(510p) Game Theory Applications to Pollution Trading

Pollution trading, also known as a Cap & Trade strategy, is both a market-based and an environmental strategy which adds flexibility to the pollution abatement decision making. This strategy aims to satisfy environmental regulations at a cost lower than that required by treatment technology implementations. The trading is based on the idea that some of the pollutant sources (firms or organizations) have enough financial capacity and infrastructure to reduce their emissions below the required limits through treatment technology implementation; hence, if the pollutant concentration of their discharges is below the allowable limits, such organizations obtain credits, which can be sold. Simultaneously, organizations which are not able reduce its emissions due to the lack of infrastructure can meet environmental regulations by purchasing credits.

This work is concerned with pollution trading in watersheds. Several mathematical programming approaches have been developed to optimize the decision making in the pollution trading strategy. In general, such approaches intend to minimize the global cost of technology implementation (the summation of technology implementation costs for every pollutant source). However, these modeling approaches are not realistic, since most of the firms are not concerned with the overall costs and benefits of the trading; they are concerned with their own costs and benefits. Therefore, the trading is really a market strategy where every pollutant source can cooperate or compete with any other source. That is, the approach should be represented as a mathematical game, where every source could be a player.

This work proposes optimization models that help pollutant sources to make optimal decisions in a polluting trading strategy. The mathematical models are formulated based on game theory. In particular, two approached have been analyzed. The first model considers a bilevel Stackelberg game; the second formulation involves a Nash equilibrium game represented as a Mixed Complementary Problem (MCP) with equilibrium constraints. Both of the formulations are solved through the GAMS modeling environment. The results include the decisions for each source of pollutants in order to satisfy global environmental regulations. To evaluate the performance of the optimization models, a mercury trading problem is considered. In particular, this presentation will discuss recent findings with respect to the impact on the solution of the numbers of leaders and followers used in the game, as well as the impact of grouping the pollutant sources in terms of the geographical location or the amount of polllutant discharges.