Design of an Ammonia Synthesis System for Producing

DESIGN OF AN AMMONIA SYNTHESIS SYSTEM FOR PRODUCING SUPERCRITICAL STEAM IN THE CONTEXT OF THERMOCHEMICAL ENERGY

STORAGE

Chen Chen Mechanical and Aerospace Engineering Dept. University of California, Los Angeles, CA, U.S.

H. Pirouz Kavehpour Mechanical and Aerospace Engineering Dept. University of California, Los Angeles, CA, U.S.

Keith Lovegrove IT Power

Canberra, Australia

Adrienne S. Lavine

Mechanical and Aerospace Engineering Dept. University of California, Los Angeles CA, U.S.

ABSTRACT Concentrating solar power plants typically incorporate

thermal energy storage, e.g. molten salt tanks. The broad

category of thermochemical energy storage, in which energy is

stored in chemical bonds, has the advantage of higher energy

density as compared to sensible energy storage. In the

ammonia-based thermal energy storage system, ammonia is

dissociated endothermically as it absorbs solar energy during

the daytime. The stored energy can be released on demand (for

electricity generation) when the supercritical hydrogen and

nitrogen react exothermically to synthesize ammonia. Using

ammonia as a thermochemical storage system was validated at

Australian National University (ANU), but ammonia synthesis

has not yet been shown to reach temperatures consistent with

the highest performance modern power blocks such as a

supercritical steam Rankine cycle requiring steam to be heated

to ~650°C. This paper explores the preliminary design of an

ammonia synthesis system that is intended to heat steam from

350°C to 650°C under pressure of 26 MPa. A two-dimensional

pseudo-homogeneous model for packed bed reactors previously

used at ANU is adopted to simulate the ammonia synthesis

reactor. The reaction kinetics are modeled using the Temkin-

Pyzhev reaction rate equation. The model is extended by

accounting for convection in the steam to predict the behavior

of the proposed synthesis reactor. A parametric investigation is

performed and the results show that heat transfer plays the

predominant role in improving reactor performance. INTRODUCTION

Today’s commercial concentrating solar power plants can

provide dispatchable power by integrating thermal energy

storage, e.g. molten salt tanks. The broad category of

thermochemical energy storage (TCES), in which energy is

stored in chemical bonds, has the advantages of higher energy

density and minimal energy losses during the storing operation

[1]. Shakeri et al. [2] indicated that an ammonia-based thermal

energy storage system was more efficient for long-term storage

than compressed air energy storage, pumped hydroelectric

energy storage, vanadium flow battery, or thermal energy

storage. In the ammonia-based thermochemical energy storage

system (Figure 1), ammonia (NH3) is dissociated

endothermically as it absorbs solar energy during the daytime.

The stored energy can be released on demand when the

supercritical hydrogen (H2) and nitrogen (N2) react

exothermically to synthesize ammonia. The released thermal

energy can be used to generate electricity. Ammonia-based

energy storage for concentrating solar power systems has been

studied and investigated since 1974 [3]. Lovegrove et al. [4, 5]

developed and experimentally validated a two-dimensional

pseudo-homogeneous steady state model of catalytic ammonia

reactors. With the same model, Kreetz et al. [6] analyzed the

exergy of an ammonia synthesis reactor in a solar

thermochemical power system. Kreetz proposed that either very

small diameter reactors or adiabatic reactors were preferable for

minimizing exergy loss [6]. Lovegrove et al. built and tested a 1

kWsol closed loop ammonia-based TCES system [7] and a 15

kWsol ammonia-based TCES system for dish power plants [8].

Their systems demonstrated ammonia dissociation on a dish

concentrator and subsequent energy recovery at temperatures

high enough for electricity generation, but did not demonstrate

heating of a working fluid. Furthermore, ammonia synthesis

has not been shown to reach temperatures consistent with

Proceedings of the ASME 2015 Power Conference POWER2015

June 28-July 2, 2015, San Diego, California

POWER2015-49190

1 Copyright © 2015 by ASME

modern power blocks such as a supercritical steam Rankine

cycle requiring steam to be heated to ~650°C.

This paper investigates the preliminary design of a pilot-

scale catalytic ammonia synthesis reactor operating at 30 MPa.

The objective of the reactor is to heat 12.5 g/s of steam at 26

MPa from 350°C to 650°C [9], corresponding to 25 kWt. We

investigate the effect of various parameters on the reactor

performance. The optimum design demands low cost electricity

generation. While the paper does not perform a

thermoeconomic analysis, it does consider the material usage

required for tube wall material and catalyst, two of the main

system costs.

For ease of expression we refer to the water, which may

be supercritical or liquid, as “steam.” The N2, H2, NH3

mixture is referred to as “the gas,” even though the species are

all supercritical within the reactor and NH3 is liquid in the

lower temperature portion of the system.

SYNTHESIS REACTOR SYSTEM Figure 2 is a schematic of an entire synthesis and steam

heating system consisting of a reactor and two heat exchangers.

As shown in this figure, the cold incoming synthesis gas enters

at the bottom right and is initially heated in a counterflow heat

exchanger from temperature T0 to temperature T1 by the gas

exiting the reactor. This “initial preheater” is only sufficient to

bring the incoming gas close to the reactor exit temperature

(T3), not to the desired reactor inlet temperature, T2. Therefore,

additional heat transfer is needed to further preheat the

incoming gas to T2. This “secondary preheating” can be

accomplished in various ways; in this paper we consider a

Figure 1. Schematic of ammonia dissociation, storage, and synthesis system [3]

Figure 2. Schematic of an ammonia synthesis system with a reactor and two heat exchangers (dashed boxes)


design in which the incoming gases are further preheated in a

gas-steam heat exchanger shown on the far left; “extra” steam

is heated in the reactor for this purpose. The total flow of steam

entering the reactor (at the top right) is heated from Tsi = 350ºC

to Tso = 650ºC by the exothermic reaction. The desired “base

steam” flow rate of 12.5 g/s exits the system to be used for

electricity generation. Then the “extra steam,” at Tso = 650ºC, is

used to preheat the incoming gas from T1 to T2.

Two possible configurations have been considered for the

design of the reactor itself. The first, shown in Figure 3, is a

concentric tube configuration with the reacting gas flowing in

the inner tube and steam in counterflow in the outer annulus.

The second, shown in Figure 4, is a shell-and-tube

configuration with the reacting gas in the tubes and steam in

counterflow in the shell. For purposes of preliminary design,

the shell-and-tube reactor is viewed as N small concentric tube

reactors, as previously illustrated in Figure 3. The fixed

parameters are listed in Table 1.

Table 1. The fixed dimensions and properties of the reactor

Fixed variables Value Steam tube wall thickness, W

s (cm) 2

Gas mass flow rate, gm (g/s) 20

Steam mass flow rate, sm (g/s) 21.6

Inlet steam temperature, Tsi (ºC) 350

Outlet steam temperature, Tso

(ºC) 650

Inlet gas temperature, T2 (ºC) 597

Steam pressure, Ps (MPa) 26 Gas mixture pressure, Pg (MPa) 30 Pre-exponential factor, ko,m (kg∙m

-3∙s

-1) 3.83×10

13

Activation energy, Ea (J/mol) 1.83×105

REACTOR MODEL As described in the preceding section, gas flows in a

tubular packed-bed catalytic reactor which exchanges heat with

steam through the outer wall. The pseudo-homogeneous steady

state model is very commonly employed for designing packed

bed reactors [10]. The model used here was originally

developed by Richardson et al. at the University of Houston for

tubular packed-bed methane reforming reactors [11]. It has

been modified and validated by Lovegrove for ammonia

dissociation via experiments on an electrically heated 1 kWchem

ammonia dissociation reactor [5].

In the pseudo-homogeneous model, the catalyst bed is

treated as a continuum with averaged properties, i.e. effective

conductivity (keff) and effective diffusivity (Deff). Also, the

model assumes: (1) negligible axial conduction, (2) gas

pressure is constant for purposes of evaluating properties and

reaction rate, (3) the average velocity and density are only

functions of axial location, (4) the radial gradients of specific

heat, effective conductivity, and effective diffusivity are

negligible. The governing equations for energy conservation

can therefore be expressed as:

�

�

Figure 4. Schematic of a tube bundle reactor. N tubes are

viewed as N concentric tube reactors.

�

�

Figure 3. Geometry and dimensions of a concentric tube reactor


1p eff

T Tvc k r r H

x r r r

(1)

The boundary conditions are [11]:

( , 0)in

eff w

wr

T r x T

Tk q

r

(2)

where rw is the inner wall radius, ( )w w cw s

q U T T , and

and w cw

U T are the overall heat transfer coefficient and catalyst

temperature at the wall, as defined in Richardson et al. [11].

The governing equations for mass conservation are:

3 3

1NH NH

eff

f fv D r r

x r r r

(3)

The boundary conditions for mass conservation are:

3 3,

3

( , 0)

0

NH NH in

NH

wr

f r x f

f

r

(4)

The Temkin-Pyzhev intrinsic rate expression is used for the

ammonia reaction [12]:

0,

123

1 (1 ) (1 )32

2 2 3

3 2

expa

m

u

NHH

p N o o

NH H

Er k

R T

ppK p p p

p p

(5)

Here po is the standard state pressure of 1 atm, explicitly

included to make the dimensions consistent; α is an empirical

parameter taken to be 0.5.

The total pressure has been assumed constant in the

model. This is typically a valid assumption; using the Ergun

equation, the pressure drop is found to be at most 3 MPa under

the conditions considered here, which is negligible compared to

the system pressure of 30 MPa.

The temperature of the steam is calculated from a cross-

sectionally averaged energy balance:

( )s

p s w

dTmc q P

dx (6)

The heat transfer coefficient for the steam flow is

calculated with the following equation:

0.8 0.4Nu

, Nu 0.023Re Prs

s

h

kh

D

(7)

where Nu is the Nusselt number, ks is the average thermal

conductivity of the steam, Dh is the hydraulic diameter, Re is

the Reynolds number based on the hydraulic diameter, and Pr is

the Prandtl number.

PARAMETRIC STUDY Effect of steam heat transfer coefficient. Similar to a

heat exchanger, the required surface area to heat the steam is

expected to decrease as the overall heat transfer coefficient, U,

increases. The overall heat transfer coefficient is related to the

thermal resistances of the catalyst bed, wall, and steam flow:

1

ln ( 2 ) /1 1

2 ( 2 )

g w s

g g g

g g w s g g

UP R R R

D W D

h D k h D W

(8)

where P is tube perimeter. In this sub-section, we consider the

effect of the steam heat transfer coefficient, hs. The heat transfer

coefficient for the steam flow can be manipulated by changing

the outer diameter of the steam tube, Ds, (see Figure 3); under

laminar conditions the Nusselt number is constant, but for

sufficiently small gap dimension the flow can be made

turbulent and the Nusselt number then increases with

decreasing gap spacing. Table 2 shows the values of the steam

tube diameter, Ds, that were selected to give steam heat transfer

coefficients hs from 33.3 to 100 W∙m-2

∙K-1

while fixing other

properties and dimensions (listed in Table 1).

Table 2. Outlet conditions and required reactor length for

different steam heat transfer coefficients

Dg

(cm)

Wg

(cm)

Ds

(cm)

hs (W∙

m-2

∙K-1

)

T3

(ºC) fNH3,out

L

(m)

Vw

(m3)

10 1 16.8 33.3 459 0.500 38.2 0.58

10 1 14.6 66.7 457 0.497 22.1 0.31

10 1 13.8 100.0 454 0.495 16.8 0.22

Figure 5 shows the bulk (cross-sectionally averaged)

temperature profiles of steam (dashed curve) and gas (solid

curve) along the x-direction (the direction of flow through the

reactor). After being preheated by steam, the stoichiometric

synthesis gas mixture (N2+3H2) enters the reactor (solid curve)

at x = 0 and begins flowing in the positive x-direction. As soon

as the gases enter the reactor, they undergo a very abrupt

temperature increase; this is because of the rapid reaction rate

caused by the elevated temperature. The steep temperature

increase brings the gases close to equilibrium conditions and


the reaction slows, but does not cease. Thereafter, the

temperature of the reacting gases decreases along the reactor

length as they lose heat to the steam. The steam (dashed curve)

enters at the right at 350°C and exits at the left at 650°C. The

slope change in the region roughly around 400°C corresponds

to the pseudocritical regime of the steam, with very large

specific heat. As the steam heat transfer coefficient hs increases,

the required reactor length to heat the steam to 650°C

decreases, as expected.

In Figure 6, the solid curve is the mass fraction of

ammonia in the gas stream as it flows through the reactor. The

dashed curve is the equilibrium mass fraction corresponding to

the bulk reactor temperature at each axial location. The

ammonia mass fraction reaches the equilibrium curve almost

immediately at the reactor inlet and then increases downstream

along with the equilibrium curve. It can be observed that under

the conditions investigated here, the synthesis process is

essentially heat-transfer-limited. That is, the reaction rate is fast

enough to maintain the reaction almost at equilibrium. The rate

at which the reaction proceeds is then dictated by the rate at

which heat is transferred to the steam, which reduces the gas

temperature and increases the corresponding equilibrium

ammonia mass fraction. The ammonia mass fraction deviates

from the equilibrium curve near the outlet; the lowered

temperature decreases the reaction rate so that it is no longer

effectively infinite. It can be seen in Figure 6 that the deviation

from the equilibrium ammonia mass fraction increases with

increasing steam heat transfer coefficient hs since greater heat

transfer tends to make the process more reaction-rate-limited

than heat-transfer-limited.

As shown in Table 2, both the outlet temperature T3

and the outlet ammonia mass fraction fNH3,out decrease slightly

as the steam heat transfer coefficient hs increases. Most

importantly, the required length decreases significantly, which

decreases material costs. The three main material costs are: the

catalyst which fills the gas tube; the wall separating the gases

and steam (inner wall), which sees a modest pressure

differential of 4 MPa; and the wall separating the steam from

the environment (outer wall), which sees a 26 MPa pressure

differential. In this case, with diameter Dg held fixed, both the

catalyst volume and inner wall material volume decrease in

proportion to the reactor length. The outer wall material volume

is given in the last column in Table 2 and is seen to decrease

rapidly as Ds decreases, due to both the reduced length and

diameter. The only disadvantage to reducing the outer diameter

is that it would also increase the pressure drop of the steam

flow. However, this pressure drop is estimated to be

significantly smaller than the pressure drop in the catalyst bed,

and would therefore not impose too great a cost for pumping

power. The conclusion is that the design should achieve a large

steam heat transfer coefficient hs by reducing the steam cross-

sectional area.

0 5 10 15 20 25 30 35 40 45 50

300

350

400

450

500

550

600

650

700

750

Tem

per

atu

res,

T (

oC

)

Reactor length, x (m)

Ds=16.8 cm: gas

Ds=16.8 cm: steam

Ds=14.6 cm: gas

Ds=14.6 cm: steam

Ds=13.8 cm: gas

Ds=13.8 cm: steam

Figure 5. Temperatures of the gas and steam along the reactor

length

0 5 10 15 20 25 30 35 40 45 50

0.0

0.2

0.4

0.6

Am

mo

nia

mas

s fr

acti

on

s, f

NH

3


Ds=16.8 cm

Ds=16.8 cm: equilibrium

Ds=14.6 cm


Ds=13.8 cm


Figure 6. The ammonia mass fraction and corresponding

equilibrium ammonia mass fraction along the reactor length

Effect of reactor tube diameter. In this sub-section, we

seek to elucidate the impact of inner tube diameter, Dg, on

reactor performance. Varying Dg would have multiple effects,

e.g. it changes both of steam and gas thermal resistances, the

gas velocity (for fixed mass flow rate), and the catalyst volume.

The impact of steam thermal resistance was established in the

previous sub-section, by varying hs while holding Dg fixed. In

this sub-section, we hold steam thermal resistance fixed, so that

we can explore the other effects of varying Dg. Since steam

thermal resistance s

R depends on the product of the steam

heat transfer coefficient and surface area, in this sub-section we

hold the product hs(Dg + 2Wg), fixed. In order to hold this

product fixed while varying Dg, we simultaneously vary Ds to

change hs. The derivation uses the standard definition of

hydraulic diameter for the outer annulus, Dh = 4Ac/Pw where Pw

= Dg + 2Wg + Ds is the wetted perimeter. Then, with fixed

steam mass flow rate, and assuming turbulent steam flow, the


following relationships can be used to find the required Ds for a

given Dg:

1 0.8

0.8

~ , ~

2( 2 ) ~ const.

2 2

s h

s w s s

s

g g

s g g

g g s s g g

h DRe P Nu Re

k

D Wh D W

D W D D D W

Table 3 lists three combinations of Dg and Ds that give fixed

hs(Dg + 2Wg) and corresponding model outputs. The last two

columns in this table are the wall material volume and catalyst

volume, respectively.

Figure 7 shows the gas temperature (solid curve) and

steam temperature (dashed curve) along the reactor length.

Referring to Table 3 and Figure 7, it can be seen that as Dg

decreases, the required reactor length decreases slightly. If the

process were limited by reaction kinetics, we would see that the

required volume would stay constant (i.e. residence time or

space velocity would remain constant), necessitating an

increase in length as diameter decreased. Instead, the

behavior is predominantly a heat transfer effect due to the

changing thermal resistance within the catalyst bed. In Table

3, the thermal resistance within the catalyst bed is seen to

decrease as Dg decreases, which largely explains the shorter

length required.

Figure 8 shows the ammonia mass fraction and the

corresponding equilibrium ammonia mass fraction. Similar to

the behavior seen previously, decreasing Dg improves heat

transfer and increases the deviation from equilibrium.

Table 3 shows that thinner reactors require less tube wall

material and catalyst (last two columns). One downside to

decreasing the reactor diameter is that the pressure drop would

increase, requiring greater pumping power. While pressure

has been held constant in the model for purposes of evaluating

properties and reaction kinetics, the pressure drop has been

calculated using the Ergun equation. Table 3 shows the

increase in pressure drop of the reacting gas as the reactor

become thinner.

0 5 10 15 20 25 30

300

350

400

450

500

550

600

650

700

750

Tem

per

ature

s, T

(oC

)


Dg=13 cm: gas

Dg=13 cm: steam

Dg=10 cm: gas

Dg=10 cm: steam

Dg=7 cm: gas

Dg=7 cm: steam

Figure 7. Temperature of gas and steam along reactor length

0 5 10 15 20 25 30

0.0

0.2

0.4

0.6

Am

monia

mas

s fr

acti

ons,

fN

H3


Ds=13 cm

Ds=13 cm: equilibrium

Ds=10 cm


Ds=7 cm


Figure 8. Ammonia mass fraction and corresponding

equilibrium ammonia mass fraction along reactor length

Tube bundle reactor. In this subsection, the use of a tube

bundle reactor is investigated. As shown in Figure 4, gas flows

in N reacting tubes while the steam flows in the opposite

direction within the gap. The number of reacting gas tubes is

varied while the cross-sectional areas for gas and steam are

fixed. Also, the wall thickness of the reactor tube, Wg,N, is fixed

to be 0.5 cm. Table 4 lists combinations of Dg,N, Ds,N, and Ds

that give fixed cross-sectional areas for different numbers of

tubes, N.

Table 3. Comparison of outlet conditions, required reactor length, and material usage for different Dg and Ds combinations

Dg

(cm)

Wg

(cm)

Ds

(cm) s

R

(W-1

∙cm∙K)

gR

(W-1

∙cm∙K)

T3

(ºC)

fNH3,out

ΔP

(MPa)

L

(m)

Vw

(m3)

Vc

(m3)

13 1 18.2 4.7 1.8 461 0.502 0.3 26.2 0.45 0.35

10 1 15.0 4.7 1.4 458 0.498 0.8 24.9 0.35 0.20

7 1 11.8 4.7 1.0 447 0.486 3 24.6 0.28 0.10


0 5 10 15 20 25 30 35

300

350

400

450

500

550

600

650

700

750

Tem

per

ature

s, T

(oC

)


N=10: gas

N=10: steam

N=8: gas

N=8: steam

N=6: gas

N=6: steam

N=1: gas

N=1: steam

Figure 9. Gas and steam temperatures along reactor length

0 5 10 15 20 25 30 35

0.0

0.2

0.4

0.6

Am

mo

nia

mas

s fr

acti

on

s, f

NH

3


N=10

N=10: equilibrium

N=8

N=8: equilibrium

N=6

N=6: equilibrium

N=1

N=1: equilibrium

Figure 10. Mass fraction of ammonia and corresponding

equilibrium mass fraction along reactor length

Figure 9 shows the gas temperature (solid curve) and

steam temperature (dashed curve) along the reactor length.

Referring to Table 4 and Figure 9, the required reactor length

decreases significantly as N increases. Once again, this is

caused by improved heat transfer: The thermal resistances,

especially the steam thermal resistance, are decreasing with

increasing N. As N increases, the surface area for heat transfer

increases and the steam heat transfer coefficient hs increases

because the gap between tubes decreases.

Figure 10 shows the ammonia mass fraction and the

corresponding equilibrium ammonia mass fraction. Increasing

the number of tubes, which increases the heat transfer surface

area, improves heat transfer and increases the deviation from

equilibrium. As shown in the last column of Table 4, the wall material

usage decreases as N increases. In this case, with the gas cross-

sectional area held fixed the catalyst volume decreases in

proportion to the reactor length. Furthermore, the gas pressure

gradient doesn’t change with tube number N since the cross-

sectional area is fixed, therefore pressure drop also decreases in

proportion to length. The preliminary conclusion is that a

bundle of tubes is preferable to a single tube reactor, and that

performance improves as N increases. However, this

conclusion will require further investigation using a more

accurate model of the steam flow and heat transfer in the shell.

CONCLUSIONS The ammonia synthesis reactor design study demonstrates

the viability of using ammonia as a thermochemical storage

system for the direct production of supercritical steam at 650°C.

The supercritical steam can be used with modern power blocks

for electricity generation.

A parametric study was made to investigate the effect of

geometry for optimum design. Under the high temperature

reaction conditions explored here, the ammonia synthesis

reaction is predominantly heat-transfer-limited. As heat transfer

is improved, the required reactor length decreases. Tube wall

material and catalyst volumes, which are associated with

reactor tube diameter and length, can be reduced by improving

heat transfer. The effects of geometry and dimensions of the

reactor on heat transfer have been studied. Either decreasing

tube reactor diameters or using tube bundles can improve heat

transfer within the reactor. This study provides guidance for

designing a reactor that can achieve the desired steam heating

while minimizing cost. Future research will establish the cost of

an ammonia-based thermochemical energy storage system.

NOMENCLATURE cp Heat capacity of the reacting gas mixture, J/(kg∙K)

D Diameter, m

Deff Effective diffusivity, m2/s

Ea Activation energy, J/mol

3NHf Ammonia mass fraction

h Heat transfer coefficient, W/(m2∙K)

H Heat of reaction, J/kg

Table 4. Comparison of outlet conditions and material usage for different tube bundle reactors

N

Dg,N

(cm)

Ds,N

(cm) s

R

(W-1

∙cm∙K)

gR

(W-1

∙cm∙K)

Ds

(cm)

T3

(ºC)

fNH3,out

Δp

(MPa)

L

(m)

Vw

(m3)

1 9.49 14.14 6.05 1.23 14.14 457 0.498 1.1 30.0 0.35

6 2.87 6.22 2.94 0.28 15.3 452 0.494 0.5 13.5 0.19

8 2.35 5.50 2.10 0.22 15.6 449 0.489 0.4 9.83 0.14

10 2.00 5.00 1.59 0.19 15.8 444 0.484 0.3 7.63 0.12


k Thermal conductivity, W/(m∙K)

ko,m Pre-exponential factor, kg/(m3∙s)

Kp Equilibrium constant

L Reactor length, m

m Mass flow rate, kg/s

Nu Nusselt number

P Perimeter of tube, m

p Pressure, Pa

pj Partial pressure, where j indicates species, Pa

q Heat transfer rate, W

wq Heat flux, W/m

2

r Radial dimension, m

r Rate of ammonia synthesis, kg/s

Ru Universal gas constant, J/(mol∙K)

T Temperature, °C

U Overall heat transfer coefficient, W/(m2∙K)

V Volume, (m3)

v Velocity of the reacting gas mixture, m/s

W Wall thickness, m

x Axial dimension, m

Greek letters

Density of the gas mixture (kg/m3)

Subscripts

c catalyst

eff effective

g gas mixture (N2+3H2 and NH3)

i inner

in inlet

o outer

out outlet

ph preheat

s steam

si inlet of steam stream

so outlet of steam stream

w wall

ACKNOWLEDGMENTS The information, data, or work presented herein was

funded in part by the Office of Energy Efficiency and

Renewable Energy (EERE), U.S. Department of Energy, under

Award Number DE- EE0006536.

REFERENCES [1] A.H. Abedin and M. A. Rosen, "A Critical Review of

Thermochemical Energy Storage Systems " The Open

Renewable Energy Journal, vol. 4, pp. 42-46, 2011.

[2] M. S. Mostafa Shakeri, R. Eric Berson and M. Keith

Sharp, "Comparison of Energy Storage Methods for Solar

Electric Production," in ASME 2014 8th International

Conference on Energy Sustainability collocated with the

ASME 2014 12th International Conference on Fuel Cell

Science, Engineering and Technology, Boston,

Massachusetts, USA, 2014.

[3] R. Dunn, K. Lovegrove, and G. Burgess, "A Review of

Ammonia-Based Thermochemical Energy Storage for

Concentrating Solar Power," Proceedings of the Ieee, vol.

100, pp. 391-400, Feb 2012.

[4] K. Lovegrove and A. Luzzi, "Endothermic reactors for an

ammonia based thermochemical solar energy storage and

transport system," Solar Energy, vol. 56, pp. 361-371, Apr

1996.

[5] K. M. Lovegrove, "High pressure ammonia dissociation

experiments for solar energy transport and storage,"

International Journal of Energy Research, vol. 20, pp.

965-978, Nov 1996.

[6] H. Kreetz and K. Lovegrove, "Exergy analysis of an

ammonia synthesis reactor in a solar thermochemical

power system," Solar Energy, vol. 73, pp. 187-194, 2002

2002.

[7] K. Lovegrove, H. Kreetz, and A. Luzzi, "The first

ammonia based solar thermochemical energy storage

demonstration," Journal De Physique Iv, vol. 9, pp. 581-

586, Mar 1999.

[8] K. Lovegrove, A. Luzzi, I. Soldiani, and H. Kreetz,

"Developing ammonia based thermochemical energy

storage for dish power plants," Solar Energy, vol. 76, pp.

331-337, 2004 2004.

[9] M. P. Naidin, I.; Duffey, R., "Supercritical water-cooled

nuclear reactors (SCWRs): Thermodynamic cycle options

and thermal aspects of pressure-channel design,"

presented at the International conference on opportunities

and challenges for water cooled reactors in the 21.

century., Vienna, Austria, 2009.

[10] D. Ramkrishna and P. Arce, "Can Pseudo-Homogeneous

Reactor Models Be Valid," Chemical Engineering Science,

vol. 44, pp. 1949-1966, 1989.

[11] J. T. Richardson, S. A. Paripatyadar, and J. C. Shen,

"Dynamics of a Sodium Heat Pipe Reforming Reactor,"

Aiche Journal, vol. 34, pp. 743-752, May 1988.

[12] J. R. Jennings, Catalytic Ammonia Synthesis

Fundamentals and Practice. New York: Plenum Press,

1991.


Documents

Design of an Ammonia Synthesis System for Producing