PHENOMENA OF DISPERSION AND EXPLOSION OF HIGH PRESSURIZED HYDROGEN

To make “Hydrogen vehicles and refueling station systems” fit for public use, research on hydrogen safety and designing mitigative measures are significant. For compact storage, it is planned to store under high pressure (40MPa) at the refueling stations, so that the safety for the handling of high-pressurized hydrogen is essential. This paper describes the experimental investigation on the hypothetical dispersion and explosion of high-pressurized hydrogen gas which leaks through a large scale break in piping and blows down to atmosphere. At first, we investigated time history of distribution of gas concentration in order to comprehend the behavior of the dispersion of high-pressurized hydrogen gas before explosion experiments. The explosion experiments were carried out with changing the time of ignition after the start of dispersion. Hydrogen gas with the initial pressure of 40MPa was released through a nozzle of 10mm diameter. Through these experiments, it was clarified that the explosion power depends not only on the concentration and volume of hydrogen/air pre-mixture, but also on the turbulence characteristics before ignition. To clarify the explosion mechanism, the numerical computer simulation about the same experimental conditions was performed. The initial conditions such as hydrogen distribution and turbulent characteristics were given by the results of the atmospheric diffusion simulation. By the verification of these experiments, the results of CFD were fully improved. 1.0 INTRODUCTION The objective of the research described here was to acquire data to be used in the reevaluation of standards in conjunction with the introduction of hydrogen refueling stations for fuel cell powered vehicles, thus serving to early practical application and popularization of fuel cells in the context of a hydrogen energy society for the future. Storage and filling of hydrogen at a high pressure of 40MPa is planned for the hydrogen refueling stations under current consideration, and the acquisition of basic data on high-pressurized hydrogen is an urgent task. After ascertaining the properties corresponding to high-pressurized hydrogen, it will be necessary to introduce appropriate safety measures, and to review various standards and regulations so as to ensure safe operation[1][2]. The research presented here accordingly envisions various accidents associated with the storage and usage of a hydrogen refueling station, considering leakage, dispersion, and explosion behavior in the respective cases. Representative cases involve either a pinhole occurring in equipment, resulting in continuous leakage at a constant flow volume (steady leakage), or rupture of the piping connected to the storage tanks, leading to a major leakage in a short time period (unsteady, large-scale leakage). Centered on these two cases, outdoor tests were conducted to ascertain the effects (concentration and blast pressure) of these scenarios. Results are reported below for experimental results and numerical simulation of major leakage of high-pressurized hydrogen gas occurring due to rupture of a 40MPa storage tank or the piping connected to the dispenser. 2.0 DISPERSION EXPERIMENTS OF HIGH-PRESSURIZED HYDROGEN 2.1 Experimental Apparatus and Test Field Hydrogen refueling stations will store high-pressurized hydrogen gas at 40MPa, and outdoor diffusionexplosion tests were conducted at the Tashiro testing facility of Mitsubishi Heavy Industries in Akita Prefecture, Japan in order to ascertain the effect of blast pressure after the leakage of high-pressurized hydrogen. The two types of conditions assumed were large-scale (unsteady) leakage from a rupture of the 10mm diameter piping connected to the high pressure storage tanks, and slight (steady) leakage from a pinhole (0.2mm~2mm in diameter). To simulate an accident, hydrogen gas was first stored in five high pressure tanks, each with a capacity of 50L, via a pressure booster. The tanks were connected to a 10mm diameter release nozzle by means of 25mm diameter piping, and a valve was fitted to the end of the piping immediately ahead of the nozzle for control of gas release. Considering the pressure loss between the high pressure tanks and the nozzle, the pressure of the hydrogen gas in the tanks was set at 65MPa to ensure a nozzle pressure of 40MPa. Assuming a large-scale leakage accident involving the rupture of 10mm diameter piping, high-pressurized hydrogen gas (40MPa) was released into the atmosphere by blow-down. 2.2 New Measurement Device for the Monitoring of varying Hydrogen Gas Concentration Because it is considered that ignition timing and ignition location have great influence on the results of diffusion-explosion tests, it is important to ascertain the time-wise changes in the behavior of leaked gas. Therefore we conducted diffusion tests in advance of the diffusion-explosion tests, in order to acquire sufficient data on a time history of spatial concentration distributions for consideration of the ignition timing and ignition location with respect to explosions in large-scale, unstable leakage. However, it is difficult to measure precise hydrogen gas concentration under the unsteady state such that the distribution of dispersed hydrogen gas changes with time, because the response time of conventional gas sensors is very slow, for example 90% response time is about 7 to 8 seconds. In order to measure more precise gas concentration, we developed a new measurement device which consists of 10 gas sensors[3]. The schematic system of the new measurement device is shown in Figure 1. Figure 1. Schematic Illustration of Measuring System of Concentration fluctuation measurement device It was named as concentration fluctuation measurement device, which was fabricated as shown in Figure 2 (i.e. an external view during assembly). In Figure 1, the sampling gas is led to each gas sensor every 1second by orderly switching each flow pass which connected to the corresponding sensor. Since the sampling gas led into the sensor remains around its sensor for 10 seconds, the slow response of the sensor will be insignificant factor in precise measurement of hydrogen gas concentration. Concentration measurement, at least along the vertical cross-section, was undertaken every 1 second by using 15 units of these concentration fluctuation measurement devices. Next, in order to confirm the measurement precision of the developed device, methane gas (which can be measured rapidly) was employed, and a concentration fluctuation measurement unit was fitted with a methane gas concentration sensor (hot wire semi-conductor type with response time of 7~8s)[4]. As illustrated in Figure 3, methane gas at 2000ppm was released from the gas release nozzle A while engaging in suitable manual left-right movement, and the methane gas was sampled by means of the sampling pipe (suction tube) and the sampling port B at the same time. Concentration of the sampled gas from the sampling pipe was analyzed at 100Hz by fast response flame ionization detector HFR-400, produced by CAMBUSTION, and concentration from the sampling port B was done every 1 second by concentration fluctuation measurement device. Accordingly, in order to match with the average time of 1 second for the concentration fluctuation measurement device, data acquired at 100Hz by HFR-400 were averaged for 1 second. Figure 4 shows a comparison between values measured using the HFR400, including two types of data consisting of 100Hz and 1s average) and using the concentration fluctuation measurement device. It can be seen that the values for the concentration fluctuation measurement device are in good agreement with those for the HFR400, indicating that this device can be practically applied for the direct ascertainment of time-wise changes in the concentration of hydrogen gas. For reference, Figure 5 presents a comparison between actual measurements of time-wise changes in concentration using a conventional concentration sensor (thermal conductivity of gases type KD-3A produced by New Cosmos Electric Company) and the concentration fluctuation measurement device. Measurement using both types of equipment was performed simultaneously at essentially the same location. Although it was physically impossible to place both pieces of equipment at exactly the same location, the Figure 2. Concentration fluctuation measurement device Unit:mm 124 262