Development of a numerical tool for simulating dust explosions

The current standards regarding dust explosion venting protecting system are based on conservative empirical models and neglecting complex geometry, which may lead to failure in estimation of explosion overpressure and therefore risk for fatalities at work places. Therefore, there is an urgent need for a reliable tool for vent protection design in the process plants.

The project aims are

• development of high-fidelity and well-validated models which address important combustion phenomena during a dust explosion such as flame expansion, turbulence generation by a flame and flame acceleration,
• development of an efficient numerical tool based on an open source toolbox for predicting consequences of dust explosions,
• simulation of dust explosions in scenarios of process industries in cooperation with the reference group members of this project.

The result of this project will be of great importance for creating a safer working environment regarding dust explosion risks in Swedish process industries. The reason is that the result of this project will improve the understanding of complicated combustion phenomena associated with dust explosions, and help the process industries in designing better vent system in case of dust explosion.

Project reports

Project part report 1

Project part report 2 

Project final report

Publications

Huang, C., Lipatnikov, A.N., Nessvi, K., 2020. Unsteady 3-D RANS simulations of dust explosion in a fan stirred explosion vessel using an open source code. J. Loss Prev. Process Ind. 67, 104237.

Huang, C., Bloching, M., Lipatnikov, A.N., (2022). A vented corn starch dust explosion in an 11.5 m3 vessel: Experimental and numerical study. J. Loss Prev. Process Ind. 75, 104707.

Popular science article: Open-source dust explosion research

Dust explosion is a continuous threat to the industries worldwide especially among countries and regions with high industrial output. The reported combustible dust explosion incidents during 2016 and 2020 lie at an almost constant level, being around 30 per year in the USA according to the Combustible Dust Incident Report 2020. Unfortunately, the situation is similar in Sweden, where the dust explosion incidents occurred repeatedly in different industries. Moreover, the dark numbers are significant since only a small portion of the incidents were reported. Therefore, dust explosions represent a big threat to the physical working environment of the employees in the process industry.

Under this circumstance, RISE and Chalmers received funding from AFA Försäkring for a three-year research project. The project aim is to develop physics-based, well-verified and well-validated models and an open-source numerical tool for studying dust explosion consequences. “It has been an exciting journey during these three years!” said Chen Huang, researcher at RISE. This project is a good example of transferring fundamental research into practical usage. The dust explosion model comes from a long-term fundamental research work by Andrei Lipatnikov research professor at Chalmers and his colleagues. The model reflects the key physics in a dust explosion. That is the flame speed increases with the increase of turbulence level; the flame speed increases with the increase of the flame size. At the same time, the dust explosion model was implemented into the open-source code/toolbox OpenFOAM within the framework of Computational Fluid Dynamics. The project result is the open-source code, FSCDustFoam, based on OpenFOAM. FSCDustFoam has been systematically verified, validated, and applied for large-scale industrial dust explosions. FSC is the abbreviation of Flame Speed Closure, which is the name of the combustion model. Moreover, this project received valuable large-scale dust explosion experimental data from IND EX^® (Intercontinental Association of Experts for Industrial Explosion Protection) for model calibration. Furthermore, a project reference group members consisting of experts, researchers, consults, energy company and pellets producer have been supporting this project during the three-year period with valuable industrial experience. They are gratefully acknowledged. Chen answers some of the frequently asked questions as follows.

For what can we use the FSCDustFoam?

The model and code are especially useful when combined with reliable experimental data. The numerical tool can be used to understand the dust explosion process, to design effective test program, to apply test results to a wider scenario. Apart from that, the tool is especially useful in scenarios where the standards for explosion protection are not valid, e.g., explosions in complicated geometries.

Who can use the open-source code FSCDustFoam?

The short answer is everyone since it is open source. At the same time, there are two groups who are most likely to use this code. The first user group is the postgraduate (master and PhD) students and the researchers in the combustion, explosion and fire community. The tool has no user interface or user guide, which makes the learning curve steeper as compared to a commercial code with user interface and dedicated support. The second user group is the risk consultants who have experience in the open-source code OpenFOAM.

What are the biggest advantages of the FSCDustFoam as compared to the current modelling program?

One of the biggest advantages is the open source, which means that the users have the greatest freedom to use, to change and to distribute the code. The path towards a predictive numerical tool for dust explosion is a tremendous task, which cannot be solved by a single organization. The current FSCDustFoam is one step towards a predictive numerical tool since it enables collaborations within the whole community. On the other side, it may take more personal hours to work on a project using FSCDustFoam as compared to a commercial code since it lacks user interphase, user guide and support.

Where can we download the FSCDustFoam?

The source code including solvers, libraries and tutorials are available on the Zenodo platform for sharing open research (https://doi.org/10.5281/zenodo.5795161).