Proceedings of the Eighth International Conference on Probabilistic Safety Assessment & Management (PSAM)
5 Oil Spill Risk Assessment for Loviisa Power Plant (PSAM-0388)
Download citation file:
- Ris (Zotero)
- Reference Manager
Preliminary oil risk estimates showed that oil transportations pose a significant core damage risk for Loviisa PWR located at the southern coast of Finland. Oil transportation has increased in the Gulf of Finland during the last few years and will further increase in the coming years enhancing the importance of this risk. Therefore, a more detailed assessment of oil spill impact frequencies has been performed for Loviisa nuclear power plant (NPP) taking into account hydrometeorological conditions, oil properties and leakage sizes as well as recent oil spill statistics.
During cold shutdown states the loss of service water due to excessive sea vegetation or other clogging material in the sea is an essential risk contributor because the normal residual heat removal system and the emergency cooling systems are cooled by the same service water system. Oil of tankers and cargo-carrying ships as well as other potentially fouling cargo may clog the sea water cooling systems. Nevertheless, the crude oil transportations in the Gulf of Finland pose maybe the biggest threat.
In the preliminary oil risk estimate for Loviisa NPP critical amount of oil was estimated to hit the plant with a frequency of 8·10−3/year. This estimate was based on the frequency in the Gulf of Finland for oil spills of tens of thousands of tons and larger, which were assumed to hit the plant with certainty. The preliminary core damage frequency due to an oil spill was 5.4·10−8/year during power and 8.6·10−6 at annual refueling outage.
The new oil spill frequencies for the Gulf of Finland were estimated based on worldwide and Gulf of Finland oil spill statistics and seaborne crude oil transportations. In this study, oil spills ≥ 136.4 tons (equivalent to approx. 1000 barrels) were taken into consideration. A uniform oil spill probability distribution was assumed for the Gulf of Finland. Furthermore, no distinction was made between the different seasons and the probability of an oil spill.
The assessment methodology of the oil impact probabilities is based on simulations of several oil spill scenarios with different intensities and durations in realistic weather and sea current conditions. Ten hypothetical spill points were selected on the main transportation route of the Gulf of Finland and five points on supplementary routes. For every spill point, different spill scenarios were studied ranging from 10 tons instantaneous spill to 144000 tons spill lasting ten days with 60 tons and 600 tons per hour spill intensities. For every spill scenario, trajectories were calculated representing space and time histories of the hypothetical spills starting from the selected spill points. Calculations were carried out for spills starting with ten minutes intervals for every day of a representative year excluding winter time when there is an ice deck on the sea. For every spill point 30240 trajectories were calculated. In the first stage of the calculations, preliminary dangerous scenarios were selected for further detailed calculations, in order to estimate the impact probabilities and average oil mass reaching the target areas. Impact probabilities were quantified for 10 μm and 50 μrn thicknesses of oil in the target areas, based on sensitivity calculations, which showed that the probabilities are negligible for oil thicknesses of more than 100 μm. As a result, probabilities of oil impact in the target areas for each spill point and spill scenario within one to ten days of the spill were obtained.
In order to cover the whole Gulf of Finland and thus, include the impact of oil spills occurring in areas beyond the calculated points, 24 additional hypothetical oil spill points were added to the main transportation route utilizing the data from the in detail calculated points. The final results were obtained by combining the determined oil spill frequencies and the impact probabilities within ten days from the 39 spill points.
The results show that when oil spills ≥ 136.4 tons are considered the average frequency for 10 μm thick oil to appear in the cooling water intake area is 4·10−3/year at power, when there is no ice deck on the sea, and slightly higher at annual refueling outage. However, for 50 μm thick oil the frequencies are over two times smaller. The estimates of the core damage frequencies seem to decrease with approximately 60% and 40% of the preliminary estimates for power and refueling outage, respectively. The effect of different variables on the result is presented in more detail in the paper.