Gas Cleaning
Gas treatment experiments were carried out under real gas conditions at four different gasifiers. Complementary, systematic tests under lab conditions employing synthetic gas mixtures were done. At all sites particulate matter and sulfur compounds were removed, employing state of the art technologies (ceramic/metal membrane filter candles, ZnO-based fixed-bed adsorption), achieving the defined target limit of below 1 ppm. Detailed investigation of the sulfur adsorption process showed a not neglectable conversion of H2S to COS under dry conditions and high gas hourly space velocity (GHSV), which must be taken to account in plant design, because of the slower adsorption kinetics of COS-adsorption and therefore the decrease of effective capacity of sulfur removal. This COS-formation can be avoided by adding steam before the adsorption process. Further parameters, that are supposed to be influencing the sulfur adsorption process were studied. Increasing CO promotes the formation of COS and therefore decreases the effective sulfur removal capacity. Increasing temperature increased outlet H2S concentrations but increased the effective removal capacity at the given set up.
Exemplary results of the measurement campaign at IEN, Warsaw
In the frame of long-term SOFC-test fueled with producer gas of the air-blown Kajot gasifier at IEN in Warsaw BEST conducted studies on desulfurization processes. Two objectives had to be fulfilled:
-
Removal of all Sulfur compounds enabling long-term operation studies of a SOFC-stack
-
The investigation of the desulfurization process under real gas condition of the gasifier with the focus on kinetic behavior of H2S-, and COS-removal and/or formation reactions of COS.
Figure 1 Schematic view of experimental setup for real syngas adsorption tests, indicating the sample points A,B,C,D, for gas characterisation.
A schematic view of the experimental set-up is shown in Figure 1. The set-up consisted of two parallel desulfurization columns, one pilot-scale adsorption column and a smaller lab-scale adsorption column, charged with different amounts of sorbent material. Dedusting is done by a commercial, heated (400 °C) metal membrane filter.
The dust free gas was split and directed to both, pilot scale adsorber and small scale adsorber by parallel pipelines at operation pressure slightly below atmosphere. All connection pipes before the adsorption columns were heated electrically up to 400°C.
The tests were performed using commercially available Zinc Oxide (ZnO)-based spherical granules (sized 2.5–4.7 mm, specific surface area is 28.3 m2/g). Key data of the tests are summarized in Table 1. The gas content (N2, CO, CO2, H2, O2, CH4, H2S, COS etc.) was automatically analyzed every 5 min by Agilent 490 Micro Gas Chromatography (µGC). The main permanent gas components of the syngas, CO2, CO, H2 and CH4 had been monitored additionally by NDIR- and TCD-based multi-component gas analyzing system from ABB.
Table 1 Real Syngas Adsorption Test Conditions.
|
Adsorption Test |
Set Temp. |
Flowrate (L/min) |
Adsorbent Amount (g) |
Inner Diameter (cm) |
GHSV (hour-1) |
|
Pilot-Scale Test |
350 |
1 |
300 |
4.7 |
240 |
|
1st Small-Scale tests |
300 |
0.7-0.75 |
2.5 |
0.9 |
26550 |
|
2nd Small-scale tests |
400 |
0.7-1.3 |
5 |
0.9 |
7000-18400 |
|
3rd Small Adsorption tests |
300-500 |
0.3-1.3 |
2 |
0.9 |
10700-46000 |
The heated pilot-scale adsorption column is employed to provide sulfur free gas for long-term Solid Oxide Fuel Cell (SOFC) operation. Therefore, the adsorption unit operated at a low GHSV of 240 h-1 and a high amount of adsorbent to ensure the complete removal of sulfur. The monitoring of H2S clean gas concentration (sample point D) was done periodically (once a day) during SOFC operation. Sulfur compounds could not be detected over the continuous test run for more than three weeks. Measurement duration was at least 30 minutes to obtain a representative gas composition. Maximum, minimum, average and average value for standard deviation are shown in Table 3.
Table 3 Gas composition values of sample point D, vol% dry base, water content from 7vol% to 10vol%
|
|
H2 [%] |
N2 [%] |
CH4 [%] |
CO [%] |
CO2 [%] |
H2S (ppm) |
COS (ppm) |
|
|
Inlet |
Maximum |
18.3 |
49.5 |
1.8 |
24.6 |
12.5 |
15.9 |
1.3 |
|
|
Minimum |
15.0 |
45.1 |
0.8 |
19.3 |
9.1 |
3.4 |
n.d |
|
|
Average |
17.1 |
47.2 |
1.3 |
22.4 |
10.6 |
10.7 |
0.8 |
|
|
Std Dev* |
1.0 |
1.3 |
0.3 |
1.7 |
1.3 |
4.9 |
0.4 |
|
Outlet |
Maximum |
20.3 |
53.1 |
1.9 |
23.9 |
11.7 |
n.d. |
n.d |
|
|
Minimum |
14.8 |
43.7 |
0.6 |
19.6 |
10.1 |
n.d |
n.d |
|
|
Average |
17.4 |
48.3 |
1.1 |
22.0 |
10.8 |
n.d |
n.d |
|
|
Std Dev* |
1.8 |
2.9 |
0.4 |
1.3 |
0.5 |
n.d |
n.d |
Although the main gas components changed during the operation of the gasifier, neither H2S nor COS were detected at the outlet. Since equilibrium constants of H2S and COS for ZnO adsorption at 350°C are quite high, a complete removal of these pollutants is possible at the given high residence times. The tests confirmed, that both, COS and H2S, can be removed simultaneously at 350°C and 240 h-1 GHSV by ZnO adsorption to clean gas concentrations below 1 ppm. It should be considered that low GHSV values had been chosen to ensure a proper SOFC-operation. These GHSV values are not suitable for an economically feasible real scale operation since very large column volumes would be required.
The smaller lab-scale adsorption unit (max. 8 g) had been assembled in parallel, using a gas pump of the Producer Gas Analyzer (PGA) for gas suction. Both of the adsorbers and all the pipelines until the adsorbers are electrically heated in order to prevent any condensation and to adjust defined adsorption operation conditions. The lab-scale adsorption column was designed for shorter residence time for observing instantaneous change in concentration of sulfur compounds to study the kinetic behavior of the desulfurization process. Three long test runs were carried out with fresh adsorbents for each time. Conditions of these runs were shown in Table 1. Due to the technical arrangement of the test set up, gas feeding of the lab-scale adsorption column had to be interrupted, while sampling at other measurement points D point and B. That’s why gas feeding in to the adsorption column was not continuous for all the time.
Due to the very fluctuating gas conditions, it was decided to feed synthetic sulfur test gas into the real syngas before the inlet of small adsorption column (3rd test campaign). The gas mixture was prepared by mixing synthetic test gas (H2S:76 ppm COS: 2 ppm balanced N2) and real syngas to have a higher sulfur concentration at the inlet of small adsorber. The tests were operated with comparatively high GHSV (up to 40.000 hour-1). From the beginning on H2S had been dedected at the outlet of the adsorber, see Figure 2.
Figure 2 3rd Adsorption Test in Lab Scale setpoint Temperature 350 °C flowrate 0,73 and 1,1 L/min
Figure 3 3rd Adsorption Test in Lab Scale setpoint Temperature 350 °C flowrate 0,73 and 1,1 L/min
Interestingly, it can be seen from Figure 3 that COS concentration was found higher than both real syngas and sulfur gas mixture for all the time. That suggests COS production occurs in the adsorption column and possible reactions could be reaction 1, 2,3 and 4. Atakul et al.,also found that COS formed during H2S removal on dolomite, possibly due to the presence of CO or CO2. Higher H2S concentration may lead to increased reaction kinetics of CO/CO2 and H2S that produce COS.
ZnS+CO ↔COS+Zn- (1)
ZnS+CO2 ↔COS+ZnO (2)
H2S+CO ↔COS+H2 (3)
H2S+CO2 ↔COS+H2O (4)
Figure 4: Calculated COS and H2S concentration results for 22.4% CO, 10.6% CO2, 17.1% H2, 39.9% N2, 11 ppm H2S, dry conditions
Detailed results of the experiments and further findings about COS formation in dry producer gases will be published in the frame of conference contributions or a journal paper (under submission).
