*3.1. Tar Reduction Efficiencies*

### 3.1.1. Tar Reduction Efficiency of Wood Shavings Filter (Treatment I)

Tar content at the inlet and outlet of wood shavings filter were 78 g/Nm<sup>3</sup> and 70 g/Nm<sup>3</sup> , respectively (Table 3). A study [18] using corn cobs as the filter medium reported outlet tar content of 2 g/Nm3. The lower outlet tar content using corn cobs may be attributed to the two layers of sponges used in combination with the corn cobs. The tar removal efficiency of the wood shavings filter obtained in this study (10%) was much lower compared to 97% reported using a water scrubber and coconut coir filter [17] and 94% using spray towers and wood shavings filter [19]. Observation of biomass filter layers (Treatment I in Figure 7) revealed that more tar was adsorbed at the upper section and around the periphery of the filter. This was attributed to the lower temperatures near the periphery of the filter.


**Table 3.** Tar contents of raw and cleaned syngas (dry basis) with the three tar cleaning treatments.

**Figure 7.** *Cont.*

**Figure 7.** Top and bottom sections of the filter for three treatments: (**I**) wood shavings filter; (**II**) heat exchanger + wood shavings filter; and (**III**) oil bubbler + wood shavings filter.

3.1.2. Tar Reduction Efficiency of Wood Shavings Filter in Combination with Heat Exchanger (Treatment II)

The inlet and outlet syngas tar contents of the filter system equipped with a wood shavings filter and heat exchanger, were 70 g/Nm<sup>3</sup> and 27 g/Nm3, respectively, resulting in tar removal efficiency of 61% (Table 3). Unlike previous studies [15,17,19] that used direct (contact) cooling systems, such as spray towers and water scrubbers; the heat exchanger used in this study was an indirect (no contact) cooling system. Compared to Treatment I (only wood shavings filter), indirect cooling of syngas using a heat exchanger (Treatment II), reduced the dew point of tar and significantly improved tar removal efficiency. Figure 7 presents the pictorial comparison of tar adsorbed at top and bottom sections of the wood shavings filter. Similar to the observation made in Treatment I, more tar deposited at the top section and around periphery of the filter in Treatment II. Figure 7 also shows that the addition of the heat exchanger clearly aided in deposition of more tar on the wood shavings filter. However, the final tar content (27 g/Nm3) obtained in this treatment was higher than those obtained with dry filter used with direct (touch) syngas cooling and cleaning systems, such as sand bed filter with water spray towers (1.5 g/Nm<sup>3</sup> final tar content) [15] and coconut coir filter with water scrubber (1.4 g/Nm<sup>3</sup> final tar content) [17]. Similarly, the final tar content of this treatment is high than those of cleaning systems with a heat exchanger with bag house filter (35 mg/Nm3) [21]; and heat exchanger with venture scrubber (10 mg/Nm3) [22]. High outlet tar content in this study can be attributed to two primary reasons: (i) indirect cooling, used in this study, has low tar removal effectiveness; and (ii) inlet tar content was high (70 g/Nm<sup>3</sup> vs. <10 g/Nm3) [15,19]. The result from Treatment II is comparable to the tar removal efficiency (61%) of the system with a cyclone, heat exchanger and oil bath filter [32]. These result, shows that indirect cooling of syngas with the heat exchanger increased the tar removal efficiency of the dry filter but further cleaning is still required if the syngas is used in an Internal Combustion (IC) engine. However, since water and some tar are condensed in the heat exchanger, the design of the heat exchanger must allow water and tars to flow out of the heat exchanger pipes. Use of a single vertical tube allowed us to collect water and tar condensed during the test. Tar deposition along the inner surface of the heat exchanger was minimal but after several runs, the inner surface required cleaning.

3.1.3. Tar Reduction Efficiency of Wood Shavings Filter with Vegetable Oil Bubbler (Treatment III)

To investigate the tar removal efficiency of wood shavings filter equipped with vegetable oil bubbler as cooling unit, the cleaning system was installed as specified in Figure 1 (Treatment III). Tar content at the inlet and outlet of the cleaning system were observed as 70 g/Nm<sup>3</sup> and 1.9 g/Nm3, respectively (Table 3). The tar content at inlet of wood shavings filter (after oil bubbler) was 3.8 g/Nm3, which suggests that a large portion of the tar was condensed in oil bubbler (as shown in Figure 8), and is also evident from Figure 4, which shows that very little tar was condensed on the wood shavings filter.

**Figure 8.** Canola oil (used in oil bubbler for treatment III) before and after the test.

High tar removal efficiency of the cleaning system equipped with oil bubbler (97%) indicates that, unlike in previous studies [24,25,27] that have used oil scrubber, an oil bubbler is also effective for removal of syngas tar. High tar removal efficiency of 98% with sunflower oil [24] and final tar content of 0.022 g/Nm<sup>3</sup> with waste palm cooking oil [27] have been reported. The high tar removal efficiency of the oil-based cleaning system is attributed to oil's lipophilicity characteristics, the ability of the oil to dissolve non-polar hydrocarbons [27]. Tar compounds are lipophilic in nature and can mix well with vegetable oils as these oils have saturated and unsaturated fatty acids.

Feasibility of removing syngas tar with biomass and oil is promising because the tar mixed oil and wood shavings can be reused in gasifier reactors as feedstock, eliminating the need to treat waste effluent. For example, the scrubbing oil was reused and tars were recycled into the gasifier by the Energy Research Center of the Netherlands [28]. Oil used in removing tar was put in the regeneration process using filtration and centrifugal sedimentation techniques and reused in the scrubber [29]. Wood shavings also removed syngas moisture depicted by the high moisture content of the filter after the test. However, the tar content at the outlet of oil bubbler was still not low enough for engine application. In conclusion, oil may have been effective in reduction of heavy tar [24], but additional cleaning is necessary for the removal of light tars.

### *3.2. Variation of Pressure Drop across Wood Shavings Filter for the Three Treatments*

Pressure drop across the wood shavings filter depends on amount of tar accumulated in the filter medium. As shown in Figure 9, pressure drop across the filter increased with time due to continued accumulation of tar on the filter medium for all three treatments. However, throughout the test duration, pressure drop was the highest for Treatment II (when heat exchanger was used before the filter) and the lowest for Treatment III (when oil bubbler was used before the filter). The trend of pressure drop indicated that tar deposition on the filter was the highest for Treatment II and the lowest for Treatment III. However, overall tar removal efficiency was the highest for Treatment III (97%) and the lowest for Treatment I (10%). These observations indicated that the highest tar removal for Treatment III is due to the oil bubbler, which removed most of the tars leaving only a small quantity of tar deposited on the filter hence the pressure drop across the filter was minimal. Syngas tar content measured at the outlet of the oil bubbler (3.8 g/Nm3) confirmed the finding that only a small quantity of tar (3.8 g/Nm3) was removed by the filter medium and most of the tar was removed by the oil bubbler (66.2 g/Nm3). For Treatment I, as the filter medium was not augmented with any other cleaning method, the pressure drop increased with increasing tar accumulation on the filter over time. In Treatment II, use of a heat exchanger before the filter medium reduced the temperature of syngas entering into the filter medium from 135 ◦C to 71 ◦C, which led to increased condensation of the tar on the filter medium. Tar deposition observed on the heat exchanger was minimal, indicating that most of the tar was removed by the filter. Higher pressure drop across the filter for Treatment II as compared to Treatment I also indicated that tar deposition on the filter for Treatment II (with heat exchanger) was higher than that for Treatment I (only filter). Hence, the use of heat exchanger reduced the syngas temperature and increased tar deposition on the filter medium. The pressure drops across the filter for all three treatments (0.2–0.5 in of H2O) were lower compared to those reported by others (0.5–2 in of H2O) [15,17] due to low condensation of tar by wood shavings. This low pressure drop is beneficial for power generation application because pressure available at the engine manifold is high and prevents high back pressure in the gasifier.

**Figure 9.** Variation of pressure drop of wood shavings filters over time.

### *3.3. Variation of Gas Temperature at Inlet and Outlet of Wood Shavings Filter for Three Treatments*

Temperature of syngas is a key parameter affecting condensation of tar. Figure 10 shows the 1 h average syngas temperature at the inlet and outlet of the wood shavings filter for the three treatments. The filter inlet temperature for Treatment I was the highest, followed by Treatment II and Treatment III. The trend shows that the cooling by the heat exchanger and the oil bubbler were effective. The syngas temperature entering into the cleaning system was the same (about 135 ◦C) for all three treatments; however, because of the use of heat exchanger and oil bubbler before the filter, the syngas temperature at the filter inlet was different. The average syngas temperature entering into the cleaning system was low (about 135 ◦C), due to the use of long piping between the cyclone and cleaning system. As a result, tar condensation could have happened along the piping. The difference between inlet and outlet syngas temperatures of the filter (Figure 10) was the highest for Treatment I, followed by Treatments II and III. This trend can be attributed to the effectiveness of heat exchanger and oil bubbler in reducing gas temperature at the filter inlet.

As expected, low syngas temperatures at the filter led to high condensation on the wood shavings medium. In addition, low temperature of 30 ◦C is desired for feeding into IC engine, [10]. The oil bubbler (Treatment III) was effective in reducing syngas temperature at the filter outlet to 27 ◦C. The heat exchanger-based cooling system also reduced the syngas temperature; however, the outlet syngas temperature (58 ◦C) was still higher than desired for most engine applications. Using a heat exchanger with multiple tubes may further reduce the temperature sufficient for engine applications.

**Figure 10.** Syngas temperature at the inlet and outlet of the wood shavings filter for the three treatments.

## *3.4. Heating Value of Syngas*

Heating value of syngas affects performance of the downstream power generation unit [21]. The lower heating values (LHVs) of syngas sampled at the outlet of the wood shavings filter for the three treatments were in the range of 5–6 MJ·Nm−3. These results are comparable to 5.3 MJ·Nm−<sup>3</sup> produced from a downdraft gasifier with olive kernel as feedstock and wet scrubber and heat exchanger as a cleaning unit [22] and 5.79 MJ·Nm−<sup>3</sup> produced from an 18 kW gasifier using hardwood chips as feedstock [33]. The average gas composition for each of the three treatments are presented in Table 4. The variation in heating values of product gas can be attributed to the difference in composition of combustible gases, such as H2, CO and CH4.

**Table 4.** Syngas composition and heating value for the three cleaning treatments.

