16:00 〜 18:00
[HDS08-P02] Risk Assessment of chemical release due to the tsunami triggered by the Nankai megathrust earthquake and measures to prevent the accident
キーワード：リスク評価、化学物質の流出、南海トラフ巨大地震、津波
It is estimated that there is a 70-80% probability that a magnitude 8-9 earthquake will occur in the Nankai Trough within 30 years. If a Nankai megathrust earthquake occurs, large tsunamis of over 10 m in height are expected to hit a wide area on the Pacific Ocean side from the Kanto region to the Kyushu region. In chemical factories, there is a risk that a tsunami has the tanks move or roll over, resulting in the chemical release. Risk assessment of storage tanks to floods has been done in previous studies (Yang et al., 2020), but that to tsunamis has not been done yet. The [IL1] purpose of this study is to evaluate the measures to prevent chemical release by the tsunami caused by the Nankai megathrust earthquake and to assess the human health effects this chemical accident happens.
Konohana-Ku, Osaka City, Osaka Prefecture was chosen as the target area, which is expected to be flooded by the tsunami, and a large amount of chemicals is handled or stored. As the target chemical substance, we chose toluene because it is handled in large quantities nationwide and is harmful to humans.
At first, we assessed the effectiveness of measures to prevent chemicals from releasing out of the tanks. Two measures were verified; i) [KR2] keeping the storage rate of the tank high, and ii) raising the storage floor of the tanks. We calculated the probability of chemical release from the rolled-down tank due to a flood by tsunami using Yang’s model (2020) to check the effectiveness of the measures. In this model, the probability of chemical release can be obtained from six parameters; diameter, height, and storage rate of the tank, flood depth, the flow velocity of the tsunami, and chemical density. The diameter and height of the tank were fixed at 15 m and 12 m, respectively. The flood depth and the flow velocity were set to 5 m (Cabinet Office, 2012) and 7.2 m/s (Osaka City, 2021), respectively, assuming the tsunami strikes Osaka Prefecture in the worst-case scenario. As for calculating the effectiveness of raising the storage floor of the tank, the flood depth was lowered, relatively. Sensitivity analysis was performed on the probability of chemical release using the storage amount in the tank and the height of the floor of tank as parameters.
As a result, it was found that raising the floor of the tank by 1.5 m would reduce the release probability to less than 10% at a storage rate of 0.7 or more. On the other hand, if the floor of the tank is not raised, the release probability was more than 99% at a storage ratio of 0.7, which means that the outflow probability was more than 10 times as high as for the same value of storage ratio. Furthermore, when the storage ratio was 0.4, the outflow probability was 80%, even if a tank was raised by 3 m. However, when the storage ratio was kept at 0.8, the outflow probability could be reduced to less than 10% only by introducing a tank by 0.5 m. Based on these results, it is adequate to implement both measures simultaneously.
Next, we assessed risk of a chemical release from a tank due to a tsunami, focusing on the human health impacts. In this study, we assumed that the release of toluene would occur from the chemical factory that handles the largest amount of toluene in Konohana-Ku. We calculated for two cases: full release and half release of the storage volume. The toluene released from the tank was assumed to diffuse on the still water surface as the waves of tsunami recedes. Fay's Oil Diffusion Model (Fay, 1971) was used to calculate the area where toluene diffuses on a still water surface. In this model, the maximum diffusion area can be calculated from the amount of spilled oil, assuming that the spilled oil diffuses in a circular pattern. The maximum diffusion area of toluene in this study was a circle with a radius of 504 m in the case of full release and a circle with a radius of 388 m in the case of half release. Therefore, we assumed that all of the released toluene would volatilize at a distance of 504 m and 388 m from the source in the case of full and half release, respectively. The Areal Locations of Hazardous Atmospheres (ALOHA) Version 5.4.7 (US EPA, 2016) was used to analyze the atmospheric diffusion of toluene, which could be inhaled by humans. The Acute Exposure Guideline Level (AEGL) (US EPA, 2020) was applied as the acute toxicity influence index, and the magnitude of damage was evaluated by the number of people exposed. In the case of a half release the number of exposed people was about half that in the case of a full release. In the latter case, the damage extended beyond Konohana-Ku and reached to the large-scale theme park "USJ". To minimize this damage, the measures such as adjusting proper storage rate or raising the floor of tanks are valid.
Konohana-Ku, Osaka City, Osaka Prefecture was chosen as the target area, which is expected to be flooded by the tsunami, and a large amount of chemicals is handled or stored. As the target chemical substance, we chose toluene because it is handled in large quantities nationwide and is harmful to humans.
At first, we assessed the effectiveness of measures to prevent chemicals from releasing out of the tanks. Two measures were verified; i) [KR2] keeping the storage rate of the tank high, and ii) raising the storage floor of the tanks. We calculated the probability of chemical release from the rolled-down tank due to a flood by tsunami using Yang’s model (2020) to check the effectiveness of the measures. In this model, the probability of chemical release can be obtained from six parameters; diameter, height, and storage rate of the tank, flood depth, the flow velocity of the tsunami, and chemical density. The diameter and height of the tank were fixed at 15 m and 12 m, respectively. The flood depth and the flow velocity were set to 5 m (Cabinet Office, 2012) and 7.2 m/s (Osaka City, 2021), respectively, assuming the tsunami strikes Osaka Prefecture in the worst-case scenario. As for calculating the effectiveness of raising the storage floor of the tank, the flood depth was lowered, relatively. Sensitivity analysis was performed on the probability of chemical release using the storage amount in the tank and the height of the floor of tank as parameters.
As a result, it was found that raising the floor of the tank by 1.5 m would reduce the release probability to less than 10% at a storage rate of 0.7 or more. On the other hand, if the floor of the tank is not raised, the release probability was more than 99% at a storage ratio of 0.7, which means that the outflow probability was more than 10 times as high as for the same value of storage ratio. Furthermore, when the storage ratio was 0.4, the outflow probability was 80%, even if a tank was raised by 3 m. However, when the storage ratio was kept at 0.8, the outflow probability could be reduced to less than 10% only by introducing a tank by 0.5 m. Based on these results, it is adequate to implement both measures simultaneously.
Next, we assessed risk of a chemical release from a tank due to a tsunami, focusing on the human health impacts. In this study, we assumed that the release of toluene would occur from the chemical factory that handles the largest amount of toluene in Konohana-Ku. We calculated for two cases: full release and half release of the storage volume. The toluene released from the tank was assumed to diffuse on the still water surface as the waves of tsunami recedes. Fay's Oil Diffusion Model (Fay, 1971) was used to calculate the area where toluene diffuses on a still water surface. In this model, the maximum diffusion area can be calculated from the amount of spilled oil, assuming that the spilled oil diffuses in a circular pattern. The maximum diffusion area of toluene in this study was a circle with a radius of 504 m in the case of full release and a circle with a radius of 388 m in the case of half release. Therefore, we assumed that all of the released toluene would volatilize at a distance of 504 m and 388 m from the source in the case of full and half release, respectively. The Areal Locations of Hazardous Atmospheres (ALOHA) Version 5.4.7 (US EPA, 2016) was used to analyze the atmospheric diffusion of toluene, which could be inhaled by humans. The Acute Exposure Guideline Level (AEGL) (US EPA, 2020) was applied as the acute toxicity influence index, and the magnitude of damage was evaluated by the number of people exposed. In the case of a half release the number of exposed people was about half that in the case of a full release. In the latter case, the damage extended beyond Konohana-Ku and reached to the large-scale theme park "USJ". To minimize this damage, the measures such as adjusting proper storage rate or raising the floor of tanks are valid.