You are on page 1of 10

Fuel Processing Technology 141 (2016) 138147

Contents lists available at ScienceDirect

Fuel Processing Technology

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

Research article

Analysis of optimization potential in commercial biomass gasication


plants using process simulation
V. Wilk a,, H. Hofbauer b
a
Bioenergy2020+ GmbH, Wienerstrae 49, 7540 Gssing, Austria
b
Institute of Chemical Engineering, Vienna University of Technology, Getreidemarkt 9/166, 1060 Vienna, Austria

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

Article history: The performance of the dual uidized bed gasication plant Oberwart was evaluated by means of an extensive
Received 28 February 2015 measurement campaign and calculation of mass and energy balances using IPSEpro. Process simulation was
Received in revised form 25 July 2015 also applied to identify future optimization potentials. Different aspects are discussed such as the reduction of
Accepted 31 July 2015
gasication temperature and the reduction of steam for gasication and air for combustion. Gasication pilot
Available online 8 August 2015
plant experience is integrated in the simulation models to increase signicance of the simulation results. The
Keywords:
mass and energy balances conrm that the performance of the CHP plant Oberwart is highly satisfactory and cur-
IPSEpro rently achieves an electrical efciency of 29%. The variations of plant parameters provide deeper insight in the
Dual uidized bed process itself and show interdependencies of different process units. With lower gasication temperatures and
Steam gasication reduction in combustion air, the electrical efciency can be increased to 31%.
Parameter variation 2015 Elsevier B.V. All rights reserved.

Symbols (continued)
Symbols (continued)

Symbol Unit

Steam-to-fuel ratio kg/kg


Throughput kg/(hm2)
Symbol Unit Steam conversion kg/kg


Q Heat ow kW
A Area m2
All Overall steam Steam for uidization
Ash stoich Stoichiometric
Bed Bed material T Temperature C
C Carbon tar Tar
C Concentration Vol.-% th Thermal
comb Combustion reactor w Mass fraction kg/kg
el Electrical waf Water and ash free
Fuel WHSV Weight hourly space velocity 1/h
gasif Gasication reactor x Load g/Nm3
Init Initial state
Excess air ratio kg/kg or m3/m3
Mass ow kg/s
meas Measured 1. Introduction
Efciency
P Power kW Solid biomass is going to play an increasing role in the energy supply
pc Post combustion chamber
of the future, as it is an important renewable source of energy and the
only source of renewable carbon. The efcient use of biomass is current-
ly receiving increasing attention. Gasication as an innovative and vi-
Abbreviations: CHP, combined heat and power; DFB, dual uidized bed; GCMS, gas
chromatograph with mass spectrometer; LCV, lower caloric value; Nm3, volume at able technology for the thermal conversion of biomass has been the
standard conditions of 101325 Pa and 273.15 K; ORC, organic Rankine cycle; RME, rape focus of considerable research for a number of decades. For gasication
seed methyl ester (biodiesel); vol.-%, volume percent (Nm3/Nm3 100%); waf, water processes, uidized bed reactors are applied by preference. The good
and ash free; wt.-%, weight percent (kg/kg 100%). mixing conditions of fuel particles, bed material, and gas phase and an
Corresponding author at: AIT Austrian Institute of Technology, Gienggasse 2, 1210
Vienna, Austria.
excellent heat transfer promote the conversion of the feedstock. De-
E-mail addresses: veronika.wilk@ait.ac.at (V. Wilk), hermann.hofbauer@tuwien.ac.at tailed reviews of different types of gasication reactors and their appli-
(H. Hofbauer). cations are available in [16]. Conventional air gasication yields a

http://dx.doi.org/10.1016/j.fuproc.2015.07.035
0378-3820/ 2015 Elsevier B.V. All rights reserved.
V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147 139

product gas which is highly diluted with nitrogen and as a consequence to changes in these process parameters, however, the range of variation
has a lower caloric value (LCV) of 46 MJ/m3. In contrast, steam gasi- is sometimes limited due to the existing plant conguration in
cation allows the generation of a nitrogen free product gas without the Oberwart. It provides deeper insight in the process itself and show in-
use of pure oxygen as a gasication agent. The LCV is in the range of terdependencies of different process units.
1214 MJ/m3. Due to absence of nitrogen, product gas from steam gas-
ication is suitable for chemical synthesis processes. Comprehensive re-
2. Materials and methods
views on uidized bed gasication with steam are provided by several
authors, [79].
2.1. Dual uidized bed gasication
At the Vienna University of Technology, the dual uidized bed (DFB)
gasication technology was developed in the 1990s in order to establish
The basic principle of DFB gasication is illustrated in Fig. 1. It is a
an efcient conversion technology to produce electricity, heat, and fuels
steam gasication process, which is carried out in two reactors: an
from solid biomass. The DFB gasication process has been successfully
allothermal gasication reactor uidized with steam and a combustion
demonstrated in Gssing (Austria), where a DFB steam gasier has
reactor uidized with air that provides heat for gasication. The reactors
been in operation at the scale of an 8 MWth demonstration plant
are connected thermally by circulating bed material. In the gasication
since late 2001. The second DFB steam gasication went into operation
reactor, a bubbling uidized bed is formed, where biomass is
in 2008 in Oberwart (Austria) and represents an important step of de-
devolatilized and partly gasied. The products of devolatilization and
velopment in DFB gasication. Today this technology is commercially
gasication react with steam to form the product gas. Ungasied bio-
available and several industrial plants ranging from 8 to 33 MW are in
mass char is transported with the bed material to the combustion reac-
operation in Europe by now.
tor, where the char is combusted with air. The combustion reactor is a
Process simulation plays an important role for the development of
fast uidized bed reactor, where the bed material is heated and
new technologies and allows for assistance in process design, plant
transported to the top of the reactor. It is separated from the ue gas
operation and optimization. IPSEpro is an equation oriented steady
stream in a cyclone and is returned to the gasication reactor. Some
state simulation tool that has been initially developed for thermal
product gas is recycled back to the combustion reactor (fuel to combus-
power cycle calculations [10]. It has been successfully used already in
tion) to fulll the energy demands of gasication and to obtain the
the phase of basic engineering of the Gssing gasier, because it can
desired temperature in the gasication reactor (e.g., 850 C). In the
be adapted as well as extended to new processes with user built models
DFB gasier, two separated gas streams are yielded, the product gas
[11]. Therefore, this simulation tool has been frequently used for design
and conventional ue gas. Due to steam gasication, the product gas
considerations and technology evaluation for several DFB gasication
has a moderate caloric value of 1214 MJ/Nm3 and is rich in hydrogen
processes. The potential application of DFB gasication plants in pulp
(N40%).
and paper processes for example is described in [12] and for waste
wood gasication in [13]. Based on IPSEpro simulations, different pro-
cess pathways for the production of renewable hydrogen from biomass 2.2. Combined heat and power plant Oberwart (CHP)
based on DFB gasication are compared in [14]. Synthesis of natural gas
from biomass is evaluated with IPSEpro based on a realized process in The CHP plant in Oberwart, Austria, is in operation since 2008 and is
[15] and based on a theoretical process that is planned to be realized an excellent example for DFB gasication at industrial scale. The plant is
in [16]. Extensive IPSEpro modeling that also considers gassolid reac- designed for a nominal fuel load of 8.5 MW of woody biomass and pro-
tions was used for the design of a chemical looping test facility [17] vides electricity and district heat. It is owned and operated by Energie
and for the design of a 10 MW chemical looping plant [18]. Burgenland GmbH, a local energy utility. In Fig. 2 the process ow dia-
Process simulation with IPSEpro is also successfully applied for gas- gram of the CHP plant is illustrated. The plant is operated on wood
ication plants in operation. In this case data from plant operation is chips that are fed to a dryer prior to gasication. In the gasication reac-
then used for balance of plant (BOP). Prll et al. analyzed the rst tor, wood chips are gasied with steam and product gas is formed that is
years of operation of the DFB gasier in Gssing with IPSEpro. By cooled in several heat exchangers. In a bag house lter, entrained parti-
means of extensive parameter variations in IPSEpro, they evaluated op- cles, char and tar are removed from the product gas stream. In the
timization potentials of the gasication process [19]. Parameter varia- product gas scrubber tar is removed by RME (biodiesel), an organic
tions on the large scale plant are normally restricted as the operator of scrubbing liquid. As the scrubber is operated at temperatures of about
the plant aims at maintaining a high availability and a high power out- 40 C, water is condensed from the product gas, which is reused for u-
put. However, process simulation reduces the risk for the operator of the idization. RME saturated with tar is combusted in the combustion reac-
gasication plant and shows the optimization potential. tor. After the product gas blower, the gas is used in two gas engines that
Similarly, IPSEpro was used at the DFB gasication plant Oberwart, produce electricity and district heat.
the second DFB gasication plant at industrial scale. The commissioning
phase and the rst years of operation of the DFB gasication plant in
Oberwart were also balanced using IPSEpro [20].
This paper presents a scientic evaluation of the current perfor-
mance of the DFB gasication plant Oberwart based on IPSEpro simula-
tion. For this purpose, an extensive measurement campaign was carried
out in 2013, which included online and discontinuous measurements of
biomass, product gas and ue gas quality. These measurements and
data retrieved from the process control system were used to calculate
mass and energy balances of the entire process. Potentials for future
optimization are identied by process parameter variations. Different
aspects are discussed in this paper: i) reduction of gasication temper-
ature, ii) reduction of steam for gasication and iii) air for combustion.
In order to simulate the inuence of these process parameters more
precisely, new model units are developed based on extensive experi-
ence gained from DFB pilot plant operation at Vienna University of
Technology. The simulation shows how the plant conguration reacts Fig. 1. Basic principle of the DFB gasication technology.
140 V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147

Fig. 2. Process ow diagram of the CHP plant Oberwart.

Flue gas from the combustion reactor is fed to the post combustion electricity. The condenser of the ORC process is cooled with water. The
chamber to ensure complete combustion. The gas is cooled in a series heat of condensation is reused to heat air for the biomass dryer.
of heat exchangers. After the ue gas lter that removes entrained par-
ticles, the ue gas is released in the stack. The heat exchangers in the 2.3. Process simulation
product gas and the ue gas stream provide heat for air preheating
and steam generation. Heat is also transferred to a thermal oil system For detailed process analysis, the simulation software IPSEpro was
connected to an ORC process, which converts low temperature heat applied. It uses an equation-oriented solver for the calculation of mass
into electricity. District heat is also provided by the CHP plant Oberwart. and energy balances of stationary processes. The process, which was im-
For peak load demands of district heat there is also a natural gas boiler. plemented in IPSEpro, corresponds to the actual process setup in Fig. 2.
This is the only unit in the CHP plant where natural gas is used. The units of the process (reactors, etc.) are connected by streams trans-
Compared to the rst industrial DFB gasication plant in Gssing, ferring mass and energy. At Vienna University of Technology a compre-
several improvements of the plant concept have been realized at the hensive model library for gasication plants has been developed. This
CHP plant Oberwart. Due to integration of a biomass dryer, inconve- set of models comprises all necessary process units of a gasication
nient uctuations in biomass moisture are eliminated and a constant plant, such as gasication and combustion reactor, lter, scrubber, and
water content of about 15% is ensured. Geometrical changes in the gas- heat exchanger. The process units are balanced according to conserva-
ication and combustion reactor, as well as in the uidization nozzles tion of mass and energy and can also contain related functional equa-
were realized to increase the mean residence time of the hot gases. tions, such as kinetic properties or geometric correlations. Mass and
Thereby the concentration of tar in the product gas and char burnout energy balances are strictly fullled for all process units. Detailed infor-
is improved. The control mechanism of the gasication temperature mation on the model library is available in [23]. Due to the model library
has been improved in order to increase the efciency of the gasication for gasication plants, IPSEpro is a well suited tool for process evalua-
process. Also the design of the heat exchangers in the product gas and tion and optimization of industrial gasiers.
ue gas was modied to lower fouling and contamination. Thus, contin- In order to determine the status of the CHP plant Oberwart, mass and
uous operation of the plant was prolonged considerably. Further infor- energy balances of stationary operation were calculated with IPSEpro.
mation on the improvement measures is available in [20] and [21]. IPSEpro allows the use of measurement data containing uncertainty.
The integration of an Organic Rankine Cycle (ORC) was an important An over-determinated equation system was formed, which contained
measure to increase the electrical output. In the CHP plant, product gas the measurement data and a range of tolerance (e.g., the standard devi-
and ue gas contain considerable amounts of sensible heat, as they ation of the measured value). The equation system was solved by the
leave the reactors at ca. 850 C and 930 C respectively. The heat is trans- Method of Least Squares. Thereby the accuracy of the measurement
ferred to two heat carrier oil cycles and is used to preheat air, to gener- equipment also can be controlled. The reconciled solution best describes
ate uidization steam and to provide district heat. The remaining heat is the actual operation of the pilot plant within the limits of the model.
valorized in an ORC process. The ORC process is similar to the conven- More information on this procedure is provided in [19].
tional steam turbine process, but uses an organic working uid instead
of water. Due to its thermodynamic properties, it can be evaporated at 3. Results and discussion
lower temperatures. The technological maturity of ORC processes are
proven in the medium power range of 2002000 kW of electricity pro- 3.1. Status of the CHP plant in 2013
duction [22]. The ORC process at the CHP plant Oberwart is connected to
the two heat carrier oil cycles that are operated at different temperature An extensive measurement campaign was carried out at the CHP
levels. Heat carrier oil with lower temperature (ca. 240 C) is used to plant Oberwart in November 2013. The composition of the wood
preheat the organic working uid of the ORC process. Heat carrier oil chips that are used as fuel at the CHP plant is provided in Table 1. Before
with temperatures of about 300 C provides heat for evaporation of drying the water content of the wood amounted to 32%, after drying the
the organic working uid. It is then used in a steam turbine to produce water content was reduced to 16%.
V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147 141

Table 1 Table 3
Fuel composition. Product gas composition.

C wt.-%, dry 49.35 H2 Vol.-%, dry 38.0


H wt.-%, dry 6.03 CO Vol.-%, dry 25.1
N wt.-%, dry 0.17 CO2 Vol.-%, dry 21.3
S wt.-%, dry 0.013 CH4 Vol.-%, dry 9.7
Cl wt.-%, dry b0.005 C2H4 Vol.-%, dry 2.2
LCV kJ/kg, dry 18,299 C2H6 Vol.-%, dry 0.2
Ash wt.-%, dry 0.98 N2 Vol.-%, dry 1.9
Volatiles wt.-%, dry 82.66 O2 Vol.-%, dry b0.1
LCV MJ/Nm3, dry 14

During the measurement campaign, two different types of operation


were performed: operation with and without ORC process. If the ORC Combustion is characterized by the excess air ratio, Eq. (5), which
process is running, the electricity output of the CHP plant is maximized amounted to about 1.5. Hence, the reaction conditions were very similar
and less district heat is available. Without the ORC process, more district for both types of operation.
heat can be provided at the expense of electricity production. The data  

basis for the mass and energy balances are data from the plant control mH2O;steam mH2O;fuel
 1
system and continuous measurements of the product gas and ue gas

mfuel 1wH2O;fuel wash;fuel
composition. Values are averaged from stationary operation of 916 h.

Table 2 provides an overview on characteristic process parameters for mfuel
the two types of operation. In the gasication reactor, the average tem- WHSV 2
mbed;gasif
perature in the uidized bed was about 850 C. In the combustion reac-
tor the temperature was ca. 80 C higher. An important parameter of 
mfuel
gasication is the steam-to-fuel ratio that relates the amount of steam 3
Agasif
used for uidization and the fuel water content to the amount of fuel
(dry and ash free) that is gasied, Eq. (1). The steam-to-fuel ratio was in   
mH2O;steam mH2O;fuel mH2O;product gas
the range of 0.50.6 kg/kg during both types of operation. During opera- water   4
tion without ORC, more biomass was gasied, therefore, more uidization mfuel 1wH2O;fuel
steam was necessary to keep the steam-to-fuel ratio relatively constant. 

The specic gas production describes how much dry product gas is pro- mO2;meas
 5
duced from the mass ow of dry fuel. It amounts to 0.9 Nm3/kg in both mO2;stoich
cases. The weight hourly space velocity (Eq. (2)) denotes the ratio of
the mass ow rate of biomass after dryer and the mass of bed material The measured product gas composition is shown in Table 3. Typical-
in the gasication reactor. At the CHP plant Oberwart, olivine is used as ly for DFB steam gasication, the product gas contained a high amount
bed material. It is a natural mineral that exhibits moderate tar cracking ac- of hydrogen (38%). The other main compounds were CO, CO2 and CH4.
tivity and has good mechanical stability [24,25]. The mass of olivine in the The lower caloric value (LCV, dry gas) amounted to 14 MJ/Nm3.
gasication reactor can be approximated based on the bed pressure and The tar, dust and char content were analyzed in the raw product gas
amounts to ca. 3000 kg. The throughput relates the biomass mass after the heat exchangers and in the clean gas that leaves the scrubber
ow to the cross section A of the gasication reactor (Eq. (3)). The (Table 4). Two different methods of tar analysis were applied: gravimet-
WHSV and the throughput remained constant for both types of operation. ric tar and GCMS tar measurement. Gravimetric tar is weighed after vac-
In the gasication reactor, steam is used for uidization. It reacts with bio- uum evaporation of the solvent: it comprises mostly tar species with a
mass and converts char into gaseous compounds. The amount of steam high molecular weight. A GCMS device (gas chromatography with a
which is consumed in the reactions is described by water (Eq. (4)). This mass spectrometer) was used to measure the content of 50 different
value is referred to the dry biomass ow to allow for easier comparison. tar species of medium molecular weight in the product gas. The mea-
It also remains constant for both types of operation. Unreacted steam re- surement range of the two techniques overlaps and therefore the two
mains in the product gas, which has a water content of 3540%. It is con- values cannot be summarized and both values are given. A detailed de-
densed in the tar scrubber and reused for uidization. scription of the tar measurement is available in [26].
The measured tar concentrations are in good agreement with mea-
Table 2 surements carried out at the DFB plant Gssing that are provided e.g.,
Overview on process parameters. in [27]. In the product gas lter and scrubber the tar concentration is
Operation with Operation without reduced considerably, which allows for unimpeded operation of the
ORC process ORC process gas engines.
Gasication temperature C 858 846 Based on the mass and energy balances, energy ow diagrams are
Combustion temperature C 945 930 set up that illustrate the in- and outgoing energy streams of the CHP
Steam uidization kg/h 450 690 plant. Fig. 3 shows operation with ORC process. A total of 8.2 MW of bio-
Steam-to-fuel ratio (waf) kg/kg 0.5 0.6 mass prior to drying is converted into 2.4 MW of electricity and 1.6 MW
Specic gas production Nm3/kg 0.9 0.9
(dry gas, dry fuel)
Weight hourly space velocity 1/h 0.73 0.71
Gasier throughput kg/(hm2) 528 512 Table 4
Steam conversion kg/kg 0.06 0.07 Dust, char and tar concentration in raw and clean product gas.
Air uidization kg/h 6632 6919
Raw product gas Clean product gas
Excess air ratio 1.5 1.5
PG temperature before lter C 144 130 Dust 3
g/Nm , dry 26.0 n.m.
PG temperature after scrubber C 36 37 Char g/Nm3, dry 32.7 n.m.
Flue gas temperature before lter C 162 160 Gravimetric tar g/Nm3, dry 1.8 0.1
Gas engines output (2 engines) kW 2155 2202 GCMS tar g/Nm3, dry 7.6 2.7
ORC power output kW 260 0
n.m. = not measured.
142 V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147

Fig. 3. Energy ow diagram of the CHP plant Oberwart (operation with ORC), all streams in kW.

Fig. 4. Energy ow diagram of the CHP plant Oberwart (operation without ORC), all streams in kW.


of district heat. This corresponds to 70% of the nominal fuel load of the Q th
CHP plant. Furthermore, 0.2 MW of RME was used in the scrubber and th 7
P fuel
1.8 MW of natural gas was red in the district heat peak load boiler to
increase the amount of district heat. Hot exhaust gases leave the


plant: the exhaust gas of the engines, of the peak load boiler and of P el Q th
all 8
the ue gas from the combustion reactor. These gas streams still P fuel
contained 1.6 MW of sensible heat. This heat ow would be available
as additional district heat. However, this heat ow cannot be used yet
because of missing heat exchangers in the exhaust gas stream of the The mass and energy balances show that the DFB gasier converts
gas engines and also because of the conguration of the existing heat biomass into electricity and heat with high efciency compared to
exchangers. Another 1.6 MW was lost to the environment in coolers. Ir- other technologies available for similar sizes of plants. If the focus of op-
reversible losses amounted to 0.8 MW. eration is on maximizing the electricity output (with ORC process), the
In Fig. 4, plant operation without ORC process is balanced. During that electrical efciency amounts to 29%. If more district heat is produced
day, more biomass was gasied (8.4 MW). Less electricity and more dis- (without ORC process), the thermal efciency increases. The low ther-
trict heat were available. The peak load boiler was not in operation, there- mal and therefore also total efciencies at the Oberwart plant suffer
fore no natural gas was consumed. Moreover, the heat ow that is lost to from a heat extraction system which is not complete and not optimized.
the environment in the coolers is signicantly smaller. This is because of The overall efciency remains constant, only the share of the different
the ORC process. The ORC process converts low temperature heat at products changes. Thus, the DFB gasier is a exible process that can
290 C into electricity. The average efciency of the ORC process is 16% be adapted to different needs. During the year 2013, the availability of
which is the usual performance of typical ORC processes. Part of the the CHP plant was about 82%, which conrms that the DFB process is ro-
heat ow, which is released in the condenser, is recovered for preheating bust and reliable.
of other process streams, but the major part is lost in coolers. In terms of
thermal efciency, operation without ORC is more favorable in case of the
Oberwart CHP plant. The sensible heat of the exhaust gases and the other Table 5
losses are in good agreement with the operation with ORC process. Overview on plant efciencies.
The efciency of the two types of operation is calculated according to
(Referred to wet Operation with ORC Operation without ORC Eq.
the following equations, Eqs. (6)-(8), and is compiled in Table 5. biomass) process process

Electrical efciency % 29 26 6
P el Thermal efciency % 19 22 7
el 6
P fuel Overall efciency % 49 49 8
V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147 143

Table 6
Parameter variations.

Parameter Current operation value (init) Variation range Comment

Gasication temperature C 858 835860 Temperatures b835 C limited because of plant control limitations
Excess air ratio in combustion reactor 1.45 1.051.45 Excess air ratios b1.15 not recommended for solid fuels
Steam-to-fuel ratio in gasication reactor 0.52 0.420.52

3.2. Variation of process parameters the temperature dependency of steam gasication of solid carbon
(char), Eq. (6).
Once the mass and energy balances are solved based on plant
data, all process variables are known, also those that cannot be mea- C H2 OH2 CO 9
sured directly such as the solid circulation rate or the amount of char
that is combusted in the combustion reactor. The simulation also Char gasication is less active at lower temperatures, therefore less
shows the interdependencies of the plant conguration and how H2 and CO are formed and less steam is consumed in the reactions. As
changes in certain process parameters affect the whole plant. In a consequence more char is available for combustion at the expense of
order to investigate the inuence of the gasication temperature, the product gas yield. It is observed that more tar is formed at lower gas-
the amount of steam and air uidization in more detail, new model ication temperatures, which is also conrmed in several other studies
units are developed and integrated into the simulation. The new [24,30,31]. Linear correlations can be assumed with good approxima-
models are based on experiments with the DFB pilot plant that is tion for the temperature range from 750870 C. The correlations are
operated for scientic purposes at Vienna University of Technology. compiled in Table 7. For the reference case the index init is used. The
Experimental results gained from the pilot plant are transferable to mass and energy balance have to be known at the gasication temper-
the industrial gasiers with good accuracy [28]. The simulation of ature Tgasif,init of the reference case. All other parameters are calculated
the CHP operation with ORC process was used as a basis for the pa- as a function of Tgasif, the new gasication temperature.
rameter variations. For the variation of gasication temperature, it was assumed that the
Table 6 shows the variation range of the parameters. The parameters power output of the gas engines remained constant. The steam-to-fuel
are selected with respect to the CHP plant in Oberwart. The selected pa- ratio and the excess air ratio were also kept constant, so that the reac-
rameters can be changed at the plant in certain ranges. To improve the tion conditions were only inuenced by temperature. The plant is
performance of the plant the current values of all three parameters have operated at ambient pressure. In the DFB gasier, the gasication
to be reduced during simulation. However, the reduction was limited temperature is controlled by the amount of clean product gas that is
because of the existing plant conguration which also was considered recycled to the combustion reactor.
in the simulation. When the CHP plant Oberwart was operated with ORC process, the
gasication temperature amounted to 858 C on average. In the simula-
3.2.1. Gasication temperature tion, the gasication temperature could only be lowered to 835 C,
To get the necessary data for the simulation program, variation of which is illustrated in Fig. 6. At the reference case of 858 C, about 10%
gasication temperature was studied rst in the DFB pilot plant at Vien- of clean product gas was used for temperature control in the combus-
na University of Technology [29]. If the temperature in the gasication tion reactor. This amount decreased considerably with lower tempera-
reactor is lowered, the temperature in the combustion reactor, the tem- tures and would become negative below 835 C. It can be seen that
perature of the circulation bed material and of the product and ue gas the gasication temperature where no product gas has to be recycled
stream decrease accordingly. As the equilibrium of gasication reactions is approximately 830 C, but for control reasons the gasication temper-
of solids and the gas phase reactions depend on temperature, the prod- ature can only be lowered to 835 C in this case. Fig. 6 also illustrates the
uct gas composition is altered as illustrated in Fig. 5. At lower tempera- effect of temperature reduction on the most important energy streams
tures the product gas contains less H2 and CO. This can be explained by and on plant efciency. As less product gas was required for tempera-
ture control in the combustion reactor, the amount of biomass that
was gasied was lower although electricity production of the gas en-
gines was kept constant. At a gasication temperature of 835 C, bio-
mass consumption was reduced by 5%. Biomass costs are an important
part of the operational costs, therefore a reduction of biomass consump-
tion at constant engine load is economically favorable. More district
heat was available, but the electricity output of the ORC process slightly
declined. This interdependency occurred because of the heat exchanger
arrangement of the CHP plant. Furthermore, less heat was lost to the en-
vironment in the coolers. As more district heat was available, the natural
gas demand of the peak load boiler decreased, which also has a positive
effect on operational costs.

3.2.2. Excess air ratio in the combustion reactor


The combustion reactor is uidized by several air streams (bottom
air, primary air, and secondary air) that create the different states of u-
idization in the reactor and provide oxygen for combustion. Bottom air
is fed at the lowest part of the combustion reactor, there a bubbling u-
idized bed is created. Primary and secondary air create a fast uidized
bed that transports the bed material to the top of the combustion reac-
Fig. 5. Inuence of gasication temperature on product gas composition based on experi- tor. The bed material circulation is controlled by the right ratios of
ments in DFB pilot plant, [29]. amounts of bottom air, primary and secondary air. If the amount of
144 V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147

Table 7
Correlations for gasication temperature based on [29].

Parameter Unit Correlation Eq.

Flue gas Tcomb C Tcomb = (Tgasif Tgasif,init) 0, 955 + Tcomb,init (10)


Bed material in combustion reactor Tbed C Tbed = (Tgasif Tgasif,init) 1, 079 + Tbed,init (11)
Flue gas after post combustion Tpc C Tpc = (Tgasif Tgasif,init) 0, 955 + Tpc,init (12)
C2H4 in product gas Vol.-%, dry cC 2 H4 T gasif T gasif ;init  0; 0025 cC 2 H4 ;init (13)
C2H6 in product gas Vol.-%, dry cC 2 H6 T gasif T gasif ;init  0; 004 cC 2 H6 ;init (14)
CH4 in product gas Vol.-%, dry cCH4 T gasif T gasif ;init  0; 023 cCH4 ;init (15)
CO in product gas Vol.-%, dry cCO = (Tgasif Tgasif,init) 0, 028 + cCO,init (16)
CO2 in product gas Vol.-%, dry cCO2 T gasif T gasif ;init  0; 037 cCO2 ;init (17)
H2 in product gas Vol.-%, dry cH2 T gasif T gasif ;init  0; 040 cH2 ;init (18)
Ungasied carbon wC kg C/kg fuel waf wC = (Tgasif Tgasif,init) 0, 1976 + wC,init (19)
Tar in product gas xtar g/Nm3 xtar = (Tgasif Tgasif,init) (0, 0142) + xtar,init (20)

combustion air is reduced, the excess air ratio decreases, Eq. (5), and too low, therefore a suitable operation point could be an excess air
higher temperatures are reached during combustion. Therefore, an in- ratio of 1.15. Then, still 3% of oxygen was present in the ue gas after cy-
crease in efciency can be expected. It was assumed again that the gas clone. Due to the reduced cooling effect of the combustion air, less prod-
engines are operated constantly at full load. The distribution of bottom uct gas had to be recycled to keep the temperature in the gasication
air, primary air and secondary air were kept constant. To be able to ob- reactor at 850 C. Therefore, biomass consumption is reduced by 6%
serve the inuence of the excess air ratio only, it was assumed that the and also more district heat was available. The power output of the
temperatures in gasication and combustion reactor remain constant. ORC process decreased, which is based on the heat exchanger arrange-
In practice, the bed material circulation is likely to decrease, if less uid- ment. The electrical efciency of the CHP plant increased to 30% and the
ization air is available. overall efciency to 53% at an excess air ratio of 1.15. This had also a pos-
As it is illustrated in Fig. 7, the oxygen concentration in the ue gas itive effect on the losses to the environment in the coolers and the nat-
decreased if the excess air ratio was lowered from 1.45 to smaller ural gas consumption of the peak load boiler.
values. Oxygen concentration in the ue gas is assessed at two locations:
at the exit of the cyclone and in the stack. After the cyclone, there is the
post combustion chamber, where additional air is supplied. In addition 3.2.3. Steam-to-fuel ratio in the gasication reactor
to that, ue gas is recycled from the stack to the post combustion cham- The steam-to-fuel ratio relates the mass ows of the uidization
ber to control the combustion temperature. That is why the oxygen con- steam and the fuel water to the mass ow of dry and ash free fuel as de-
centration in the stack is higher than after the cyclone. To ensure ned in Eq. (1). At constant fuel mass ow and water content, the
complete combustion of solid fuels an excess air ratio of 1.05 would be steam-to-fuel ratio only depends on the amount of uidization steam.

Fig. 6. Inuence of gasication temperature on the CHP plant Oberwart.


V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147 145

Fig. 7. Inuence of excess air ratio on the CHP plant Oberwart.

Experiments in the DFB plant are the basis to simulate the inuence The correlations were applied to the simulation of the CHP plant
of uidization steam on gasication more accurately [32]. Fig. 8 shows Oberwart with ORC operation. The initial steam-to-fuel ratio amounted
how the product gas composition is effected by the amount of steam to 0.52 kg/kg. The experiments in the pilot plant only cover the range of
that is used for gasication. The watergas shift reaction, Eq. (21), has 0.51.0 kg/kg. For the simulation linear extrapolation was applied. The
an important inuence on the product gas composition. If less steam boundary conditions for the variation were constant power output of
is available, less H2 and CO2 are formed. The effect of the steam-to-fuel the gas engines, constant gasication temperature and constant condi-
ratio can be approximated with good accuracy by linear correlations tions in the combustion reactor. As the bed material circulation is pre-
that are compiled in Table 8. Starting from the conditions of the refer- dominantly inuenced by the uidization of the combustion reactor, it
ence case at init, the product gas composition, tar content and amount can be assumed that a reduction of uidization steam in the gasication
of char for combustion are calculated for a new steam-to-fuel ratio . reactor will not change the circulation signicantly. This is also based on
investigations in cold ow models of the DFB system.
CO H2 OH2 CO2 21 Fig. 9 shows the effect of the steam-to-fuel ratio on the water bal-
ance in the gasication reactor. Water from biomass remained constant
as it was xed in the boundary conditions, with decreasing steam-to-
fuel ratio the amount of uidization steam decreased. Thereby the
amount of water consumed in the gasication reactions only decreased
slightly. Most notably the excess steam that was not used in the
reactions was reduced. Accordingly, a smaller heat ow was required
for steam production and was available for the ORC process, which

Table 8
Correlations for steam-to-fuel ratio based on experiments in DFB pilot plant [32].

Parameter Unit Correlation Eq.

C2H4 in product gas Vol.-%, dry cC 2 H4 init  0; 2787 cC 2 H4 ;init (22)


CH4 in product gas Vol.-%, dry cCH4 init  1; 2706 cCH4 ;init (23)
CO in product gas Vol.-%, dry cCO = ( init) (7, 4992) + (24)
cCO,init
CO2 in product gas Vol.-%, dry cCO2 init  4; 8921 cCO2 ;init (25)
H2 in product gas Vol.-%, dry cH2 init  3; 7318 cH2 ;init (26)
Ungasied carbon kg C/kg fuel wC = ( init) 1, 2599 + wC,init (27)
wC waf
Tar in product gas g/Nm3 xtar = ( init) (1, 0039) + (28)
xtar xtar,init
Fig. 8. Inuence of steam-to-fuel ratio on product gas composition, [32].
146 V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147

Fig. 9. Inuence of steam-to-fuel ratio on the CHP plant Oberwart.

generated more electricity. However the amount of district heat de- Acknowledgments
creased considerably. This was closely related to the heat exchanger
arrangement. The reduced thermal efciency led to a lower overall ef- The authors would like to thank the Test Laboratory for Combustion
ciency. Therefore a reduction of the steam-to-fuel ratio is not suitable to Systems at the Institute of Chemical Engineering at Vienna University of
increase the efciency of this CHP plant. Technology for their support. Bioenergy2020 + is funded within the
Austrian COMET program managed by the Austrian Research Promotion
4. Conclusions Agency (FFG). The nancial support from the funding association FFG
and the industrial partners Energie Burgenland AG and Binder-
The mass and energy balances conrm that the performance of the Industrieanlagenbau GesmbH is gratefully acknowledged.
CHP plant Oberwart is highly satisfactory and currently achieves an
electrical efciency of 29%. The CHP plant Oberwart proves that DFB gas- References
ication is a viable process route to provide power and heat from bio-
[1] A. Bridgwater, Renewable fuels and chemicals by thermal processing of biomass,
mass (wood chips). Chem. Eng. J. 91 (23) (2003) 87102 (http://www.sciencedirect.com/science/arti-
Parameter variations show optimization potentials in the CHP plant. cle/pii/S1385894702001420).
A decrease of the steam-to-fuel ratio lowered the amount of excess [2] C. Higman, M. van der Burgt, Gasication, Gulf Professional Publishing, Burlington,
2003, ISBN 0-7506-7707-4.
steam in the product gas but did not lead to an increase in overall ef- [3] P. Basu, Biomass Gasication and Pyrolysis Practical Design and Theory, Academic
ciency. This is mainly due to the arrangement of the heat exchangers Press Elsevier, Burlington, 2010.
in the CHP plant. In order to achieve an efciency increase by this mea- [4] J. Zhang, R. Wu, G. Zhang, C. Yao, Y. Zhang, Y. Wang, G. Xu, Recent studies on chem-
ical engineering fundamentals for fuel pyrolysis and gasication in dual uidized
sure, a detailed analysis and rearrangement of the heat exchangers is re- bed, Ind. Eng. Chem. Res. 52 (19) (2013) 62836302.
quired. It is shown that a reduction of gasication temperature and a [5] E.G. Pereira, J.N. da Silva, J.L. de Oliveira, C.S. Machado, Sustainable energy: a review
lower excess air ratio will further increase the efciency of the process. of gasication technologies, Renew. Sust. Energ. Rev. 16 (7) (2012) 47534762
(http://www.sciencedirect.com/science/article/pii/S1364032112002985).
The simulation results are in good agreement with simulations and ex- [6] Z.A.B.Z. Alauddin, P. Lahijani, M. Mohammadi, A.R. Mohamed, Gasication of ligno-
periences from the Gssing gasier, where the gasication temperature cellulosic biomass in uidized beds for renewable energy development: a review,
was already lowered successfully. These measures would not require Renew. Sust. Energ. Rev. 14 (9) (2010) 28522862 (http://www.sciencedirect.
com/science/article/pii/S136403211000206).
additional investment for plant equipment, but would allow for savings [7] J. Corella, J.M. Toledo, G. Molina, A review on dual uidized-bed biomass gasiers,
in operational costs as less biomass and natural gas would be consumed. Ind. Eng. Chem. Res. 46 (21) (2007) 68316839, http://dx.doi.org/10.1021/
In addition to that, these measures would increase the electrical ef- ie0705507.
[8] K. Gransson, U. Sderlind, J. He, W. Zhang, Review of syngas production via bio-
ciency to about 31% which means to increase the electrical output by
mass {DFBGs}, Renew. Sust. Energ. Rev. 15 (1) (2011) 482492 (http://www.
nearly 10%. sciencedirect.com/science/article/pii/S1364032110003199).
The variations of plant parameters provide deeper insight in the pro- [9] C. Pfeifer, S. Koppatz, H. Hofbauer, Steam gasication of various feedstocks at a dual
cess itself and show interdependencies of different process units. Due to uidised bed gasier: impacts of operation conditions and bed materials, Biomass
Convers. Biorenery 1 (2011) 3953, http://dx.doi.org/10.1007/s13399-011-0007-1.
integration of pilot plant experience the variation of parameters be- [10] E. Perz, A computer method for thermal power cycle calculation, J. Eng. Gas Turbines
comes more descriptive and reliable. Power 113 (2) (1991) 184189 (http://link.aip.org/link/?GTP/113/184/1).
V. Wilk, H. Hofbauer / Fuel Processing Technology 141 (2016) 138147 147

[11] S. Kaiser, K. Weigl, G. Schuster, H. Tremmel, A. Friedl, H. Hofbauer, Simulation and opti- [22] I. Obernberger, G. Thek, Combustion and gasication of solid biomass for heat and
mization of a biomass gasication process, in: S.e.a. Kyrisits (Ed.), Proceedings of the power production in Europe state of the art and relevant future development,
First World Conference on Biomass for Energy and Industry, James&James, London 8th European Conference on Industrial Furnaces and Boilers, 2008.
2000, pp. 19221925 (http://members.aon.at/biomasse/sevilla-simulation.pdf). [23] T. Prll, H. Hofbauer, Development and application of a simulation tool for biomass
[12] M. Stidl, Prozesssimulation von spezischen Anwendungsfllen der gasication based processes, Int. J. Chem. React. Eng. 6 (2008) A89 (http://www.
Zweibett-Wirbelschicht-Dampfvergasungs-Technologie fr die Papier- und bepress.com/ijcre/vol6/A89).
Zellstofndustrie(Ph.D. thesis) Vienna University of Technology, 2012. [24] S. Koppatz, C. Pfeifer, H. Hofbauer, Comparison of the performance behaviour of sil-
[13] V. Wilk, H. Hofbauer, Waste wood: from gasication tests in the pilot plant to sim- ica sand and olivine in a dual uidised bed reactor system for steam gasication of
ulation of industrial applications, Proceedings of the 19th European Biomass Confer- biomass at pilot plant scale, Chem. Eng. J. 175 (2011) 468483 (http://www.
ence and Exhibition, Berlin, Germany, 2011. sciencedirect.com/science/article/pii/S1385894711011442).
[14] A. Miltner, W. Wukovits, T. Prll, A. Friedl, Renewable hydrogen production: a tech- [25] R. Rauch, C. Pfeifer, K. Bosch, H. Hofbauer, D. Swierczynski, C. Courson, A.
nical evaluation based on process simulation, J. Clean. Prod. 18 (Supplement 1(0)) Kiennemann, Comparison of different olivines for biomass steam gasication, in:
(2010) S51S62 (sustainable Hydrogen from, Biomass. http://www.sciencedirect. A. Bridgwater, D. Boocock (Eds.), Proceedings of the Conference for Science in Ther-
com/science/article/pii/S0959652610002118). mal and Chemical Biomass Conversion, CPL Press, Newbury, Berks UK, Victoria,
[15] B. Rehling, H. Hofbauer, R. Rauch, C. Aichernig, BioSNG process simulation and com- Canada 2004, pp. 799809.
parison with rst results from a 1-MW demonstration plant, Biomass Convers. [26] U. Wolfesberger, I. Aigner, H. Hofbauer, Tar content and composition in producer gas
Biorenery 1 (2011) 111119, http://dx.doi.org/10.1007/s13399-011-0013-3. of uidized bed gasication of wood inuence of temperature and pressure, Envi-
[16] T. Grbl, H. Walter, M. Haider, Biomass steam gasication for production of ron. Prog. Sustainable Energy 28 (3) (2009) 372379.
SNG process design and sensitivity analysis, Appl. Energy 97 (0) (2012) [27] T. Prll, I. Siefert, A. Friedl, H. Hofbauer, Removal of NH3 from biomass gasication
451461 (energy Solutions for a Sustainable World Proceedings of the Third producer gas by water condensing in an organic solvent scrubber, Ind. Eng. Chem.
International Conference on Applied Energy, May 1618, 2011 - Perugia, Italy. Res. 44 (5) (2005) 15761584 (http://pubs.acs.org/doi/abs/10.1021/ie049669v).
http://www.sciencedirect.com/science/article/pii/S030626191200044X). [28] F. Kirnbauer, V. Wilk, H. Kitzler, S. Kern, H. Hofbauer, The positive effects of bed ma-
[17] J. Bolhr-Nordenkampf, T. Prll, P. Kolbitsch, H. Hofbauer, Comprehensive modeling terial coating on tar reduction in a dual uidized bed gasier, Fuel 95 (2012)
tool for chemical looping based processes, Chem. Eng. Technol. 32 (3) (2009) 553562 (http://www.sciencedirect.com/science/article/pii/S0016236111006843).
410417, http://dx.doi.org/10.1002/ceat.200800568. [29] F. Kirnbauer, V. Wilk, H. Hofbauer, Performance improvement of dual uidized bed
[18] K. Marx, J. Bolhr-Nordenkampf, T. Prll, H. Hofbauer, Chemical looping combustion gasiers by temperature reduction: the behavior of tar species in the product gas,
for power generation concept study for a 10 mwth demonstration plant, Int. J. Fuel 108 (2013) 534542 (http://www.sciencedirect.com/science/article/pii/
Greenhouse Gas Control 5 (5) (2011) 11991205 (http://www.sciencedirect. S0016236112009982).
com/science/article/pii/S175058361100072). [30] S. Rapagn, N. Jand, A. Kiennemann, P. Foscolo, Steam-gasication of biomass in a
[19] T. Prll, R. Rauch, C. Aichernig, H. Hofbauer, Fluidized bed steam gasication of solid uidised-bed of olivine particles, Biomass Bioenergy 19 (3) (2000) 187197
biomass performance characteristics of an 8 MWth combined heat and power (http://www.sciencedirect.com/science/article/pii/S0961953400000313).
plant, Int. J. Chem. React. Eng. 5 (2007) A54 (http://www.bepress.com/ijcre/vol5/ [31] H. Hofbauer, R. Rauch, Stoichiometric water consumption of steam gasication by
A54). the FICFB-gasication process, in: A.V. Bridgwater (Ed.)Progress of Thermochemical
[20] J. Kotik, ber den Einsatz von Kraft-Wrme-Kopplungsanlagen auf Basis der Biomass conversion, Vol. 1, Blackwell Science Ltd., Oxford, UK, Innsbruck, Austria
Wirbelschicht-Dampfvergasung fester Biomasse am Beispiel des Biomassekraftwerks 2000, pp. 199208.
Oberwart(Ph.D. thesis) Vienna University of Technology, 2010. [32] S. Koppatz, J.C. Schmid, C. Pfeifer, H. Hofbauer, The effect of bed particle inventories
[21] J. Kotik, T. Prll, H. Hofbauer, Advanced concept for a next generation biomass gas- with different particle sizes in a dual uidized bed pilot plant for biomass steam gas-
ication chp plant basic engineering and cold ow model results, Circulating ication, Ind. Eng. Chem. Res. 51 (31) (2012) 1049210502, http://dx.doi.org/10.
Fluidized Bed Technology, IX 2008, pp. 697702. 1021/ie202353b.