(563e) Breaking Barriers: Reactive Study of a Revolutionary Turbo-Device for Defossilizing the Process Industry

Our world is facing huge challenges. Climate change, a problem created by humankind that goes far beyond national borders, requires a coordinated solution and demands a drastic change in our mindset and behavior. It is undebatable that our current society needs to develop an economic vision in which renewable energy and circularity thrive. In this respect, a crucial role is reserved for the (petro-)chemical industry, as it allows for the production of more sustainable products through changes in their processes and/or feedstocks. But before this chemical industry can become the key enabler of a sustainable society, it must become CO₂-neutral itself.

With a global energy consumption of 15 EJ of process energy annually, without considering feedstock-related energy, the chemical industry is responsible for a staggering 1.24 Gtonnes of CO₂, which is more than 18% of the total industrial greenhouse gas [1]. Even more surprising is that ~50% of this process energy consumption is related to only 7 base chemicals, which are ammonia, methanol, ethylene, propylene, benzene, toluene, and mixed xylenes. Therefore, by reinventing the production process of these 7 high-volume chemicals, there is potential to curb global greenhouse gas emissions by approximately 680 Mtonne on an annual basis, reducing global chemical and petrochemical process-related emissions by more than 54% annually [1].

The RotoDynamic Reactor™ (RDR™) developed by Coolbrook [2], offers the opportunity to drastically reshape the chemical skyline by reinventing how energy is transferred to the process fluid. Using meticulously placed turbomachinery blades, the fluid is accelerated beyond the speed of sound generating thereby shockwaves across which fluid properties like static temperature, static pressure, and density change almost instantaneously. Hence, gas heating becomes volumetric in nature rather than conductive/convective as in a conventional furnace.

To accelerate the development and widespread adoption of RDR™ within the (chemical) process industry, Computational Fluid Dynamics (CFD) simulations have been conducted. These simulations aim to provide a deeper understanding of the fluid dynamics within these groundbreaking turbo-devices. Moreover, by implementing a detailed pressure-dependent kinetic model into the CFD framework, in-depth insights are obtained about the product distribution within these devices. Pyrolytic coking rates in different zones of the reactor have for instance been analyzed, and the impact of chemistry on the reactor behavior has been studied through comparison with non-reactive results. Despite the massive potential of the RDR™ in the energy-intensive industry, this work focuses on the light-olefin industry, which is responsible for producing 5 of the 7 key building blocks - ethylene, propylene, benzene, toluene, and xylene – and therefore is the second-largest greenhouse gas emitter in the petrochemical industry.

Methodology

The complex geometry of the RDR™ is captured by utilizing a hybrid meshing approach containing both hexahedral and polyhedral cells. By leveraging the natural periodicity observed in the RDR™, only half the geometry must be modeled, which enables proper flux transfer across the domain through rotational periodic boundary conditions. In this work, a total of 18 million computational cells are used with adequate wall resolution (y⁺ < 5).

The inherent transient nature of the RDR™ is captured through unsteady Reynolds Average Navier-Stokes simulations (URANS) with sliding mesh in the commercial software tool Ansys Fluent (v22R1). Given the highly turbulent flow within this device, an Explicit Algebraic Reynolds Stress Model (EARSM) proposed by Wallin and Johansson [3] is used for turbulent closure. This extended two-equation model accounts for a non-linear relation between the Reynolds stresses and the mean strain rate and vorticity tensors, thereby accurately accounting for anisotropy in the Reynolds stress tensor and enhanced secondary flow prediction due to improved pressure recovery in cases of strong adverse pressure gradients. Such gradients are notably observed within the RDR™ [4], thereby underpinning the selection for EARSM turbulence closure.

Although it is theoretically feasible to generate a reactive model by combining a microkinetic model with a suitable non-reactive model, practical limitations arise as the computational overhead increases more than linearly with the number of species in the gas phase. Hence, three chemistry acceleration methods have been utilized to allow a first-of-its-kind reactive simulation model of the RDR™, including a-prior model reduction, in-situ adaptive tabulation (ISAT), and chemistry agglomeration. Unlike pre-tabulation, which is prohibitive due to large table dimensions, ISAT stores thermophysical compositions of the gas phase only on the relevant composition space. When an already stored thermophysical composition is encountered, the output is retrieved from the table by passing the need for computationally demanding chemistry solution steps. Chemistry agglomeration further reduces computational load by binning cells with similar compositions, averaging their properties, and performing chemistry integration with the averaged composition, thereby reducing the number of costly chemistry calls.

Results and Discussion

The unique multi-regenerative design of the RDR™ utilizes a single electrically driven impeller section through which the fluid passes multiple times when flowing from the inlet toward the outlet. Each time the flow passes through the impeller, energy is added to the flow which is then dissipated into internal energy through a series of complex shock systems. Consequently, the RDR™ can increase the temperature above 1000 °C - see Figure 1(A) - in an order of magnitude faster than conventional fossil fuel-based furnaces.

Due to the intrinsically slow chemistry of steam cracking compared to the time scale of turbulent mixing (Da << 1), these reactions benefit from significant turbulent (micro)mixing. Hence, the diffuser within the RDR™ is specifically tailored to induce large-scale flow separation which progresses into the vaneless space where these large turbulent vortical structures will homogenize the flow-conditions which should favor the occurrence of steam cracking reactions. To study this, a pressure-dependent microkinetic model for ethane cracking is implemented, in which the added enthalpy will be used to enable and sustain the chemical reactions within the RDR™. This will result in overall lower temperatures within the RDR™ (see Figure 1(B)). Through this model, yields of economically interesting components can be assessed, see Figure 2(A), and the presence of pyrolytic coke-precursing components like benzene and butadiene can be observed, see Figure 2(B). This allows for an estimation of pyrolytic coking rates in these innovative devices [5]. Although there are inherent coking challenges in steam cracking processes, the RDR™ mitigates coke deposition through two key mechanisms: (1) direct mechanical energy transfer reducing wall temperatures, and (2) supersonic gas velocities promoting high shear stress, an order of magnitude higher compared to conventional steam cracking, thus inhibiting coke accumulation on the reactor surfaces. This translates to extended operational uptime and reduced maintenance concerns.

Furthermore, through a comparison of reactive and (hypothetical) non-reactive scenarios, it becomes evident that as the fluid undergoes chemical reactions, leading to changes in properties such as average molecular weight, its compressibility decreases. Consequently, this alters its tendency to achieve supersonic velocities. Moreover, the specific heat changes under the influence of the occurring reaction, which has a direct impact on the achievable temperature increase per stage. Indeed, the work applied to the fluid at a fixed rotational speed is directly proportional to the temperature increase per stage (W = c_p ΔT). Even though chemistry significantly alters fluid properties throughout the RDR™, it displays remarkable resilience to these changing fluid conditions.

Conclusion

Addressing global challenges such as climate change demands a shift towards renewable energy and circularity, with the (petro-)chemical industry playing a pivotal role. However, before this industry can facilitate a sustainable society, it must become CO₂-neutral itself. The RDR™ by Coolbrook emerges as a transformative solution, as it reinvents energy transfer in chemical processes. Through Computational Fluid Dynamics simulations, this study achieves insights into the fluid dynamics and chemical behavior of the RDR™, particularly for light-olefin production. It is shown that this device enhances mixing, facilitating efficient steam cracking reactions while the resulting high shear stresses mitigate coke deposition. Its swift heating, above 1000 °C, which is an order of magnitude faster than current state-of-the-art, results in a highly selective production of economically interesting products like ethylene and propylene. Through comparative analysis between reactive and non-reactive simulations, it is shown that despite changes in fluid properties, the RDR™ displays remarkable resilience, underscoring its potential as a robust and sustainable technology.

References

[1] IEA, "The Future of Petrochemicals Towards more sustainable plastics and fertilisers," 2018. [Online]. Available: https://iea.blob.core.windows.net/assets/bee4ef3a-8876-4566-98cf-7a130c…

[2] Coolbrook. "Electrification Solutions." https://coolbrook.com/electrification-solutions/

[3] S. Wallin and A. V. Johansson, "An explicit algebraic Reynolds stress model for incompressible and compressible turbulent flows," Journal of Fluid Mechanics, vol. 403, pp. 89-132, 2000, doi: 10.1017/S0022112099007004.

[4] D. Rubini, N. Karefyllidis, L. Xu, B. Rosic, and H. Johannesdahl, "A new robust regenerative turbo-reactor concept for clean hydrocarbon cracking," J. Glob. Power Propuls. Soc., vol. 6, pp. 135-150, 2022 2022, doi: 10.33737/jgpps/150550.

[5] K. M. Sundaram, P. S. Van Damme, and G. F. Froment, "Coke deposition in the thermal cracking of ethane," AIChE Journal, vol. 27, no. 6, pp. 946-951, 1981/11/01 1981, doi: https://doi.org/10.1002/aic.690270610.