How Much Catalyst is needed for

Synthesizing Ammonia
S. SINGH, A. NAHAR
KBR Inc.
Houston, TX, USA

While comparing operating performance of the ammonia synthesis loops of similar production
capacity, plant owners compare the total catalyst volume and the size and number of converters
being used in their synthesis loops. Ammonia converter is one of the major equipment items that
decide installed cost of a synthesis loop besides syngas compressor and refrigeration
compressor. Per pass ammonia conversion, total compression power and heat recovery are key
process parameters often compared by the owners.
KBR recently concluded a study and compared performance of the ammonia synthesis loops
of 2200 MTPD capacity based on make-up synthesis gas produced from (a) Partial Oxidation
Process post nitrogen wash (b) Steam Natural Gas Reforming Process based on Purifier
technology and (c) Steam Natural Gas Reforming Process based on conventional technology.
Catalyst volume, per pass conversion, compression power and steam make are all compared for
these three cases. Capacity utilization of the synthesis loop and catalyst life over the plant life
cycle is also compared.
The study concludes that the inert level in make-up gas, stability and purity of the make-up
gas composition, are the key factors that demarcate the synthesis loops based on three types of
make-up gases into the three distinct classes that determine capital and operating cost of these
synthesis loops.

INTRODUCTION
Owners of various existing ammonia plants compare the synthesis loop in their plants with each other to
reconcile their opinion on cost effectiveness of specific synthesis loop considering the installed and operating
costs. Some of these loops may be unchanged from their original design but others would have gone through
one or more revamps through the life cycle.

- 229 -
 S. Singh, A. Nahar

While comparing operating performance of their ammonia synthesis loops considering the production capacity,
typically owners compare the total catalyst volume and the size and number of converters being used in their
synthesis loops. Besides the synthesis and refrigeration compressors, ammonia converter is the major
equipment item that determines loop installed cost. Per pass ammonia conversion, total compression power
and heat recovery as high pressure steam are the key process parameters often compared by the owners
when evaluating operating cost.
While revamping various existing plants that were originally licensed from KBR or from other technology
providers, KBR has noted wide variations in the existing converter catalyst volumes even after considering
plant capacity differences. To enable owners to reconcile cost effectiveness of their synthesis loop vs. others,
KBR recently concluded a study, by modeling and comparing performance of the ammonia synthesis loops of
2200 MTPD capacity plant based on make-up synthesis gas produced from:
A. Partial Oxidation Process post nitrogen wash
B. Steam Natural Gas Reforming Process based on PurifierTM technology and
C. Steam Natural Gas Reforming Process based on conventional technology
Catalyst volume, per pass ammonia conversion, compression power and steam make are all compared for
these three cases. Capacity utilization of the synthesis loop and catalyst life over the plant life cycle is also
compared in this paper. This paper is based on only the promoted iron catalyst since it is the predominant
catalyst used in the ammonia industry. The study compares only the synthesis loop details, not their syngas
generation frontends.
PROCESS OVERVIEW
Ammonia Synthesis Loop
Ammonia synthesis loops are typically configured to produce warm and/or cold ammonia product. Hydrogen
and nitrogen contents of synthesis gas react in the presence of a catalyst at moderate to high pressure to
produce ammonia per the following exothermic reaction:

3H2 + N2  2NH3 + HEAT

As shown in figure-1, make up synthesis gas is compressed in a centrifugal compressor and combined with
recycle gas and circulated through ammonia converter. The recycle and make-up stages are part of a common
synthesis gas compressor train. The circulated stream is preheated in a feed effluent exchanger and fed to the
ammonia converter. Ammonia rich synthesis gas from the converter is passed through steam raising waste
heat boiler and then through feed effluent exchanger again, followed by its cooling using cooling water. A
portion of product ammonia is condensed in the cooler before the gas is chilled to low temperature using
refrigerant ammonia to recover maximum ammonia by further condensation. Liquid ammonia product is
separated and vapor is recycled after preheating it. A purge stream is drawn to control inert gas buildup in the
synthesis loop before recycling the main gas flow stream to the syngas compressor. An ammonia refrigeration
system including refrigeration compressor is provided to meet refrigeration duty of the synthesis loop.

230 Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012)
 How much catalyst is needed for synthesizing ammonia

Fig 1: Ammonia Synthesis Loop

Make Up Gas

SYNTHESIS GAS
COMPRESSOR REFRIGERATION
COMPRESSOR

HEAT
RECOVERY

Ammonia
Product

HORIZONTAL MAGNETITE CONVERTER

Purge
Gas

Synthesis Gas Generation

Ammonia is commercially synthesized using make-up synthesis gas from three distinct processes used in the
syngas generation section also called frontend of the plant:
A. Partial oxidation (POX) or gasification of heavy oil, coal or pet coke
B. PurifierTM process based steam methane reforming
C. Conventional steam methane reforming
All processes can be broadly included in either type A, B or C for the purpose of this paper.
A. Pox Process
When syngas is to be produced from coal, heavy oil, pet coke or similar feedstock, partial oxidation (POX) or
gasifier process is used to produce syngas. Such plants typically have an air separation unit (ASU) that
supplies oxygen to the POX unit. After purifying the raw syngas through various units, a mixture of H2/N2
syngas in desired molar ratio of 3:1 is produced by using nitrogen from the ASU through nitrogen wash as
shown in Figure-2. Make-up syngas from such front ends have very high purity with stable gas composition.
Air AIR SEPARATION Nitrogen
UNIT

NH3
O2 Product

GASI - PURIFI- SHIFT & N2 COM- MAGNETITE

FICATION CATION CO2 REM
WASH PRESSION SYNTHESIS

Purge
Feed Stock

Fig 2: Syngas production by POX and N2 wash

Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012) 231
 S. Singh, A. Nahar

B. PurifierTM Process
When syngas is to be produced from natural gas or from naphtha or such feed-stocks, Purifier process having
mild steam methane reforming (SMR) and a cryogenic purifier cold box is used to produce syngas. Such
plants typically have a mild primary reformer furnace followed by a secondary reformer using excess process
air flow, 50% more than conventional plants. Flowsheets using PurifierPlusTM process should also be
included in this category. After processing the raw reformed gas through various units, a mixture of H2/N2
syngas in desired molar ratio of 3:1 is produced at the purifier outlet as shown in Figure-3.
Make-up syngas from such front ends has high purity with no methane and minimum argon inert. The
composition is stable since the syngas is washed with liquid nitrogen inside the purifier cold box. Inert gas
content inlet of ammonia converter is maintained by purging gas that is fed to the purifier cold box to maximize
H2 recovery.
C. Conventional Steam Methane Reforming Process
When syngas is to be produced from natural gas or from naphtha or such feed-stocks, conventional steam
methane reforming (SMR) process can be used to produce syngas. Such plants typically have a primary
reformer furnace followed by secondary reformer having process air supply from a compressor. Flowsheet
having all type of heat exchanger based steam methane reformers should also be included in this category.
After purifying the raw reformed gas through various units, a mixture of H2/N2 syngas in desired molar ratio of
just below 3:1 is produced by fixing process air flow to secondary reformer as shown in Figure-4.
Make-up syngas from such front ends have lower purity with significant methane and argon inert and have
composition (including H2/N2 ratio) variations during operation. Significant inert contents of make-up gas
require much larger purging from the synloop to avoid their accumulation.

Excess Air Recycle Purge

PRIMARY SECONDARY SHIFT & METHANATION COM-

REFORMER REFORMER CO2 & DRYERS PURIFIER PRESSION SYNTHESIS
REMOVAL

NH3
Waste Gas to Fuel Product
Feed & Steam

Fig 3: Syngas production by purifier process based steam methane reforming

Many synthesis loops are provided with purge gas recovery unit (PGRU) that recovers H2 from the purge gas
that is recycled to the make-up syngas. Two types of PGRU are typically used (a) Membrane (b) Cryogenic.
Variations in H2/N2 ratio inlet ammonia converter from the optimum value due to variations in the make-up gas
compositions are corrected gradually by adjusting flow rate of process air to secondary reformer. Variations in
inert content inlet ammonia converter from the optimum value due to variations in the make-up gas
compositions are corrected gradually by adjusting purge gas flow rate.
Air

PRIMARY SECONDARY SHIFT & METHANATION COM-

REFORMER REFORMER CO2 & DRYERS PRESSION SYNTHESIS
REMOVAL
NH3
Product
OPTIONAL PURGE
To Fuel GAS RECOVERY UNIT
Loop Purge
Feed & Steam

Fig 4: Syngas production by convention steam methane reforming

232 Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012)
 How much catalyst is needed for synthesizing ammonia

SYNTHESIS LOOP CONFIGURATION

Using make-up synthesis gas from the different frontend syngas generation processes, ammonia synthesis
loops were configured following these criteria-
1. All the synthesis loops produce 2200 MTPD warm ammonia product. Cooling water supply
temperature is 34 degC.
2. Syn-loop configurations in all cases are “dry” type. Make-up gas is already purified (oxides removed)
during/before compression so it can be directly fed to ammonia converter.
3. Higher per pass ammonia conversion is preferred in each case as applicable, since it improves energy
efficiency. Additional loop configurations are made to study sensitivity.
4. Syngas gas compressor suction pressures and temperatures are hypothetically assumed the same for
all the cases and machine efficiencies are assumed the same to simplify the comparison exercise.
5. Compression power is reported as total of syngas and refrigeration compressors.
6. Three stage intercooled adiabatic converter (KBR horizontal) is used in all the cases. All the gas is
passed through all the beds and small size 1.5-3.0 mm catalyst is used to achieve maximum
conversion in given converter volume. While maximizing per pass ammonia conversion in
conventional cases, the converter size is restricted to maximum justifiable size to keep the synloop
cost effective. The use of additional KBR radial converter as 4th bed in series is avoided as far as
possible. However 4th bed has to be used for the sensitivity cases for conventional frontends if the
ammonia conversion has to match with the level of the other synthesis loops. The sensitivity cases are
provided for comparison purposes.
7. Heat of reaction is recovered by raising high pressure steam at 124 kg/cm2 and preheating boiler feed
water.
MAKE-UP GAS COMPOSITIONS
Make-up gas compositions as shown in Table-1 are supplied by respective frontend sections to produce 2200
MTPD ammonia in synthesis loops configured based on the uniform criteria described above.
Table 1
Make-up gas comparison
POX Purifier Conventional Conventional Conventional
Process SMR – No SMR – SMR –
PGRU Membrane Cryogenic
PGRU PGRU
H2/N2 molar ratio 2.99 2.99 2.99 2.85 2.85
Inert CH4+Ar mol% <0.01 0.3 1.3 1.2 1.2
Oxide impurities ppm small negligible <5ppm <5ppm <5ppm
Flow rate KNM3/h Note-1 242.8 245.1 269.1 255.4 254.8
Composition stability in High High Low Low Low
operation
Note-1: Makeup gas flow is given at suction of syngas machine excluding H2 stream recovered from PGRU. Actual flow in Purifier is 257.6
KNm3/h including H2 recovered through the coldbox.
Inert in the make-up gas are significantly higher in the case of conventional process with or without PGRU
since methane content exit the methanator is not washed with liquid nitrogen; the same reason applies for
relatively higher oxide impurities (i.e. H2O, CO and CO2). High oxide impurities adversely affect the synthesis
catalyst in long run. Overall stability of the make-up gas composition (i.e. H2/N2 ratio and inert) over time will
also be significantly less in case of conventional process caused by various external changes in the frontend

Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012) 233
 S. Singh, A. Nahar

like ambient and feedstock variations. Such external changes affect the operating condition of process air
compressor and consequent process air flow and primary reformer furnace firing situation. Such variations
may change from plant to plant but they are unavoidable as known from operating experience of so many
plants. Such external variations do not affect in case of POX and Purifier processes since make-up gas is
supplied after N2 wash that keeps the makeup gas composition smooth.
CONVERTER SIZE – CATALYST VOLUME
There are number of synthesis loop parameters which can affect the converter size and loop efficiency. These
include pressure, inert gas level and inlet ammonia content (separator temperature). The following will explore
the effect of these on catalyst volume.
Table-2 shows synthesis parameters based on the different make-up gas conditions from the related frontend
sections in each case.
Table 2
Effect of inerts on catalyst volume & conversion at constant pressure
Ammonia Converter & POX Purifier Conventional Conventional Conventional
Synloop Process SMR – No SMR – SMR – Cryogenic
PGRU Membrane PGRU
PGRU
Pressure Kg/cm2 155 155 155 155 155
Inert mol% 0.8 3.5 10 8 7
H2/N2 ratio 3.0 3.0 3.0 3.0 3.0
Per pass NH3 18.0 18.0 15.6 16.4 17.2
conversion % (Note-1)
Catalyst Volume % 100 119 141 141 141
Circulation Flow % 100 100 113 108 104
Compression Power % 100 103 110 110 109
Heat Recovery to HP 100 100 99 99 100
Steam %
Note-1: Per pass conversion is converter outlet minus inlet NH3 mol% keeping inlet NH3 % about the same.
Although larger size three stage converters has been used in case of the conventional frontend, per pass
ammonia conversion is significantly less than that in the other two cases due to high inert contents of makeup
gas present in the conventional case. Converter pressure is kept same in all the cases to keep them on
comparative basis. With a base catalyst requirement using three stage KBR adiabatic horizontal converters for
the POX frontend of 100, purifier frontend requires 19% more catalyst. For the conventional cases, desired
conversion can not be reached in spite of increasing the catalyst volume by 41% over the POX case. This
significant performance difference is clearly due to presence of high inert in the makeup gas of the
conventional cases that is inherent for such frontend technology and can not be matched with the first two
cases. Expectedly the other key parameters for conventional cases, like total compression power, are also
unfavorable.
Catalyst volume sensitivity cases
To make the three conventional cases comparable with the POX and Purifier cases, additional loop
configurations were prepared where per pass ammonia conversion has been matched by either (I) adding
additional converter in series or by (II) increasing loop pressure.

234 Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012)
 How much catalyst is needed for synthesizing ammonia

Option-(i) Add More Catalyst:

Table-3 provides a comparison when catalyst volume is varied keeping the synloop pressure same for all the
cases. As can be seen, a fourth stage adiabatic bed is required to achieve an equal ammonia conversion. As
seen, the catalyst volumes are significantly higher in conventional cases requiring 98%, 72% and 61% more
catalyst for conventional frontends without PGRU, with membrane PGRU and with cryogenic PGRU cases
respectively.
Significant difference in catalyst volume is clearly due to presence of high inert in the makeup gas of the
conventional cases that is inherent for such frontend technology and can not be matched with the first two
cases. Purpose of this sensitivity case is to illustrate effect of inert in makeup gas on catalyst volume if per
pass conversion has to be the same for all the cases.
Option-II Increase Pressure:
Table-4 provides a comparison of results when pressure is varied while ammonia conversion is equal for all
the cases. As seen, the loop pressures are significantly higher in conventional cases at 163 to 180 Kg/cm2
compared to that in POX and Purifier cases of 155 Kg/cm2 both. This significant difference in loop pressure is
clearly due to presence of high inert in the makeup gas of the conventional cases.

Table 3
Effect of inerts on catalyst volume at constant conversion & pressure
Ammonia Converter & POX Purifier Conventional Conventional Conventional
Synloop Process SMR – No SMR – SMR –
PGRU Membrane Cryogenic
PGRU PGRU
Pressure Kg/cm2 155 155 155 155 155
Inerts mol% 0.8 3.5 10 8 7
H2/N2 ratio 3.0 3.0 3.0 3.0 3.0
Per pass NH3 conversion 18.0 18.0 18.0 18.0 18.0
% (Note-1)
Catalyst Volume Main 100 119 141 141 141
Converter %
Catalyst Volume 4th Bed % Not Not 57 31 21
require required
d
Total Catalyst Volume % 100 119 198 172 161
Circulation % 100 100 100 99 99
Compression power % 100 103 107 108 108
Heat Recovery to HP 100 100 101 100 100
Steam %
Note-1: Per pass conversion is converter outlet minus inlet NH3 mol% keeping inlet NH3 % about the same

Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012) 235
 S. Singh, A. Nahar

Table 4
Effect of inerts on catalyst volume & pressure at constant conversion
Ammonia Converter & POX Purifier Conventional Conventional Conventional
Synloop Process SMR – No SMR – SMR –
PGRU Membrane Cryogenic
PGRU PGRU
Pressure Kg/cm2 155 155 180 168 163
Inerts mol% 0.8 3.5 10 8 7
H2/N2 ratio 3.0 3.0 3.0 3.0 3.0
Per pass NH3 18.0 18.0 18.0 18.0 18.0
conversion % (Note-1)
Catalyst Volume Main 100 119 141 141 141
Converter %
Circulation % 100 100 99 99 99
Compression power % 100 103 112 110 109
Heat Recovery to HP 100 100 102 101 101
Steam %
Note-1: Per pass conversion is converter outlet minus inlet NH3 mol% keeping inlet NH3 % about the same

AMMONIA CONVERSION WITH VARYING MAKE-UP GAS COMPOSITION

Steam methane reforming based frontend of an ammonia plant routinely experiences variations in parameters
due to factors like ambient temperature changes, foaming in the carbon dioxide removal unit, feed flow
variations, furnace firing variations with fuel composition and ambient air changes, steam header pressure
fluctuations and so forth. Make-up gas quality exit the methanator, i.e. H/N ratio and methane content can not
always be kept constant. These changes drive synloop parameters such as inert, purge flow, and pressure, to
a suboptimal state during transitions before an operator optimizes the front end again and smoothes out the
synloop. In the case of a conventional technology based frontend, such variations in synloop are inevitable. In
contrast, the synloop of a POX or Purifier technology based frontend is decoupled from its front end as H/N
ratio and methane/inert of makeup gas are controlled independently of air flow to the secondary reformer and
methane exit the methanator. More details are described in previous papers1,2, so will not be covered here.
In ammonia synthesis loops operated to maximize daily ammonia production, the above inherent difference
leads to a lower daily production in plants having conventional steam methane reforming frontends. This
fundamental difference is less glaring in plants that are not pushed to maximize daily production, but otherwise
the actual production is estimated to be higher in a Purifier Plant with the same degree of operator
intervention. Additional catalyst provided in the synloops will not avoid such production loss as variations in
suboptimal H/N ratio and inert inside the loop will affect ammonia production during transient operation. The
impact of routine front end variations on ammonia production in case of conventional frontend is illustrated
below. Figure 5 shows increased methane concentration and reduced H/N ratio in the make-up gas. The
operator intervenes by adjusting purge flow and troubleshooting the front end problem. The results are
illustrated in Figure 6. Ammonia production is lost during such upset in the conventional case in spite of having
larger catalyst volume.

236 Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012)
 How much catalyst is needed for synthesizing ammonia

120.0%
HN Ratio
115.0%
CH4%
110.0%
105.0%
100.0%
95.0%
1 2 3 4 5 6 7 8

Fig 5: Upset Conditions in Methanator Effluent Composition

120.0%
110.0%
100.0%
90.0%
80.0% HN Ratio
70.0% Inert%
60.0% NH3T/D
1 2 3 4 5 6 7 8

Fig 6: Response of Synloop to Upset Conditions in Figure 5 for a Conventional Plant

CATALYST LIFE
Based on feedback from various plants, it is found that in spite of starting with much smaller catalyst volume at
the beginning of run, the performance of ammonia synthesis catalyst does not deteriorates as quickly over
time in synthesis loops being fed from the POX and Purifier based frontend due to high purity of their make-up
gas and inherent removal of oxide poisons. Catalyst life is significantly longer in such plants in spite of the fact
that the conventional ones start with much more catalyst volume at the beginning of run.
CONCLUSION
The findings of this KBR study assist the owners of various existing ammonia plants who may compare the
synthesis loops in their plants with each other to reconcile their opinion on cost effectiveness of specific
synthesis loop considering the installed and operating costs. The study concludes that
(1) Inert level of make-up gas,
(2) Stability of the make-up gas composition, and
(3) Sustained purity of the make-up gas (i.e. oxides)
are the three most important factors that demarcate performance of synthesis loops based on the three types
of make-up gases depending upon type of their frontend design used.
Synthesis loop based on a frontend using partial oxidation (POX) process with nitrogen wash has the minimum
catalyst requirement compared to the other two cases due to negligible inert in the makeup gas. This case has
all the advantages over the rest with lower capital and operating costs.
While the Purifier and conventional processes produce synthesis gas from natural gas in frontend by steam
reforming, synthesis loops based on Purifier process based frontend have significant capital and operating
cost advantages over synthesis loops with conventional process frontends, i.e.
o Catalyst volume is significantly lower in case of Purifier process, due to lower inert in makeup gas
o Compression power hence energy consumption is lower as well in Purifier case
o Utilization of the design production capacity on life cycle basis is higher for Purifier case due to stable
make-up gas composition

Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012) 237
 S. Singh, A. Nahar

o Catalyst life is longer in spite of starting with a lower catalyst volume in case of Purifier and POX due
to better make-up gas purity
Thus clearly the ammonia synthesis loops receiving make-up gas from POX and Purifier process based
frontends have lower capital and operating costs than the conventional cases.
References:
1.Grotz, B.J.; “The Purifier; A Key to Smooth Operation”, Nitrogen, September/October, 1995
2.Singh, S. & Gosnell, J.: “Life Cycle Benefits of PurifierTM Technology based Ammonia Plants”, Nitrogen & Syngas
Conference 2010

238 Nitrogen + Syngas 2012 International Conference (Athens, Greece, 20-23 February 2012)