** This page is divided into 4 web pages due to length.

Use the navigation bar above or at the end of this section to view remainder of the document **

SYSTEM EVALUATION EXPERIMENTS

Individual processes of the system (adsorption, electrothermal regeneration, and condensation) were characterized by separate experiments. Integration tests were also conducted to demonstrate performance of the system. The experiments performed are discussed in this section.

Evaluation of Adsorption Process

Several adsorption experiments were performed to characterize the adsorption dynamics within the ACFC fixed bed. Some experiments were also performed with a packed bed of ACFC to determine adsorption capacity of ACFC for higher concentration levels than the available measured data (Cal, 1995). The packed-bed results can also be compared against the fixed-bed results to quantify the effect of packing density on the dispersion of concentration front and utilization of ACFC. Adsorption equilibrium tests using a gravimetric balance were also performed to provide an extra means for checking the fixed-bed and packed-bed adsorption capacity measured.

Adsorption breakthrough tests for acetone with ACFC fixed-bed

The experimental setup for these tests was the same as shown in Figure 3.8. The initial steps were to regenerate the ACFC fixed-bed, cool the bed to the ambient temperature, generate the sample gas, calibrate and monitor the concentration for a steady-state condition. The fixed-bed was electrothermally regenerated under a stream of UHP N_2. The regeneration step was shifted to the cooling step when no contaminant could be detected by GC/FID. Cooling was achieved by heat transfer from the walls of the adsorber to the surroundings and heat sweeping by forced convection of a low flow rate UHP N₂ gas through the fixed-bed. The gas generation system was calibrated using draw samples and the GC/MS. In the calibration steps the bed was by-passed. The first calibration point was for UHP N₂, and the second for 1% acetone in a Matheson calibrated gas standard. A linear relationship was assumed between the MS counts and the related vapor concentration values. Prior to the start of each experiment, the GC/FID was also calibrated in a similar fashion.

After the calibrations, the generated TVOC gas mixture was passed through the ACFC fixed-bed. The effluent acetone concentration from the ACFC bed was monitored both continuously by the GC-FID and intermittently by draw sample using a gas-tight syringe and injection into the GC-MS. The temperature upstream and downstream of the bed fluctuated slightly around room temperature throughout the experiment. Experimental breakthrough results describing how acetone concentration changed with time and location are presented in Figures 4.1 to 4.4. For each test, the total gas flow rate through the bed was 5 lpm at standard temperature and pressure (STP: 25EC, 101,325 Pa). Superficial gas velocity was 5.21 cm/s. The bed contained three modules of pleated ACC-5092-20 for a total mass of 27.05 g ACFC at a packing density of 94.5 mg/cm³.

Figure 4.1 shows how the concentration at each sampling port along the ACFC bed changed with respect to time. Ports A, C, and E are located in the middle of each module. Ports B, D, and F are located in the void spaces between the modules. Sampling from port D was not performed because that was the location for temperature monitoring during the initial electrothermal regeneration runs. The concentration profiles are in the form of diffusion-limited S shape. Due to similarity in the positions of ports A, C, and E, their concentration profiles show some similarity. Concentration profiles for ports B and F also show some similarities due to their positions. The absence of complete agreement between the concentration profiles is due to several factors, including the entrance transient condition, the temperature rise in the adsorption bed, mixing effects and dilution, and differing contact times and heat transfer rates.

Figure 4.2 demonstrates development and movement of MTZ inside the ACFC bed.

Behind the MTZ, the ACFC is saturated with acetone, whereas in front of MTZ, the bed is virtually free of acetone. Exponential concentration decay at initial stages of the test is observed from the concentration profile at 10 min. The concentration profile at 70 min shows the concentration distribution close to break point. It is observed that the MTZ thickness is almost equal to the length of 1.5 modules (the length of each module is 2.18 in).

Figure 4.3 shows the BTC results for four different runs. Table 4.1 summarizes the key points derived or calculated from the experimental results.

Reproducibility of these results can be seen from the presentation of these curves in a non-dimensional form (Figure 4.4). In the figure, where t is the adsorption time and t_s is the stoichiometric time.

The mass of acetone adsorbed was calculated by numerical integration of the breakthrough curve in the form of :

Where (M_ads)_t = cumulative mass of adsorption at time t, M_w = molecular weight of acetone, P = total pressure, (dv / dt)_in = acetone inlet volume flow rate, (dv / dt)_out = acetone outlet volume flow rate. In terms of mole fraction, the above equations can be written in the form of:

Where C is the volume or mole fraction of acetone, and Q_N2 is the N₂ flow rate. The specific length of unused bed (LUB) was calculated as:

where M_b = M_ads at break point, and M_sat = M_ads at saturation. The throughput ratio (TPR), the ratio of actual effluent volume to the feed volume that could ideally saturate the fixed-bed, is calculated as the ratio of elapsed run time to breakthrough to stoichiometric time (t_s).

In our experiments, the throughput ratio of the bed was approximately 78% for a 10,000 ppmv acetone in N₂ gas mixture.

For modeling purposes, adsorption breakthrough experiments were performed in a packed-bed of ACFC to determine the adsorption capacities in the range of 1,000 to 10,000 ppmv. Additionally the packed-bed results for 10,000 ppmv were used to research the effect of packing density on the dispersion of MTZ. These tests are described in the next section.

Adsorption Breakthrough Tests for Acetone with the ACFC Packed-Bed

Adsorption breakthrough tests were performed with a packed bed of ACFC to determine the adsorption capacity of ACFC for concentration levels up to 10,000 ppmv. A schematic diagram of the experimental setup given in Figure 4.5. A stream of 1% by volume acetone gas (Matheson Inc. calibrated standard) was diluted with a stream of UHP N₂ to provide a desired concentration of acetone. The flow rates of standard gas and dilution gas were controlled by two Tylan mass flow controllers with 500 and 200 sccm ranges. The prepared gas stream can be passed through the packed-bed or through the by-pass line for calibration purposes. The temperature of the bed was controlled at a near isothermal (NIT) condition by circulating water in a cooling coil that surrounds the bed. Water was supplied from an automatic temperature-controlled water bath unit. The temperature of the bed was monitored by a type K thermocouple connected with a digital readout. The concentration of the gas mixture upstream and downstream of the bed were measured by the GC/FID. Packed-bed configuration was cylindrical with an internal diameter of 1.1cm. Figure 4.6 shows the results of breakthrough tests for 10,000, 5,000, and 3,333 ppmv acetone.

Figure 4.5 Packed bed experimental set-up

Figure 4.6 Acetone breakthough curves for packed-bed experiments, M_c is the mass of carbon.

 

To be able to compare the adsorption dynamic behavior of the packed bed with the ACFC fixed bed, the superficial gas velocity for the 10,000 ppmv packed-bed test was set at the same velocity as the fixed bed (5.2 cm/s). The ratio of the ACFC weight in the packed-bed and in the fixed-bed was equal to the ratio of their gas flow rates.

A summary of the experimental conditions and test results is given in Table 4.2. Figure 4.7 shows the BTCs in terms of time normalized by the stoichiometric time. The form of the BTCs illustrate that higher concentration levels and flow rates provide less dispersion and shorter mass transfer zones.

 

Table 4.2 Summary of acetone adsorption breakthrough tests with ACFC packed-bed.
Test Number	1	2	3
Inlet gas concentration (ppmv)	3,333	5,000	10,000
Ambient temperature (EC)	24 " 1	24 " 1	24 " 1
Mass of ACFC (g)	0.62	0.62	1.6
Total flow rate (slpm)	0.3	0.4	0.297
Stoichiometric time (t50%) (min)	90	45.4	98
Mass adsorbed (mg)	203	231	694
Adsorption capacity (mg/g)	327	372	434

Figure 4.7 Normalized acetone breakthrough curves for packed-bed experiments.

The adsorption capacity for 10,000 ppmv acetone found from packed-bed tests was 512 mg/g while the average adsorption capacity from the fixed-bed experiments was 434 mg/g. To check these results, the adsorption capacity of ACC-5092-20 was measured by a gravimetric method using the same experimental setup used by Cal (1993, 1995). The equilibrium adsorption capacity of acetone for 10,040 ppmv concentration was determined by mass gain of an ACC-5092-20 sample that is exposed to a Matheson calibrated standard acetone gas in a Cahn gravimetric balance (Cahn Model C-2000). From the adsorption data, equilibrium capacity was determined to be 504 mg/g ACFC. From the desorption data, total adsorption capacity was measured to be 489 mg/g ACFC.

Figure 4.8 shows the mass gain sample as a function of time. Two distinct mechanisms are likely to be responsible for total adsorption capacity, one fast and one slow mechanism. The slow mechanism might be due to capillary condensation in the transitional pores between supermicropore and mesopore regions or diffusion into ultra micropores.

Figure 4.8 Mass gain and loss of ACC-5092-20 using gravimetric techniques.

The equilibrium isotherm for adsorption and desorption of acetone and ACC-5092-20 can be classified as Brunauer, Deming, Deming, and Teller (BDDT) type IV isotherm. The equilibrium micropore adsorption and capillary condensation capacities of the ACFC for 1% by volume acetone were measured from the mass gain of the fast mechanism and the mass gain of the slow mechanism to be 401 mg/g and 93.6 mg/g, respectively.

Effect of packing density on BTC

The effect of packing density on adsorption dynamics is presented in Figure 4.9. Smaller packing density results in a shorter breakthrough time and a larger MTZ due to axial dispersion and mixing effects. Increasing packing density from 94.5 mg/cm³ to 450 mg/cm³ increases the breakthrough time from 75 min to 91.7 min. TPR was increased from 78% to 94%. While these are clearly benefits, increased packing density also produces a larger (and more costly) pressure drop. It is possible, however, to find a packing density for optimization of adsorption and pressure drop.

 

Figure 4.9 Effect of packing density of the characteristics of BTC of acetone adsorbed by ACFC and comparison with BTC of a commercially available activated carbon

To compare adsorption dynamics for the ACFC and a commercially available activated carbon, breakthrough tests were performed with ACFC and then with Calgon BPL 12x30 mesh activated carbon in packed bed configuration (Figure 4.9). To keep conditions similar, the superficial gas velocity through the bed was kept at 5.2 cm/sec for each test. The ratio of adsorbent mass to sample gas flow rate was also kept constant for the tests. The packed bed dynamic adsorption capacity results for the Calgon BPL and the ACFC were determined to be 241 mg/g and 434 mg/g, respectively. The adsorption capacity of the ACFC was almost twice of the adsorption capacity of the BPL. Breakthrough times for the ACFC and BPL adsorbents were 91.7 min and 47.2 min, respectively. The measured adsorption capacity for the BPL sample was in good agreement with modeled values (236 mg/g); using data by McCabe et al. (1993) for a 1% by volume acetone stream at ambient conditions.

Most of the difference in breakthrough times for the two adsorbents is due to the differences in adsorption capacity. The ACFC does demonstrate a sharper slope at the leading edge of the MTZ suggesting faster mass transfer of the TVOC to the ACFC. At higher flow velocities this rate difference is expected to be amplified.

Adsorption Breakthrough Tests for MEK with the ACFC Fixed-Bed

The experimental setup and conditions for the MEK tests were the same as those for acetone experiments described in previous sections, except that the gas generation system was calibrated for 5,000 ppmv MEK. Figure 4.10 shows the resulting MEK BTCs. The BTCs in normalized form are given in Figure 4.11. The form of the normalized BTCs provides a suitable means for checking the performance of adsorber. For example, the curves with open symbols demonstrate a higher longitudinal dispersion rate than the other tests. This observed dispersion was likely due to a small channeling inside the bed in one case, and the room temperature variability in the other case, as will be discussed later. Channeling inside the bed occurs if the nuts and bolts of the adsorber are not tightened completely. The close overlap of the normalized MEK BTCs in the concentration range of 5,000 to 15,000 ppmv shows that dispersion due to axial molecular diffusion is not an important factor in this concentration range.

Figure 4.10 Adsorption breakthrough curves for MEK experiments.

Table 4.3 summarizes the conditions of experiments and key points derived or calculated from the experimental results. The TPRs for the three properly performed tests are greater than 80%. TPRs greater than 90% can be achieved by increasing the bed depth. Figures 4.12 to 4.14 show the normalized temperature variation versus normalized time for three different runs. Positions of thermocouples are shown in Figure 3.10. Thermocouple tips, for ports B, D, and G, are located at the middle of bed cross section and after each ACFC module. A, C, and E thermocouples are located close to the wall and in the middle of ACFC modules. At the locations where adsorption takes place, the heat of adsorption is released and results in a rise of temperature. Movement of the concentration wave generates a thermal wave inside the adsorber. For a normal run, the maximum temperature measured at port B is higher than those at the other ports, because it is located in the region of minimum bed length (MBL) and away from the entrance region. Temperatures measured at ports A, C, and E are lower than B, D, and G because they are close to the wall and heat transfer is faster.

Figure 4.11 Normalized breakthrough curves for MEK experiments

The occurrence of maximum temperature at point D and the high noise seen on the temperature curves that is seen in Figure 4.12 is indicative of channeling in the bed. Channeling shifts the location of maximum temperatures toward the end of the fixed-bed. This temperature shift and the mixing effects from channeling disperse the BTC more than its normal form (as seen in the normalized concentration curves). Dispersion effects on the BTC of test number 4 are not likely from channeling. This can be concluded from temperature data for this run as presented in Figure 4.13. On the day of the experiment, the ambient laboratory temperature changed from a starting temperature of 27EC to an ending temperature of 30.5EC. This problem contributes to the dispersion of the BTC by reduction of adsorption capacity, decrease in throughput ratio, and increase in length of unused bed. A well behaved temperature profile, as is seen in Figure 4.14 and Table 4.3, accompanies a less dispersive breakthrough curve, shorter MTZ and LUB, as well as a higher TPR.

Figure 4.12 Temperature history of adsorption breakthrough experiment for 5,300 ppmv. Letters identify sampling adsorption bed sampling ports in Figure 3.8.

Figure 4.13 Temperature history of adsorption breakthrough experiment for 5,012 ppmv MEK. Letters identify sampling adsorption bed sampling ports in Figure 3.12.

Figure 4.14 Temperature history of adsorption breakthrough experiment for 5,250 ppmv MEK.. Letters identify sampling adsorption bed sampling ports in Figure 3.12.

Click Here to Continue