The electricity sector in most CGE models is often highly aggregate lacking in the technology details that can describe the substitution between different energy inputs in the sector. To introduce these details into a CGE model, the first step is to disaggregate the total output of the electricity sector into outputs of different technologies, then ‘re-combining’ these back into the total output of the sector. Some issues arise during these processes: (1) how the cost structures of different technologies can be adequately represented, (2) how the substitution between these technologies can be explained. A conventional approach with regard to (1) is to assume that technology costs are represented simply by a ‘levelised’ cost index, but this masks the distinction between running costs and capacity utilization costs. With regard to (2) the conventional approach is to treat technology substitution as though competition between ‘intermediate inputs’ in an aggregate production function, but this ignores the fact that electricity outputs are homogenous and must be considered as near perfect substitutes. Near perfect substitutes, however, can lead to corner solutions, therefore to overcome this problem, capacity constraints such as in a mathematical programming approach must be introduced. In this paper, we propose an alternative method to the conventional approach which can be simpler and also more effective in handling capacity issues and technology output substitutions in the electricity generation sector. The proposed new method can be considered as though a hybrid between the top-down aggregate production function approach and a bottom-up mathematical programming approach but which can combine the important characteristics of both. The new approach is implemented to an existing CGE model (GTAP-E) to arrive at a new model called GTAP-E2. The new model in then applied in two simulation experiments to illustrate the usefulness of the new approach. The first experiment looks at the short run impacts of the Fukushima nuclear electricity accidents in Japan in 2011 and the second experiment looks at the long run impacts of the imposition of Japan's Post-Kyoto climate change and energy policies on the Japanese electricity sector for the period between 2013 and 2030.

大多数 CGE 模型中的电力部门通常是高度聚合的，缺乏可以描述该部门不同能源输入之间替代的技术细节。要将这些细节引入 CGE 模型，第一步是将电力部门的总产出分解为不同技术的产出，然后将这些“重新组合”为该部门的总产出。在这些过程中出现了一些问题：（1）如何充分表示不同技术的成本结构，（2）如何解释这些技术之间的替代。关于 (1) 的传统方法是假设技术成本仅由“平准化”成本指数表示，但这掩盖了运行成本和产能利用成本之间的区别。关于（2），传统方法是将技术替代视为总生产函数中“中间投入”之间的竞争，但这忽略了电力输出是同质的并且必须被视为近乎完美的替代这一事实。然而，近乎完美的替代可能会导致角落解决方案，因此为了克服这个问题，必须引入容量限制，例如数学规划方法。在本文中，我们提出了一种替代传统方法的方法，它可以更简单、更有效地处理发电部门的容量问题和技术输出替代。所提出的新方法可以被认为是自上而下的聚合生产函数方法和自下而上的数学规划方法之间的混合，但可以结合两者的重要特征。将新方法应用于现有的 CGE 模型 (GTAP-E) 以得出称为 GTAP-E2 的新模型。然后将新模型应用于两个模拟实验，以说明新方法的有效性。第一个实验着眼于2011 年日本福岛核电事故的短期影响，第二个实验着眼于日本后京都气候变化和能源政策在 2013 年至 2030 年期间对日本电力部门的长期影响。