken
overhead), the percent extra heat input to the primary furnace would drop
from 3.1 percent to 0.5 percent. Under these conditions, a 907 m. ton/day
ammonia plant would require heat input of 14,300 m /day rather than 12,300
m /day.
fo reduce the stripper overhead from 6.73 percent to 1.92 percent of
the feed would require that the stripper be almost doubled in height. This
added height would at least double the cost of the stripper installation.
Further, any additional heat input would require the installation of another
furnace to add the lost heat to the process system,
Additional Equipment for Stripper Overhead Reinjection
Equipment purchases in addition to the steam stripping unit required
for primary furnace reinjection would include:
Overhead Condenser Unit-
Based on the 2268 kg/hr, the heat load would be 1,329,000 Kcal/hr,
utilizing approximately 16 ot2 of condenser surface with a throughput of 1100
1/min cooling water., If a reflux of 2:1 were used, the surface area would
be increased to 59 m with approximately 3,906,000 Kcal/hr heat load utilizing
over 3,400 1/min of cooling water.
Feed Pump—
A high pressure, low capacity feed pump is needed. This, pump has to
deliver from 4 to 40 1/min at 50 kg/cm to the sparger.
Auxiliary Tubular Furnace—
This furnace would preheat the stripper overhead to the process condi-
tions existing at the furnace inlet. Based on 6.73 percent overhead, the
furnace duty would be 6,300,000 Kcal/hr with an efficiency of 75 percent.
55
-------
INJECTION OF STRIPPER OVERHEAD VAPOR INTO THE FURNACE STACK
Theoretical Decompositions
Injection of the stripper overhead vapor containing ammonia and
methanol offered an interesting possibility. At the stack temperature of
200°C to 260°C, ammonia and methanol would largely decompose. In order to
evaluate this method of disposal, the thermodynamic equilibrium of ammonia
and methanol in the presence of stack exhaust gases was calculated for the
200°C to 260°C temperature range.
To calculate the free energy of decomposition, the products of decomposi-
tion from the ammonia and methanol were defined as shown in the following
equations:
NH3^ 1/2 N2 + 3/2 H2 (1)
2 NH3 + 5/2 02^±2 NO + 3 H20 (3)
2 NH3 + 7/2 02=? 2 N02 + 3 H20 (4)
CH2OH + 3/2 02^± C02 + 2 H20 (5)
Table 19 presents the free energy of the assumed reactions as a function
of temperature. Using the data developed in this table, it was possible to
plot the free energy for the assumed decomposition equations as a function of
temperature. This plot would indicate the potential decomposition to those
products which might be expected in the furnace stack. This plot is shown in
Figure 17. If oxygen was not present in the stack, 90 percent of the ammonia
would decompose to nitrogen and hydrogen at a temperature of approximately
254°C. Also at this temperature, 99 percent of the methanol would decompose
into carbon monoxide and hydrogen. If oxygen was present, then the decomposi-
tion of the ammonia would be through the mechanism of equations (3) and (4).
In general, furnace stack gases contain very little excess air, as this
condition is not in the interest of maximum heat utilization of available
fuel.
PRECIPITATION OF THE AMMONIA WITH MAGNESIUM PHOSPHATE AND BIOTREATMENT OF
THE METHANOL CONTAMINATED WASH WATER
No experimental data was determined for this nrocess technique. However,
a cost-benefit evaluation based on assumed operating conditions was developed.
Data for this process were furnished by Dr. R. Swank of the Environmental
Protection Agency. The cost-benefit evaluation appears in Section 10.
56
-------
TABLE 19. GIBE'S FREE ENERGY CALCULATIONS
Temperature
Reaction (°C)
CHLOH -> CO + 2H9 149
j z 15?
177
204
232
260
NH_ •»• 1/2N0 + 3/2 H0 149
1 77
204
246
254
2NH_ + 5/2 00 -> 2NO + 3H,0 149
J z z 204
260
2NH0 + 7/2 00 •* 2N00 + 3H_0 149
O • / •• / ^— • ^ rj
260
CH,OH + 3/2 0,, •> CO, + 2H,0 149
204
260
Free Energy
cal/g mole
- 635.4
-1058.5
-2161.2
-4121.7
-5166
-6724
925.5
218.3
- 506.9
-1567.2
-1802.9
-91,849
-97,825
-93,712
-101,465
-100,887
-100,804.8
-130,014
-120,540
-131,074.9
57
-------
o
X
3 3)
p I—1
o> o
B
Qj
a) 60
m a
•33
C5
-100
-120
-110 |-
-130
s
60
3
_j___j
180
• -Ji i. \i. -
L_— L _ 1-
200
- --— ^ ou ^ ZH ,j
j i . .. I
220 240
260
Temperature °C
Figure 17. Gibb's free energy for ammonia and methanol reactions at furnace
stack temperatures.
58
-------
SECTION 9
EVALUATION OF COMMERCIAL STEAM STRIPPER WITH OVERHEAD INJECTION
INTO THE FURNACE STACK
INTRODUCTION . . , .
The theoretical thermodynamic analysis of the decomposition of the
ammonia and methanol within the furnace stack offered an attractive, economi-
cal solution to the disposal of the stripper overhead, The stripper bottoms
could be recycled to the boiler feedwater systems. During this program, one
of the ammonia plants installed a stripper which diverted the overhead into
the primary furnace stack. To corroborate the data obtained earlier with
the bench unit and the thermodynamic equilibrium calculations, field tests
were conducted on this commercial unit.
COMMERCIAL STRIPPER PROCESSING CONDITIONS
The commercial column was designed -with a 901 m stripping section and
packed with Pall rings. The overhead stripping line entered the furnance
stack approximately 8 m from the ground. The stack stood 32 m high. A
schematic of the stripping towers, overhead vapor line, and furnace stack is
shown in Figure 18,
Since the intent of the field test work was to determine the amount of
ammonia and methanol decomposed in the stack,, sample points were installed
to measure flow and obtain representative samples. These points are indi-
cated in Figure 18. Sampling of the furnace stack below and above the
stripper vapor entrance gave an indication of the components added to the
furnace exhaust gases. Measurement of the vapor from the stripper overhead
would permit determination of the amount of stripper components being decom-
posed,
_S amp ling Steam Stripper
A number of test rims were made on the steam stripper to obtain reliable
operating data. Because the overhead was vapor, a cooler had to be Installed
to condense and sub-cool this condensate to insure total recovery. A diagram
of the feed test equipment is shown in Figure 19.
Sampling S_tack Analysis
The furnace stack was sampled at two points; (1) at the furnace outlet
and (2)' above the stripper overhead injection point. With this sampling
59
-------
Stack Aanubar
i
-------
Stripper Overhead
To Differential
Pressure Cell
Feed In
>otf oms
Overhead
Condensate
Water
Out
0.009K H0SO,
2L-J
Ammonia—Methanol
Trap (lea Batb)
Ammonia—Methanol
Condensate Trap
Figure 19. Annnonia/methanol sample train for stripper
overhead analysis.
61
-------
procedure., the status of the ammonia and methanol constituents could be
accurately determined. The sampling train used for this analysis is shown
in Figure 19.
Data Collection
A total of 74 runs were made on the steam stripper (runs 1-9 ware no
data-familiarization runs); determinations were made of the flow conditions
and the individual component analysis. Alsos appropriate measurements were
taken at the two sampling points in the stack during these stripper tests.
These data are presented in Tables 21 and 22, The measurements were made
over a 3-month period from September 1976 to January 1977, A new catalyst
had been installed in the primary reformer prior to this evaluation period.
The catalyst is far more selective during the initial stages of operation,,
and as the catalyst ageS;, the concentrations of ammonia and methanol gradually
approach the values experienced during the bench-scale evaluations.
Analy_si_s__gf_j)ataj ^ Stripper Material Balance
There were 65 tests performed on the stripper overhead, with analyses
made for ammonia and methanol. These data are presented in Table 21. Flow
conditions are shown in Table 22. For comparative purposes, the values were
averaged to determine the efficiency of the stripper in removing ammonia and
methanol from the process condensate. The average chemical analysis for all
runs is shown in Table 20, The field tests found an average of 487 mg/1
ammonia in the feed to the stripper. To reduce the bottoms to 7 mg/1 required
a removal efficiency of 96,8 percent.
TABLE 20. A¥ERAG! CHEMICAL ANALYSIS FOR ALL RUNS ON THE STEAM STRIPPES
Steam
Stripper
Feed
Overhead
Bottom
% Reduction of
both products
Ammonia
(mg/1)
487
4750
7
(kg/hr)
39.2
37,9
1.3
96,8
Methanol
(mg/D
262
2610
3,
(kg/hr)
21.1
20.8
4 0,3
98.8
During this same periods the process conditions on the stripper towers
were recorded. The average values are shown in Table 22 „ Daily averages of
the flow conditions are shown in Table 20. These data indicated'an overhead-
to-feed ratio of 9.9 percent. Comparison of these data with those obtained
from the bench unit was difficult. The amount of ammonia and methanol in
the process condensate was about half that which was found during the bench
scale test work,,
62
-------
Sample from
Duct/Stack Probe"""-53""'
Filter Holder
Flow
Control
Valve
Temperature
Indicator
Rotameter
Dry Gas Meter
Impingers
Impinger 1 - 100 ml 0.009N K
Impinger 2
Impinger 3 - Dry
Impinger 4
100 ml 0.009W H0SO,
2 4
Anhydrous CaSO, (Drierite)
Figure 20. Ammonia/methanol sampling train for stack analysis.
-------
TABLE 21. FIELD DATA ON PROCESS CONDENSATE STRIPPER AND STACK ANALYSIS
Condensata S
Condensed
Feed O^ethsads
Run
No.
10
11
12
13
14
Ti
16
17
IS
19
20
21
22
23
24
25
26
27
28
29
30
11
32
33
34
35
36
38
39
41
42
43
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
70
74
* D<
** f)(
( — 5
Cmg/1) Cm
SH
480
4E5
600
600
425
600
650
500
450
700
350
300
500
500
450
400
560
560
500
400
500
400
450
800
450
500
aoo
300
550
350
450
390
360
380
560
480
400
440
450
275
no
650
300
300
550
600
650
800
700
430
750
650
300
500
500
350
350
600
•notes
motes
Indie
cH3oa NH3
340 4680
288 5000
288 4660
297 2592
297 4395
287 3400
258 5400
273 5000
253 4150
250 4000
277 4000
319 5000
291 5050
310 5000
301 5000
296 5000
251 5400
228 5200
235 5510
240 4000
238 5000
245 6000
252 6000
248 4000
257 3200
261 5000
239 3000
246 3800
203 5600
733 6000
205 5150
191 5100
204 5250
204 5000
172 4800
203 5200
283 4500
228 5000
211 5000
204 4750
232 4200
211 4200
229 4500
238 4200
301 5000
289 4000
281 4300
278 4000
219 5500
224 4500
213 4SOO
268 4000
223 4500
226 5000
234 5500
242 4800
274 4300
318 5800
244 3500
ppm In top of
ppni at sample
ates sample not
g/1)
CiijOH
2091
2208
2822
2592
2503
2730
2678
2680
2341
2755
2251
2824
2658
2716
2760
2721
2575
2567
2523
2550
2635
2685
2719
2466
2568
2901
2340
2182
2145
2H74
2131
2230
2228
2430
2461
2523
2355
2538
2478
2564
2564
2605
2582
2754
2756
2736
2774
2787
2627
2577
2509
2853
2616
2644
2746
2311
3000
2662
2562
trigger ,
Analysis
BotCOtns Impinge,.' "A"
(mg/1}
HH3
2.5
1.5
4.0
4.5
0.5
0.6
80.0
1.2
4,0
6.0
4.0
3,0
3.0
3.0
7.0
0.3
0.5
2.4
16.0
20.0
7.0
0.5
0.9
8
10.5
11
18
25
18
2.0
2.7
1.2
1.0
10.2
11, 5
30,0
1Z.O
38,0
8,5
8.0
7,0
4.0
5.0
1.2
1.0
J5.4
3.0
10,6
2.4
5.4
15
5.0
4,5
3
5
5
3,0
9.0
reformer acock
point ft
taken
5
3
3
3
3
1
2
2
3
2
4
2
i
4
2
3
0
0
4,7
4,4
(sample
to stea
Cms/1)
HH3 CHjOH
-
-,
3.5
0.6 17,5
0.5 15,9
0.3 19,0
4.0 17,0
0,1 37. J
1,0 0
1,0 24,9
1.5 16.9
1,2 22,1
80.0 25.6
100.0 32,4
40.0 20.1
1.5 14.9
2.5 17.8
0.1 18, A
12.0
5,0
4
0,7
1.0
7
5.6
1,5
8.5
10 16.5
10
-?
8.5 10.8
9.2 14.1
9,0 17.fi
10.0 16.4
10,0 14.9
1.0 11.9
1.0 0
3.0 21,9
32.0 18.6
4.0 23.6
28.0 20.3
1.5 22.4
4.5 22.9
3.0 27.6
3.0 10.9
1.0 12.4
4.5 15.4
3.0 16.2
5.0 14.1
2 0
3 0
4.0 16,4
3.0 7.6
~
point #2).
Impin;
S :
ack Ana'
5er"-i" l:npir,ger"2"
(rag/1) fe|
NH3
12.0
3.5
14.0
18.0
12.5
1.3
5.0
20.0
6.0
70,0
—
10.0
30.0
16.0
40.0
—
17.0
11.0
-~
30.0
10.0
—
9,0
4,0
—
20
70
28
70
—
53
21.0
—
19.0
20.0
—
18.0
—
72.0
_-
12.0
—
14,8
^
8,0
—
11.6
1.1.0
20.0
—
10.0
—
5
—
5.0*
0.0**
45 . 0*
0.0**
m stripper injection in
CHjOH MH,j
0,4 0,5
0,8 6.0
2.1 0.5
0.2 0.3
0 1.6
5.7 'i.O
5.9 5.0
5.4 0, 5
5.8 4,0
2.2 1.0
4.4 1.3
7.7 1,3
11.4 i.O
—
5.9 0.1
6,2 0.1
—
0.5
0.1
—
5.3 5.5
0.0 4
10
8
—
5
~_
6.4
—
8.4
7
—
7 1.0
, —
7.9
10.0 1.0
—
10.9 1.0
—
13.9
—
5.3
4,2
0 19.0
—
0 1
—
10.0* 3
0.0"
—
11.3* 0.3*
0.0** 0.0**
6.3* 0.4*
0.0** 0.0**
stack.
./I)
CR3OH
0
0
0
0
0
0
0,5
0
0
0
~_
0
0
0
0
—
0
—
S
5
—
-_
0.0
0.0
„
—
0,0
—
0,0
a
^~
0
—
a
—
3
—
0.0
~_
0
—
0
0
0
—
0
--
0.0*
C,Q«
—
0.0*
O.o*i
0.0*
0.0**
L-reJ3 —
{.Alkaline
-------
TABLE 22. FIELD DATA
Material Flow around Condensatc
Steam Stripper
Rur
No.
from Reference Stack
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
23
26
27
28
29
30
31
i2
33
34
35
'»>
37
IK
'.9
40
41
42
43
4'.
43
46
47
'48
'*9
71
79
81
71
79
80
77
79
79
81
81
73
78
77
75
76
82
83
86
87
88
88
90
83
82
85
SK
77
K2
70
XI
79
75
76
71
81
80
8H
76
79
,587
,845
222
,586
,845
,535
,782
,157
,151
,909
,238
,663
,377
,1 10
,716
,430
,599
,973
,729
,420
,246
, 105
,382
,973
,874
,210
, i(15
,092
,399
,209
, ]<)3
,832
,716
,385
, 386
,975
, 3 I ,.'
,087
,385
,923
7,802
8,754
9,616
7,620
8,165
8,618
8,301
8,573
8,605
8,800
8,845
7,938
8,709
8,459
7,346
K,391
8,981
9,163
9,457
9,553
9,618
9 , 60 i
9,707
8,800
8,664
8,936
9,593
8,550
8,«22
7,684
8,3413
8,369
S.lh'i
8 . 39 1
7,802
9 , 1 6 i
8,698
9,593
8,210
8,709
7
7
8
7
8
8
7
8
8
8
8
7
7
7
7
7
8
8
8
8
8
8
8
7
7
8
8
7
7
6
7
7
7
7
7
8
7
8
7
a
,167
.979
,70
-------
TABLE 23. PRODUCTION UNIT AVERAGE PROCESS CONDITIONS
Steam Stripper
Flow (kg/fax)
Feed measured
Steam measured
Subtotal
Overhead measured
Bottoms by difference
80,500
8., 680
89,180
Components T?re3ent_in .the JFurnace Stack_ Exh_aus_t
The three components of interest in the furnace exhaust stack are
ammonia,, methanol and nitrogen oxides,, Potential sources of these components
are listed In fable 24 and discussed below.
TABLE 24. OF
Ammonia
Purge gas from synthesis loop Yes
les
Ho
Natural Gas to Fire Box—
A component analysis of the natural gas used for fire box combustion
revealed that It contained traces of nitrogen. Combustion of this nitrogen
would be a source of nitrons' oxides-.
Total Fuel to Furnace Fire Box-
la addition, to the natufal gas used for fuel, the purge gas from the
synthesis loop of the ammonia plant Is buttled, in the fire box. It is mixed
with the natural gas, and the mixture burns under the same conditions,. The
total component analysis of the furnace fuel gas,is given in. Table 25,
TABLE 25 <, TOTAL ANALYSIS OF THE FuMACE FUEL GAS
Component
Natural gas
Hydrogen
Ammonia
Total Furnace Fuel Components
Mole % Moles /hr
77,92 2500
21.33 684
0.75 24
kg/hr
19 „ 088
620
185
100.00
3208
19 ,,893
66
-------
Conversion of Atmospheric Nitrogen and Ammonia to Nitrogen Oxides in the
Furnace--
At the fire box temperatures in the reformer furnace, the conversion of
atmospheric nitrogen in the intake air to nitrogen oxides is negligible.
Several tests on a reformer furnace operating only on. natural gas corroborate
this statement. These measurements indicated that the KOx in the exhaust'
gases from reformer furnaces was around 35 ppm. Since the conversion of the
nitrogen in the combustion air to nitrogen oxides is minimal, the source of
the nitrogen oxides in the furnace stack must be the purge gases. The
furnace outlet was analyzed for NOx, ammonia,, and methanol. The data for
each of these runs are summarized in Table 26. A total of 8 runs were made
on the furnace exhaust gases prior to the stripper overhead injection.
Previously reported values on similar .plants have found NQx concentrations
ranging from 255 to 320 ppm,
TABLE 26, STACK GAS ANALYSIS PRIOR TO STRIPPER OVERHEAD INJECTION
Run
Number
1
2
3
4
5
6
7
8
Average
Aimaonia
(ppm)
0
0
0
0
0
0
0
0
0
Methanol
(ppm)
0
0
0
0
0
0
0
0
0
NOx
(ppm)
156
233
163
244
160
115
148
157
172
Stripper Overhead Theoretical Conversion to NOx-—
The data shown in Tables 21 and 22 are the average values of all the
data presented in Tables 19 and 20 for the amounts and concentrations of
ammonia and methanol in the stripper overhead. If all the ammonia were
oxidized to NOx in the furnace stack, the quantities indicated in Table 2?
would be expected.
TABLE 27, THEORETICAL CONVERSION OF AMMONIA IN STRIPPER OVERHEAD TO NOx
Anmonia_inOverhead NOxin Stack
mg/1 kg/hr ppm kg/hr
Overhead 4750 37.9 260=7 102.6
Furnace Outlet — — 172 67.7
Total 432.7 170.3
'67
-------
Conversion of Ammonia in Stripper Overhead to NOx—
Previously, the thermodynamics of degradation of ammonia and methanol were
examined as a function of temperature. Based on those calculations, it was
predicted that 99 percent of the methanol and 90 percent of the ammonia would
decompose at the operating stack temperature of approximately 221°C. Table 28
summarizes the average of all stack analyses made to determine the individual
.components of ammonia, methanol, and NOx. This actual stack gas analysis shows
that the ammonia and methanol from the stripper overhead have been reduced
by 59,3 and 74.7 percent, respectively. This amount of degradation for
those two products was somewhat less than that projected from theory. The
partial pressure effects of the other components could account for this
difference.
TABLE 28. AVERAGE STACK EMISSIOM VALUES WITH STRIPPER OVERHEAD IRJECTIOM
TotalStack Gases
ppm kg/hr
Ammonia
Methanol
lOx
39.3
13.4
242.0
15.5
5.3
95.3
According to the Gibb's free energy calculations (Pig. 17)» the decom-
position of ammonia (NH~) to nitrogen dioxide (N0_) in the furnace stack in
the presence of oxygen 102) is highly probable. If 100% of the ammonia
(37.9 kg/hr of NEL) out of the steam stripper overhead were converted to
nitrogen dioxide in the primary reformer furnace stack by the following
equation
2 NH3 + 7/2 02 -> 2 N0'2 + 3 H20
then the 37.9 kg/hr of ammonia would be converted into 102.6 kg/hr (260.7
ppm) of nitrogen dioxide. However, 15.5 kg/hr of ammonia was detected in
the primary reformer stack discharge outlet, indicating a reduction of 22.5
kg/hr and/or a decomposition of 59.2% of ammonia. Also found in the primary
furnace stack discharge point was 5.3 kg/hr'of methanol, indicating a reduc-
tion of 15.6 kg/hr and/or a decomposition of some 74.7% of the methanol into
carbon dioxide and water. However, an increase from 67.6 kg/hr (172 ppm) to
95.3 kg/hr (242 ppm) of nitrogen oxide was observed in the primary furnace '
stack discharge point. This increase of 27.7 kg/hr (70 ppm) and/or 40.9% of
nitrogen oxide can be related to the observed reduction of ammonia. For
example, if the 22.5 kg/hr decomposition of ammonia were converted into
nitrogen dioxide, this would represent an increase of 60.7 kg/hr of NO .
Since only a 27.7 kg/hr increase of N02 was observed, it is assumed that
some of the ammonia decomposed into N, and H_.
68
-------
SECTION 10
ECONOMIC COMPARISONS OF SELECTED TREATMENT SCHEMES FOR REMOVAL
OF AMMONIA FROM PROCESS CONDENSATE
INTRODUCTION
The installation of any process equipment must be justified economically
and environmentally. The economic reasons for installation of capital
equipment are usually dictated by corporate policy. The cost of additional
equipment to recover traces of residual intermediates or product can be
analyzed against actual value in making any economic justification for this
recovery.
Government regulations and popular demand for environmental improve-
ments necessitate the recovery of process wastewater streams through opera-
tional changes and capital additions. Certain criteria should be evaluated
to ensure that the removal of these contaminants from prtfcess wastewater
streams does not cause undue economic hardship on the operations or result
in an unacceptable price increase.
Cost comparisons were made for selected process schemes to reduce the
ammonia and methanol in the process condensate. These process schemes were
as follows:
• Atmospheric steam stripping of process condensate with vapor
injection into furnace stack
• Reinjection steam stripper with injection of the condensed
overhead into the primary furnace inlet*
• Absorption of ammonia on vanadium pentoxide catalysts to produce
aqueous ammonia (28%) and/or anhydrous ammonia by-product upon
regeneration of the catalyst.
• Addition of phosphates and potassium magnesium sulfate to the
process condensate stream to produce a marketable by-product of
magnesium ammonium phosphate fertilizer.
The economic evaluation of these processes was based on capital cost,
raw materials, and operating cost. The amount of land usage for the various
processes was not taken into account in the economic evaluation. In cases
where by-products' were formed, a product credit was given to the process and
deducted from the annual cost of production. In all process cases, the
following conditions were utilized in the economic evaluations:
69
-------
• Straight-line, 10-year depreciation.
• Eight percent interest rate, averaged over the 10-year period.
• Approximately 331 days/year operation.
» Approximately 8,000 hr/yr operation.
• 760 1/min process cotidensate stream with 1000 ppm NH,.,
PROCESS CHARACTERIZATION FOR ECONOMIC EVALUATION
Atmospheric Steam Stripping
Atmospheric steam stripping is a. process that utilizes live steam as
the driving force to strip out the ammonia in the condensate effluent via a
packed column. In this case, the overheads are vented via the furnace stack
of the primary reformer prior to atmospheric discharge. There are several
advantages of atmospheric steam stripping via the furnace stack: it is the
least expensive of all processes to operate; it is the simplest process
scheme and requires least supervision; it requires only a small amount of
process land area. However, with atmospheric steam stripping, there is a
possibility of air pollution by NOx.
Figure 21 shows the process scheme from which the economic evaluation
was made.
Equipment Cost Estimate
Equipment Cost
1. Feed storage tank $12,000
2. Stripping column (packed) 28,000
3. Feed pre^heater - 99 m
§ $376/ni „ 35,000
4. Bottoms - 99 m @ $376/n 35,000
5. Pumps (feed, bottoms, storage) 10,000
6. Assorted piping 10,000
Total $130,000
Total installed cost $350,000
Operating Cost for Steam Stripper-—
The items which make up the operational cost of such a facility are
listed below. The basis of utility cost was set forth in the design criteria.
70
-------
„ _ Feed In
law Process ^_
Ctmdensate (757 1/minT
Packing
Steam In
Stripped Overhead
Steam Stripper
Primary Reformer Stack
(2i5-260*C)
Bottoms Out (10 pptn - 20 pptn NH_)
Figure 21. Atmospheric steam stripper discharge via primary reformer stack.
-------
Operating Cost Per Year
1. Steam Consumption ($6.61/1000 kg @
5,440 kg/hr) $288,000
2. Electricity (50 KWH @ 2.5C/KMH) 10,000
3. Labor (2 men @ $15,000/man) 30,000
4. Supplies and chemicals (1.5%
capital investment) 5,000
5. Supervision (2.0% of capital
investment) 7,000
6. Maintenance and materials (8.0% of
capital investment) 28,000
Total $368,000
Fixed Cost for Atmospheric Steam Stripper—
Two significant costs, depreciation and interest on capital, are
applicable to an economic evaluation of this operation. Start-up ex-
penses, working capital, and general and administrative corporate expenses
increase initial capital requirements. These expenses are outlined
below.
Fixed Cost Per Year
1. Depreciation - 10-year straight-line $35,000
2. Interest - 10 years at 8 percent 15,000
3. Start-up expenses - 10-year amortization 1,000
4. Working capital - 10-year amortization 1,000
5. General and administrative, insurance
and taxes (3% of capital investment) 10,000
Total $62,000
Cost-Benefit Value of Atmospheric Steam Stripping—
Cost Per Year
1. Variable costs (operations) $368,000
2. Fixed cost (depreciation, etc.) 62,000
Total $430,000
3. Recovered credit None
4. Total annual Cost $430,000
5. Cost per liter of water treated $ 0.0012
6. Cost per m. ton of ammonia $ 1.50
Atmospheric Steam Stripping with Reinjection of_ the Condensed Stripper
Overhead into the Primary Furnace Inlet
In the reinjection process, instead of being discharged to the atmosphere,
the overheads of the steam stripper are condensed, pressurized, and reinjected
into the primary reformer. Although there would be practically zero discharge
of pollutant to the environment, there would be an increase in natural gas
requirements; an increase in cooling water and cooling tower usage; and an
increase in tower height, packing material, and foundation strength. Figure 22
72
-------
Stripped Overhead
(757 l/n»in)
Feed In —
Ammonia Process
Conuensate
Cooling
Water
Out
T
Heat Exchanger
Bottoms Out
Vent
Condenser
Cooling Water In.
Product
Compressor
Steam Line
Primary Re former!
Figure 22. Relnjeetion of steam stripped process condensate into primary reformer via
steam injection.
-------
shows the process scheme on which the economic evaluation of this process
was made. Conventional techniques of major equipment pricing and factoring
were used to derive the values listed below.
Equipment Cost Estimate
Equipment Cost
1. Feed storage tank $12,000
2. Stripping column (packed) 28,000
3. Feed pre-heater 35,000
4. Condenser 7,000
5. Bottoms cooler 20,000
6. Separator drum, overheads drum
(collection) 3,000
7. Sparger 2,000
8. Pumps (feed, bottoms, overheads,
storage, sparger) 35,000
.. 9. Assorted piping 20,000
Total $162,000
Total installed cost (also boiler
for heating reinjection to primary
reformer) $600,000
Operating Cost for Reinjection of Overheads—
Items contributing to the operational cost of such a facility are
listed below. The basis of utility cost were set forth in the design criteria.
Operating Cost Per Year
1. Steam consumption ($6.61/1000 kg
@ 7,260 kg/hr) $384,000
2. Electricity (150 KWH @ 2.5
-------
Fixed Costs Per Year
1. Depreciation - 10-year straight line $ 60,000
2. Interest - 10 years @ 8 percent 26,400
3. ; Start-up expenses 15,000
4. Working capital 15^000
5. General and administrative, insurance
and taxes (3% of capital investment) 18,000
Total $134,400
Cost-Benefit Value of Reinjection of Overheads —
Cost Per Year
1. Variable costs (operations) $814,500
2. Fixed cost (depreciation, etc.) 134,400
Total $948,900
3. Recovered credit None
4. Total annual cost $948,900
5. Cost per liter of water treated $ 0.0026
6. Cost per m.> ton of ammonia $ 3.20
Vanadium Pentoxide Catalyst Absorption
In this process, air stripping of the process condensate stream is the
first step. The overhead vapor (NH-, H-0 vapor, and air) is passed through
a bed of vanadium pentoxide catalyst, resulting in the chemical reaction
illustrated by the following equation.
~
(Vanadium (Ammonia) (Water) (Ammonium
Pentoxide) VanaJate)
After absorption of ammonia as ammonium vanadate, the catalyst can be
regenerated by heating to 450°C or steam regeneration at 200°C. However, at
the lower temperature, stable intermediates are formed. According to the
literature (11) , the regeneration of 200 °C indicated about two-thirds of the
ammonia was evolved. Under these conditions, the reaction proceeds as
follows :
The evolved ammonia can be condensed, with enough water subsequently
added to produce aqueous ammonia of commercial strength (28% NH3) ; or the
gaseous stream can be dried over caustic or soda lime, and the resulting
anhydrous ammonia stored as a liquid under pressure.
The process shown in Figure 23 utilizes a triple sequence of fixed
catalyst beds for the ammonia absorption. The air stripped ammonia goes
into the bottom, is absorbed to the V 0 catalyst, and exits the top of the
absorbers (A, B and C) . It can eitheJ oe vented to the air or recycled back
75
-------
(200°C) Steam In
Raw Process
Condensate
Feed In
(757
Process
Air
(23,000 I/sec)-,.-
Bottoms
Out
Steam Out * Steam Out
(A,B,C-Catalyst Beds)
Catalyst
Absorber
-g Cooling
Water In
Cooling Water Out
28% Aqueous Ammonia
Product Out
Figure 23. Vanadium pentoxide catalyst absorption.
-------
into the stripper. This process was designed on a three-day cycle. While
absorber A is being used for the absorption, absorber B is being regenerated
and cooled to room temperature for reuse after absorber A has completed its
cycle. The air stream from the unit being used as the absorber could be
used to cool down the unit after regeneration and thus speed up the cycle.
A fixed bed (semibatch) instead of a continuous absorber, was used in
the economic evaluation(12,13). This process would require a smaller capital
investment, but possibly would suffer a greater loss of catalyst. In the
vanadium pentoxide absorber system, there is practically zero discharge of
pollutant to the environment. Further, the system reclaims a product which
was once discarded to the environment. It is a relatively simple process,
and the absorbency power of catalysts upon regeneration is very good.
Disadvantages of the vanadium pentoxide absorber are: (1) the cost of
catalysts is relatively high ($6,75/kg) j (2) regeneration losses could
occur; (3) power requirement for air stipper is high (the process requires
approximately 2.2 m of air per liter of condensate stripper); (3) only
two-thirds of the NH« in catalysts bed is removed per regeneration when 2QQ°C
steam is used, and (4; no removal of metnanol is indicated.
Capital Cost for Vanadium Pentoxide Absorber—
Basts:
1. 331 days/year and/or 8,000 hours/year.
2. Straight-line 10-year depreciation.
3. Interest on capital investment at 8 percent with a 10-year
payout.
4. 760 1/min ammonia process condensate treatment (1000 ppa
ammonia).
5. Initial charge of catalysts included in capital investment
($150,000).
6. Assume 1% loss of catalyst per regeneration. 3
7. Using air stripping prior to catalysts absorbers (2.2 m
air/liter of treated water).
Equipment Cost Estimate
Equipment Cost
1. Feed storage tank $12,000
2. Stripping column (packed) ,S nnn
3. Initial catalysts charge 75000
4. Absorbing column X3) 2s i/'nnn
5. Condenser (11,3 «2 ft $1236/m ) M.OOO
6. Product holding and storage tank 20,000
I'. FaT(2.2 .3 air/liter of treated water) IMjjO
9. Assorted piping §4bo'oOO
cost $1,600,000
77
-------
Operating Cost for Vanadium Pentoxide Absorber—
These calculations followed the pattern used in assessment of the
operating cost of the steam stripper.
Operating Cost Per Year
1. Air stripper power (2047 KWH @ 2.5
-------
Cost-Benefit Value for Vanadium Pentoxide Absorber-
Cost Per Year
1. friable costs (operations) §890,500
^. Fixed cost (depreciation, etc.) 348 400
* « T°t!1 $1,238,900
J. Recovered credit 61 000
A. Total annual cost
j>. u»st per liter or water treated $ 0.003
6. Cost per HI. ton of ammonia $ 3.94
Conversion of gJ3 to Magnesium
rK^-r conversion of ffl to magnesium aomotila phosphate utilizes several
chemicals. The following equation shows the chemical reaction which occurs:
+ 2Ca(OH>2 + 4H2Q
(Sewage, Water) (K-nag) (Lime) (Water)
> 2
(Strivite) (Syngenite) (Gypsum)
This process has been focused on recovery of ansttonia through treatment
of municipal sewage and should be applicable to recovery of aamonia from
the process condensate.
In applying the above equation to the ammonia process, ammonia water
and Ca(H2P04)» would replace the sewage water and Ca(OH)2> respectively.
This modified equation was used in the evaluation of this process. Figure
24 shows the process scheme for which the economic evaluation was nade.
Advantages are that there is practically zero discharge of pollutants to the
environment and a product which was once discarded to. the environment could
be reclaimed. However, a large capital investment is required; a by-product
has to be marketed, a large land area and product storage facilities are
needed.
79
-------
Feed In
Raw
Process-
Condensate
Feed In
(757 1/min
•OS
o
Ca(HPO
K-Mag
To River
or Sewe
Mixing Tank
(30 min retention time)
Bagging
and
Storage
d
Granulator
Heat Out
Dust
Filte
Settling Tank
(30 rain retention time)
Dryer
Heat In ~
Filter
Presa
Recycle Process Water
NH,,Mg P0..6H-0
CaSO, (Slightly Soluble)
Figure 24. Magnesjijm, ammonium phosphate process.
-------
The procures described above .era used to make the followiag
Capital Cost for Magnesium Ammonia Phosphate—
Equipment.Cost 'Estimate/
Equipment Cost
1. feed storage tank $12 000
*3 1x.* - * " *"*" t ** vW
2- Mixing tank 15 oon
3. Settling tank "'**
4. Filter press 5o;ooo
!' ^ryer1 20,000
6. Granulator 60 000
7. Pumps 40j000
«. Assorted piping 60,000
9« Stirrers and mixers 40 000
10. Bagging and storage 100[OOP
Total $404^000
Total installed cost $1,500,000
Operating Cost for &naonit« Phosphate—
Operating Cost Per Year '
1- law materials
2. R-aag $235,500
3. Ca
-------
Magnesium Ammonium Phosphate Recovered—
1. Flow
2. NH3
3. Phosphate
4. Magnesium (K-mag)"'
5. Magnesium ammonium phosphate
6. Magnesium ammonium phosphate
at 5.5c/kg
$45,360 kg/hr
45.4 kg/hr
449 kg/hr
725 kg/hr
5,226,000 kg per year
$288,000 per year
Cost-Benefit Value for Magnesium Ammonium Phosphate—
Cost Per Year
1. Variable costs (operations)
2. Fixed cost (depreciation, etc.)
Total
3. Recovered credit
4. Total annual cost
5. Cost per liter of water treated
6. Cost per m. ton of ammonia
$1,380,900
325.400
$1,706,300
288.000
$1,418,300
$ 0.004
$ 4.41
SUMMARY OF ECONOMIC EVALUATION
A summary of the economic evaluations for each process is presented in
Tables 29 and 30. As indicated in Table 29, the atmospheric steam stripper
is the least expensive to operate; the magnesium ammonium phosphate process
is the most expensive. Using the atmospheric steam stripper as the basis,
Table 30 shows the cost ratio of each process and also the cost per liter to
treat process condensate prior to river discharge.
TABLE 29. ECONOMIC EVALUATION OF VARIOUS PROCESSES
Process
Subtotal Product Total
Cost/Yr Credit/Yr Cost/yr
Atmospheric Steam Stripper $ 449,150 None $ 449,150
Reinjection into Primary
Reformer $ 948,000 None $ 948,000
Vanadium Pentoxide
Phosphate $1,706,300 $ 61,000 $1,177,900
Magnesium Ammonium ' •-
Phosphate $1,706,300 $288,000 $1,418,300
82
-------
TABLE 30 . PROCESS COST RATIOS AND COST PER LITER OF INFLUENT^
Liters/Yr
Process Cost Ratio Processed Cost/Liter
Atmospheric Steam Stripper 1 363,360,000 $0.0012
Reinjaction into Primary
Reformer 2.11 363,360,000 $0.0026
Vanadium Pentoxide Catalysts 2.76 363,360,000 $0.003
Magnesium Ammonium Phosphate 3.79 363,360,000 $0.004
83
-------
REFERENCES
1. Samples, W.R. , Chem. Eng. Prog., Symp. Ser. (53:78 (1967).
2. Bingham, E.G., ejt al., Chem. Eng. Prog., Sym. Ser. 6J:107 (1970).
3. Eckenfelder, W.W., Chem. Eng. Prog., Sym. Ser. 63:78 (1967).
4. Johnson, W.K. , purdue University Eng. Exptl. Sta. Bull., No. 96, 151-162
(1959).
5. Bingham, E.G. and R.C. Chopra, International Water Conference, The
Engineers' Society of W. Pa., 32nd Annual Meeting, Pittsburgh, Pa.,
November 4, 1971.
6. Atkins, P.F. and U.A. Scherger, Ammonia Trmobsl in a. Physical Chemical
Wastewater Treatment Plant, presented at the 27th Purdue Industrial
Waste Conference, Lafayette, Indiana, May 1972.
7. Rohlich, G.A. and R.A. Taft, Sanitary Eng. Center Report W 61-3,
130-135 (1961).
8. Gulp, Gordon, and Selecta, Bull. Calif. Water Pollution Control Assn. 3^
10-24 (1967).
9. Public Works 97 90-92 (1966).
10. Finneran, J.A. and P.H. Whelchel, Industrial Process Design for Water
Pollution Control, Chem. Eng. Prog., Sym. Ser. &5:79 (1971).
11. Envirogenics Technical Brief, "Envirogenics Process for the Removal and
Recovery of Ammonia from Wastewater," March 1976.
12. Chindgren, C.J., L.C. Bauerle, and B.K. Shibler, "Calcium Vanadate
Precipitation in Processing," Bureau of Mines Report of Investigation
No. 7058, December 1967.
13. Envirogenics Technical Brief, March 23, 1973.
84
-------
tp, TECHNICAL REPORT DATA
("g<"c read Inslructions on the reverse before completing)
EPA-600/2-77-200
3. RECIPIENT'S ACCESSION NO.
[.TITLE AND SUBTITLE ry, , , _ A ~~ — '
I reatment of Ammonia Plant Process
Condensate Effluent
5. REPORT DATE
September 1977
6. PERFORMING ORGANIZATION CODE
AU
C.J.Romero, F.Yocum, J.H.Mayes, and
D. A. Brown (Gulf South Research Institute)
8. PERFORMING ORGANIZATION REPORT NO.
9, PERFORMING ORGANIZATION NAME AND ADDRESS
Louisiana Chemical Association
251 Florida Street
300 Taylor Building
Baton Rouge. Louisiana 70801
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
Grant S 802908
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 7/74-8/77
14. SPONSORING AGENCY CODE
EPA/600/13
s. SUPPLEMENTARY NOTES
Mail Drop 62, 919/541-2547.
ject
thig rt ^ Ronald A Venezia,
16. ABSTRACT
The report gives results of an examination of contaminant content and
selected treatment techniques for process condensate from seven different ammonia
plants. Field tests were performed and data collected on an in-plant steam stripping
column with vapor injection into the reformer furnace stack. Bench scale steam strip-
ping was studied on several different plant process condensates for comparative
purposes. Data for design of a commercial steam stripper were obtained on the bench
scale unit. Design conditions for the commercial unit were given. Four different
methods of treating the stripper overhead were compared. The results indicate that
stripping the process condensate and injecting the vapor into the reformer stack
offers a viable control technology for reducing the amount of ammonia and methanol
discharged to the environment.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
. COS AT I Field/Group
Pollution
Ammonia
Industrial Processes
Condensates
Treatment
Steam
Stripping (Distil-
lation)
Carbinols
Pollution Control
Stationary Sources
Steam Stripping
Reformer Furnace
Methanols
13B
07B
13H
07D
07A
07C
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report 1
Unclassified
!1. NO. OF PAGES
93
Unlimited
20. SECURITY CLASS (Till*page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
85
------- |