In addition to advancing the industry’s sustainability, the electrification of chemical processes can offer substantial performance benefits. Electrothermal reactors allow flexibility in reactor design and geometry. They can also provide a greater degree of control over the reaction compared to traditional combustion-fired reactors, leading to improved selectivity and catalyst preservation. Targeted and volumetric catalyst heating, especially from electromagnetically driven heating methods, can lower waste heat and facilitate heterogeneous catalysis.
Electrified chemical reactors can be powered by resistive (Joule) heating, microwave heating, induction heating, or by plasma. Induction heating has been highlighted as a potentially highly effective and scalable solution to power thermocatalytic reactors. Induction heating uses an alternating magnetic field to generate eddy currents in conductive media and, simultaneously, hysteresis losses in ferromagnetic media. Induction furnaces have long been employed for the heating and processing of metals such as steel, and systems in these applications can reach impressively large throughput and scale.
The scalability and wealth of industrial experience tied to induction heating in the steel industry are encouraging to its potential success as a power source for electrified chemical reactors. However, the academic research into induction heating for chemicals production has generally been constrained to small scales or select reactions, providing limited insight into broad design considerations and scalability challenges. In addition, a fluidized bed reactor, composed of fluidized catalyst particles that are heated directly by electromagnetic induction has never been reported in the context of chemicals production.
In this work, we apply induction heating to directly heat a fluidized thermocatalytic reactor for the first time, demonstrating the system’s effectiveness and potential for scalability by performing methane pyrolysis for electrified, clean hydrogen production. Five different reactor configurations are explored to provide insight over a broad range of processes and conditions: 1) quartz tubular reactor filled with carbon particles, 2) quartz tubular reactor filled with iron particles, 3) steel tubular reactor filled with iron particles, 4) quartz tubular reactor filled with carbon particles and an iron susceptor along the reactor centerline, and 5) steel tubular reactor filled with carbon particles and an iron susceptor along the centreline. The induction system was purchased from UltraFlex Power Technologies (UPT n5) and consisted of a water-cooled copper coil (7 turns, 5 cm diameter, and 7 cm height) wrapped around the tubular reactors and connected to an alternating current power supply. The system operates at 300 kHz and can provide up to 5 kW of power. All the tubular reactors are 4 cm in diameter and filled with about 100 g of catalyst particles, with insulation placed in the gap between the reactor and the induction coil. Bed temperature is measured via a 1/16” thermocouple while surface temperature is measured via an infrared pyrometer.
First, the results reveal that carbon particles (activated carbon, petcoke, and graphite) in configuration 5 are able to be heated to reaction temperature (1000 °C) when the bed is packed but not when the bed is fluidized. The constant particle motion in the fluidized bed disrupts the conductive network between particles and necessitates that each carbon particle be heated individually by the alternating magnetic field. However, the carbon particles are not conductive enough to be heated individually when the bed is fluidized.
In configuration 2 (iron particles, quartz reactor), the alternating magnetic field from the induction coil causes strong electromagnetic cohesive forces between the iron particles, causing particle clustering and severely inhibiting fluidization. These cohesive forces increase the minimum fluidization velocity by a factor of 8 compared to when no magnetic field is present. Operating under a packed bed causes rapid and reactor-wide sintering of the iron particles, preventing the progress of the reactions.
To obtain fluidization without increasing the inlet flow beyond the case with no magnetic field, we pulse the inlet power to the coil. In this pulsed field setup, short periods of “field-on” allow the bed to be heated while the following periods of “field-off” allow temporary fluidization. This scheme proves effective in minimizing particle sintering while simultaneously allowing the reactor to reach the high temperatures required for methane pyrolysis (>700 °C). The effects of pulsing amplitude and frequency are explored, showing a wide range of conditions that were effective. The power input when the field is on is on the order of 500 W. This compares favorably to other forms of electromagnetic heating. For example, our previous microwave-driven setup consumed >1000 W and heated about half the mass of catalyst as this induction heater (50 g to 100 g, respectively).
The axial and radial temperature gradients in reactor configurations 2-5 are then mapped and compared in detail. The system shows high degree of effectiveness in heating iron and steel materials. Sharp temperature gradients are observed along the axial dimension, especially approaching the axial center. The radial temperature gradient and power absorption in the configurations with steel reactors are investigated and compared to the quartz configurations. Steel, a reactor material highly relevant to industry, is heated effectively via electromagnetic induction, implying important consequences on power absorption and reactor temperature. In addition, the ability of the iron susceptors to raise the overall bed temperature and their impact on fluidization is explored.
Finally, methane pyrolysis is performed to understand the system’s ability to carry out chemical processes. Methane pyrolysis requires high temperature (>700 °C) and is known to have severe coking, making it an illustrative example. The methane conversion rate and hydrogen yield are tracked over time for different reactor temperatures. The methane conversion rate is shown to be highly sensitive to the peak reactor temperature, which rises sharply near the axial center. Unlike in certain microwave-heated setups, the electromagnetic penetration is not sensitive to coke deposition on the reactor walls, allowing controlled operation for long periods of time.
In conclusion, we demonstrated for the first time a fluidized bed thermocatalytic reactor system where the catalyst particles are directly heated by electromagnetic induction. The application of induction heating for chemical production at this reactor scale (4 cm diameter and 10 cm height) enabled novel observations that can become highly insightful when extrapolated to an industrial scale. The induction system showed highly effective heating of ferromagnetic catalysts and susceptors. However, the cohesive forces on ferromagnetic particles caused by the magnetic field severely inhibited fluidization, and packed bed configurations exhibited rapid particle sintering. To allow effective heating but avoid sintering, the power to the induction coil must be pulsed, giving alternating periods of heating and fluidization. Reactor temperature and behavior for a variety of configurations were examined in detail, highlighting the system’s broad applicability. Commentary was provided on the effect of industrially relevant materials such as steel reactor tubes. The system was successfully demonstrated in high-temperature methane pyrolysis to produce clean hydrogen and showed efficient hydrogen production capability. The development of fluidized bed induction-heated chemical reactors at this scale has provided landmark data and performance characteristics. This work represents a significant push towards the large-scale use of induction heating to decarbonize process heat in the chemicals production industry.