| 摘要 | 1. Introduction - Research Necessity and Goal
Admittedly, oil has had a great influence on human life style since the Industrial Revolution. Oil has become a necessity in human life because of its merits such as diverse traits, facility of use and low prices, so humankind has depended on oil considerably. On the other hand, oil has also faults of irrenewability in a short period of time and finiteness of reserves. This ambivalence accounts for oil supply disruption such as temporary supply interruption caused by oilwell accidents, oil nationalization or wars around oil deposits, frequently leading to gettting global economy into trouble. This study aims to analyze previous cases of oil supply disruption by type in a statistical method and deduce political implications.
Oil has been used as a light fuel since the last half of 19th century. In 1912 petroleum products such as gasoline, kerosene, diesel and heavy petroleum were produced with oil being processed according to principles of difference in boiling points. Given the continuous development of oil refining technology and the spread of automobiles and necessities from petroleum, oil has contributed to rapid development of civilization, and its demand also increased. As a result, it may fairly be said that humankind is 'oil-addicted' because modern civilization can't go without oil.
In addition, given finiteness of oil and lack of alternative energy, oil supply disruption brings about the temporary imbalance of oil demand and supply and the increase of oil prices globally. Most countries with high dependence on oil, including South Korea, take measures against oil supply disruption such as oil stockpiling or international coordination system. Oil stockpile aims to secure oil supply against economical, social and political disorders to come about in future because of short-term oil supply disruption.
To enhance effectivity of these measures, basic study on oil supply disruption is necessary. This study makes use of statistical method in estimating the spread of oil supply disruption, which will be the starting point of analysis on oil stockpiling benefits. Especially, from 1951 to 2003 there were 21 cases of oil supply disruption, some of which in succession show considerable changes in disruption types, starting from specific point of time. The changes require a new approach of Basian method as a solution to the problem, which will be applied to estimation of change points and parameters.
2. Summary
To presume types of oil supply disturbance is crucial in statistically estimating appropriate stockpiling amount in future. Previous studies were based on the assumption that the rate of dividing daily average supply deficit by daily average global supply amount should correspond to Weibull Distribution. But given that rate of daily average deficit may be influenced by supply amount, the rate has a risk of distorting types of oil supply disruption. On the assumption that daily average supply deficit per se is deduced from Weibull Distribution, this study presumes types of oil supply disruption.
Previous studies estimated 21 cases of oil supply disruption from 1951 to 2004 in a method of Weibull Distribution. It is not so reliable because it is in a considerably early stage. To apply all 21 cases to only one Weibull Distribution in spite of noticeable changes of types of oil supply deficit at a specific point of time may result in miscalculation of appropriate oil stockpiling amount.
To resolve it this study introduces a conception of 'single change-point' making it possible to classify types of oil supply disruption. Types of oil supply disruption classified approximately at this change point of time show different Weibull Distributions respectively, and the number of oil supply disruption is less than 11. Thus, traditional statistical method may lower the reliability in estimating parameters of Weibul Distribution.
This study introduced Basian method to enhance reliability of estimation of small samples. This study grasps differences between parameter method and Basian method in estimating distribution by Monte Carlo Simulation and change aspects of distribution types according to whether change point of time is present or not.
3. Study Result and Implications for Policy-Making
Given the fact that with middle part of 1970s as a starting point, types of oil supply disruption were changed, distribution should be estimated with possible change points as a standard. Otherwise, the estimation can lead to distorted result, consequently causing improper energy policies to be taken and political and economical disorders in future. In addition, since the first oil shock in 1973 the width of disruption has been enlarged and the aspects diversified.
Oil supply disruption is one of themes laying a cornerstone in study on energy risk control. This study can contribute to laying the foundation for statistical analysis of oil supply disruption and supplying data base concerning oil stockpiling, self-development and early signal system. Furthermore, at the present time when international society is making efforts to reduce consumption of fossil energy blamed for environmental pollution in compliance with Kyoto Protocol, our government should take diverse measures for energy conservation, alternative energy development, etc. and invest in them in order to reduce our dependence on oil. |
