1317. CONCLUSIONS AND RECOMMENDATIONS The objectives of this research were to develop and apply a computational framework to evaluate the potential air quality benefits of NOBxB emission trading as well as to examine the spatial variability of ozone productivity associated with elevated NOBxB point sources. The dissertation includes a literature review that develops the context and need for this research, and illustrates the computational framework development by describing the development of the ozone productivity database, the quantitative impact indices and their integration into the NOBxB trading tool. Additionally, this dissertation described the results of the trading tool application to a case study of fifty-one point sources in the eastern Texas region, and an analysis of the impacts of time varying emissions. 7.1 Summary of Research Tasks This research fulfilled the following objectives: • A series of over 150 CAMx photochemical air quality modeling sensitivity analyses simulations were conducted to develop a database that contains ozone productivity data for fifty-one elevated point sources in the eastern Texas region. • Quantitative ozone impact indices were developed to measure and compare the relative ozone productivity of elevated NOBxB sources in selected regions. The two indices applied in this research measure the average change in ozone and the average change in ozone over a threshold within an area of impact.132• A Visual Basic Application “NOBxB trading tool” was created to systematically calculate hourly and time-averaged impact index values for scenarios involving variable facilities and regions of interest. • Impact index results were used to evaluate the potential air quality benefits of a scenario in which net NOBxB emissions are reduced by 50% by the fifty-one point sources in the eastern Texas region, two sub-regional analyses, and a case study of facilities with variable temporal emissions. • A set of CAMx modeling runs and impact indices were developed for four facilities with modified time varying emission profiles that more accurately reflect peak-loading facilities. 7.2 Key Conclusions The results of this research can be summarized by the following main points: • The eastern Texas case study results showed that emission trading can result in significantly better or worse impacts than equivalent reductions in NOBxB emissions distributed uniformly across all facilities (across the board reductions). For example, trading scenarios have a 64% chance of resulting in air quality impacts larger than +/- 10%, and a 19% probability of resulting in air quality impact larger than +/- 25% different than across the board reductions. • Meteorological uncertainty and choice of selected index lead to variable impact index values, but emissions from certain facilities can consistently be classified as relatively high or low impact independent of index and modeling day. • Restricting emissions trading to not allow a few high-impact facilities to purchase emission allowances greatly increases the likelihood that133emission trading will result in greater air quality benefits than across the board emission reductions. By targeting NOBxB reductions at “high-impact” facilities, it is possible to achieve a targeted air quality benefit with substantially lower NOBxB reductions than would be required if NOBxB reductions were applied uniformly across all facilities. • The difference between time varying and constant emission indices was small compared to the difference in indices among facilities, and time varying indices were consistently slightly lower than the constant emission indices. In terms of trading, these results indicate that if base loading (constant emissions) and peak loading (variable temporal emissions) facilities have similar impact indices as measured by constant emission modeling, trading emissions between the two facilities would not greatly impact air quality. The contribution of this research has been to develop a framework for quantitatively evaluating relative impact of air quality control strategies and for applying this framework to design NOBxB trading strategies that result in a greater air quality benefit. 7.3 Recommendations Several opportunities exist for extensions and modifications to this research. For example: • Additional sensitivity analyses could be conducted to augment the ozone productivity database, such as: o extensions of the time varying emissions modifications discussed in Chapter 6 o including additional modeling episodes (to add to the range of meteorological conditions included in the results)134o creating new database files with updated model inputs (such as updated point source or biogenic emission inventories) o developing new database files for mobile and area sources • The trading tool is not limited to evaluating NOBxB emission trading. It could be applied to quantify the relative air quality benefit of different emission reduction strategies (e.g., 50% mobile reductions vs. 90% point source reductions in Houston) • The tool could be modified to include different types of indices, such as population-weighted, maximum ozone level, or area of exceedence. • A relationship between index value and distance from the source could be developed by using the tool to calculate indices for a series of increasing areas of impact. • Cost constraints could be integrated into the tool to test the relationship between or develop a multi-objective optimization to identify lowest cost and greatest air quality benefit trading scenarios. • Industrial systems could be integrated into the tool to evaluate and optimize environmental dispatching (i.e., producing power at the times and/or locations on a power generation grid that result in the lowest air quality impacts). This research fulfilled the goal of developing a tool that can be used by decision makers to aid in evaluating and designing effective NOBxB emission trading strategies. Initial application of the tool has proven that spatial ozone productivity should be considered in order to maximize the benefit of trading programs. Future applications and modifications of this research can be used to develop more efficient and effective air pollution control