The international literature has paid much attention to so-called ‘fringe benefits’ of greenhouse gas (GHG) policies. The concept is that GHG emission reduction in developing countries will also reduce local air pollution. On the basis of this concept, it is argued that it is possible to achieve simultaneously a reduction of local air pollution and a reduction of greenhouse gas emissions. In reality, however, there is a sequence to the policy agenda. First the apparent local air pollution problems are tackled then the more distant greenhouse gas problem is considered. This sequence has consequences for the optimal policy selection. Moreover, most studies that focus on fringe benefits of GHG policies neglect the existence of cost-effective dedicated abatement technology for local air pollutants. This paper analyses the optimal set of policies for reduction of SO 2, NO x and CO 2 in Shanghai for the period of 2000–2020. The analysis is based on a linear programming MARKAL model for the Shanghai energy system. The results show that the relevance of no-regret options is limited because Shanghai has improved its energy efficiency significantly in recent years. The model calculations suggest that this trend will persist if current policies are sustained. This energy efficiency improvement and the planned introduction of natural gas have important benefits from a GHG emission point of view. These benefits have received little attention as yet. Local air pollution reduction can result in additional GHG emission reduction up to 2010. After 2010, however, its CO 2 emission co-benefits are limited. Dedicated abatement technology is the most cost-effective way to reduce local air pollution. An additional incentive of 100 Yuan/t CO 2 emission reduction (12.5 Euro/t) results in an additional emission reduction of 11% (22 Mt CO 2), and it results in a significantly different technology mix than stand-alone local air pollution policies. The total potential for GHG emission reduction amounts to 66 Mt in 2010 and to 49 Mt in 2020 compared to base case levels without policies.
Read full abstract