Applied Thermal Engineering 109 (2016) 542–559

Contents lists available at ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Research Paper

Material and energy flows in rotary kiln-electric furnace smelting

of ferronickel alloy with energy saving
Peng Liu a, Baokuan Li a,⇑, Sherman C.P. Cheung b, Wenyuan Wu a
a
School of Metallurgy, Northeastern University, Shenyang 110819, China
b
School of Aerospace, Mechanical and Manufacturing Engineering, RMIT University, Victoria 3083, Australia

h i g h l i g h t s

Establish the synergy relationship of material and energy in key RKEF processes.

Develop an analysis model to study energy saving with internal cycling of energy.

Analyze material and energy flow parameters and assess its associated synergy effect.

A methodology to evaluate the synergy and design indices of RKEF processes.

a r t i c l e i n f o a b s t r a c t

Article history: An energy saving strategy with two energy saving measures has been proposed for reducing energy loss
Received 11 April 2016 in the rotary kiln-electric furnace (RKEF) for the smelting of ferronickel alloy. One of the measures is to
Revised 27 July 2016 recover the waste heat of exhaust gas from the rotary kiln for preheating and dehydrating the wet laterite
Accepted 18 August 2016
ores in the rotary dryer. Another measure is to recycle the furnace gas from the electric furnace into the
Available online 20 August 2016
rotary kiln as fuel. Based on the mass conservation and energy conservation laws, an analysis model of
material and energy flows has been developed to understand the potential energy saving with the inter-
Keywords:
nal cycling of material and energy in the RKEF process. The analysis model not only considers the energy
Rotary kiln-electric furnace
Ferronickel smelting
efficiency but also assess the synergy degree of system. Furthermore, the model also predicts the ratio of
Material and energy flows raw materials and the energy flow distribution to investigate residual heat and energy and analyze the
Energy efficiency analysis effects of nickel content on energy flow. Finally, the evaluation methodology of synergy and the technic
Synergetic theory indices are also presented. Through the investigation of the synergy effect, the performance of the RKEF
process can be evaluated and quantified for performance optimization in future.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction
With the increasing demand for the ferronickel alloy, the tech-
nology of ferronickel smelting with high Ni grade has attracted rea-
sonable attention in the past decades. Currently, the current
methodology to alter the original mineralogy of laterite ores is to
apply some chemical processes (e.g. pyrometallurgical or
hydrometallurgical) that include pressure acid leaching [1–3],
Caron process [4,5], atmospheric leaching [6,7], and rotary kiln-
electric furnace (RKEF) process [8,9]. Among these methods, the
rotary kiln-electric furnace (RKEF) process has the advantages of
yielding high nickel/iron grade from crude ferronickel products,
less harmful elements, raw materials with strong adaptability, high
production efficiency and mature process. Nonetheless, the RKEF
⇑ Corresponding author.
E-mail address: libk@smm.neu.edu.cn (B. Li).
process also suffers from the high energy consumption and a large
amount of by-product (e.g. slag and off-gas). Reducing the energy
consumption while achieving the same product quality has
become a pressing need for the current ferronickel smelting indus-
try. In theory, considering the overall energy consumption, the
supply energy can be categorized into three main components.
The main portion of the supply energy is utilized to drive the fer-
ronickel smelting process which is normally referred as available
energy. The other portion is the unavoidable and irreversible
energy loss due to the limitations of existing smelting technology.
For example, the electricity loss and heat dissipation loss to the
cooling water and walls are unavoidable in the RKEF process. The
rest of supply energy is the avoidable energy loss which is dis-
charged with the waste gas or heat throughout the whole process.
In order to improve the energy efficiency, minimizing the avoid-
able energy loss or recover such waste energy could be a promising
approach.

http://dx.doi.org/10.1016/j.applthermaleng.2016.08.095
1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559 543

Nomenclature

a air factor dr dissociation reaction

C0 continuation degree of process E electric furnace
cp specific heat (kJ/(kg °C)) e evaporation
DR mismatching degree el electric loss
E energy (GJ/h) F fuel
e order parameter Fa ferronickel alloy
EC efficacy coefficient Fg furnace gas
G the group fg flue gas
gd dissipation quantity of energy flow (GJ/h) Fgc furnace gas combustion
gk output flow of energy (GJ/h) Fgl furnace gas heat loss
K number of output flow of energy fge flue gas entered into electric furnace
k variable fgl flue gas heat loss
L number of the order parameter index fm free moisture
l variable fml free moisture heat loss
L0 theoretical air requirement (kg) hl heat loss
Ln actual amount of air supply (kg) ht haulage time
M mass flows (t/h) in input flow (s)/inlet
N number of the order parameter J Joule heat
OD order degree L limestone
q energy consumption per unit (kJ/kg) lhm latent heat of melting
R recovery ratio m number of groups
Sen entropy (kJ/K) mt maintenance time
SA coordination ability n number of order parameter components
SD synergy degree oc laterite ores calcined
T temperature (K) ocl laterite ores calcined heat loss
t time (s) out output flow (s)/outlet
w weighting ppt processing-time
V volume (m3) R rotary kiln
rr reduction reaction
Greek letters S sensible heat
q density (kg/m3) sl slag
ge energy efficiency (%) so semi-dry laterite ores
gep discharge rate of energy flow (%) sol laterite ores heat loss
gsy recycling rate of residual heat and energy sr slagging reaction
gbp recycling rate of by-product energy Un unavailable
gws recycling rate of waste solid Unavo unavoidable
gw recirculation rate of waste w waste
c fill rate wc water-cooling
w void fraction whl waste heat loss
h normalized data wo wet laterite ores
ws waste solid
Subscripts wt waiting time
A available
a anthracite Superscripts
Avo avoidable ca combustion air
bc bituminous coal (bitumite) Fg furnace gas
bcc bituminous coal combustion fg flue gas
ca combustion air ll lower limit
D rotary dryer ul upper limit
d dust
dl dust heat loss

Several promising energy saving strategies have been proposed
for other smelting fields, including ironmaking, steel making and
limekilns industries. Lee and Sohn [10] adopted the strategy of
recycling the exhaust gas and the slag for preheating the scrap
and granulation. Besides, using the wasted heat of the exhaust
gas, the CO2 gas is also recycled to produce CO gases as fuel or
reducing agent. Belt [11] used the furnace benchmark tests to mea-
sure and analyze the melt loss and melt rate, and the energy distri-
bution or heat losses of the system. Nevertheless, a comprehensive
energy saving strategy for the RKEF systems remains outstanding
in the literature. In order to improve the energy efficiency of the
RKEF, a complete mathematical model to describe the underlying
material and energy flows within the system is very crucial. Previ-
ously, mathematical models for other applications have been pro-
posed. Sohn and Oliver-Martinez [12] presented a mathematical
model based on the material and energy flows of ironmaking pro-
cess to analyze the chemical reaction heat and energy requirement
of the system. Sagastume [13] adopted the energy and exergy anal-
ysis methods to determine the mass, energy and exergy balance
and identify the dominant factors affecting the thermal efficiency
of lime kiln. Several researchers adopted the synergetic theory to
analyze the cooperativity between material and energy flows,
544 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559
and evaluate the order degree among subsystems. Zhang and Zhu
[14] proposed a network selection algorithm based on Analytic
Hierarchy Process (AHP) and synergetic theory, and evaluate syn-
ergy degree of network by calculating order degree and entropy
of system. In addition, previous studies have also investigated
the essential mechanisms of the ferronickel smelting process. Pick-
les [15] described, from the fundamentals of drying, the results of a
thermogravimetric study of the drying of a nickeliferous limonitic
laterite ore are analyzed. Kotzé [16] proposed the process of pro-
ducing ferronickel from nickel oxide ores and dust particles by
DC arc furnace technology. Guo [17] carried out ferronickel smelt-
ing experiment to investigate the selective reduction and melting-
separation processes, and to analyze the influence parameters of
metallization rate.
Although considerable efforts have been devoted to investigate
the energy efficiency in other systems, to the best of our knowl-
edge, a thorough energy analysis and its possible energy saving
measures for the rotary kiln–electric furnace is yet to be fully
explored. It is therefore essential to study the distribution of avail-
able energy and energy loss throughout the RKEF; to identify the
energy saving potential and analyze the energy efficiency of the
process. The objective of this paper is therefore to present a com-
plete and comprehensive material and energy flows analysis
methodology; leading to a thorough characterization of system
performance and identification of potential energy savings. Partic-
ular attention has been directed towards discussing the effect of
the operation parameters on the distribution of material and
energy flows as well as the associated impact on the product qual-
ity. Synergy and technic index analyses are also presented to fur-
ther assess the performance of the system under various scenarios.
2. The essential mechanisms of the technological process of
RKEF
The rotary kiln-electric furnace process is the one of the most
widely adopted technological process for ferronickel smelting pro-
duction. A schematic of the RKEF process with rotary dryer, rotary
Fig. 1. Schematic of the RKEF process with rotary dryer, rotary kiln and electric furnace.
kiln and electric furnace is shown in Fig. 1. As depicted, the RKFF
process involves three main processes: the dehydration process
in rotary dryer; the roasting and pre-reduction process in rotary
kiln and the reduction smelting process in electric furnace. The
whole process involves many complex physical and chemical phe-
nomena: including dehydrating (removing free moisture, crystal
water and chemically bonded water), oxydoreduction reaction, dis-
sociation reaction, replacement reaction, slagging reaction and
separation of slag/alloy. The details of the technological process
of each subsystem are shown in Fig. 2 and discussed below.
2.1. Technological process of the rotary dryer
In the rotary dryer process, the wet laterite ores (typically with
23% free moisture and 12% chemically bonded water) are firstly
broken into small particles by screening and crushing process
[18,19] and then fed into the rotary dryer. To remove the free mois-
ture from the wet ores, the flue gas waste heat of rotary kiln and
bituminous coal combustion are used for heating the ores up to
its dehydration temperature (i.e. 220 °C). After the rotary dryer
process, the semi-dry laterite ores (with 7% residual moisture
and 12% chemically bonded water) are mixed with the limestone
and the anthracite in raw proportioning station. Furthermore,
using agglomeration technique, the nickeliferous laterite ores par-
ticles contained in dust are recovered.
2.2. Technological process of rotary kiln
In the process of rotary kiln, using the furnace-gas of electric
furnace and bituminous coal as a fuel, the residual free moisture
is completely evaporated. Afterwards, in the pre-reduction and
roasting processes, the crystal water and chemically bonded water
of laterite ores are removed at high temperature (i.e. 290–820 °C)
and the dissociation of limestone is occurred within the tempera-
ture range of 820–1000 °C [18–20]. The laterite ores calcined are
then transported into electric furnace for reduction smelting.
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559 545

Fig. 2. Schematic of the ferronickel smelting technological process.

Meanwhile, the flue gas of rotary kiln is transported into rotary Table 2
dryer for drying the wet ores. The chemical reactions occurred in the electric furnace.

Moreover, carbon dioxide will be reduced by anthracite into Reaction Equation

carbon monoxide which further reduces the metallic oxide at Reduction reaction NiO þ C ¼ Ni þ CO (r10)
about 220 °C [21]. Therefore, the majority of the metallic oxide is 3Fe2 O3 þ C ¼ 2Fe3 O4 þ CO (r11)
reduced by the carbon monoxide in the rotary kiln. Only a small Fe3 O4 þ C ¼ 3FeO þ CO (r12)
portion of the metallic oxide is reduced by the anthracite. The FeO þ C ¼ Fe þ CO (r13)

chemical reactions occurred in the rotary kiln are shown in Table 1. Slagging reaction 2CaO þ MgO þ 2SiO2 ¼ CaSi2 O7 (r14)
2CaO þ 2C þ Fe2 SiO4 ¼ 2Fe þ Ca2 SiO4 þ 2CO (r15)

2.3. Technological process of electric furnace

In the electric furnace process, due to the effect of anthracite

and quicklime, metallic oxide will undergo the reduction reaction
and slagging reaction (Table 2). The ferronickel and slag are then
generated and separated. Meanwhile, the furnace gas (containing
carbon monoxide) is also produced. After undergoing the purifica-
tion treatment, these furnace gases are recycled into rotary kiln as
fuel. Besides, in order to produce a high nickel grade in ferronickel
alloy, the reduction of nickel oxide is completely with limited
reduction of iron oxide [19,22]. Meanwhile, quicklime can permute
the nickel/iron oxide from mineral which enhances the activity of
metallic oxide [18,19], improve the recovery ratio of nickel/iron
[23,24], and contribute to the separation of slag/alloy through
reducing the melting point of slag [24–26].

3. Theory of material and energy flows in RKEF process

3.1. Mathematical model of material and energy flow in rotary dryer

The material flow conservation of the rotary dryer can be

expressed as follow (see also in Fig. 3): Fig. 3. Schematic of material and energy flows in the rotary dryer.

M D;in ¼ MD;bc þ MD;ca þ M R;fg þ M D;wo ð1Þ

M D;out ¼ M Dso þ MD;fg þ MD;d
where MD,in and MD,out denote the total mass of input flows and out-
put flows in rotary dryer, respectively; MD,bc, MD,ca, MD,wo, MD,so and
Table 1
The chemical reactions occurred in the rotary kiln.
MD,d are the mass of bituminous coal, combustion air, wet laterite
ð2Þ ores, semi-dry laterite ores (with 7% free moisture) and dust (con-
taining Ni) in rotary dryer, respectively; MR,fg is the exhaust gas
mass of rotary kiln; MD,fg is the flue gas mass of rotary dryer.
In terms of energy balance, there are six energy inputs and nine
energy outputs are given as:
ED;in ¼ ED;bcc þ ED;bc þ ED;wo þ ED;fm þ ED;ca þ ER;fg ð3Þ

Reaction Equation ED;out ¼ ED;fg þ ED;so þ ED;fm þ ED;e þ ED;hl þ ED;d þ ED;fml þ ED;sol
Reduction reaction C þ CO2 ¼ 2CO (r1) þ ED;dl ð4Þ
NiO þ CO ¼ Ni þ CO2 (r2)
3Fe2 O3 þ CO ¼ 2Fe3 O4 þ CO2 (r3) where ED,in and ED,out denote the total energy of input flows and
Fe3 O4 þ CO ¼ 3FeO þ CO2 (r4) output flows in rotary dryer, respectively; ED,bbc is the heat from
FeO þ CO ¼ Fe þ CO2 (r5)
bituminous coal combustion where its energy equation is given
NiO  SiO2 þ C ¼ Ni þ CO þ SiO2 (r6)
by [27]; ED,bc, ED,wo, ED,fm and ED,ca are the sensible heat of bitumi-
Decomposition reaction CaCO3 ¼ CaO þ CO2 (r7)
nous coal, wet laterite ores (without free moisture), free moisture
Combustion reaction C þ O2 ¼ CO2 (r8) in input materials and combustion air, respectively; ER,fg is the
2CO þ O2 ¼ 2CO2 (r9)
waste heat of flue-gas recovery from rotary kiln; ED,e is the heat
546 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559
consumption of free moisture evaporation; ED,hl is the heat loss of
rotary dryer and its calculation equation is given by [27]; ED, fm,
ED,d and ED,so are the waste heat of residual 7% free moisture, dust
and semi-dry laterite ores entered in rotary kiln, respectively; ED,
fml, ED,dl, ED,sol and ED,fg are the waste heat of residual 7% free mois-
ture, dust, semi-dry laterite ores and flue gas loss to the surround-
ings, respectively. For simplicity, the free moisture is calculated
separately from the raw materials.
3.2. Mathematical model of material and energy flow in rotary kiln
Similarly, the material conservation of the rotary kiln is
expressed as (see also in Fig. 4):
M R;in ¼ M R;bc þ ME;Fg þ M R;ca þ M R;L þ M R;a þ M D;so þ M D;d
M R;out ¼ M R;oc þ M R;fg þ MR;fge
where MR,in and MR,out denote the total mass of input flows and out-
put flows in rotary kiln, respectively; MR,bc, ME,Fg, MR,ca, MR,L, MR,a,
MR,oc and MR,fge are the mass of bituminous coal, furnace gas, com-
bustion air, limestone, anthracite, laterite ores calcined and flue gas
entered into electric furnace, respectively.
For the energy balance, there are ten energy inputs and nine
energy outputs. The corresponding energy conservation equations
are given by:
ER;in ¼ ER;bcc þ ER;bc þ ER;Fgc þ EE;Fg þ ER;ca þ ED;so þ ER;fm þ ER;L
þ ER;a þ ED;d
ER;out ¼ ER;fg þ ER;oc þ ER;rr þ ER;e þ ER;dr þ ER;hl þ ER;fgl þ ER;ocl
þ ER;fge
where ER,in and ER,out denote the total energy of input flows and out-
put flows in rotary kiln, respectively; ER,bcc and ER,Fgc are the heat
from bituminous coal and furnace gas combustion, respectively;
ER,bc, ER,ca, ER,fm and ER,L are the sensible heat of bituminous coal,
combustion air, free moisture in input materials and limestone,
respectively; EE,Fg is the waste heat of furnace gas entered into
rotary kiln; ER,a is the heat of anthracite reaction; ER,oc and ER,fge
are the waste heat of laterite ores calcined and flue gas entered into
electric furnace; ER,rr, ER,e and ER,dr are the heat consumption
of reduction reaction, moisture evaporation and dissociation
Fig. 4. Schematic of material and energy flows in the rotary kiln.
reaction of limestone, respectively; ER,hl is the heat loss of rotary
kiln and its calculation equation is given by [27]; ER,fgl and ER,ocl
are the waste heat of flue gas and laterite ores calcined loss to the
surroundings, respectively.
3.3. Mathematical model of material and energy flow in electric
furnace
As depicted in Fig. 5, the material conservation of the electric
furnace can be expressed as:
ME;in ¼ M R;oc þ M R;fge ð9Þ
ME;out ¼ M E;Fa þ M E;sl þ ME;Fg ð10Þ
ð5Þ
where ME,in and ME,out denote the total mass of input flows and out-
ð6Þ put flows in electric furnace, respectively; ME,Fa and ME,sl are the
mass of ferronickel alloy and slag, respectively.
Meanwhile, the corresponding energy balance can be calculated
from the following equations:
EE;in ¼ EE;J þ ER;oc þ EE;CO þ EE;sr þ ER;fge ð11Þ
EE;out ¼ EE;Fa þ EE;sl þ EE;rr þ EE;Fg þ EE;hl þ EE;lhm þ EE;el þ EE;wc
þ EE;Fgl ð12Þ
where EE,in and EE,out denote the total energy of input flows and out-
put flows in electric furnace, respectively; EE,J is the Joule heat; EE,CO
ð7Þ and EE,sr are the heat of exothermic reaction of producing CO and
slagging reaction, respectively; EE,Fa, EE,sl and EE,Fgl are the waste
heat of ferronickel alloy, slag and furnace gas loss to the environ-
ment, respectively; EE,rr is the heat consumption of endothermic
ð8Þ reduction reaction; EE,hl is the heat loss of furnace; EE,lhm is the
latent heat of melting; EE,el is the heat loss due to electric loss;
EE,wc is the heat loss due to water-cooling. Detail calculation proce-
dures of the EE,hl, EE,lhm, EE,el and EE,wc can be found in the [27,28]
and the reference therein.
3.4. Recycling flue gas waste heat of rotary kiln
To reduce the energy consumption of smelting process, the flue
gas waste heat of rotary kiln is recycled into rotary dryer for pre-
heating and dehydrating. To simplify calculation, it is assumed that
Fig. 5. Schematic of material and energy flows in the electric furnace.
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559
the mass loss of rotary kiln flue gas is negligible in the transporta-
tion process which consists of the mass of carbon dioxide, sulfur
dioxide, oxygen, nitrogen and steam.
For the flue gas in rotary kiln, the carbon dioxide gas is mainly
generated from reduction reaction, limestone dissociation and
combustion of bituminous coal and furnace gas. The rest of carbon
dioxide gas is released from furnace gas. Similarly, the sulfur diox-
ide gas is released by the coal combustion product and furnace gas.
The oxygen gas is released by the residual combustion air and fur-
nace gas which is given by:
ca Fg
M fg
R;O2 ¼ O2  ððLn  L0 Þbc þ ðLn  L0 ÞFg Þ þ M E;Fg  O2
where Ln is the actual amount of air supply; L0 is theoretical air
requirement; Oca
2 is the mass percentage of oxygen in combustion
air; OFg
2 is the mass percentage of oxygen in furnace gas.
In the rotary kiln, the excess air mass of bituminous coal
combustion is as follows:
ðLn  L0 Þbc ¼ ða  L0  L0 Þbc
¼ 0:2  ð0:0889  C þ 0:2667  H þ 0:0333  ðS  OÞÞ
where a is air factor, for the bituminous coal combustion, a = 1.2, for
furnace gas combustion, a = 1.1 [27,29]; H, S and O are the mass
percentages of H, S and O in as-received bituminous coal.
The excess air mass of furnace gas combustion:
0:5  COFg  OFg
ðLn  L0 ÞFg ¼ ða  L0  L0 ÞFg 0:1  2
21
where COFg and OFg
2 are the mass percentages of carbon monoxide
and oxygen in furnace gas, respectively.
The nitrogen gas is released by combustion air and furnace gas:
M fg ca Fg
R;N2 ¼ M R;ca  N2 þ M E;Fg  N2
where Nca Fg
2 and N2 are the mass percentages of nitrogen in combus-
tion air and furnace gas, respectively.
The vapour is released by dehydrating and furnace gas:
M fg
R;H2 O ¼ M D;so  H2 Oso þ M R;bc  H2 Obc þ M R;L  H2 OL þ M R;ca
 H2 Oa þ M E;Fg  H2 OFg
where H2 OL , H2 OR;bc , H2 Oso and H2 Oa are the mass percentages of
the moisture in limestone, bituminous coal and semi-dry ores and
anthracite respectively; H2 OFg is the mass percentage of vapour in
furnace gas.
3.5. Recycling the furnace gas of electric furnace
In electric furnace, due to the vast amount of carbon monoxide
in the furnace gas, the furnace gas is recycled into rotary kiln as
fuel for additional heat source. The heat of furnace gas combustion
is given by [27]:
ER;Fgc ¼ M E;Fg  ½126:2CO þ 107:8H2 þ 359:1CH4 þ   
þ 231:2H2 S
where CO, H2, CH4 and H2S are the mass percentages of carbon
monoxide, hydrogen, methane and hydrogen sulfide in furnace
gas. Moreover, the furnace gas is composed of flue gas entered into
electric furnace and carbon monoxide, in which the carbon monox-
ide is generated by reduction reaction. Therefore, the mass of fur-
nace gas consists of the carbon dioxide, sulfur dioxide, oxygen,
nitrogen and steam in flue gas and carbon monoxide in electric fur-
nace. The mass of flue gas enters into electric furnace is given as:
M R;fge ¼ V fge  qfge ¼ V R  c  w  qfge
547
where Vfge is the volume of flue gas enters into electric furnace; VR is
the volume of rotary kiln; c is the fill rate of material in rotary kiln;
w is the void fraction.
3.6. The waste heat of rotary dryer flue gas
In rotary dryer, the flue gas of rotary dryer is composed of bitu-
minous coal combustion product, flue gas of rotary kiln and water
vapour gas of 16% free moisture. Besides, the mass of combustion
production in rotary dryer depends on the mass of bituminous coal
and combustion air:
ð13Þ
MD;fg ¼ M R;fg þ 16%  M D;wo  H2 Owo þ M D;bc þ M D;ca ð20Þ
where H2 Owo is the mass percentage of free moisture in laterite ores
of rotary dryer.
3.7. Efficiency analysis of RKEF process
An effective way to evaluate the energy efficiency in rotary
ð14Þ kiln-electric furnace process is by means of energy flow analysis.
The energy flow analysis could identify most of the energy lost
to the surroundings as well as the energy degradation because of
the entropy generation within the system. In addition, it could also
be adopted in distinguishing and quantifying both the available
energy used to drive the system as well as the unavailable energy
(i.e. avoidable and unavoidable energy loss) loss to the
ð15Þ surroundings.
The energy flow balance of a process takes into consideration
that the energy supply is partly obtained by materials for driving
a system (available energy), partly loss by energy transport to
the surroundings (unavailable energy) [13,26]:
ð16Þ EF;E þ ES ¼ EA þ EUn ð21Þ
where EF,E is the supply energy by fuel combustion and electric
energy; ES is the sensible heat of input materials; EA is the available
energy; EUn is the unavailable energy.
The available energy contains the energy consumption of chem-
ical reaction, of dehydrating and of melting (alloy and slag) as
ð17Þ shown in Table 3. The unavailable energy takes into consideration
that avoidable and unavoidable energy loss as shown in Table 3.
The avoidable energy loss contains the waste heat of output flows
loss to the environment, which can be reutilized for improving
energy efficiency. The unavoidable energy loss contains heat losses
of furnace, cooling-water and electric loss.
The energy efficiency (ge) is defined as the ratio between avail-
able energy and energy supply (EF,E + ES):
ge ¼ EA =ðEF;E þ ES Þ ð22Þ
If the avoidable energy loss can be reused as heat source for pre-
heating and driving the process, the energy efficiency will be
improved significantly. Therefore, discharge rate of energy flow
(gep) is defined as the ratio between avoidable energy loss and
energy supply (EF,E + ES):
ð18Þ
gep ¼ EAvo =ðEF;E þ ES Þ ð23Þ
3.8. The technic index of RKEF process
The technic index is help to evaluate the technological level of
production process. For example, in material flow network, the
nickel and iron contents are very important character parameters,
which can directly affect the metallicity of alloy. The recovery ratio
of nickel/iron (Rfe/Rni) is defined as the ratio between the metal
ð19Þ
mass (i.e. nickel and iron) in alloy and total mass of metal in
548 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559

Table 3
The equations of available energy, avoidable and unavoidable energy loss.

Rotary dryer Rotary kiln Electric furnace

EA EA;D ¼ ED;e EA;R ¼ ER;rr þ ER;e þ ER;dr EA;E ¼ EE;rr þ EE;lhm
EAvo EAvo;D ¼ ED;fg þ ED;fml þ ED;sol þ ED;dl EAvo;R ¼ ER;fgl þ ER;ocl EAvo;E ¼ EE;fa þ EE;sl þ EE;Fgl
EUnavo EUnavo;D ¼ ED;hl EUnavo;R ¼ ER;hl EUnavo;E ¼ EE;hl þ EE;el þ EE;wc

laterite ores [24]. The basicity of metal can directly impact on the
viscosity of slag and slag composition. Furthermore, the basicity
depends on the contents of CaO, MgO and SiO2 [24].
For improving the resource utilization, the slag is reutilized as
construction material and the nickel is recycled from dust into
rotary kiln. Moreover, the flue gas and furnace gas are recycled
as energy supply to increase efficiency. The recycling rate of waste
solid is defined as the ratio between the mass of recycled waste
solid (dust and slag) and total mass of waste solid of process:
gws ¼ ðMd þ Msl Þ=ðMD;d þ ME;sl Þ
where Md and Msl are mass of recycled dust and slag respectively.
The recirculation rate of waste is given by:
Md þ MFg þ M fg
gw ¼
MD;d þ M D;fg þ M R;fg þ M E;sl þ ME;Fg
where MFg and Mfg are mass of recycled furnace gas and flue gas,
respectively. The continuation degree of process is defined as the
ratio between pure process time of material flow in the process
and total consumption time of ferronickel produced [30,31]:
P dimension is the energy flow network (G2). Each network contains
tppt
C0 ¼ P P P P ð26Þ
t ppt þ t wt þ t ht þ t mt
P P
where tppt is sum-of-processing-time of material flow; twt is
P
sum of waiting time of material flow; tht is sum of haulage time
P > em;n ell
of material flow; tmt is sum of maintenance time.
every network is considered as a compound system composed of
three subsystems (rotary dryer, rotary kiln and electric furnace),
which works in coordination with each other to present the overall
performance of the network. The larger the synergy degree is, the
higher matching degree between M-E-N is; implying a better per-
formance of the compound system. In other words, the energy effi-
ciency of system is higher and the corresponding energy loss is
lower. Therefore, the synergy method can reflect the cooperation
of subsystems. The recent research of synergetics [32–34] pointed
out that the order parameter can dominate the evolution of the
ð24Þ
process, and assist the disorder system transform into a new
ordered structure by self-organizing. In other words, the order
parameter can govern the final structure of the system and synergy
degree. Meanwhile, according to Sun [35], the order parameter
ð25Þ could incur positive and negative effect on the subsystem order
degree. The efficacy coefficient is introduced to indicate the contri-
bution degree of order parameter to the order of system. To ana-
lyze the synergy degree between M-E-N in RKEF process, define
the system in two dimensions (Gm, where m = 1, 2). The first
dimension is the material flow network (G1) and the second
5 indices of the order parameters, for n = 1, 2 . . . N as shown in
Table 4. The efficacy coefficient of the order parameter is defined
as follows:
8
< eul em;n ; em;n 2 fThe order parameter with positive effectg
ll
EC m ðem;n Þ ¼
m;n m;n

In energy flow network, some indices are used to assess the > ll e
: em;n ; em;n 2 fThe order parameter with negative effectg
m;n
eul ell
energy utilization efficiency. For instant, the recycling rate of resid- m;n m;n

ual heat and energy can reveal the energy saving ability of system. ð30Þ
The recycling energy contains chemical energy and waste heat of
output flow. The recycling rate of residual heat and energy is where em;n is the order parameter component; eul ll
m;n and em;n are the

m;n 6 em;n 6 em;n .

upper limit and lower limit of em;n , viz eul ll
defined as the ratio between the recycling energy and the total
residual heat and energy: The efficacy coefficient and the weights of order parameters are
calculated to obtain the order degree. Based on the linear weighted
ðER;Fgc Þ þ ðEE;Fg þ ER;oc þ ER;fg Þ
gsy ¼ ð27Þ summation method, the order degree can be written as:
ER;Fgc þ EE;Fg þ ER;oc þ ER;fg þ EAvo;D þ EAvo;R þ EAvo;E
X
N
where the chemical energy and waste heat recovery are referenced ODm ðGm Þ ¼ ½wm;n  EC m ðem;n Þ ð31Þ
in first and second parts of the numerator in Eq. (27), respectively. n¼1
The recycling rate of by-product energy can express the reuti-
lization degree of by-product. The by-product of RKEF process con- where ODm(Gm) is the order degree of Gm, for 0 6 ODm(Gm) 6 1; wm,n
tains flue gas, furnace gas and slag. The recycling rate of by-product is the weighting of the efficacy coefficient of the order parameter
PN
energy is defined as the ratio between the energy of the recycled em,n, for wm;n P 0 and n¼1 wm;n ¼ 1. In the case ODm(Gm)
by-product and the total energy of the by-product: approaches to 0, it indicates that the subsystem order degree is at
ðER;Fgc Þ þ ðER;fg þ EE;Fg Þ the lowest value; while ODm(Gm) approaches to 1, it means that
gbp ¼ ð28Þ
ER;Fgc þ EE;Fg þ ER;fg þ ED;fg þ ER;fgl þ EE;Fgl þ EE;sl
Table 4
The specific energy consumption can reveal the comprehensive
The indices of order parameter (em,n).
energy consumption per ton product:
The indices of Material Flow (G1) The indices of Energy Flow (G2)
qFa ¼ EF;E =M E;Fa ð29Þ
e1,1 Ni recovery rate (Rni) e2,1 Specific energy consumption
where qFa is specific energy consumption. (qNiFe)
e1,2 Fe recovery rate (Rfe) e2,2 Energy efficiency of system (ge)
e1,3 Continuation degree of e2,3 Recycling rate of residual heat
3.9. Synergy degree of RKEF process process (C0 ) and energy (gsy)
e1,4 Recirculation rate of waste e2,4 Discharge rate of energy flow
To further assess the efficiency of the RKEF process, the syner- (gw) (gep)
e1,5 Re-resourcization rate of e2,5 Recycling rate of by-product
getic theory is introduced to analyze the synergy degree between
waste solid (gws) energy (gbp)
Material flow and Energy flow Networks (M-E-N). In this paper,
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559 549

the subsystem order degree is the highest. The entropy value is then Table 5
converted into weighting: Constraint condition.

Parameter Value
1  Sen
m;n
wm;n ¼ ð32Þ Production per hour, t/h 12
XN
Nickel content in crude ferronickel alloy,% 12
N Sen
m;n
Rotary dryer
n¼1
The inlet temperature of rotary kiln flue gas, °C 260
The entropy value method is a kind of the objective weighting The outlet temperature of rotary dryer flue gas, °C 105
methods, which is based on the information from every index to The outlet temperature of semi-dry laterite ores, °C 85

determine the weighting. The entropy value of the order Rotary kiln
parameter: The inlet temperature of semi-dry laterite ores, °C 21
The inlet temperature of furnace gas, °C 230
The outlet temperature of laterite ores calcined, °C 880
1 X L
The outlet temperature of rotary kiln flue gas, °C
Sen
m;n ¼  ðhl  lnhlm;n Þ ð33Þ 375
ln L l¼1 m;n Electric furnace
The inlet temperature of laterite ores calcined, °C 400
where Sen l The outlet temperature of furnace gas, °C 950
m;n is the entropy value of em,n; hm;n is the normalized data of
The outlet temperature of alloy, °C 1556
em,n. The outlet temperature of slag, °C 1534
Based on the min-max normalization method, the normalized
data is: (1) In rotary dryer, through dust-collection technology, there
8 are no the nickel/iron components in laterite ores loss with
> elm;n minfelm;n g
>
> flue gas.
>
>
l
; em;n 2 fThe order parameter with positive effectg
< maxfelm;n gminfelm;n g
hlm;n ¼ l l (2) In rotary kiln, the laterite ores is completely dehydrated.
>
> maxfelm;n gelm;n (3) In rotary kiln, the waste heat loss of the flue gas entered into
>
>
: maxfelm;n gminfelm;n g ; em;n 2 fThe order parameter with negative effectg
l
>
l l
electric furnace is ignored.
ð34Þ (4) In electric furnace, due to the enclosed electric furnace, there
is no air entering into the electric furnace during smelting
Since every network interacts and collaborates with each other, ferronickel.
the synergy degree represents the coordination ability and match-
ing degree. Therefore, the synergy degree between material flow Moreover, for calculation, the assumptions are as follows:
and energy flow is given by:
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (i) The Basicity of material is 0.8.
u 2
uY (ii) Nickel recovery is 95%.
SD ¼ ð1  DRÞ  t ODm ðGm Þ ð35Þ
(iii) The dehydration rate of laterite ores in rotary dryer is 16%.
m¼1

where DR represents the mismatching degree between M-E-N,
on the contrary, (1-DR) represents the matching degree between
M-E-N.
Furthermore, the performance of system depends not only on
the value of order degree of every network, but also the combina-
tion form of networks. Thus, the coordination ability is got by using
geometric method:
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 2
uY
SA ¼ t ODm ðGm Þ
m¼1
Due to the difference between the order degrees of M-E-N, the
coordinated operation of two networks will be influenced; leading
to energy dissipation. Therefore, the energy flow dissipation rate is
used to indicate the mismatching degree between the material
flow and energy flow:
X
K
DR ¼ g d g k
k¼1
where g 
k is the k-th the energy of output flow, for k = 1,2, . . . , K.
3.10. The coupling algorithm for solving material and energy flows
Throughout the complicated the smelting process of rotary kiln-
electric furnace, the crude ferronickel is produced for further
refinement. The constraint condition was made for some of the
process parameters as shown in Table 5. Based on the process
requirements, it is necessary to take into account some considera-
tions, which are described below:
To resolve the interaction among subsystems, the mathematical
model requires coupling method to resolve the material and
energy flows of each subsystem. During solution process, the order
of calculation is adjusted according to the constraint relation as
shown in Fig. 6. The procedures of couple algorithm are briefly
described as follows:
1. Obtain the wet ores mass and chemical composition mass based
on the constraint condition. Calculate the mass of dehydration
ð36Þ of rotary dryer and rotary kiln.
2. Since the recycling the flue gas of rotary kiln and furnace gas of
electric furnace will impact the flue gas flow rate, the coupling
calculation of material and energy flows is firstly carried out
without recycling flue gas into rotary dryer and furnace gas into
rotary kiln. The mass flow rates of bituminous coal, flue gas and
furnace gas are calculated as well as related energy flows, which
will be used as the boundary conditions for further calculation.
3. Determine the amount of flue gas transported to rotary dryer
ð37Þ and the furnace gas transported to rotary kiln and update the
values of material and energy flows using iterative procedures.
4. Calculate the technic indices based on the result of material and
energy flows.
5. Obtain the efficacy coefficient, entropy value and weighting of
the order parameter.
6. Finally, evaluate the order degree, coordination ability, match-
ing degree and synergy degree.
4. Results and discussion
Through the calculation of material and energy flows, the tech-
nic index can be obtained. The calculation results of material and
550 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559

Fig. 6. Procedure schematic of energy efficiency analysis and synergy degree evaluation.

energy flows balance of each subsystem are shown in Tables 6
and 7.
4.1. Material flow analysis
In this present work, assuming the production rate of 12 t/h fer-
ronickel alloy with 12% of mass Ni, the Sankey diagram of material
flow in rotary dryer is shown in Fig. 7a. Due to the additional mass
of the combustion gases and vapour, the mass percentage of flue
gas increases by 14.87%. Moreover, the mass of dehydration
accounts for 8.70% of output flows. Fig. 7b shows the Sankey dia-
gram of the material flow in rotary kiln. The mass percentage of
raw materials (the ores, limestone, anthracite and dust) is
decreased by 18.05% due to the dehydration and chemical reac-
tions. Meanwhile, the mass of dehydration accounts for 9.02% of
the output mass flow. Finally, the Sankey diagram of material flow
in electric furnace is shown in Fig. 7c. As depicted, the mass per-
centage of furnace gas generated accounts for 7.60% of output mass
flow.
4.2. Energy flow analysis
The Sankey diagram of energy flow in the rotary dryer is shown
in Fig. 8a. As shown in the figure, the majority of the energy supply
for the rotary dryer comes from the bituminous coal combustion
(51.39% of energy supply) and the waste heat of flue gas (43.79%
of energy supply). The available energy of rotary dryer for the
16% free moisture evaporation is corresponding to the energy effi-
ciency (ge) of 63.59%. Nevertheless, 29.92% of the energy output is
lost through the waste heat loss of the semi-dry ores (i.e. 4.07%),
free moisture (i.e. 1.99%), dust (i.e. 0.99%) and flue gas
(i.e. 22.87%). One should notice that the flue gas waste heat loss

Table 6
Material flow balance.

Rotary dryer Rotary kiln Electric furnace

MD-in (t/h) MD-out (t/h) MR-in (t/h) MR-out (t/h) ME-in (t/h) ME-out (t/h)
MD,bc 1.93 MD,so 80.75 MR,bc 5.22 MR,oc 96.53 MR,oc 96.53 ME,Fa 12.00
MD,ca 12.63 MD,fg 137.13 MR,ca 53.69 MR,fg 102.08 MR,fge 0.07 ME,sl 77.27
MR,fg 102.08 MD,d 18.03 MR,L 26.13 MR,fge 0.07 ME,Fg 7.34
MD,wo 119.31 ME,Fg 7.34
MD,so 80.75
MD,d 18.03
MR,a 7.48
Total 235.95 Total 235.91 Total 198.64 Total 198.68 Total 96.60 Total 96.61
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559 551

Table 7
Energy flow balance.

Rotary dryer Rotary kiln Electric furnace

ED,in (GJ/h) ED,out (GJ/h) ER,in (GJ/h) ER,out (GJ/h) EE,in (GJ/h) EE,out (GJ/h)
ED,bcc 44.760 ED,fg 19.918 ER,bcc 121.210 ER,fg 38.139 EE,J 133.833 EE,Fa 11.555
ED,bc 0.051 ED,so 1.163 ER,bc 0.137 ER,oc 38.059 ER,oc 38.059 EE,sl 106.047
ED,wo 1.447 ED,fm 0.735 ER,Fgc 72.658 ER,rr 7.266 EE,CO 18.369 EE,rr 53.662
ED,fm 2.434 ED,e 55.391 EE,Fg 1.765 ER,e 50.750 EE,sr 51.138 EE,Fg 1.765
ED,ca 0.273 ED,hl 3.469 ER,ca 1.141 ER,dr 43.378 ER,fge 0.036 EE,hl 4.405
ER,fg 38.139 ED,d 0.284 ED,so 1.163 ER,hl 2.011 EE,lhm 31.098
ED,sol 3.545 ER,fm 0.699 ER,fgl 16.646 EE,el 3.229
ED,fml 1.734 ER,L 0.458 ER,ocl 45.672 EE,wc 23.922
ED,dl 0.865 ER,a 42.441 ER,fge 0.036 EE,Fgl 5.753
ED,d 0.284
Total 87.104 Total 87.104 Total 241.956 Total 241.957 Total 241.435 Total 241.436

is considerably substantial. Through the processes of dedusting
and waste gas cleaning, the flue gas is directly discharged into
the environment without recycling its waste heat. Therefore, the
waste heat of flue gas is completely lost to the surroundings (con-
tribute to the heat loss of 19.918 GJ/h – see also in Table 7). How-
ever, it is also worth noting that the temperature of the flue gas is
approximately 85 °C. Such a low temperature has a relatively low
recovery value; causing it becomes economically not viable to
scavenge the waste heat from the flue gas.
In Fig. 8b, it appears that the energy supply from the bitumi-
nous coal combustion (i.e. 50.1%) is the dominant factor in com-
pared to the anthracite reaction (i.e. 17.54%) and furnace gas
combustion (i.e. 30.03%). The available energy of rotary kiln is used
to supply heat source for reduction reaction (i.e. 3.00%), limestone
decomposition (i.e. 17.93%) and dehydrating (i.e. 20.97%) with the
energy efficiency (ge) of 41.90%. Meanwhile, the discharge rate of
energy flow (gep) is of 25.76%. As depicted, the avoidable energy
loss is essentially caused by the heat dissipation occurs during
the transport process of flue gas and the ores calcined, in which
the waste heat loss of flue gas and the ores calcined account for
6.88% and 18.88% respectively. Therefore, by recycling the waste
heat of flue gas, the coal consumption of rotary dryer could be
reduced by 0.72 t/h. In addition, by reducing heat dissipation loss
of the ores calcined, the specific energy consumption could be also
reduced in electric furnace.
Similarly, as shown in Fig. 8c, the main energy is supplied by
Joule heating (55.43% of energy supply), which is around three
time as much of the waste heat of the ores calcined (i.e. 15.76%)
and more than double of the slagging reaction (i.e. 21.18%). In
the electric furnace, the available energy is used to supply heat
source for chemical reactions and slag/alloy melting. According
to the calculation, the resultant energy efficiency (ge) is of
35.11%. However, the discharge rate of energy flow (gep) is of
51.09%. This is because of the slag is reutilized as construction
material without recycling its waste heat so that its waste heat is
lost to the environment by heat dissipation. Moreover, the heat
dissipation also occurs during the transport process of furnace
gas and ferronickel alloy. As depicted, the waste heat loss of slag,
ferronickel alloy and furnace gas account for 43.92%, 4.79% and
2.38% of energy output respectively. With the waste heat of fur-
nace gas recovery, the coal consumption of rotary kiln could be
reduced by 0.25 t/h.
Based on the above energy flow analysis, it is important to pro-
pose some energy conservation measures for the process. In sum-
mary, the energy efficiency for the rotary dryer, rotary kiln and
electric furnace is of 63.59%, 41.90% and 35.11% respectively (see
Table 8). The rotary dryer shows a higher energy efficiency due
to its relatively lower exit temperatures for flue gas and the ores;
causing a reduction of heat dissipation loss. On the other hand,
the discharge rate of energy flow for the rotary dryer, rotary kiln
and electric furnace is of 29.92%, 25.76% and 51.09% respectively.
The electric furnace shows a higher discharge rate of energy flow.
One of the reasons is due to the high exit temperatures for the slag
and alloy; leading to a substantial of heat dissipation loss. Another
reason is the waste heat of slag and alloy loss to the environment
without recycling the waste heat. Under these conditions, to
increase the energy efficiency, it is necessary to reduce exit tem-
peratures of output flows and reutilize waste heat in order to min-
imize the heat dissipation loss.
In addition to the heat dissipation loss, it is also crucial to pay
attention to the heat loss of furnace body, preheating loss and elec-
tric loss. As stated in Table 7, it appears that the heat loss of rotary
dryer, rotary kiln and electric furnace body are 3.469 GJ/h,
2.011 GJ/h and 4.405 GJ/h, respectively. In the other words, it could
be also worthwhile to minimize heat loss by improving the ther-
mal insulation of furnace body. The air leakage of the feed inlet
and discharge outlet could also be the primary cause of heat loss
by convection and radiation. As feeding material and discharging,
the heat will lost with off-gas from the feed inlet and discharge
outlet or dissipated to the surroundings by radiation. The time of
feeding material and discharging could be adjusted to reduce the
associated heat loss. Moreover, it is necessary to minimize the bus-
bar and transformer losses [28] which are the major causes of the
electric loss. However, measures or strategy to reduce the afore-
mentioned energy loss is subject to further investigation.
4.3. Energy saving for RKEF process
One of the main energy saving measures in the rotary kiln-
electric furnace process is to recycle the rotary kiln flue gas into
rotary dryer for preheating bituminous coal and combustion air.
As depicted in Fig. 8b, flue gas waste heat accounts for 22.66% of
energy output in rotary kiln, which contains waste heat loss (i.e.
6.88%), residual waste heat entered into rotary dryer (i.e. 15.76%)
and residual waste heat entered into electric furnace (i.e. 0.02%).
Furthermore, 30.36% of the waste heat is lost to the environment
where only 69.64% of the waste heat is scavenged as heat source.
The reutilization of the waste heat could save 1.64 t/h bituminous
coal corresponding to around 45.94% of the bituminous coal con-
sumption. Another energy saving measure is to recycle the electric
furnace gas as fuel in rotary kiln. The furnace gas combustion
accounts for 30.03% of energy supply (Fig. 8b). Similarly, the reuti-
lization of furnace gas could save approximately 3.21 t/h bitumi-
nous coal; reducing the coal consumption by approximately 38.00%.
Nevertheless, lots of waste heat is still not reutilized. For exam-
ple, the waste heat of flue gas in the rotary dryer could be reused to
preheat mixed material in raw proportioning station. As shown in
Table 9, the waste heat of the flue gas can be recovered by the
water heat boiler with a potential saving of 2.589 GJ/h. Further-
more, the efficiency of the rotary dryer could be enhanced by
552 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559

Fig. 7. Sankey diagram of the material flow: (a) Rotary dryer; (b) Rotary kiln; (c) Electric furnace.
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559
Fig. 7 (continued)
2.97%; reducing the associated avoidable energy loss by 9.93%.
Nonetheless, decreasing the temperature in the transportation pro-
cess could increase the resultant waste heat loss and the associated
energy loss. As shown in Table 9, decreasing the temperature of
semi-dry ores by 10 °C will increase the waste heat loss by
15.62%; where the corresponding discharge rate of energy flow
and the avoidable energy loss in rotary dryer are increased by
0.64% and 2.13% respectively. Similarly, decreasing temperature
of laterite ores calcined and flue gas by 10 °C will increase the
resultant waste heat loss by 2.08% and 8.76% respectively. There-
fore, it is necessary to improve the thermal insulation property of
flue for minimizing the heat dissipation loss of flue gas.
For the electric furnace, the analysis also shows that a total of
59.386 GJ/h of waste heat could be recovered from the slag by
power generation through waste heat recovery. As a result, the
energy efficiency of electric furnace will increase by 24.60%; reduc-
ing the avoidable energy loss by 48.14%. In contrast, caused by
reduction of furnace gas temperature (i.e. 10 °C less), the resultant
waste heat loss will slightly increase by 1.42%. The discharge rate
of energy flow and the avoidable energy loss of electric furnace will
only increase by 0.03% and 0.07% respectively. Meanwhile, by
using mold flux, the temperature of ferronickel alloy could be
effectively controlled (i.e. only reducing approximately 15 °C).
Although such temperature reduction will incur additional waste
heat loss by 0.116 GJ/h, the mold flux could save a total of
11.439 GJ/h waste heat; reducing the discharge rate of energy flow
the avoidable energy loss of electric furnace by 4.74% and 9.27%.
4.4. Influence of Ni content in ferronickel alloy
The effect of nickel content in ferronickel alloy on material and
energy flows is shown in Figs. 9–11. As shown in
Figs. 9b, 10b and 11b, with the higher Ni grade in the ferronickel
alloy, the mass of related material flow increases in the three sub-
systems. This is owing to the fact that enhancing nickel grade in
553
alloy increases the demand of nickel quantity. As a result, more
raw materials (i.e. laterite ores, anthracite and etc.) are required
for smelting which consumes more fuel eventually; leading to a
higher output flows (e.g. flue gas and slag). From Figs. 9b and 10b
it appears that the bituminous coal of rotary dryer slightly increases
from 1.66 t/h to 2.28 t/h and bituminous coal of rotary kiln also
slightly increases from 4.26 t/h to 7.30 t/h. Meanwhile, an increase
of raw materials will increase the thickness of the material layer,
which leads to reducing the temperature of materials at the bottom
layer. Moreover, as shown in Figs. 9b, 10b and 11b, the growth rate
of the by-product is considerably larger than the product in the
three subsystems. The increment of raw material will produce more
slag (see Fig. 11b) and more vapour and carbon oxide (CO or CO2);
resulting more flue gas output.
Similarly, by providing more fuel (i.e. bituminous coal and fur-
nace gas) in the two subsystems, it will increase the amount of
combustion air and eventually affect the efficiency of rotary dryer
and rotary kiln. There are two main possible reasons: Firstly, the
increase of combustion air will also result in more flue gas in the
process; leading to more associated heat loss through the flue
gas. Secondly, the temperature in the kiln will increase which
sequentially enhance the heat loss to the surroundings through
conduction of furnace body or convection and radiation. Mean-
while, with the increase of the material layer thickness, heat trans-
fer to the bottom material layer will be hindered which ultimately
reduce the conversion degree of calcination. The effect of nickel
content in ferronickel alloy on energy efficiency and discharge rate
of energy flow is shown in Fig. 15a.
4.5. Influence of Ni content in laterite ores
The material and energy flows with respect to the nickel con-
tent in laterite ores are shown in Figs. 12–14. As depicted, it
appears that the mass of related material flows decreases in the
554 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559

Fig. 8. Sankey diagram of the energy flow: (a) Rotary dryer; (b) Rotary kiln; (c) Electric furnace.

three subsystems with the increasing nickel content in laterite
ores. With the nickel grade in alloy remains the same, the
increased nickel content in laterite ores will reduce the raw mate-
rials consumption. The associated fuel consumption and the
amount of output flows will be also reduced. On the other hand,
the available energy and avoidable energy loss are also decreased
as shown in Figs. 12a, 13a and 14a respectively. This is essentially
due to the reduction of energy consumption for chemical reaction
and dehydration. Meanwhile, a reduction of raw materials supply
will also enhance the continuation degree, reduce the discharge
of output flows quantities and its associated energy flow dissipa-
tion loss. The effect of nickel content in laterite ores on energy effi-
ciency and discharge rate of energy flow is shown in Fig. 15b.
4.6. Assessment of synergy degree for RKEF process
In order to assess the performance of the RKEF process, technic
indices and synergy analysis were carried out in the present study.
Two test cases (referred as Test 1 and 2 hereafter) corresponding to
produce 12 t/h ferronickel alloy containing 12 mass% Ni and
14 mass% Ni respectively were adopted for assessment. The pro-
duction parameters of the cases are then utilized to calculate the
technic indices. The values of proposed indices are tabulated in
Tables 10 and 11.
In general, one may notice from the table that Test 1 shows a
better technic index than Test 2. For example, in the material flow
network, the iron recovery rate (Rfe) and continuation degree (C00 )
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559 555

Fig. 8 (continued)

Table 8
Energy flow analysis.

EA (GJ/h) EAv (GJ/h) EUnavo (GJ/h) ge (%) gep (%)

Rotary dryer 55.391 26.062 3.469 63.59 29.92
Rotary kiln 101.394 62.318 2.011 41.90 25.76
Electric furnace 84.760 123.355 31.556 35.11 51.09

Table 9
Energy saving analysis.

DEa Ewhl D ge Dgep DgAvb Dgwhlc

(GJ/h) (GJ/h) (%) (%) (%) (%)
Flue gas (D) 2.589 19.918 2.97 – 9.93 13.00
Semi-dry 0.554 3.545 – 0.64 2.13 15.62
ores
Flue gas (R) 1.461 16.646 – 0.60 2.34 8.76
Ores 0.951 45.672 – 0.39 1.53 2.08
calcined
Slag 59.386 106.047 24.60 – 48.14 56.00
Furnace gas 0.082 5.753 – 0.03 0.07 1.42
Alloy 11.439 11.555 – 4.74 9.27 98.99
E0 E0
DE ¼ M  cp  DT; bDgAv ¼ AVEAV AV ¼ EDAVE ; cDgwhl ¼ whlEwhl whl ¼ EDwhl
E E
a E
; where DE denotes
the energy change of material flow with different temperature; DgAv and Dgwhl
denote the change rate of avoidable energy loss and waste heat loss. The avoidable
energy loss is presented in Table 8.

of Test 1 is higher than Test 2. As discussed, this is caused by the

increased nickel content in alloy (in Test 2) leads to a higher mate-
rial (i.e. wet laterite ores, limestone, and anthracite) consumption,
longer production time and lower continuation degree (C00 ). Simi-
larly, in energy flow network, Test 1 shows a higher efficiency than
Test 2. Again, this is owing to the higher material consumption in Fig. 9. The influence of Ni content in the alloy on material and energy flows in the
Test 2; causing higher energy consumption, avoidable energy loss, rotary dryer. (a) Available energy and avoidable energy loss, (b) Material flow.
556 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559

Fig. 10. The influence of Ni content in the alloy on material and energy flows in the Fig. 12. The influence of Ni content in the ores on material and energy flows in the
rotary kiln. (a) Available energy and avoidable energy loss, (b) Material flow. rotary dryer. (a) Available energy and avoidable energy loss, (b) Material flow.

Fig. 11. The influence of Ni content in the alloy on material and energy flows in the Fig. 13. The influence of Ni content in the ores on material and energy flows in the
electric furnace. (a) Available energy and avoidable energy loss, (b) Material flow. rotary kiln. (a) Available energy and avoidable energy loss, (b) Material flow.
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559
Fig. 14. The influence of Ni content in the ores on material and energy flows in the
electric furnace. (a) Available energy and avoidable energy loss, (b) Material flow.
and energy discharge rate (gep). The resultant energy efficiency is
therefore reduced in Test 2. Moreover, with a lower continuation
degree (C00 ) in Test 2, energy loss due to heat dissipation will also
increase causing a lower recovery rate of waste heat (i.e. gsy and
gsy).
As shown in Tables 10 and 11, Test 1 shows a higher order
degree (ODm) than the Test 2. This could contribute to the fact that
the order parameters (em,n) of Test 1 are larger than Test 2, and the
combination form of order parameters of Test 1 is more optimized.
In other words, the larger the order parameter (em,n) of Test 1 is,
the higher order degree (ODm) of networks for Test 1 could obtain.
As a result, the overall order degree (ODm) between M-E-N for Test
1 is higher than Test 2; indicating that the coordination ability of
Test 1 is larger than Test 2. Meanwhile, Test 2 exhibits a higher
mismatching degree (DR) in compared to the Test 1. This could
be caused by the continuation degree (C00 ) of Test 2 which is lower
than the Test 1; representing more energy dissipation loss. In other
words, the system of Test 1 is more efficient and the networks of
Test 1 coordinates better as indicated by the higher matching
degree. Under these conditions, the synergy degree (SD) of Test 1
is higher than Test 2.
Through the analysis of energy flows distribution of rotary kiln-
electric furnace process, several energy conservation measures for
enhancing the energy efficiency of the process are identified as fol-
lowing: Firstly, recycling the waste heat of output flows for pre-
heating and dehydrating could improve energy efficiency and
minimize the energy loss to the surroundings. Secondly, the recov-
ery of furnace gas as fuel could contribute to reducing significant
fuel consumption. Thirdly, reducing the exit temperatures of out-
put flows could decrease the energy loss due to heat dissipation.
Fourthly, by power generation technology for waste heat recovery,
the waste heat loss of flue gas and slag could be reduced; leading to
557
Fig. 15. The energy efficiency of the RKEF process. (a) the influence of Ni content
change in the alloy on energy efficiency and discharge rate of energy flow, (b) the
influence of Ni content change in the ores on energy efficiency and discharge rate of
energy flow.
improving energy efficiency. Fifthly, by using mold flux, the waste
heat loss of ferronickel alloy could also be reduced. Sixthly,
improving the time of feeding material and discharging, busbar
loss and transformer loss could minimize energy loss. Finally,
improving the thermal insulation property of flue could minimize
the heat dissipation loss of exhaust gas. These measures could be
implemented to optimize the system efficiency while ensuring
the quality of the product. In addition, synergy and energy effi-
ciency analysis were also carried out to quantify the performance
of the RKEF process under two working conditions. From the above
analysis, improving the continuation degree and recovery rate of
waste heat contribute to reducing specific energy consumption
and raw material consumption, and enhancing the productivity
and energy efficiency. In other words, optimizing the structure of
the system can help the subsystems to be in coordinated work
and ensure RKEF process to meet the requirement of energy saving
for practical working conditions.
5. Conclusion
Based on the material and energy flows analysis methods, the
effect of operation parameters on the product quality and energy
distribution in ferronickel smelting process has been analyzed. As
discussed, the analysis shows that the energy efficiency of the pro-
cess could be significantly enhanced by optimizing the raw mate-
rial and energy consumption and recovery of residues and energy.
558 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559

Table 10
The technic index of ferronickel alloy containing 12 mass% Ni-Test 1.

Ni (12%) e1,1 e1,2 e1,3 e1,4 e1,5 e2,1 e2,2 e2,3 e2,4 e2,5
em,n 0.95 0.399 0.723 0.338 0.95 1.045 0.423 0.413 0.371 0.727
Sen
m;n
0.6857 0.9375 0.8062 0.9307 0.8930 0.9615 0.9952 0.9419 0.9258 0.9146
wm,n 0.4208 0.0836 0.2595 0.0928 0.1433 0.1474 0.0184 0.2226 0.2844 0.3273
ECm(em,n) 0.9515 0.4071 0.4100 0.3558 0.9596 0.2613 0.3800 0.3682 0.3188 0.8077
ODm(Gm) 0.7147 0.4824
SA 0.5872
DR 0.371
SD 0.3694

Table 11
The technic index of ferronickel alloy containing 14 mass% Ni-Test 2.

Ni (14%) e1,1 e1,2 e1,3 e1,4 e1,5 e2,1 e2,2 e2,3 e2,4 e2,5
em,n 0.949 0.335 0.688 0.341 0.936 1.295 0.411 0.396 0.39 0.695
Sen
m;n
0.6996 0.7637 0.6679 0.7769 0.9280 0.7867 0.8196 0.7701 0.8408 0.7430
wm,n 0.2581 0.2030 0.2853 0.1917 0.0619 0.2052 0.1735 0.2211 0.1531 0.2471
ECm(em,n) 0.9586 0.3418 0.2933 0.3589 0.9455 0.3238 0.3659 0.3482 0.3412 0.7722
ODm(Gm) 0.5278 0.4500
SA 0.4873
DR 0.39
SD 0.2973

1. With the waste heat recovery of flue gas in rotary dryer, the
consumption of bituminous coal will be reduced by about
45.94%. Recycling the furnace gas as fuel causes a reduction of
about 38.00% bituminous coal consumption in rotary kiln.
2. Through the comparison of the efficiency of each subsystem,
increasing the energy efficiency is necessary to decrease exit
temperatures of output flows and the sequential energy loss.
3. The reutilization of waste heat of output flows is helpful to opti-
mize energy distribution and to eliminate energy loss due to
heat dissipation.
4. With enhancing Ni grade in ferronickel alloy, the available
energy and avoidable energy loss are increased in each subsys-
tem. Meanwhile, using the laterite ores with higher Ni content,
the available energy and avoidable energy loss are decreased in
each subsystem.
5. Minimizing the heat dissipation loss of output flows in the
transportation process, by using mold flux and improving the
thermal insulation property of flue, contributes to improving
the energy flow discharge rate.
6. The system with a higher synergy degree can show a better per-
formance for energy efficiency and energy saving.
Acknowledgements
The authors are especially grateful to key programme of
National Natural Science Foundation of China (Grant No.
51210007).
References
[1] K. Liu, Q.Y. Chen, H.P. Hu, Z.L. Yin, B.K. Wu, Pressure acid leaching of a Chinese
laterite ore containing mainly maghemite and magnetite, Hydrometallurgy
104 (2010) 32–38.
[2] X.Y. Guo, W.T. Shi, D. Li, Q.H. Tian, Leaching behavior of metals from limonitic
laterite ore by high pressure acid leaching, Trans. Nonferr. Metal. Soc. 21
(2011) 191–195.
[3] R.R. Moskalyk, A.M. Alfantazi, Nickel laterite processing and electrowinning
practice, Miner. Eng. 15 (2002) 593–605.
[4] G. Senanayake, A. Senaputra, M.J. Nicol, Effect of thiosulfate, sulfide, copper(II),
cobalt(II)/(III) and iron oxides on the ammoniacal carbonate leaching of nickel
and ferronickel in the Caron process, Hydrometallurgy 105 (2010) 60–68.
[5] A.N. Nikoloski, M.J. Nicol, The electrochemistry of the leaching reactions in the
Caron process II. Cathodic processes, Hydrometallurgy 105 (2010) 54–59.
[6] R.G. McDonald, B.I. Whittington, Atmospheric acid leaching of nickel laterites
review. Part I. Sulphuric acid technologies, Hydrometallurgy 91 (2008) 35–
55.
[7] R.G. McDonald, B.I. Whittington, Atmospheric acid leaching of nickel laterites
review. Part II. Chloride and bio-technologies, Hydrometallurgy 91 (2008) 56–
69.
[8] J.C. Nayak, Production of ferronickel from sukinda laterites in rotary kiln-
electric furnace, Trans. Indian Inst. Met. 38 (1985) 241–247.
[9] C. Walker, Kashani-Nejad, A.D. Dalvi, N. Voermann, I. Candy, B. Wasmund,
Future of rotary kiln- electric furnace (RKEF) processing of nickel laterites, in:
Proceedings of the European Metallurgical Conference, EMC, 2009.
[10] B. Lee, I. Sohn, Review of innovative energy savings technology for the electric
furnace, JOM 66 (2014) 1581–1594.
[11] C. Belt, Challenges in melt furnace tests, JOM 66 (2014) 1612–1615.
[12] H.Y. Sohn, M. Olivas-Martinez, Methods for calculating energy requirements
for processes in which a reactant is also a fuel: need for standardization, JOM
66 (2014) 1557–1564.
[13] A. Sagastume, J.B. Cogollos, C. Vandecasteele, Energy and exergy assessments
of a lime shaft kiln, Appl. Therm. Eng. 51 (2013) 273–280.
[14] L. Zhang, Q. Zhu, Multiple attribute network selection algorithm based on AHP
and synergetic theory for heterogeneous wireless networks, J. Electron. (China)
31 (2014) 29–40 (in Chinese).
[15] C.A. Pickles, Drying kinetics of nickeliferous limonitic laterite ores, Miner. Eng.
16 (2003) 1327–1338.
[16] I.J. Kotzé, Pilot plant production of ferronickel from nickel oxide ores and dusts
in a DC arc furnace, Miner. Eng. 15 (2002) 1017–1022.
[17] Y.G. Guo, R. Zhu, M. Lv, M.W. Guo, Y.W. Wang, C.F. Zhou, Extraction of a nickel-
iron alloy from nickel laterite ore through selective reduction and smelting
process, J. Univ. Sci. Technol. Beijing 36 (2014) 584–591 (in Chinese).
[18] H. Wang, B. Li, The Drying and Pre-reduction Technology of Nickel Laterite
Ores, Science Press, Beijing, 2012, pp. 26–71.
[19] X.H. Luan, L. Tang, X.M. Li, S.B. Hou, The Metallurgy Technology and Equipment
on Ferronickel, Metallurgical industry press, Beijing, 2010, pp. 6–73.
[20] B. Li, Y.G. Wei, H. Wang, Action mechanism and phase transformation
characteristics of garnierite in drying process, Chin. J. Nonferrous Met. 23
(2013) 1440–1446 (in Chinese).
[21] M.J. Rao, G.H. Li, T. Jiang, J. Luo, Y.B. Zhang, X.H. Fan, Carbothermic reduction of
nickeliferous laterite ores for nickel pig iron production in China: a review,
JOM 65 (2013) 1573–1583.
[22] H.B. Lu, Thermodynamic research on production of ferronickel alloy by electric
furnace reduction from lateritic nickel ore, Chin. J. Rare Met. 36 (2012) 785–
790 (in Chinese).
[23] H.B. Lu, Research progress on the production of ferronickel alloy by electric
furnace reduction from lateritic nickel ore, Min. Metall. 21 (2012) 60–64 (in
Chinese).
[24] M. Liu, X.W. Lv, E.G. Guo, P. Chen, Q.G. Yuan, Novel process of ferronickel
nugget production from nickel laterite by semi-molten state reduction, ISIJ Int.
54 (2014) 1749–1754.
[25] Z.H. Liu, X.B. Ma, D.Q. Zhu, Y.H. Li, Q.H. Li, Preparation of ferronickel from
laterite ore in reduction smelting process, J. Cent. South Univ. Sci. Technol. 42
(2011) 2905–2910 (in Chinese).
[26] X.X. Fan, H.G. Dong, Y.H. Wang, Y.D. Wu, B.Y. Li, B.J. Li, J.C. Zhao, L.H. Zan,
Production of Ni-Fe alloy from nickeliferous laterite ore by prereduction-
 P. Liu et al. / Applied Thermal Engineering 109 (2016) 542–559 559

electric furnace smelting separation, J. Cent. South Univ. Sci. Technol. 43 [31] R.Y. Yin, Metallurgical Process Engineering, second ed., Metallurgical Industry
(2012) 3344–3348 (in Chinese). Press, Beijing, 2009, pp. 210–223.
[27] Z. Mei, Nonferrous Metallurgy Furnace Design Manual, Metallurgical Industry [32] D.F. He, A.J. Xu, N.Y. Tian, Analysis on degree of continuation of thin slab
Press, Beijing, 2007, pp. 11–151. casting and rolling, Steelmaking 25 (2009) 45–47 (in Chinese).
[28] B.Y. Zheng, Measurement and calculation of busbarloss of EAF for steelmaking, [33] Y. Long, Study on the Large-scale Systems based on the Synergy Among
Energy Metall. Ind. 15 (1996) 53–57 (in Chinese). Material Flow, Energy Flow and Information Flow PhD Thesis, Huazhang
[29] Z.M. Yi, H. Xiao, J.L. Song, G.B. Ma, J.M. Zhou, Mathematic simulation of heat University of Science and Technology, WuHan, China, 2009 (in Chinese).
transfer and operating optimization in alumina rotary kiln, J. Cent. South Univ. [34] J. Li, Y.A. Zhang, Study on logistics network management based on order
20 (2013) 2775–2780. parameter, Logis. Technol. 31 (2012) 133–136 (in Chinese).
[30] A. Sagastume, C. Vandecasteele, Exergy-based indicators to evaluate the [35] W.Q. Sun, Fundamental Research on Synergy of Steel Manufacturing Process
possibilities to reduce fuel consumption in lime production, Energy 36 (2011) and Energy Flow Network PhD Thesis, Northeastern University, ShenYang,
2820–2827. China, 2013 (in Chinese).