You are here:
REGENERATION AND REACTIVATION OF CARBON ADSORBENTS BY RADIO FREQUENCY INDUCTION HEATING
Impact/Purpose:
Carbonaceous adsorbents are used in many and diverse industrial processes to absorb organic solvent vapors from liquid and gas streams. When saturated they are regenerated by steam heating or by a heated stripping gas at moderate temperatures. After several saturation/regeneration cycles the adsorbent must be re-activated at high temperatures. Hot gas regeneration methods result in large volumes of relatively concentrated desorbed organic contaminant; yet the contaminant is still diluted in a large volume of stripping substance and the off-gas is usually incinerated. Steam regeneration with a low volume of stripping gas results in concentrated condensed solvent (which may be reused); but steam is only capable of moderate temperatures in practical systems. Electromagnetic heating with microwaves or radio frequency fields can accomplish both regeneration at moderate temperatures and re-activation at high temperatures. Simple and efficient collection of the desorbed organic solvent vapors is possible by condensation during regeneration, and the condensed solvent can be reused in the original process. The relatively high electrical conductivity of activated carbon particles often makes them poor candidates for microwave heating, however.
We will study radio frequency (RF) induction-coupled heating in a cylindrical adsorbent bed geometry, as is typically-used in industry. We propose a program of experimental studies at both laboratory and industrial scale and analytical numerical models to study and optimize RF induction adsorbent regeneration parameters in these important materials. We will study the feasibility of re-activation with RF induction. We will work closely with two industrial partners; Sorbent Control Technologies Inc., Elgin IL and 3M Austin Center, Austin IX.
Description:
1. Electrical Properties of Adsorbents: We measured the electric permittivity of four commercially available carbon adsorbents (supplied by Wesvaco Inc) over the radio frequency range (1 to 40 MHz). Westvaco is by far the largest volume supplier of activated carbon adsorbents, and the four types had quite different electrical properties: BX-7540 and CX-HC40 are semiconductors and WV-B and BAX-1500 are lossy dielectrics. Semiconductors are best heated with induction applicators, as was described in the project proposal. Lossy dielectrics are best heated by electric field applicators. Consequently, the project plan was modified to include experiments and numerical model studies of both types. 2. Develop and Calibrate Improved Measurement Devices: We designed, constructed and calibrated a capacitive differential voltage divider to measure the voltage applied to both E-field capacitive plate electrodes and induction coils. The inductance of the coils was measured prior to experiments (loaded and unloaded) and used to calculate inductor current. Other instrumentation included fiber optic temperature sensors, purge gas flow meters and a sensitive mass balance to determine water loss rates. 3. Design Coils for More Uniform RF Heating: Induction coil designs were evaluated by numerical models using a commercial High Frequency Finite Element Modeling program (Ansoft, Inc.). The effect of operating RF frequency and coil dimensions on the heating pattern was determined over the ISM frequencies: 6.78, 13.56 and 27.12 MHz. Each model required approximately 20 hours to solve on a Dell Precision 420 Pentium III computer. 4. Small Scale Experimental Studies: Experiments were conducted at 27.12 MHz for powers between 150 and 300 W between capacitive plates and in RF induction coils. Transient temperatures and carbon mass measurements were collected for comparison to model studies. Applied power was estimated for measurements of applicator voltage. The experiments showed an interesting phenomenon, apparently unique to RF heating: at low coverage, the rate of water elimination from the carbon depended on RF power density alone, not on local temperature. This surprising observation suggests that it might be possible to use low level RF heating to prevent water from collecting in pores while allowing non-polar VOCs to deposit, thus extending the useful life and/or the cycle time of the adsorbent. 5. Numerical Model Analytical Studies: Two numerical models were constructed to predict transient temperatures and water loss during RF regeneration: one for the capacitive plate E-field applicator and the other for the induction coils. The electric field in the capacitive plate applicator is nearly uniform and relatively simple to estimate. The electric field induced by the induction coil is very difficult to calculate, and the FEM model program was used to estimate the RF heating field. Anempirical fit was used to include a direct RF water ejection process, as suggested by the experiments (see 4. above). The numerical model results compare well with the experimental results in both applicators. 6. Perform Larger Scale Experiments:This objective could not be achieved in the time and with the resources available. The original project plan was substantially altered to accommodate suggestions from the SAC and to include E-field heating as well as the original induction heating investigation. The inclusion of E-field heating was made necessary by the electrical property measurement results (see 1. above).