Abowerbung
Home » Using a Hierarchical Approach for the Dilution of Gases

Using a Hierarchical Approach for the Dilution of Gases

In order to ensure safe underground working conditions, especially in case of gas outbursts, it is essential to develop ventilation concepts for the necessary dilution of harmful gases with respect to the occurring gases, gas concentrations and general operational conditions as to cross-sectional areas and existing ventilation systems. Therefore, a concept for a hierarchical approach for the dilution of gases was developed, combining Computational Fluid Dynamics (CFD) and Ventilation Network modeling. For the two-phase gas/air flow that occurs during a gas burst, a CFD simulation was performed and the behavior of the gas burst and its impact on the existing ventilation were analysed. Based on the simulation, various measures for the dilution of air dependent upon the severity of the gas burst as well as the geometrical conditions were tested and evaluated according to their efficiency. An integration of the simulation results into the existing superior ventilation network calculations was then performed with the Mine Ventilation Simulation Software VentsimTM, a tool that provides compressible ventilation system calculations and is capable of delivering the requirements for the development of quick, effective and safe ventilation concepts.

Authors:
Dr.-Ing. Elisabeth Clausen und M.Tech. Amit Agasty, Institut für Bergbau, Technische Universität (TU) Clausthal, Clausthal-Zellerfeld

1  Problem Definition

In order to provide safe working conditions underground it is essential to develop and evaluate suitable ventilation concepts and strategies for the necessary dilution of harmful gases. The reasons and sources for gases in underground could be diverse, e. g. gas emissions due to blasting, the usage of combustion engines or sudden release of gas during an outburst. Though most outbursts in the world are associated with methane, gas outbursts associated with carbon dioxide “are more violent, more difficult to control and more dangerous because of the greater sorption capacity for carbon dioxide” (1). In Germany carbon dioxide could occur in salt deposits due to volcanic activities resulting in an inclusion of carbon dioxide within the crystal lattice, leading to potentially sudden releases (2). Based on the amount of gas released the required airflow for a proper dilution can be determined. This will be more complicated in case of specific areas with complex geometry, which, depending on gas characteristics, could lead to an accumulation of gases in floor depressions or underneath the roof and needs to be integrated in ventilation network calculation. As ventilation network calculations need to make several assumptions, such as perfect mixing of gases or one dimensional flow pattern in the airways a combination with CFD modeling, taking into consideration i. a. geometrical structures as well as gas characteristics, though only possible for spatially limited areas, may be beneficial.

This paper will present a hierarchical approach for the development of ventilation strategies in case of the occurrence of gases and after gas outburst by generating and evaluating dilution strategies for specific, spatially limited areas whilst integrating the results into the overall ventilation network model. The methodology, models and its interaction will be explained and applied to an exemplary case study. In the conclusion, the suitability of the approach will be assessed and an outlook into current advancements and future work be given.

2  Methodology

The methodology proposed is based on a hierarchical and iterative approach taking into consideration the mine layout and overall ventilation situation on the one hand, and individual requirements in the network related to complex geometrical structures and specific gas characteristics on the other hand.

Fig. 1. Hierarchical approach for the dilution of gases. // Bild 1. Hierarchischer Ansatz für die Verdünnung von auftretenden schädlichen Gasen.

Fig. 1. Hierarchical approach for the dilution of gases. // Bild 1. Hierarchischer Ansatz für die Verdünnung von auftretenden schädlichen Gasen.

The primary stage using Ventilation Network Analysis is applied for simulating the airflow through the entire mine network, allowing for ventilation design through airflow as well as pollutant dispersion and comprising following elements and aspects (Figure 1, left):

  • mine layout and ventilation situation;
  • demonstration and analysis of the concentration distribution within the ventilation network under the assumption that natural buoyancies remain unmodified.

Such a simulation, however, makes various simplifications such as perfect mixing of gases and flow pattern in the airways as being one dimensional, among others.

Therefore, at the secondary stage a CFD model is applied, which is capable of simulating multiphase dynamic flows for complex geometries with respect to the specific gas characteristics and their respective interactions. Nevertheless, such a simulation is based on geometry discretization and the quality of the results depends on the convergence and boundary condition specifications. Furthermore, due to computation capacity and costs, CFD modelling is not feasible for complete mine network. In the hierarchical approach, following elements and aspects will be considered and evaluated for individual mine areas using CFD modeling (Figure 1, right):

  • concentration distribution where natural buoyancy is taken into consideration;
  • demonstration of the concentration distribution for specific sections, where the geometrical structure, specific behavior of the gases and their interactions are considered and
  • simulation and evaluation of ventilation concepts and strategies with regards to gas dilution for specific areas.

Coupling the individual solutions through the hierarchical and iterative approach proposed assures that the independent results are coherent with each other, so that

  • ventilation concepts developed at the secondary stage could be simulated and evaluated with regards to gas dilution for the entire network in the primary stage as well as
  • ventilation parameters could be modified within the overall ventilation and evaluated with regards to the efficiency of gas dilution, plausibility and viability.

3  Model Explanation

The Ventilation Network Analysis was conducted using the software VentsimTM from the Australian company Chasm Consulting. VentsimTM is an underground mine ventilation simulation software package designed to model and simulate ventilation, airflows, pressures, heat, gases, financials, radon, fire and many other types of ventilation data from a model of tunnels and shafts (3). Applying an iterative estimation method referred to as Hardy Cross Method, this software allows the calculation of airflows in a model by progressively adjusting the airflow values until estimation errors lie within acceptable limits (4). Simulation in VentsimTM can perform airflow distribution and contaminant dispersion studies within a ventilation network, which is described as a definite structure of branches and nodes, defined with specific properties and connectivity. The purpose of using this tool for the hierarchical approach proposed is to be able to identify the distribution of gas after an outburst as well as to identify further dispersion resulting from targeted dilution from complex geometrical structures (Figure 2).

Fig. 2. Airflow and contaminant simulation in VentsimTM. // Bild 2. Nachbildung der Volumenstromverteilung und Schadstoffausbreitung in VentsimTM.

Fig. 2. Airflow and contaminant simulation in VentsimTM. // Bild 2. Nachbildung der Volumenstromverteilung und Schadstoffausbreitung in VentsimTM.

The gas dispersion analysis at this primary stage with VentsimTM is performed using Contaminant Simulation tool of the software. This steady state simulation tool allows the user to trace different types of contaminants, such as gases, dust, and smoke among others, through a mine network. By distributing the contaminants linearly in airways and considering perfect mixing at nodes, the software solves an otherwise much more difficult problem of multiphase flows with relative ease (4). However, it is important to note that is kind of assumption is only valid for high abstraction level studies, with consideration of large networks. A study requiring diffusion of one gas relative to the other in a 3 dimensional airway cannot be sufficiently addressed by this method.

VentsimTM provides users with two different tools for simulating contaminant spread in a network: Steady State Simulation and Dynamic Simulation. The former method allows for modeling a spread of contaminant in the mine network based on the location of source and prevalent airflow. The simulation works until the distribution of contaminant in the network has reached an equilibrium concentration. In case of the latter method a time dependent spread of contaminant can be simulated and additionally allows the user to pause the calculations at any time and make changes in the network, which are taken into account for further calculations.

A gas source can be defined by changing the properties of an airway with contaminant concentration, either as percent initial concentration or in any other volumetric units. The simulated concentrations in the rest of the network pertain to the initial concentration at source diluted as a result of perfect mixing at junctions. In a dynamic contaminant simulation, the gas source can be defined with gas concentration for an airway as a function of time, for example fixed rate release, linear or logarithmic decay.

In addition, VentsimTM provides simultaneous plausibility checks of important conditions within the ventilation network, e. g. plausibility check for the entire ventilation, dispersion of gases in the network and dilution as well as check for any resulting recirculation.

The modeling of the 3-Dimensional Gas Dispersion was conducted using the CFD software ANSYS. ANSYS CFD is a software package that uses numerical techniques and algorithms to solve and analyze fluid flow problems (5). ANSYS CFD codes basically solve the fluid flow problem, mathematically expressed by the Navier-Stokes equations, by additional simplifications and constraints put on by boundary conditions. The flow domain itself is divided into smaller volume elements and the fluid flow equations are solved iteratively for each of them. Another issue while simulating industrial flows is modeling of turbulence, which is solved in ANSYS by defining separate equations for turbulence related variables along with averaged Navier-Stokes equations.

A CFD simulation can incorporate multiple phases and components and solve for dispersion effects based on natural buoyancy and induced flows. A CFD code can be used to perform flow simulations in regular or irregular 3-dimensional geometrical domains as well as in a time-dependent manner. Two main methodologies exist for solving such multiphase problems, depending on the phase configurations. In case of dispersed flows, with particulate components in continuous medium, the Lagrangian method is used, whereas for multiple continuous fluids simultaneously existing in a medium, Eulerien method is used (6).

In case of the hierarchical approach proposed for effective dilution of gas after outburst, CFD simulation is implemented to model complex geometrical flow domain. Transient and multiphase Eulerien flow simulations are performed based on the input conditions obtained from primary stage of solution approach, from where gas volume and airflow in the gallery can be determined. CFD simulations are used to test different ventilation strategies for dilution of gas related to complex geometrical structure.

4  Example Situation

Fig. 3. Ventilation network for the example situation. // Bild 3. Wetternetzmodell für die exemplarische Betriebssituation.

Fig. 3. Ventilation network for the example situation. // Bild 3. Wetternetzmodell für die exemplarische Betriebssituation.

An example situation is presented in order to demonstrate the process and effectiveness of the proposed hierarchical approach for the dilution of gas after an outburst. Figure 3 represents the ventilation network, with airflow distribution among different working regions. The example situation consists of two vertical shafts, intake and upcast, connecting different working areas of the mine between them. The mine network specifically consists of two major working districts at one level and additional areas where the development of headings for further working areas is carried out. The two main districts are connected to the intake fresh air shaft directly and get supplied with the air in a parallel fashion. The remaining advancing workings are supplied with fresh air as splits from these districts. The ventilation in the network for a normal situation is provided by an exhaust fan located at the top of the upcast shaft`s ventilation drive. The connection between the upcast shaft and rest of the mine network is through a cross cut drive at the bottom level which splits into the upper level. In a normal situation, all the exhaust air flows through the upper level split and out of the upcast shaft.

Fig. 4 Input geometry of the trough structure and meshing for CFD model. // Bild 4. Eingangsgrößen für die Nachbildung der geometrischen Struktur sowie Vernetzung innerhalb des CFD-Modells.

Fig. 4 Input geometry of the trough structure and meshing for CFD model. // Bild 4. Eingangsgrößen für die Nachbildung der geometrischen Struktur sowie Vernetzung innerhalb des CFD-Modells.

As a result of the above network analysis a critical location can be identified, where the gas gets accumulated after the occurrence of a gas outburst. Based on this information, the geometry and the airflow conditions at the location of concern, a basic CFD mesh is constructed as shown in Figure 4. The constructed geometry represents one symmetric half of the mine airway of concern, along its axis, in order to reduce the computational cost. The total length of airway with depression is 80 m and the depression depth is 12 m. The constructed geometry of the flow domain is discretized with unstructured tetrahedral mesh, which is then exported to the solver of ANSYS CFD code.

For the solution of the problem, a scenario with CFD a multiphase Euler-Euler mixture model is used to simulate air and carbon dioxide in the flow domain and to represent the fluid interactions and momentum transfer. A homogenous k-epsilon turbulence model is used for both of the fluids. Interphase momentum transfer is implemented with the help of drag coefficient. By enabling gravity for the flow domain body forces are accounted for, which act against the drag forces exerted by the momentum of air and thus resist motion of carbon dioxide. Only a strong enough drag between the two phases would be capable of causing diffusion and thus dilution of heavier carbon dioxide from the depression shaped airway.

Fig. 5. Boundaries for the CFD model of the geometrical depression – flow domain. // Bild 5. Randbedigungen für die CFD-Modellierung der Senke – Strömungsbereich.

Fig. 5. Boundaries for the CFD model of the geometrical depression – flow domain. // Bild 5. Randbedigungen für die CFD-Modellierung der Senke – Strömungsbereich.

To simulate drag forces, different ventilation scenarios were modeled in transient manner, representing normal operation and operation of a booster fan to dilute the gas out of the depression. The applied boundary types can be seen in Figure 5. The intake for the domain is velocity inlet boundary condition; different velocities represent distinct mass flow rates and thus simulate normal airflow situation and situation with booster fan application. The outlet boundary is defined with average pressure and thus, together with a rough wall boundary condition, defines a specific pressure drop in the domain resulting from flow of air. The transient simulation represents a time after the gas outburst has taken place and gas has accumulated at the bottom of the depression, on the basis of which the domain is initialized with a representative 1000 m3 of carbon dioxide gas.

Fig. 6. Volume fraction contours for gas (a) initial condition; (b) normal flow after 4 minutes; (c) booster dilution after 4 minutes. // Bild 6. Volumenanteil CO2 (a) Ausgangssituation, (b) reguläre Strömungsverhältnisse nach 4 Minuten, (c) Strömungsverhältnisse bei Verwendung eines Zusatzlüfters nach 4 Minuten.

Fig. 6. Volume fraction contours for gas (a) initial condition; (b) normal flow after 4 minutes; (c) booster dilution after 4 minutes. // Bild 6. Volumenanteil CO2 (a) Ausgangssituation, (b) reguläre Strömungsverhältnisse nach 4 Minuten, (c) Strömungsverhältnisse bei Verwendung eines Zusatzlüfters nach 4 Minuten.

The simulated scenarios provide an in-depth understanding of the flow in the depression region as a result of normal flow situation and a clear comparison is possible based on simulation of flow with booster fan installation. Figure 6 represents a volume fraction contour for CO2 gas at initial condition and in case of normal as well as booster flow after four minutes of dilution. The transient simulation of the normal flow situation shows that the prevailing air velocities remain insufficient to disperse the gas out of the depression. On the other hand, a stronger air flow, resulting from installation of booster fan in the airway, causes the complete dispersion of gas within four minutes of ventilation. This is exported as a result from secondary CFD modeling stage of the hierarchical model back to the primary stage.

Fig. 7. Ventilation network modified by booster fan installation. // Bild 7. Modifiziertes Wetternetzmodell bei Verwendung eines Zusatzlüfters.

Fig. 7. Ventilation network modified by booster fan installation. // Bild 7. Modifiziertes Wetternetzmodell bei Verwendung eines Zusatzlüfters.

In order to check implementability of the solution proposed by CFD simulation, a modified network simulation is performed at the primary stage of the hierarchical approach as a back-coupling. Figure 7 and 9 present the modified ventilation situation in the network along with the critical depression zone location and gas monitoring locations subject to the following scenarios considered:

  • installation of booster fan to improve gas dispersion and
  • installation of booster fan in combination with mine door.
Fig. 8. Airflow in (a) original case and (b) booster fan scenario. // Bild 8. Volumenstromverteilung im (a) ursprünglichen Fall und bei (b) Verwendung eines Zusatzlüfters.

Fig. 8. Airflow in (a) original case and (b) booster fan scenario. // Bild 8. Volumenstromverteilung im (a) ursprünglichen Fall und bei (b) Verwendung eines Zusatzlüfters.

Booster fan switching results in effective dilution of gases from the source airway as proved by CFD simulation. However, booster fan operation results in airflow changes in the network, resulting in greater spread and slower overall dispersion out of the mine (Figure 8). The CFD solution is then modified in VN stage, by introducing mine doors along with booster fan operation as can be seen in Figure 9. The idea behind this modification is to provide the gas dispersed out of the critical airway a shorter path out of the mine network, and thus faster and effective dilution.

Fig. 9. Modified solution for gas dispersion in the mine network. // Bild 9. Modifizierte Gasverteilung innerhalb des Wetternetzmodells bei Verwendung eines Zusatzlüfters sowie von Wettertüren.

Fig. 9. Modified solution for gas dispersion in the mine network. // Bild 9. Modifizierte Gasverteilung innerhalb des Wetternetzmodells bei Verwendung eines Zusatzlüfters sowie von Wettertüren.

The modified solution shows that the booster fan operation results in improved dilution of gas out of the depression and activation of mine doors results in directed and faster removal of gases from the mine. By means of the activation of mine doors, the recirculation and thus slower dilution of gases is avoided.

The solution procedure for this example scenario with the application of hierarchical approach demonstrates a step by step addressal of the problem and utilization of the available modeling capabilities to the best of their abilities as follows:

  1. i) Primary stage ventilation network analysis: With the know-ledge of mine and geometrical structures along with a gas dispersion simulation at the primary stage after the gas outburst, the critical location is identified where the carbon dioxide gas accumulates.
  2. ii) Secondary stage CFD simulation: The results and information gained from the primary stage is feed into the CFD software. A representative geometry and boundary conditions are designed and solved for two-phase gas dilution. Different ventilation scenarios are modeled and a booster implementation is chosen on the basis of its effectiveness for gas dilution.

iii) Feedback to primary stage: The selected ventilation scenario is used to modify the original ventilation network to test the plausibility of its application and further implication on the entire network. A further adjustment to the solution is performed, in which the mine doors are activated along with booster application to induce faster and more effective gas dispersion out of the mine.

5  Conclusion and Outlook

This paper presented a hierarchical approach for the effective dilution of gases, esp. after gas outbursts. The methodology proposed comprises of two individual models – Ventilation Network Calculations and CFD Modeling – taking advantage of the strength of each individual simulation tool.

Summarized, the functions of the Ventilation Software VentsimTM within the hierarchical approach are as follows:

  • compressible ventilation network calculation program;
  • demonstration, modification and analysis of ventilation situations;
  • aid in the development of ventilation concepts through the internal analysis of plausibility of the ventilation network and objectives;
  • capable of simulating paths and concentrations of contaminants (e. g. gases) for planning or emergency situations;
  • assumption: distribution of contaminants in linear velocity fashion and perfect mixing at the junctions;
  • steady state and dynamic simulation resulting from the changing ventilation conditions during the contaminant spread can be performed;
  • reference frame: entire ventilation network.

Summarized, the functions of the CFD modeling, using Ansys, within the hierarchical approach are as follows:

  • two-phase flow simulation of the behavior of an air/gas mixture after a gas burst;
  • analysis and evaluation of the required ventilation output through variations in air flow for the dilution;
  • determination of parameter values to be integrated into -VentsimTM;
  • reference frame: specific (spatially limited) area.

It could be demonstrated that the hierarchical approach proposed is suitable for supporting the development of ventilation concepts in case of occurrence of gases and gas outbursts, respectively. However, the implementation and effectiveness of the approach depends significantly on the amount and quality of data provided. Current advancements deal with the enhancement of the approach and models for including various, partly dynamic, gas sources for the development of ventilation-on-demand concepts.

References / Quellenverzeichnis

References / Quellenverzeichnis

(1) Lama, R., Saghafi, A.: Overview of Gas Outbursts and Unusual Emissions. Coal Operators Conference, 6 – 8th February 2002, pp. 74 – 88.

(2) Marschall, V.: Schlechte Luft unter Tage? – IPA untersucht Kohlen-dioxidbelastung in Kalibergbau: IPA-Journal 03/2011, S. 26-29.

(3) Chasm Consulting (a), www.ventsim.com, 2016.

(4) Chasm Consulting (b), “Ventsim VisualTM User Guide”, Version 4.0, QLD Australia, 2016.

(5) Ansys Inc. “ANSYS Academic Documentation”, Release 13.0, November 2010.

(6) Höhne, T.: CFD simulations for single and multi-phase flows. FZD Theory Seminar Series, Institute of Safety Research, Dresden-Rossendorf/Deutschland, März 2010 (Präsentation).

Authors:
Dr.-Ing. Elisabeth Clausen und M.Tech. Amit Agasty, Institut für Bergbau, Technische Universität (TU) Clausthal, Clausthal-Zellerfeld