Evaluation of Green House Gases (GHGs) Reduction Plan in Combination with Air Pollutants Reduction in Busan Metropolitan City in Korea

2. 1 Selection of DEA Model

DEA which is an extension of Farrell’s (1957) relative efficiency concept was initially proposed by Charnes et al. (1978) and was established as a basis for the efficiency analysis. DEA is a non-parametric method that measures the efficiency of multiple decision-making units (DMUs) which produce multiple outputs using multiple inputs.

Some of the advantages of DEA are:

• (a) Capability of handling multiple inputs and outputs
• (b) No need to explicitly specify a mathematical relation between inputs and outputs
• (c) Direct comparison of a evaluated DMU with a peer group
• (d) Outputs can have different units regardless of inputs unit

Some of the disadvantages of DEA are:

• (a) Difficulty in measuring the absolute efficiency
• (b) Results are sensitive to the selection of inputs and outputs (Berg, 2010)
• (c) The number of efficient DMUs on the frontier tends to increase with the number of inputs and output variables (Berg, 2010)

Of the DEA methodology, a model developed by Charnes, Cooper and Rhodes (CCR model) defines the efficiency as a sum of weighted outputs to a sum of weighted inputs, where the weight structure is calculated by means of mathematical programming and constant returns to scale (CRS) are assumed (Charnes et al., 1978). Other model developed by Banker, Charnes and Cooper (BCC model) considers variable returns to scale (VRS) (Banker et al., 1984). Depending on efficiency assessment measures, the DEA methodology is divided into the input-oriented measures and output-oriented measures. There are other available models such as the super-efficiency model which overcomes drawbacks of the initial DEA models that are incapable of providing order information among efficient DMUs (Anderson and Petersen, 1993).

This study used the input-oriented super-efficiency CCR model to assess the order among efficient DMUs. The super-efficiency CCR model measures the radial distance from a target DMU to the efficient frontier which is estimated without the target DMU. The equation for the super-efficiency CCR model is expressed as follows (Cooper et al., 2007):

where, x_ij is ith input of DMU j and y_rj is rth output of DMU j. θ, λ_j, s_i^-, and s_r⁺ are model’s decision variables. ε>0 is the usual non-Archimedean element. m is number of inputs and n is number of outputs. J is number of DMUs. p is the index of target DMU. We can determine DEA efficiencies of all DMUs by solving Equation 1 J times, designating each DMU as a target DMU p in turn. Here, a DMU p with the optimum θ*≥1 is judged DEA efficient, while a DMU p with θ*<1 is DEA inefficient. A DMU having greater θ* is more efficient than a DMU having smaller θ*. θ* is termed as efficiency score.

We defined each GHGs reduction action plan as a DMU. We defined costs according to each GHGs reduction action plan as an input, and reduction potential of GHGs and air pollutants according to each GHGs reduction action plan as an output. Using these data, we calculated efficiency scores.

2. 2 Selection of GHGs Action Plan

Referring to the Guideline for Local Governments’ Establishment of Green House Gases Reduction Master Plan (ver. 1) by the National Institute of Environmental Research (Hong et al., 2010) and Master Plan for Climate Change Response in Busan Metropolitan City-Establishment of Climate Change Response Master Plan and Yearly Implementation Plan (Yang et al., 2010) to select GHGs reduction action plans, we chose GHGs reduction action plans from the residential, commercial and public, and transport sectors since these sectors have relatively high GHGs emissions and various reduction plans were already developed for these sectors. After further considering possibility of costs quantification along with GHGs and air pollutants reduction potential estimates, we chose 32 GHGs reduction action plans. The GHGs action plans selected for the study is presented in Table 1.

Table 1.

GHGs reduction action plans.

2. 3 GHGs and Air Pollutants Reduction Estimates

Estimates of GHGs and air pollutants reduction potentials in the target year, 2020, according to the reduction action plans in Busan are presented in Table 2. The GHGs and air pollutants reduction potential due to the electricity usage decrease was estimated by calculating indirect emissions reduction due to the decreased electricity usage.

Table 2.

GHGs reduction and air pollutants reduction potentials according to action plans.

2. 4 Cost Estimates Relative to Reduction Action Plans

Table 3 shows estimated total costs relative to the GHGs reduction action plans. Several papers and reports (Electric Power Statistics Information System, 2011; Korea City Gas Association, 2011; Hong et al., 2010; Korea Electric Power Corporation, 2010; Ministry of Land, Transport and Maritime Affairs, 2010; Yang et al., 2010; Kwon et al., 2008) have been reviewed and consulted for the estimates of the costs relative to the GHGs reduction action plans. We considered installation and operation costs for the assessment of the reduction action plan costs, and obtained yearly costs for equipment by dividing by running years. For action plans where construction period is more than 1 year, costs were converted to the present value. Action plans presented with negative costs are cases where financial profits are achieved with cost saving measures such as energy saving. Utility decrease which is difficult to express as monetary value was not considered in the study.

Table 3.

Total costs relative to reduction action plans.

2. 5 Modification to Input Data

In assessing effects of the GHGs reduction on the air pollutants reduction, each air pollutant to be reduced was assigned with its weighted fraction based on environmental characteristics of the pollutant in Busan since each air pollutant has different influences from other air pollutants, on the environment and public health. To assign the weighted fraction for each air pollutant, we calculated probability that each air pollutant would exceed 80% of the Standard in the Clean Air Conservation Act using mean, standard deviation, and the Standard, and compared the probability with the probabilities of the other air pollutants. Table 4 and 5 present data used to calculate the weighted fraction for each air pollutant and its results.

Table 4.

Air pollutants levels and standards (2008).

Table 5.

Weighted fraction adopted.

Since the level of CO was much less than the Standard and the probability to exceed 80% of the Standard was 0 as shown in the Table 4 and 5, CO was excluded from further analysis. In case of VOCs, since there was no data available, these compounds were also excluded from further analysis.

The GHGs reduction action plans having negative costs are relatively cost-effective comparing to the reduction action plans having positive costs. We applied the DEA model to cost-negative and cost-positive cases separately. For negative costs, the value was first converted to absolute value and then, converted to inverse value for application to the DEA model. In this case, priority order of profits (i.e., negative costs) times reduction potentials were evaluated using the DEA model. This is a good tool for this study since the objective of this study is to obtain relative order among the GHGs abatement action plans.

2. 6 DEA Application and Analysis of the Results

Priority orders of the GHGs reduction action plans acquired using the DEA model are shown in Table 6 through Table 9. We put the costs of the GHGs action plans as inputs, and the GHGs and air pollutants reduction potentials as outputs. The results were obtained for cost-negative group and cost-positive group, separately, and we divided the results for each cost-negative and cost-positive group to GHGs reduction cases, and simultaneous GHGs and air pollutants reduction cases, respectively.

Table 6.

Priority oder of cost-positive cases - only for GHGs reduction.

Table 7.

Priority order of cost-positive cases - for simultaneous GHGs and air pollutants removal.

Table 8.

Priority order of cost-negative cases - only for GHGs removal.

Table 9.

Priority order of cost-negative cases - for simultaneous GHGs and air pollutants removal.

In case of cost-positive group, there were no differences of the priority orders between the GHGs reduction action plans, and the simultaneous GHGs and air pollutants reduction action plans except that the priority orders of “green home supply-insulation improvement” and “intelligent transportation system (ITS)” were switched with each other between the GHGs reduction action plans, and the simultaneous GHGs and air pollutants reduction action plans. When the simultaneous reduction of GHGs and air pollutants was counted, “green home supply-insulation improvement” appeared to be most cost-effective, and this indicates that the simultaneous GHGs and air pollutants reduction could be effectively achieved with relatively low costs for this category. Contrary to that, “solar light house supply” and “solar light generator installation” showed low priority. The results were similar to those of Park et al. (2010). Comparing investment return periods among renewable energies, they showed the longest investment return period was for solar light. Their findings of the investment return period of the renewable energies are as follows: small hydropower<solar power<wastes<thermal<bio<wind power<solar light.

There are two types of reduction action plans which have negative costs due to energy savings: they are voluntary energy saving plans, and plans where energy saving is greater than yearly cost, such as wastes recycling and urban natural gas network expansion. For the cost-negative group, the priority order of the action plans considering only GHGs reduction is as follows: green start network formation and running>daily car use control program>employing stricter low carbon design standards for building construction and renovation>switch to high efficiency LED lighting>getting energies from municipal wastes and constructing heatonly boilers. The voluntary energy saving action plans such as reducing heating time, lowering heating temperature, reducing TV watching time, cleaning up boilers regularly, and limiting indoor cooling and heating temperatures, followed the priority order. The results were different from those of the simultaneous GHGs and air pollutants reduction action plans.

For the cost-negative group, when the simultaneous GHGs and air pollutants reduction was considered, the urban natural gas network expansion showed lower priority order than utilizing extra heat from solid wastes incinerators, and thermal energy use showed relatively high priority order. The results are attributed to differences in air pollutants reduction potentials according to energy sources which could be saved by the reduction action plans.

It is important to consider the GHGs reduction potentials in addition to the priority order to be obtained through efficiency comparison when choosing reduction action plans for establishing and implementing GHGs reduction action plans. Accordingly, among negative-cost action plans, green start network formation and running, urban natural gas network expansion, daily car use control program, and employment of stricter low carbon design standards are proper measures taking into account both the GHGs reduction potentials and cost effectiveness. Contrary to that, high efficiency LED lighting usage is an action plan having relatively low reduction potential although it has high cost effectiveness. Saengok LFG to energy project, utilization of methane from digesters, and construction of small hydroelectricity power plants have low reduction potentials and relatively low cost effectiveness.

Among cost-positive action plans, ITS project and insulation improvement showed relatively high reduction potentials and high cost effectiveness. The ITS project is considered to be a proper reduction action plan since the project is considered as a good alternative to the current transportation system to solve traffic problems. Solar light house supply and solar light power generator installation had relatively low cost effectiveness.