Open Access

Bei der Freisetzung von Wasserstoff aus Ausbläsern kann nicht ausgeschlossen werden, dass es zu einer Selbstentzündung und der Bildung einer Freistrahlflamme kommt. Als Auswirkungen für die Umgebung sind der Explosionsüberdruck bei der Zündung und die Wärmestrahlung im Rahmen einer Gefährdungsbeurteilung zu betrachten. In der Literatur sind verschiedene Ansätze für die Modellierung einer Wasserstoff-Freistrahlflamme zu finden, wobei viele sich mit der Berechnung der Flammenlänge befassen. Experimentelle Freifeld-Untersuchungen unter Windeinfluss und entsprechende Modellierungen der Wärmestrahlung in der Umgebung einer Wasserstoff-Freistrahlflamme liegen nur wenig vor. In der Bundesanstalt für Materialforschung und -prüfung BAM sind solche Experimente durchgeführt worden und die daraus abgeleitete Modellierung wird im Folgenden vorgestellt.

For a comprehensive safety assessment of stationary lithium-ion-battery applications, it is necessary to better understand the consequences of thermal runaway (TR). In this study, experimental tests comprising twelve TR experiments including four single-cell tests, two cell stack tests and six second-life module tests (2.65 kW h and 6.85 kW h) with an NMC-cathode under similar initial conditions were conducted. The temperature (direct at cells/modules and in near field), mass loss, cell/module voltage, and qualitative vent gas composition (Fourier transform infrared (FTIR) and diode laser spectroscopy (DLS) for HF) were measured. The results of the tests showed that the battery TR is accompanied by severe and in some cases violent chemical reactions. In most cases, TR was not accompanied by pregassing of the modules. Jet flames up to a length of 5 m and fragment throwing to distances to more than 30 m were detected. The TR of the tested modules was accompanied by significant mass loss of up to 82%. The maximum HF concentration measured was 76 ppm, whereby the measured HF concentrations in the module tests were not necessarily higher than that in the cell stack tests. Subsequently, an explosion of the released vent gas occurred in one of the tests, resulting in the intensification of the negative consequences. According to the evaluation of the gas measurements with regard to toxicity base on the “Acute Exposure Guideline Levels” (AEGL), there is some concern with regards to CO, which may be equally as important to consider as the release of HF.

For industrial applications dealing with hydrogen, the definition of safety distances and the assessment of possible hazards emanating from releases is mandatory. Since hydrogen is usually stored and transported under pressure, one scenario to be considered is the momentum driven release of hydrogen from a leakage with subsequent ignition. In this scenario, the emitted heat radiation from the resulting jet flame to the surroundings has to be determined to define adequate safety distances. For hydrocarbon flames, different jet flame models are available to assess the hazards resulting from an ignited jet release. Since hydrogen flames differ from hydrocarbon flames in their combustion behavior, it has to be checked if these models are also applicable for hydrogen. To evaluate the accuracy of these models for hydrogen jet flames, tests at real-scale are carried out at the BAM Test Site for Technical Safety (BAM-TTS). Herein, the flame geometry and the heat radiation at defined locations in the surroundings are recorded for varying release parameters such as leakage diameter (currently up to 30 mm), release pressure (currently up to max. 250 bar) and mass flow (up to max. 0.5 kg/s). The challenge here is the characterization of the flame geometry in an open environment and its impact on the thermal radiation. Existing heat radiation data from the literature are mostly based on unsteady outflow conditions. For a better comparability with the steady state jet flame models, the experiments presented here are focused on ensuring a constant mass flow over the release duration to obtain a (quasi) stationary jet flame. In addition, stationary outflow tests with hydrocarbons (methane) were also carried out, which are intended to serve as reference tests for checking flame models based on hydrocarbon data.

In order to assess the thermal radiation emanating from hydrogen jet flames, experiments under real scale conditions were carried out at the Test Site Technical Safety of BAM. Herein, the behavior of inclined hydrogen jet flames was investigated. The aim of the work is to determine the Surface Emissive Power and radiant heat fraction of these flames and to provide a reliable dataset for model evaluation purposes. Since the aforementioned values are not directly measurable, the incident heat radiation was measured at defined distances from the flame, as well as the flame's shape and size. The required values were then derived from these measurements. The hydrogen releases ranged from 0.005 kg/s to 0.175 kg/s with a 30 mm orifice. The mass flows were held constant during the releases, nevertheless a transient behavior of the flame could be observed since the experiments were carried out under open field conditions, with unsteady wind fields. In the literature, the flame lengths are often determined using visible light imaging, either by injecting coloring substances in the low light emitting hydrogen jet flame or by carrying out the measurements in darkness. In this work the jet flames were visualized using infrared (IR) and OH* imaging. The recorded flame shapes and resulting flame lengths are compared. Results from this showed that the flame lengths determined with OH* and IR recordings differ greatly. A flame length ratio l_(f OH*)/l_(f IR) in the range of 0.47-0.62 can be found. In addition, the SEP differ also in the range of 10 kW/m²-16 kW/m² (IR) and 40 kW/m²-80 kW/m² (OH*) for hydrogen jet flames due to differences in the determined flame surface. Conclusions regarding the determined x_(RAD )values for IR and OH* result in approximately the same range of 0.031-0.043.

In order to reduce the human footprint of CO2 emissions and limit global warming effects hydrogen combustion is becoming increasingly important. To enable fuel cells and gas turbines to operates this carbon free fuel, unprecedently large amounts of hydrogen need to be produced and safely transported and stored. The investigation of the effects of accidents involving hydrogen is therefore becoming of outmost importance. Since hydrogen is usually stored and transported under pressure, one scenario to be considered is the release of hydrogen from a leakage with subsequent ignition. The resulting jet flame must be characterized with respect to the thermal radiation emitted into the environment to define safety regulations. Various models that characterize the resulting flame shape and radiation already exist in the literature, but these are mainly based on empirical data from hydrocarbon jet flames.[1-4] To verify these models, a H2 Jet Flame project conducted at BAM, is investigating the safety of momentum driven hydrogen jet flames. For this purpose, large-scale tests are carried out at the Test Site Technical Safety (BAM-TTS). The object of the investigations is to assess the effects of real scale release scenarios regarding flame geometry and the thermal radiation emitted. Parameters such as release angle, leakage diameter (currently 1 mm to 10 mm), pressure (currently up to max. 250 bar) and mass flow (up to max. 0.5 kg/s) are varied. In addition, influences such as the type of ignition, ignition location as well as delayed ignition can also be investigated. The gained knowledge will be compared with existing jet flame models, to validate these and identify a possible need for further development. In particular, the focus will be laid on the thermal radiation of hydrogen flames. The challenge here is the visualization and characterization of the flame geometry in an open environment. Visualization is performed using infrared (IR) camera systems from at least two viewing angles. Measurements of the heat radiation of jet flames, which can be found in the literature, are mostly based on unsteady outflow conditions.The experimental setup used here allows for the generation of a steady-state outflow for several minutes and thus a direct comparability with existing (steady-state) models. Furthermore, the tests can be carried out for comparative measurements with hydrocarbons (methane, etc.) as well as mixtures of hydrogen and hydrocarbons.