Based on the basic experiment on the pyrolysis and the reforming, we have developed the simulator by which the gaseous yields and/or the energy efficiency through BT process can be estimated. Here, we compared the practice data through the demo-plant with the result of the simulator.

In general, there would be somewhat deviation between the practice data and the estimated one due to the simulator. That is, it would be extremely significant to identify the deviation from the viewpoint of the reliable plant operation.

The calculation logic of the simulator which we developed is as follows (Dowaki et al. 2007):

1. The reaction temperature in each furnace (pyrolyzer and reformer) and the steam feeding rate are fixed.

2. Based on the gaseous yields in the pyrolysis and/or the reforming reactions, which were analysed by gas chromatograph (GC-8A, Shimadzu), the gaseous components in

each furnace were estimated due to the following two equilibrium reactions. In our experiments, Shincarbon-ST was used as a measure column, and the six kinds of gases which can be measured include H2, CO, CO2 CH4, C2H4 and C2H6, respectively.

Note that we considered the approach temperature difference between the theoretical one and the reaction one on these equilibrium reactions.

Here, in the gasification process after the pyrolysis reaction, it is usually thought that the following two reactions take place.

CH4 + H2O о CO + 3H2 (1)

CO + H2O о CO2 + H2 (2)

Using our experimental apparatus, the approach temperature differences between theoretical temperature and the actual temperature was measured. This temperature difference is known as the approach temperature.

In Eqs. (1) and (2), the equilibrium constant of shift reaction and that of methanation reaction are represented as follows:

Where, AT, R and p are the approach temperature [K], gas constant [J/molK] and the partial pressure of i-component [Pa], respectively.

Next, we compared the measurement data in Izumo plant (1t/d) with the estimated results using the simulator (see Table 1). Here, the tasks of Run 11 and Run 11.2 in which the decomposed reactions are assumed to be completed in the pyrolyzer are described. Both tasks were executed on March, 2009 (Kameyama et al. 2010).

	Deviatoin [-]	Temparature at Refomer [°C]
Measured	Estimated
Run 11.1	0.134	820	801
Run 11.2	0.367	750	774

Table 1. Comparison of the measured data and the simulated ones

In this verification, we focused on the molar fractions of H2, CO, CH4 and CO2, and found the average reaction temperature so that the total deviation on molar fraction between the measurement data and the estimated one due to Eqs. (3) and (4) is a minimum. That is, we investigated if the gaseous components based on the temperature which was measured in the plant corresponded to the estimated ones due to the simulator. The reason why we verified the gaseous components using the temperature as a variable is as follows; in the demo-plant, we did not know the temperature profile on the vertical and/or horizontal directions precisely since the sampling point of the temperature in reformer is one position. Thus, assumed that the estimated temperature based on the measured gaseous components would represent the average one of reformer, we made sure that the process simulator would be more suitable.

Next, based on the process simulator which has a precision to some extent, we describe the example due to the biomass feedstock of waste Japanese cedar. With regard to the gasification performance, since gaseous yields and concentrations are dependent upon the kind of materials, the operating temperature, and the inner pressure, they were examined using the gasifier apparatus which has a reformer and a pyrolyzer.

Here, Table 2 shows the ultimate analysis of the waste Japanese cedar.

C* H* O* S* N* Cl* Ash*	46.660% 5.480% 47.351% 0.000% 0.120% 0.000% 0.389%	wt.% wt.% wt.% wt.% wt.% wt.% wt.%
HHV*	18,348	kJ/kg
Moisture Content	20.0	wt.%
Volatile Matter	86.21%	wt.%
Bulk density	0.14	t/m3

Table 2. Ultimate analysis of the waste Japanese cedar

^ Dry-Base

Through the tests, the syngas components and the equilibrium constants were obtained. For instance, Fig.3 illustrates the gaseous yields on the pyrolysis at 550 °C with variation of S/C =0.14 to 0.98, and the reforming reaction at S/C=1.0 with variation of 800 to 950 °C, respectively. Here, a steam carbon ratio is defined as the following equation.

Note that the gaseous components are modified at 20% moisture content. Also, the approach temperature for each reaction is shown as Table 3.

Reaction	AT	Unit
Pyrolysis	78.3	°C
Reforming	252.0	°C

Table 3. Approach temperature for each reaction (estimated)

Fig. 3. Gaseous yields of pyrolysis (a) and reforming reaction (b).

Based on the above experimental results, we estimated the following material balance: C31.078H43.495 O23.677 N0.069 + 31.078#2O ^

24.052H2 + 11.682CO + 2.435CH4 + 9.442CO2 + 23.228H 2O (6)

+2.92 X 10 N2 + 4.40 X 10 NH3 + C2.686H0.482O0.343N0.004

C2 686H0482O0343N0004 is the chemical component of char, and its heating value was 32.0 MJ/kg. In our simulator, the energy performance would be solved so that the input and the output on heat and materials would be balanced.

Next, using ф=9.5mm ball, we measured the temperature profiles at the surface of ball and the center of it. In the phase of absorption of heat, the ball was kept at each designed temperature between 200 and 950 °С. At the time, there was difference between the surface temperature and the center one, and the temperature differences were measured. Inversely, in the phase of heat radiation, the ball was heated up to 1,000 °С in the furnace, and it was put in a room temperature. Simultaneously, the temperature differences were measured. Note that these temperature profiles are time series data.

As a result, the thermal conductivities can be obtained. Also, since the thermal circulation time has to be the same as the reacting time on a pyrolysis and a steam reforming reaction, the optimal size of the ball is decided. Thus, the adequate auxiliary power for the circulation of HC would be obtained. Due to this result, we can estimate the suitable residence time in each reactor for the temperature profile which would be led by the simulator. Based on the above concept, we could estimate the syngas through BT process (Dowaki et al., 2008a, Dowaki, 2011a).