*2.9. Simulation Algorithm*

In the SEG system, internal solid circulation (mass flow from the Regenerator *MRegOut*), the biomass feed stream, and the steam input are important parameters defining the gasification temperature. Therefore, in the simulation model, the molar calcium looping ratio (*FCaO*/*FC*) is used to set the gasification temperature. In this ratio, *FC* considers the molar flow of carbon contained in the biomass. The corresponding flow chart of the model is shown in Figure 5.

After setting boundary conditions and model parameters, the average CO2 carrying capacity is calculated using the make-up flow of fresh limestone and the looping ratio. Starting from an initial temperature, first the pyrolysis products and then the fluid dynamics of the gasifier are calculated. After solving the mass and energy balances, a new gasification temperature is calculated, and the pyrolysis step is updated for a certain looping ratio. This is iterated until the change of the empty reactor velocity, *uempty*, is smaller than a defined value δ; then, the results of the operation point are saved in a file.

**Figure 5.** Simulation flow chart for SEG fluidized bed gasifier model.

#### **3. Results and Discussion**

In Section 3.1, the model parameters (e.g., heat transfer coefficients) are verified against experimental temperature data measured in the 200 kW fluidized bed gasifier [11]. Section 3.2 contains a model validation with measured gas compositions (H2, CO, CO2, CH4, C2H4), the lower heating value (*LHV*) of the syngas, and the *M* module from the 200 kW fluidized bed gasifier [11,19] in a temperature range of 600–850 ◦C. According to [48,49], the accuracy of the parameter verification and model validation was quantified by the sum squared deviation method:

$$\text{MeanError}: \overline{E} = \sqrt{\frac{\sum\_{n=1}^{N} \left(\frac{q\_{exp} - q\_{model}}{q\_{exp}}\right)^2}{N}} \tag{36}$$

For parameter verification, a low mean error of the temperature distribution along the reactor height of 10.6% was found. The model validation was carried out with a limestone make-up flow rate (*MU*) of 6.6 kg/h, according to experiments [11]. To characterize the effect of limestone make-up flow rate, a parametric study with 0.2 kg/h, 6.6 kg/h, and 15 kg/h is also included in the result diagrams. Mean errors to quantify the model prediction accuracy are listed in Table A2.

#### *3.1. Verification of Model Parameter*

In Figure 6, simulated temperature profiles from variations of the parameter α are compared with temperatures measured on different positions in the fluidized bed and the freeboard (available in [11]).

**Figure 6.** Simulated temperature profiles of the gasifier for various α (ranging from 0.8 to 0.98); . *MBM*,*<sup>w</sup> f* = 29.7 kg/h, *S*/*C* = 2.2 mol/mol, *WHSV* (weight hourly space velocity) = 0.68 1/h, limestone make-up 6.6 kg/h, and comparison with experimental data from 200 kW dual fluidized bed (DFB) pilot plant [11].

The profile of the measured temperatures (circles in Figure 6) over the height of the gasifier can be explained as follows: due to a good mixing of solids in the fluidized bed (0 m to 1.15 m) the temperatures were close to 640 ◦C, followed by an inflection, which was caused by the inlet of hot solids at 1.7 m (available thermocouple was at 1.5 m). In the freeboard above the solid inlet, a decrease of the gas temperature was observed due to heat losses through the reactor wall. A variation of the model parameter α in the range of 0.8–0.98 was carried out to adjust the fluidized bed temperature by changing the proportion of wake mixed within each discretization cell. From Figure 6, it can be seen that, with a value of α = 0.95 (red line), a temperature profile could be achieved, which corresponds to the measured temperature values. An almost vertical profile confirmed a homogeneous particle mixing in the fluidized bed. In the model, a heat transfer coefficient of *kFluidB* = 12.9 Wm−2K−<sup>1</sup> was considered for the fluidized bed area. For the particle–gas heat transfer, a heat transfer coefficient value *kp* = 160.7 Wm−2K−<sup>1</sup> and for the freeboard *kf* = 3.4 Wm−2K−<sup>1</sup> were determined to describe the given temperature profile.

#### *3.2. Validation with Experimental Data*

After setting the model parameters, we carried out a verification based on temperature along the gasifier height, the syngas composition, *LHV*, reactions rates, and the *M* module, in order to compare the simulation results with experimental data over a temperature range from 600 ◦C to 850 ◦C.

#### 3.2.1. Effect of Gasification Temperature on Fractions of Syngas Components

In Figure 7a–e, the simulated volume fractions of synthesis gas components (H2, CO, CO2, and CH4) and non-condensable hydrocarbons (in the form of C2H4) are plotted over a gasification temperature between 600–850 ◦C. For each synthesis gas component, the curves of three different make-up flows (0.2 kg/h, 6.6 kg/h, and 15 kg/h) are shown to demonstrate its influence on the gas

volume fraction. Additionally, experimental results derived from both 200 kW (Experiment 1) and 20 kW (Experiment 2) DFB systems are included to evaluate the simulated volume fractions.

**Figure 7.** (**<sup>a</sup>**–**<sup>e</sup>**) Simulated syngas components: (**a**) H2, (**b**) CO, (**c**) CO2, (**d**) CH4, and (**e**) C2H4 for gasification temperatures in the range of 600–850 ◦C (lines for limestone make-up of 0.2, 6.6, and 15 kg/h); comparison with experimental results from 200 kW (Experiment 1) and 20 kW (Experiment 2) DFB systems (data points).

For sorption enhanced gasification, one important characteristic is the strong dependency of the gas composition on the gasification temperature, which is a ffected by CO2 capture through the carbonation reaction [19,22]. Beside gasification reactions, the pyrolysis step also has an important impact on the initial gas composition in the fluidized bed [34]. The results from a 20 kW system were additionally included in the present work, as it enables operation temperatures up to 850 ◦C due to its electrical heating system. For low gasification temperatures, there is a larger distance between the actual CO2 concentration and the equilibrium curve for the carbonation/calcination regime [37,38]. This leads to a strong capture of CO2 and, hence, low CO2 concentrations in the syngas. With higher temperatures, the CO2 capture rate decreases and, consequently, the CO2 concentration in the syngas rises. Furthermore, this influences the water–gas shift reaction, resulting in decreased H2 concentrations and increased CO concentrations. When the CO2 concentration reaches the equilibrium concentration at around 750 ◦C, an inflection in the concentrations of CO2 and H2 can be observed. This demonstrates the strong coupling of the carbonation reaction with the water–gas shift reaction. Particularly for low gasification temperatures, there is also a distinctive influence of the make-up flow. Presumably, the reason for this e ffect is a reduced circulation mass flow of fresh CaO from the regenerator with a simultaneously higher CO2 capture rate due to lower temperatures. For instance, at a gasification temperature of 600 ◦C, the delivered circulation mass flow is around ten times lower, compared to that when operating at 850 ◦C. This can lead to a higher content of carbonated particles if the bed inventory is hardly exchanged. In this operating range, an increase of the limestone make-up rate can increase the activity in the bed and, thus, the CO2 capture rate.

#### 3.2.2. E ffect of Gasification Temperature on LHV

As seen in Section 3.2.1, the gas composition is strongly a ffected by gasification temperature, due to the temperature dependence of (i) products released from pyrolysis, (ii) Arrhenius approaches to describe the gasification reactions, and (iii) carbonation/calcination equilibrium. Based on the lower heating values (*LHV*) of pure syngas components, the *LHV* of the gas mixture was calculated and compared with experimental data. In Figure 8, simulation results are shown over a temperature range of 600–850 ◦C and for di fferent make-up flow rates.

**Figure 8.** Simulated lower heating value (*LHV*) of syngas for gasification temperatures between 600 ◦C and 850 ◦C (lines for limestone make-up of 0.2, 6.6, and 15 kg/h); comparison with experimental results from a 200 kW (Experiment 1) and 20 kW (Experiment 2) DFB system.

The results show a good correlation with the experimental data and only a slight variation for different make-up flow rates is noticed. Furthermore, the effect of CO2 capture and its limitation at around 750 ◦C, recognizable as an inflection, can be described with this model. For higher temperatures, the *LHV* remains the same at a value of 10.9 MJ/m<sup>3</sup> @STP. The highest values of *LHV*—around 14.5 MJ/m<sup>3</sup> @ STP—can be reached at temperatures lower than 650 ◦C.

#### 3.2.3. Effect of Make-Up Flow on Reaction Rate

In Figure 9, reaction rates (in mol m<sup>−</sup><sup>3</sup> s<sup>−</sup>1) are shown over a temperature range from 600 ◦C to 850 ◦C. Additionally, a variation of the make-up flows (0.2 kg/h, 6.6 kg/h, and 15 kg/h) was included to investigate the influence of sorbent deactivation on all considered reaction rates for the gasification process.

**Figure 9.** Reaction rates over temperature; influence of effective make-up of fresh limestone (lines for 0.2, 6.6, and 15 kg/h) to bed activity.

As the gasification temperature varies with changes of the circulation mass flow, the sorbent residence time differs and, according to Equation (35), the sorbent activity is also affected. It can be seen that deactivation mostly influenced the carbonation (reaction 4) and water–gas shift (reaction 5) reactions. For instance, for a constant temperature and a constant circulation mass flow, the carbonation reaction rate is higher with larger amounts of fresh limestone.

Considering the influence of the gasification temperature on the reaction rates, it can be seen that the reaction rates of the water–gas shift (reaction 5), the heterogeneous water–gas (reaction 2), and the Boudouard (reaction 3) reactions increased with higher temperatures. In contrast, for the pyrolysis (reaction 1) reaction, a minor decrease can be observed, even though an increase should be expected with higher temperatures. The reason for this behavior can be explained as follows: The amount of biomass in the fluidized bed system is limited by a constant fuel input flow. When the gasification temperature is increased, the reaction rate increases but, at the same time, a higher solid circulation is required for the higher temperature. Hence, more unreacted biomass is extracted from the gasifier through the loop seal, thus reducing the gas yield. Kinetic parameters of the Arrhenius approach influence the gradient of reactions for increasing temperatures. Comparing the Arrhenius parameters listed in Table 1, it can be seen that the influence of temperature on the heterogeneous water–gas shift reaction (reaction 2) was higher than that of the pyrolysis reaction (reaction 1).

3.2.4. Effect of Gasification Temperature and Make-Up of Fresh Limestone on M Module

The *M* module from Equation (37) relates the gas concentrations of hydrogen, carbon dioxide, and carbon monoxide and has been considered as an important parameter which dictates the application of syngas [50]. For validity of an ideal gas, it can be written with volume fractions.

$$M = \frac{y\_{H\_2} - y\_{CO\_2}}{y\_{CO\_2} + y\_{CO}} \tag{37}$$

For instance, a *M* module of two is required for full stoichiometric conversion into dimethyl ether [50] and, for methane synthesis, a value of three can be derived from methanation reactions. Higher values are mostly interesting for hydrogen production. Figure 10 shows the influence of the gasification temperature on the *M* module. Additionally, three different simulation results, with a limestone make-up of 0.2 kg/h, 6.6 kg/h, and 15 kg/h, are compared with experimental data. It can be seen that a higher make-up flow enables higher *M* modules under the same gasification temperature.

**Figure 10.** Influence of gasification temperature and make-up mass flow (lines for 0.2, 6.6, and 15 kg/h) on *M* module.

For a low gasification temperature, the experimental data can be reached with a make-up flow of 0.2 kg/h. At a gasification temperature above 650 ◦C, the simulation model with a make-up flow of 6.6 kg/h precisely describes the experimental data. This parametric study reveals that the make-up flow is an important factor when optimizing the gasification process for a certain application. When considered from an economic point of view, operation strategies with low make-up rates are preferable.

#### **4. Sensitivity Analysis of Fuel Feeding Rate**

For a realistic and flexible operation of the biomass gasification process, it is important that the process allows for safe operations under a wide load range. In this section, the effects of different fuel feeding rates on the bed height, the fluidization level (gas velocity ratio: superficial gas velocity *u* based on minimum fluidization velocity *umf*), and the power of the syngas, depending on the gasification temperature, are investigated.

#### *4.1. E*ff*ect of Biomass Feeding Rate on Bed Height and Gas Velocity Ratio*

The model was also used to study the influence of different biomass feeding rates on the hydrodynamics in the fluidized bed. In Figure 11, the gas velocity ratio (superficial gas velocity/minimum fluidization gas velocity) is shown over a height of 3 m of the gasifier focusing on the fluidized bed and lower part of the freeboard. In this figure, the biomass feeding rate is considered as curve parameter in the range of 22–40 kg/h, whereby the operating point considered in Section 3.1 (feeding rate: 29.7 kg/h) was additionally drawn as a red line. Since the fluidized bed expands with larger fluidization volume flows, the diagram also shows the height of the fluidized bed for the different biomass feeding rates as blue dots. This allows the gas velocity at the surface of the fluidized bed to be read directly from the diagram.

**Figure 11.** Simulated effect of biomass feeding rate (22–40 kg/h) on gas velocity ratio (superficial gas velocity/minimum fluidization gas velocity) over reactor height and position of bed height (*H*FluidB).

It can be seen that with a higher biomass feeding rate, the gas velocity ratio increased at any position in the reactor. The reason is that when biomass particles pyrolyze in a fluidized bed, the released gas contributes to the reactor fluidization. In addition, a constant *S*/*C* ratio of 2.2 was selected for the simulations, in order to maintain a stable syngas quality [51] and to ensure comparability with the experimental data [11]. This changes the amount of steam supplied and, hence, the gas velocity ratio. In Figure 11, the red line corresponds to the same case which was considered for model verification in Section 3.1, with the reactor temperature in the axial direction. At the zero position in *y*-axis, the primary steam inlet is located. Due to the smallest cross-section in this area (compare Figure 4a), the highest gas velocity ratios were found here. On higher levels of the gasifier, the diameter of the reactor increases, which leads to a decrease of the gas velocity ratio due to the continuity equation. However, increasing velocities can be observed resulting from the gas release due to biomass pyrolysis (*h* = 0.2 m) and the secondary steam inlet (*h* = 0.285 m). At *h* ≥ 0.35 m, the reactor has a cylindrical shape and from the constant cross section in combination with further biomass pyrolysis, the gas velocity ratio slightly increases. In the freeboard above the inlet of solids, the temperatures decrease due to heat losses, and hence, the gas velocity ratio decreases. This is indicated by an inflection in the gas velocity ratio curves at a height of 1.7 m. Looking at the fluidized bed height for operation with 22 kg/h and 40 kg/h biomass feeding rate, one can see that the height doubles. For the case with 40 kg/h, the bed height (blue dots) almost reaches the area of the inlet for the solid circulation at 1.7 m. In order to ensure stable operation with this reactor geometry, operating modes that lead to a further increase in bed height should be avoided.

#### *4.2. Performance Diagram for Gasifier Operation*

Based on the results derived from this work, a performance diagram of the bubbling fluidized bed gasifier was created in Figure 12. This relates the selected gasification temperature (based on downstream requirements of the syngas composition) and fuel feeding rate to the power of the syngas. Additionally, the gas velocity ratio from the superficial gas velocity at the fluidized bed surface and the gas velocity ratio for the lowest velocity in the fluidized bed are depicted.

**Figure 12.** Power of syngas for gasification temperatures between 650 and 750 ◦C for water-free fuel input of 25 kg/h, 30 kg/h, and 36 kg/h (left side); Corresponding gas velocity ratio (related to *umf*) at the top of the fluidized bed and for the position where the lowest velocity in the fluidized bed occurs (right side).

By combining the gasification chemistry (which lead to the syngas power) and the hydrodynamic information (represented as the gas velocity ratio), it is possible to identify a realistic operational range of the gasifier.

With a gasification temperature of 650 ◦C and 25 kg/h fuel input, the lowest gas velocity ratio of 20 could be identified. This means that, in any position of the fluidized bed, the superficial gas velocity was 20 times higher than the minimum fluidization velocity. Hence, good mixing of the bed inventory can be assured. At the bed surface, the superficial gas velocity was 26 times higher than the minimum gas velocity and only a little particle extraction can be expected. Increasing the fuel feeding rate to 36 kg/h at 650 ◦C, the syngas power increased and reached almost 100 kW. At this point, the velocity at the surface of the fluidized bed increased by a factor of 37 (related to *umf*) and by a factor of 27 (related to *umf*) for the lowest velocity in the bed. By increasing the gasification temperature and the gasifier feeding rate up to 750 ◦C and 36 kg/h, respectively, good mixing in the fluidized bed is guaranteed. However, due to the high velocity at the bed surface (a factor of 47 related to *umf*), a high particle extraction has to be accounted for. To avoid this negative effect, the fuel input or the gasification temperature can be modified. In that case, the syngas composition is not very important for downstream applications, an identical syngas power (around 100 kW) can be reached for a gasification temperature of 750 ◦C and 30 kg/h fuel input or for an operation with 650 ◦C and with a fuel input of 36 kg/h. However, the operational point at 650 ◦C reduced particle extraction from the gasifier due to the non-linear behavior of the gas velocity ratio curves.
