An Experimental Study on Emission Trading Behaviors of Generation Companies
jie huang,yusheng xue,chao jiang,fushuan wen,feng xue,ke meng,zhao yang dong
DOI: https://doi.org/10.1109/TPWRS.2014.2366767
2015-01-01
Abstract:The overall performance of emission trading (ET), a market-based emission regulation tool, strongly relies on participants' participation and responses. In order to improve market design, it is important for policy makers to understand the participants' trading behaviors in different market environments. However, human behaviors cannot be easily modeled with conventional analytical methods due to its “bounded rationality” characteristics. In this paper, based on the complementary features between experimental and agent-based computational methods, a hybrid interactive simulation methodology is proposed to solve human behaviors related problems. Human-subjected experiment based on European Union Emissions Trading System price data in 2006 is conducted, the results show that there is no fixed emission trading interval for generation companies, and the strategic behaviors of market participants are observed. Major driving factors of emission trading are categorized into emission price, emission quantity and time related factors, which are in accordance with empirical analysis results on EU ETS 2005-2006 transaction dataset. Furthermore, more human-subjected experiments are conducted under different emission price scenarios to obtain samples for quantitative analysis. Based on thousands of samples obtained, the joint influences of driving factors on emission trading behaviors are analyzed. The quantitative analysis results obtained can reflect the trading patterns of human participants, which provide basis for constructing computer agents that can act as useful substitutes for human participants.