Supercritical Fluid Dynamics

Fluid dynamics of supercritical fluids for process and energy applications

Researchers:

 

 

 

  • Ir. John Harinck
  •  

     

     

     

     

     

  • Dr. ir. Piero Colonna
  •  

     

     

     

     

     

  • Ing. Teus van der Stelt
  •  

     

     

     

     

     

  • Dr. Alberto Guardone (Visiting scholar - Politecnico di Milano)
  •  

     

     

     

     

     

  • Dr. ir. Joop ter Horst
  •  

     

     

    Project manager:

     

     

     

  • Dr. ir. Piero Colonna
  •  

     

     

    Funding:

     

     

     

  • Delft Centre for Sustainable Industrial Processes (SIP)
  •  

     

     

    Background

    Computational fluid dynamics (CFD) has become a trusted and valuable design tool in many engineering applications that involve fluid flow. Currently, CFD methods rely on simple (ideal) gas models used for the thermodynamic modeling of the flow. The use of CFD is therefore limited to fluid flows in easy to handle thermodynamic states like the ideal gas state, which is valid only under a limited range of conditions of pressure and temperature: generally at low pressures and high temperatures. At thermodynamic states close to the critical point or close to the saturated vapor line, the gas behavior is far from ideal. This region is called the real gas, dense gas or close-to-critical region. Commercial CFD software (based on a simple ideal gas model) fail to correctly simulate fluid flows in this dense gas region.


    The supercritical region in the phase diagram

    There are promising engineering applications in the process and energy field which can or would benefit from fluid flows in the dense gas region. In this region several peculiar phenomena occur: on the vapor side the compressibility is very high, while on the liquid side the vapor behaves liquid-like. Moreover on the vapor side, the caloric and transport properties of fluids have a very different characteristic with respect to the low pressure states. In some applications involving dense gas flows, a purely empirical approach is taken while for other potential applications no advances are made because of the lack of suitable simulation tools.


    Supercritical Fluid (SCF) expansion and Organic Rankine Cycle (ORC) expansions in a schematic pressure-volume diagram. Also indicated is the region where the ideal gas law is valid.

    It is clear that in both cases, validated CFD software could provide a fundamental help in either obtaining a deeper understanding of the physical phenomena and/or improving the process itself. In the energy conversion field, processes which can involve transformations in the dense gas region are: supercritical steam power cycles, Organic Rankine Cycles, Stirling Cycles, carbon dioxide (CO2) refrigeration cycles and CO2 power cycles. In the process industry field, dense gas flows occur in supercritical extraction, supercritical dyeing processes and supercritical particle technology using CO2 as the process medium. The volatile compound carbon dioxide is non-toxic, inexpensive and easily separated, purified and reused. Supercritical carbon dioxide is a particularly good solvent for hydrophobic components, like many active pharmaceutical ingredients. The solvent properties together with the gas-like fluid properties such as a low viscosity and high diffusion coefficients ensure a rapid nucleation upon expansion over a nozzle. Volatile compounds like carbon dioxide can therefore be used to develop green and sustainable (nano)particle production processes. All these processes would benefit from a greater understanding of dense gas fluid dynamics and from a modeling and simulation tool capable of predicting the peculiar behavior of fluids in that thermodynamic region.

    The group of P. Colonna recently developed zFlow, an advanced CFD code which can model inviscid flows of dense gases. It fills a demand for a fluid dynamic code capable of simulating the flow of a broad range of fluids in conditions that are currently difficult or impossible to model. The code is the result of the merging of a sophisticated CFD code developed in Italy by S. Rebay et al. with a wealth of thermodynamic library developed or further developed by P. Colonna et al.

    Objectives

    The project aims at providing a state-of-the-art CFD simulation tool for dense gas (non-ideal) fluid dynamics that is of much interest in the energy and process industries.
    The principal objectives are:

     

     

     

  • Characterize and understand real-gas effects in dense-gas flows and supercritical CO2 flows.
  •  

     

     

     

     

     

  • Experimental validation of the code for nonideal flows.
  •  

     

     

     

     

     

  • Further development of the zFlow CFD code in order to make it a tool suitable for non-ideal process simulations with particular emphasis on supercritical CO2 fluid flows and dense gas flows.
  •  

     

     

    Work programme

    In order for a CFD code to be effectively used as a tool to gain physical insights in a complex process or for it to be used directly as a design tool, it must be capable of treating 3-dimensional problems involving boundary layer effects and turbulence. Moreover the process to be studied can be stationary or dynamic. In order to reach these goals, the zFlow numerical scheme will be extended to the solution of Reynolds Averaged Navier-Stokes Equations (RANS) and an engineering turbulence model (for instance k-e) will be included. To this purpose a transport property model capable of treating the dense gas region has to be included for several of the fluids libraries available to zFlow. Another useful improvement will be the capability to treat axisymmetric problems in a computational efficient way. To accomplish objectives 3 to 6, the fluid library should be enriched with the calculation of several thermodynamic functions for multi-phase and multicomponent fluids. These are already available for single-phase pure fluids and partly for mixtures.

    The CFD software will be validated against experimental data with two test cases. Within the scope of this project an experimental setup will be designed in the P&E laboratory, in order to produce flows through a cm to mm scale nozzle from the high pressure vessel to the low pressure vessel. Another source of experimental data to validate zFlow will come from ORC turbine manufacturers, namely Triogen b.v. in the Netherlands, Turboden s.r.l. in Italy and Carrier Inc. in the U.S., who already agreed to provide geometry, operating conditions and flow and pressure measurements for their plants.

    Upon expanding or compressing a dense fluid, the phase boundary between liquid and gas may be encountered. Upon crossing this phase boundary the fluid becomes metastable (supersaturated) and nucleation of the gas (bubble formation, boiling) or liquid (droplet formation, condensation) can occur. Similarly upon expanding a solution the solubility may decrease so that a supersaturated solution is obtained and nucleation of a solid or liquid phase may occur. For nucleation processes an energy barrier exists which decreases in a highly non-linear way with increasing supersaturation. It is therefore extremely important to accurately know supersaturation profiles upon e.g., expansion over a nozzle. These supersaturation profiles are the first step towards understanding of nucleation processes in dense fluids. Models to describe the kinetics of this phenomena should be developed and coupled with the code.

    Utilization

    The zFlow code in its final development state will be able to correctly describe the flow of a large variety of fluids even under non-ideal conditions (supercritical, close-to-critical, close to the saturation vapor line, etc).

    Owing to this unique capability, it will allow for the design of new and improved turbomachinery cascades under highly non-ideal flow conditions. Some examples are high pressure turbomachinery cascades, Organic Rankine Cycle turbines (including BZT fluid turbines), compressors and expanders in (transcritical) refrigeration cycles, compressors and expanders for the chemical and petrochemical industry. After a successful validation of zFlow, it will finally be able to correctly evaluate of the mass flow and of the losses in such fixed and rotating cascades. The systematic application of the program in the design stage will increase the load per stage, reduce losses and thus reduce manufacturing and operational costs.

    In process industry research a large effort is put into the formation of (nano)particle products using supercritical fluid technology. The development of the zFlow code is the first step towards a systematic and fundamental understanding of the supersaturation generated upon expansion of a supercritical fluid solution. The supersaturation profile upon expansion determines the nucleation process and therefore the (nano)particle product quality (crystal size distribution, polymorph). After validation the zFlow code therefore can be used to optimize product quality by optimizing supersaturation profiles, nozzle design and process design. Within the framework of this research, zFlow would also be used to improve the fluid dynamic design of other applications for which STW projects are on-going in the field of supercritical CO2 processes. A close cooperation with Prof. Witkamp and dr. ir. Joop Ter Horst of the Process Equipment section is planned: zFlow will be used to study novel processes in which close-to-critical CO2 is employed to produce crystals during a rapid expansion.

    Project partners:

     

     

     

  • Prof. G. Witkamp, Dr. Ir. J.H. ter Horst, Process Equipment Section, TU Delft, The Netherlands
  •  

     

     

     

     

     

  • Prof. S. Rebay, Mech. Eng. Dept.- Univ. of Brescia, Italy
  •  

     

     

     

     

     

  • Turboden s.r.l., Italy
  •  

     

     

     

     

     

  • Tri-O-gen B.V.
  •  

     

     

     

     

     

  • Carrier Corp. (UTCPower)
  •  

     

     

     

     

     

  • FeyeCon B.V.
  •  

     

     

     

     

     

  • Fluid Dynamics Section at TU Delft, The Netherlands
  •  

     

     

    Publications:

     

     

     

  • P. Colonna, J. Harinck, S. Rebay, and A. Guardone, ``Real-gas effects in organic Rankine cycle turbine nozzles,'' J. Propul. Power, vol. 24, pp. 282-294, March-April 2008.
  •  

     

     

     

     

     

  • Harinck, J., Turunen-Saaresti, T., and Colonna, P., Predictions of performance and flow fields of a highexpansion-ratio radial orc turbine. Technical report ET-2262, Delft University of Technology, Process & Energy department, Leeghwaterstraat 44, 2628 CA Delft, The Netherlands, July, 2007.
  •  

     

     

     

     

     

  • P. Colonna, S. Rebay, J. Harinck and A. Guardone, Real-Gas Effects in ORC Turbine Flow Simulations: Influence of Thermodynamic Models on Flow Fields and Performance Parameters. In: Proceedings of the European Conference on Computational Fluid Dynamics 2006, Egmond aan Zee, The Netherlands, September 2006.
  •  

     

     

     

     

     

  • P. Colonna, A. Guardone, J. Harinck, and S. Rebay, Numerical Investigation of Dense Gas Effects in Turbine Cascades, In: Proceedings of the 15th U.S. National Congress on Theoretical and Applied Mechanics, Boulder (CO), US, June 2006.
  •  

     

    Chair:

    Involved People:

    Facilities used: