You are on page 1of 6

CO 2 emissions reduction

Improving energy
efficiency in an
ammonia plant
D. Velzquez, F. Rossi and J. Rodrguez of DVA Global Energy Services and F.Galindo of Fertiberia
present the results of an energy study carried out in an ammonia plant in Spain. A comprehensive
energy audit of the main equipment (furnaces, syngas, air and ammonia compressors, steam
turbines, cooling towers and refrigeration systems), pinch analysis and steam modelling resulted
in the proposal of 34 projects, some of which have cost savings of more than one million euros
per year with payback times of less than one year.

T
he directives 2003/87/EC and
2009/29/EC1,2 set strict CO2 emis-
sion limits for some chemical indus-
tries and can have a significant impact on
companies profits. Because of this, as well
as the increase in the price of natural gas
in recent years, many European ammonia
producers have been forced to carry out
important energy efficiency improvements,
in order to maintain their competitiveness
within the international market.
During recent years, the average spe-
cific consumption of ammonia production
has been globally quantified as 36.6 GJ/
tNH3 (LHV base). Performance indicators
of ammonia plants located in Canada and
western Europe show the best values3.
Within the most efficient regions, natu-
ral gas costs represent more than 80% of
total ammonia production costs, reaching
90% in some cases. Fertiberias ammonia plant in Spain.
Current best available technologies
for ammonia production from natural gas mal energy recovery in the process and The ammonia produced is either exported
allow specific consumption levels of about hence reduction of fuel consumption. or consumed by a urea plant located in the
28 GJ/tNH3 to be achieved4. same facility. Its utility services are also
Among the methodologies aimed at Plant description connected to a nitric acid plant. Figure 1
finding energy saving opportunities, pinch shows a simplified diagram of the process.
analysis linked to power and steam model- Process description
ling has proved to be a powerful way for The ammonia site for which this energy Energy structure
determining projects to improve the over- audit has been carried out was designed Steam reforming is an endothermic proc-
all energy efficiency of industrial sites. This to produce 1,130 t/d of ammonia, via ess, which is carried out at high tempera-
procedure has been applied successfully catalytic steam reforming of natural gas, tures. Thermal energy demand is supplied
in many industrial facilities, allowing opti- using Kellogg technology with ICI license. by a furnace located in the first reform-

Nitrogen+Syngas 322 | March-April 2013  1


CO 2 emissions reduction

Fig 1: Simplified diagram of the ammonia production process

air
reformer I reformer II
steam conversion
combustion CO CO2
air NG NG
aux boiler

NG

NG boiler feed super- steam HRSG


water heated
steam HRSG

desulphurization

solution

synthesis steam CO2 removal


compressor
methanation
COx CH4
solution + CO2

separation
NH3
NH3 water

Fig 2: Main inlet and outlet energy flows involved in the ammonia production process
Stream Energy losses 3%
Feed (NG) 58,3% 60.2%
process
Inlet Fuel (NG) 39,8% 58.3% NH3
98.1%
Electricity 1,9% natural gas
1.9% 39.8%
total
Ammonia 60,2% energy electricity
Cooling water 25,9% combustion

Steam export 6,1% flue gas 4.9%


Outlet
Flue gas 4,8%
steam cooling
steam export 6.1% system
Other loses 3,0% water
25.9%

ing step. This furnace is the main energy erate high pressure steam downstream of In addition to natural gas, the purge
consumer at the site, consuming more the reforming unit, as well as after the CO- stream of the synthesis loop is also used
than 70% of the overall fuel supplied to to-CO2 conversion unit, where exothermic in the fuel feed network, as it contains high
the factory. Combustion gases from this reactions take place. levels of hydrogen. Due to the exothermic
equipment are sent to a gases channel Additional high pressure steam is reaction of ammonia production, the ther-
where thermal energy is recovered through generated in an auxiliary boiler, whose mal energy contained in the ammonia
several heat exchangers that preheat proc- combustion gases are sent to the same synthesis reactor outlet stream is used to
ess streams and generate steam at differ- combustion exhaust gas channel of the preheat the boiler feed water.
ent pressure levels before gases are sent reforming furnace. The auxiliary boiler is As shown in Fig. 2, heat supplied to the
to the stack. There is no requirement for the second largest energy consumer of process through the combustion of natural
thermal energy in the second stage of the the facility, representing over 20% of total gas is subsequently recovered for steam
reformer (secondary reformer), since it is energy consumption. The third-largest production (power production) and heating
supplied by the combustion reactions pro- thermal energy consumer is the furnace combustion air and other process streams.
duced by introducing process air into the upstream of the desulphurisation reactor, Unrecovered heat is removed by cooling tow-
syngas stream. Heat recovery boilers gen- where natural gas is heated. ers and air coolers. High pressure steam is

2 Nitrogen+Syngas 322 | March-April 2013


CO 2 emissions reduction

Fig 3: Steam model


reformer

aux boiler GROSS heat 36,441 MW 3,4 t/h


total all NG users 30,012 aux boiler absorbed heat 25,417 MW BlowDn
process heat to HPS generation 150,1 MW
total heat to HPS generation 175,61 MW aux steam 69,5 t/h
process generation 159,9 t/h
HS total SS generation 229,4 t/h
100,2 bara 3207
437,3 C
3207 kJ/kg 6,0 t/h 223,4 t/h
2742 kJ/kg dH 11,935 MW 103-JTB 2,57 MW
103-JTA total 103 JT
232,8 t/h 9,36 MW
220,9 t/h delta P 36,7 bara 16,1 t/h
3054 3,0 bar 331,1 C 3054 2479
leakage
204,8 t/h
MS 6,0 t/h 3054 kJ/kg
39,7 bara 3207 kJ/kg
335,9 C
3058 kJ/kg 6,7 t/h 13,8 t/h 0,9 t/h 81,3 t/h 108,0 t/h
0,14 MW 14,12 MW
118 JT 101 J 39,31 t/h MS traps
0,50 MW 116 JT 102 VTFT 105 J 26,2 t/h MS users 3,6 t/h
1180 JT 104 J 9,1 t/h
105 JLOT 2527 107 A 3,4 t/h 2434 flash
103 JLOT 2,5 t/h 107 B 3,4 t/h
2,1 t/h
ES 6,7 t/h 13,8 t/h 3207 kJ/kg
2741 kJ/kg
4,2 bara 3058 kJ/kg 2929 kJ/kg
255,8 C
2976 kJ/kg 5,4 t/h 0,0 t/h 5,3 t/h 14,3 t/h
control T 1,0 t/h
vent 2568 145,7 C desup 15,3 t/h
0,60 MW ES users
101 BJT 0,1 bara 2562 0,1 bara 2441 urea plant 2,5
6,3 t/h 97,4 t/h 111-CA/B 8,0
4187 kW 61765 kW 1160-C 2,0
1180-E 103-L 1161-C 0,6
5,0 t/h
0,6 t/h NEW 0,0
614 kJ/kg
159 kJ/kg 159 kJ/kg traps 2,2
106-C 38,0 C 38,0 C
95,7 C 181,1 t/h 95,7 C 166,7 t/h 100,1 C 42,0 C
401 kJ/kg 401 420 176
0 kW 12240 kW
desuperheaters
1,0 t/h 47,4 C
14,0 t/h 109-C
233,8 t/h to REF jackets 199 20 C 76,9 t/h treated water
465 kJ/kg 84 kJ/kg
109,1 C 2446 kW
47,9 t/h
934 kW 411 kJ/kg 44,0 C
steam driver 14,4 t/h
184 kJ/kg 223,4 t/h 11,935 MW 103-JTB 2,57 MW
from urea plant

103-JTA total 103 JT


37,3 t/h stripper condensate
9,36 MW
delta P 36,7 bara 16,1 t/h
99,0 C 415 kJ/kg heat loss 634 kW
220,9 t/h 2479
10,6 t/h
95,0 C
ES condensate
398 kJ/kg leakage 3054 3,0 bar 331,1 C 3054
145,7 C 614 kJ/kg

204,8 t/h

mainly consumed by turbocompressors for


Table 1: Projects summary
process air, syngas, the ammonia cooling
system and injection to the reforming proc-
Area Number of Maximum saving Payback
ess. A significant amount of the steam is projects (ke/year) (years)
also exported to the urea plant.
Pinch projects 9 1.214 0.3
The natural gas consumption for this
site represents more than 95% of the Natural gas saturation 1 437 <1
overall energy consumption of the factory. Cooling towers enhancement 1 63 0.7
The main objective of the proposed energy Natural gas expander 1 285 1.9
efficiency assessment is to reduce natural
Auxiliary boiler combustion control 1 209 <1
gas fuel consumption, corresponding to
39.8% of the total energy inlet. Insulation maintenance 1 17 <0.5
Steam trap repairing 1 78 <0.5
Approach Compressors and absorption cooling 19 106 >7
Equipment audit Total 34 2.409 <2
The largest energy consumers were identi-
fied, with the aim of performing an energy
efficiency analysis and subsequent efficiency to the steam model that was developed of the different energy savings projects
enhancement assessments for each one of for the thermal and power systems of the identified in the previous stages of the
them. Furnaces, boilers, compressors and factory. The study also considered the study. The inputs to the model are linked
cooling systems in the facility were evalu- identification of appropriate residual heat to the results obtained from pinch analy-
ated. A steam trap and insulation infrared sources, with the purpose of assessing sis. All steam turbines, heat exchangers,
thermography surveys were also carried out. the feasibility of implementing an absorp- boilers, steam generators and consumers
tion refrigeration system. of the facility are included in the model. In
Pinch analysis addition, the model allows the specifica-
The aim of the pinch analysis is to improve Thermal and power model tion of steam turbine performance curves,
the existing energy recovery from the proc- A model was developed of the steam gen- which are needed for correct prediction of
ess streams to reduce natural gas con- eration and power systems, in order to pre- their performance under different operating
sumption. This analysis is closely linked dict and economically assess the effects conditions.

Nitrogen+Syngas 322 | March-April 2013  3


CO 2 emissions reduction

the target situation, in which the energy


Fig 4: Composite curves
recovery is improved at 20C minimum
approach. Although these values should
2000 pinch: 20C heating: 31 MW cooling: 242 MW not be considered as strict objectives
hot composite to be fulfilled, they still provide a useful
cold composite guideline for energy recovery enhance-
ment from the process.
1500
For the same natural gas consump-
tion in the auxiliary boiler and desulph
temperature, C

urisation furnace, high pressure steam


1000 generation could be increased by 7.9 t/h.

Auxiliary boiler
Savings shown in Table 2 suggest the
500
potential for reducing natural gas con-
sumption in the boiler while maintaining the
same total steam production (229.6 t/h).
0 Load reduction in the boiler can be
0 100 200 300 400 compensated by enhancing thermal energy
enthalpy, x 000kW recovery from the process through the
installation of new heat exchangers.

The model was developed in Excel for- grand composite curves which are shown Thermal and power effects in the
mat, which makes it easy to implement in in Fig. 4. The composite curves of the proc- primary reformer
any industrial facility, and represents a fast ess represent the process energy demand All combustion gases of the plant are
and accurate calculating tool that can be (heating and cooling) versus their tempera- sent to the primary reformer gases chan-
used for economic assessment of changes tures. These curves are obtained by adding nel, where thermal recovery takes places
introcuced in the plant. The reliability of the the demand of hot streams, which need to for the heating of different streams. The
model is based on the possibility of using be cooled, and cold ones, which oppositely stack of the desulphurisation furnace is
real operating data to adjust the efficiencies need to be heated, for each temperature connected to the gases channel in the
and consumptions of each unit. Figure 3 interval. last section, affecting only air heating. It
shows the main interface of the model. In spite of the proper level of exist- is worth noting that a reduction of natural
ing energy recovery in the process, a big gas burned in the boiler and/or primary
Results potential improvement was detected by reformer will have important effects on the
means of retrofitting the existing heat performance of heat exchangers used for
Table 1 shows a summary of the main exchanger network. For the analysis, a energy recovery in this duct. The gas chan-
energy saving projects identified and minimum approach of 20C between com- nel was simulated and it was observed
assessed in the analysis. posite curves was chosen. Figure 4 shows that the exchangers most affected are the
the optimised energy recovery case, corre- boiler feed water and combustion air pre-
Pinch analysis sponding to 20C minimum approach. heaters. Since steam generation is linked
to the whole steam and power model, the
Pinch analysis allows the quantitative Potential savings global steam balance is affected once the
assessment of existing energy recovery, Table 2 shows potential savings and amount of burned natural gas is changed.
graphically represented by hot and cold objectives of energy consumption for This also affects the flow rate of high pres-

Table 2: Potential savings

Utility Base Target energy use Energy saving potential


kW t/h kW t/h KW t/h
Natural gas 30.012 30.012 0
Low pressure steam (consumed) 6.254 10,6 0 0 6.254 10,6
High pressure steam (generated) 175.543 229,6 181.583 237,5 6.040 7,9
Condensate 13.798 13.798 0
Air 14.562 14.562 0
Cooling water 109.167 101.955 5.913
NH3 13.682 12.383 1.299

4 Nitrogen+Syngas 322 | March-April 2013


CO 2 emissions reduction

sure steam expanding through the tur-


Table 3: Pinch analysis projects summary
bines and therefore generated shaft work.
Heat recovery in the gases channel was
N Project Saving Payback
therefore carefully modelled with the aim (ke/year) (years)
of accurately quantifying these effects and
P1 Low pressure steam (LPS) generation in gases 729 0,2
linking it to the general steam and power channel to reduce auxiliary boiler natural gas
model. consumption
P2 Increasing combustion air preheating 191 1,2
Projects
Table 3 summarises the identified and P3 P1+P2 874 0,6
proposed projects resulting from pinch P4 Natural gas preheating 281 1,4
analysis. P5 P1+P4 738 0,7
Nine different projects were identified,
P6 Increasing boiling feed water preheating 445 0,6
which indicated relevant reductions of
natural gas consumption at the site. Eco- P7 P1+P6 1.214 0,4
nomic savings achieved by single projects P8 P4+P6 725 1,4
are between e200,000 and e1,200,000, P9 Increasing boiling feed water heating 628 0,8
with payback times of less than one year
for most of them. The high profitability of
one of the proposed projects is worth high-
lighting and is described in the following tor in the gases channel with the aim of tion, total savings increase up to e801,000
section. This project considers the instal- achieving 9% reduction of fuel consump- (CO2 emission tax: 15e/t).
lation of a new low-pressure steam genera- tion, while maintaining steam generation. The project investment for a new LPS
tor in the gases channel. The amount of LPS generation predicted by generator, piping and insulation is of the
the steam model was 7.2 t/h. The appro- order of e155,000, which means a pay-
Low-pressure steam generator in gas priate place for the installation of the LPS back time of less than four months.
channel generator in the gases channel is before
As previously mentioned, pinch analysis the last thermal recovery section, dedi- Equipment audit
showed the possibility of increasing high cated to combustion air preheaters.
Compressors
pressure steam generation and reduc- The reduction of fuel consumption affects
ing the natural gas consumption by 9% the energy recovery through the gases chan- Two energy efficiency enhancement
in the auxiliary boiler. Further reduction of nel downstream of the auxiliary boiler, alter- projects resulted from the compressors
natural gas consumption would affect the ing the operating conditions of all the coils assessment, both of them aimed at reduc-
heat recovery process through the exhaust downstream. Figure 5 shows these new ing the consumption of their driving steam
gases channel, preventing the possibil- operating conditions. In the proposed situa- turbines:
ity of achieving a new thermal and power tion, combustion gases will be exhausted at l Surge control improvement for the
equilibrium in the whole plant unless major a lower temperature (approximately 125C) ammonia refrigeration centrifugal
investments were carried out to revamp the than the current 154C. A total cost saving compressor. During periods of cold
current heat exchanger network. value of e729,000 per year for the project weather (5 months per year) the com-
This project proposes the installation of has been calculated. Considering cost sav- pressor operates at reduced loads,
a new LPS (low pressure steam) genera- ings associated with CO2 emissions reduc- which causes the surge control to act

Fig 5: Simulation of gases channel and steam generation

212,0 C 227,0 C 278,1 C


91,38 89,08 311,8 C
SR-Air2 New LPS new Air %O2 Basis 6,0%
227,9 t/h Hact = 2729 %O2 Now 5,0%
new BFW
4552 kW 0 kW Reformer Eff'y 50,0%
109,1 C 227,9 t/h MCP 2,30 109,1 C 147,1 t/h 286,4 C 27272 kW 101-CA
465 kJ/kg 465 1123-C 1269 25327 kW 101-CB
157,7 C 458,0 C
0 kW 32856 kW 62,3 t/h 7404 kW 102-C
103-B Flue Gas 8344 kW 103-C
29,2 t/h 1409
114-C blowdown to HP gen 149028 kW excluding Aux Boiler
6520 kW aux Boiler 3,4 t/h to HP gen 171883 kW including Aux Boiler
losses 22855 kW aux Boiler
143255 kg/h to rad 1901 kW 1269 0 kW compensating NG for lost heat to combn Air, SR-airProc & SR-GN
60489 kg/h to aux 51,5 t/h 286,8 C 22855 kW 224,5 t/h 0 kg/h extra NG for reformer combustion
60,56 MCP 157,7 C 575,0C 5,0% O2 438,2C 625,6C 473,0C
63242,35 kg/h 3207 0 kW 0 kW required compensating NG fuel

126,7 C 212,0 C 227,0 C 278,1 C 278,1C 271779 kg/h 407,2 C 542,0C 532,0C 770,5C 869,4C 993,3C
MCP 91,38 89,08 89,08 89,08 74,61 74,61 74,61 MCP
SR-Air1 SR-Air2 New LPS new Air SR-BFW SR-Reh1 SR-Reh2 SR-airProc SR-GN 208537
7792 kW 4552 kW 0 kW 11507 kW 12005 kW 17793 kW 7378 kW 9244 kW
lost heat pickup= 1740 kW MCP 2,30 2922 16,37 61,26
29,0 C 157,7 C 458,0 C 350,0 C 175,0C 322,1C
103-B Flue Gas

Nitrogen+Syngas 322 | March-April 2013  5


CO 2 emissions reduction

to prevent it from malfunctions. The


Fig 6: iManergy energy optimisation of operation
installation of an automatic surge con-
trol system was proposed that mini-
mises operating hours with opened
recycle valve. This system adjusts the
set-point of these valves as a function
of current operating conditions and
dynamic surge curves.
l Recovering residual heat to produce
cold glycol water in LiBr/H2O or NH3/
H2O absorption chillers. Results from
the economic assessment of these
projects showed high payback times
(>7 years), mainly due to the high
investment required for heat recovery
equipment, absorption chillers and
dedicated cooling towers.

Cooling water system


The cooling water system is properly oper- demanded by steam reforming and thus Future steps
ated. Nevertheless, it was proposed that the possibility of reducing steam genera-
the existing cooling tower fans be replaced tion of the auxiliary boiler. Estimated sav- The installation of an energy management
with more efficient and better designed ings for this project are e437,000, with a system (EMS) is recommended with the
ones. This results in a reduction of the payback time less than one year. aim of maintaining the achieved savings
electric consumption of the driving motors. predicted in the study. The objective of the
The estimated saving was e63,000/year Natural gas expander EMS is to keep track of the energy effi-
with a payback time of less than one year. Natural gas fuel is currently expanded from ciency of the whole site, main equipment
Improvements were also detected in 45-50 barg pressure to 2 barg through a and areas, defining achievable objectives
the cooling network design; a new configu- valve system. Therefore, the possibility of for energy efficiency and acting to correct
ration was proposed, in which warm water using this expansion to produce electric the deviation from those defined targets.
is substituted with cold water for steam power was evaluated. The nominal electric In an energy management system it
condensing. power of the proposed expander is 450 is of paramount importance to properly
kW, which generates economic benefit of define the main energy efficiency indica-
Combustion control e285,000/year. The investment associ- tors (EEI) and its variables of influence (VI).
Combustion control was operated manu- ated with the acquisition and installation The company therefore developed a pow-
ally. Combustion optimisation is not of the expander results in a payback time erful tool, iManergy (Fig. 6), that allows
a priority and was not performed on a of two years. The feasibility of the project the definition of dynamic references for the
continuous basis, the main focus of the depends on the speed and reliability of EEI correlated to its variables of influence.
operators being to maintain the produc- the control system when opening the relief The software allows the client to compare
tion of the plant. Therefore an automatic bypass in case of failure of the expander. its performance with best historical values
control system was proposed for the com- under homogeneus conditions, which sig-
bustion equipment in the auxiliary boiler Maintenance nificantly helps to improve the daily opera-
and primary reformer. The installation of Two main areas were assessed: tional performance of the plant from an
an automatic control system could be dif- l Thermal losses, caused by insula- energy point of view.
ficult considering the age of the primary tion conditions, mainly located in the Ammonia producers, considering the
reformer and the number of burners. The reformer furnace and main pipelines need for continuous improvement in its
proposed project may therefore require a were evaluated. The economic losses operations should consider the installation
complete revamp of the combustion sys- correspond to e17,000/year. of energy manager software. n
tem, thus requiring careful evaluation from l Steam traps do not usually draw the
the company. Total savings of e200,000/ attention of maintenance person-
year were assessed for the auxiliary boiler nel, either because they are not suf-
References
1. Directive 2003/87/EC
and primary reformer, with a payback time ficiently accessible or because they
2. Directive 2009/29/EC
of less than one year for the auxiliary are not considered to have a big effect
3. Benchmarking energy efficiency and car-
boiler. on global energy efficiency. Through bon dioxide emissions. Canadian ammonia
an exhaustive ultrasound analysis of producers.
Natural gas saturator steam traps, among the 300 steam 4. Methodology for the free allocation of emis-
The saturation of natural gas using hot con- traps in the facility, 20 were found to sion allowances in the EU ETS post 2012.
densates upstream of the primary reformer be not working properly, which trans- Ecofys, Fraunhofer ISI, ko-Institut
was simulated and assessed. This meas- lates to economic losses of e78,000 5. Pinch Analysis: For the efficient use of energy,
ure leads to a reduction of steam injection per year. water and hydrogen. Natural resources Canada.

6 Nitrogen+Syngas 322 | March-April 2013