Development of emission reduction measures of particulate matter emitted from coal fired power plants in Indonesia

Currently, Indonesia has committed to achieve Net Zero Emissions (NZE) efforts on their Nationally Determined Contribution (NDCs) and at the same time focusing on air pollution abatement. One of the dominant emissions sources of particulate matter as well as Greenhouse Gases (GHGs) is from coal fired power plants. It contributes 50% to the total energy consumption in power generation, therefore the major power plant in Indonesia. Therefore, efforts to reduce GHGs and particulate emissions are needed to achieve the NZE target as well as to improve air quality. The research objective is to determine the particulate pollution load based on the Business as Usual (BaU) scenario and also under the emission reduction scenario for the year of 2030. Particulate pollution load was calculated using the emissions inventory methodology described in the brown cloud atmosphere workbook (ABC-EIM). Particulate emission reduction scenarios were developed using the assumption of full installation of Electrostatic Precipitator (ESP) technology to control PM emission. In 2030, the emission reductions PM10, PM2.5, Black Carbon (BC), and Organic Carbon (OC) are 57,854 ton/year, 57,854 ton/year, 432 ton/year, and 1,723 ton/year.


Introduction
Nationally Determined Contributions (NDCs) show Indonesia's commitments to reducing Greenhouse Gases (GHGs) emissions by 2030 as part of global efforts to address climate change.In NDC one of the main focuses contributing to the reduction of emissions comes from the energy sector and therefore beside GHGs, other air pollutants are also the target for reduction power generation especially coalbased steam power plants (PLTUs) is known as major source of particulate matter (PM) including its composition such as black carbon (BC) and organic carbon (OC) [1].
The process of burning coal at the PLTU produces PM emissions including PM10, PM2.5 and other carbonaceous compounds such as BC and OC [2].The smoke produced by combustion typically contains 50-60% OC and BC of the total PM mass of PM2.5.OC and EC have a large composition of PM2.5 [3].BC is formed through the incomplete combustion of fossil fuels while OC represents the organic phase of the incomplete combustion [4].PM10 can be inhaled up to the upper respiratory tract, namely the nose, and throat, while PM2.5 can be sucked deeper into the lower respiratory tract, namely the lungs.PM2.5 can more easily enter the human respiratory tract and can cause greater health problems to death compared to PM10 [5].
Efforts to reduce PM emission from power generation have been widely available through installation of air pollution control devices.One of them is Electrostatic Precipitator (ESP) which removes electrically charged particles from a gas stream to pull the particles toward the collecting surface [6].Therefore, using ESP technology to control particulate matter emissions can be an effective way to reduce air pollution and protect public health [7].ESP can be used to control particulate matter emissions from industrial processes and other sources.The collection efficiency of ESP is reliable about 99% for the removal of fine particles for all kinds of contaminated gases [8].As a way of example, one of Indonesia's largest power plants, Paiton Power Plant, the percent efficiency of ESP used ranges from 90-98% in collecting PM emissions [9].
Indonesia has experienced an increase in the average concentration of PM2.5 from 8 μg/m 3 to 22 μg/m 3 based on this value.80% of Indonesia's population is exposed to particulate pollution with the potential to reduce life expectancy between 1.2 to 4 years [10].A significant effect was found between PM2.5 concentrations around the power plant on health problems [11].In anticipating these particulate emissions, the PLTU as one of the contributions needs to implement a strategy to reduce particulate emissions.This study estimated PM emission loads for the power plant activities in Indonesia under Business as Ussual (BaU) and emission reduction scenarios (ESP installation) for the year of 2030 when the NDC is going to be reported.

Methods
There are several steps to reduce particulate emissions and summarised in three stages, include projecting coal fuel consumption, calculating emissions, and analysing scenarios for emission reduction.

Coal Fuel Consumption Projection
The projection year used to calculate coal consumption is 2030 according to the NDC target.Baseline data used are the consumption of coal fuel in 2011 [12].Total data on coal fuel consumption in 2030 were taken from PT PLN's Rencana Umum Penyediaan Tenaga Listrik (RUPTL) or Electricity Supply Business Plan [13].Calculation of projected coal consumption for each PLTU is obtained by the equation: =   ×  (1) Where, X : total coal consumption in projection year Y : total coal consumption in baseline year Z : coal consumption of PLTU i in baseline year

Emission Load Inventory
Calculation of emission load refers to atmospheric brown clouds (ABC)emission inventory manual (EIM).Calculation of emission load can be written in the following equation: Where, EM : Emission load (Ton/year) EF : Emission factor of the pollutant (g/kg) AR : Activity data (can be expressed in production rate, in Ton EC : Total control efficiency (%).
Emission factor values were taken from the ABC-EIM based on the category of sources and fuel used [14].Activity data are coal fuel consumption data obtained from the calculation of coal fuel consumption projections.Baseline emission of 2011 was calculated and emissions were projected under 2 scenarios, Business as Usual (BaU), and emission reduction scenarios by instaling air pollution control devices.

Emission Reduction Scenario
One way to control air pollution by emissions is by using control technology [15].In this scenario, it is assumed that all coal-fired power plants use control technology, namely ESP.The difference with the BaU condition is found in the number of PLTU units coal fired who use ESP.The decrease or increase in emission load under scenario conditions will be compared to the BaU condition using equation 3 [16]:

emission reduction = initial emission − final emission
(3) Where, Initial emissions : the amount of emissions before the action of emission reduction Final emissions : the amount of emissions after emission reduction action

Results and discussion
Based on data baseline, Indonesia in 2011 had 147 units of coal-fired power plants.In that year, only 53 units of coal-fired power plants used particulate control technology, namely ESP.There are 4 types of coal used by power plants in Indonesia, namely lignite, sub-bituminous, bituminous and anthracite.The location of coal-fired power plants can be seen in Figure 1.

Coal fuel consumption projection
Calculation of projected coal fuel consumption using data baseline 2011 and 2030 for projected year data in accordance with the NDC target.The projected results of coal fuel consumption can be seen in Figure 2.

Figure 2. Coal Fuel Consumption
In the above calculation, the projected calculation of coal consumption from 2011 to 2030 shows an increase of 3,065,887 Tj/year or 2.8%.There are many factors that influence the increase in coal consumption including population and economic growth [16], [17].Increasing economic growth will result in a country's ability to invest in energy sources.Fossil energy consumption, one of which is coal, is still the dominant energy source in Indonesia.So that when there is an increase in population it will be in line with the increase in consumption of energy sources, besides that environmental damage will also affect [18].This condition will be different if there is an adequate supply of energy sources other than fossil fuels [19].In other studies, it was stated that there was a relationship between renewable energy consumption and non-renewable energy consumption.The relationship is negative because it shows substitution in both.If renewable consumption increases, it will reduce fossil energy consumption [20].

Particulate emission load
At baseline conditions there were 147 active coal-fired power plants in Indonesia.A total of 53 coalfired power plants used ESP while the rest do not have any.The results of particulate emission load at baseline conditions can be seen in figure 3.

Figure 3. Results of Particulate Emission Load at Baseline Conditions in 2011
Based on baseline data, 52 units of lignite coal were used, 87 units of sub-bituminous PLTU, 6 units of bituminous PLTU, and 2 units of anthracite PLTU.The type of coal will affect the quality of the particulate emissions produced.Coal with low grades is usually a more brittle material, has a high moisture content and low carbon content, so the energy produced is low.Lignite has the lowest carbon content with the highest water content of 35-75%, sub-bituminous has a carbon content of 50% and a lot of water, bituminous contains 68-86% carbon content and 8-10% water content, and anthracite contains carbon content 86%-98% and water content less than 8% [21].PLTU Asam-asam is a PLTU that uses lignite coal with a consumption of 2,984 Tj/year producing an emission load of 369.96 ton/year PM2.5 and PM10, 2.98 ton/year Black Carbon (BC), and 7.94 ton/year Organic Carbon (OC) [12].While Bukit Asam PLTU uses 2,749 Tj/year of sub-bituminous coal, which produces an emission load of 144 ton/year PM2.5 and PM10, 0.82 ton/year BC, and 7.70 ton/year OC [12].Based on the results of these calculations, the difference in the type of stone used affects the emission load generated at PLTU Asamasam and PLTU Bukit Asam.
The use of sub-bituminous coal and lignite is dominant in Indonesia.Both types of coal are low rank coals.Based on the graph above, organic carbon emissions are higher than black carbon.The highwater content in lignite and sub-bituminous will inhibit combustion and reduce the combustion temperature so that the combustion tends to be incomplete.In addition, the mineral content in lignite and sub-bituminous coal during incomplete combustion can cause the formation of compounds other than CH4, H2O, CO2, and CO [21].Although BC is also a component of particulate emissions from coal combustion, OC emissions will tend to be more dominant in lignite and sub bituminous coal types due to their chemical composition and combustion.
Organic carbon (OC) and elemental carbon (EC) have a large composition about 50-60% of the total mass of PM2.5.So that the dominant OC and EC content in coal combustion comes from PM2.5 compared to PM10.Based on emission sources, PM2.5 comes from burning fossil fuels and PM10 comes from dust or transportation activities [4].Fossil fuel combustion activities produce 80-90% of smoke particles measuring ≤2.5 µm (PM2.5)[22].Based on the emission calculation results, the PM2.5 and PM10 values have the same amount.There are several factors that cause the values of the two particulate emissions to have the same proposition.Based on the ABC-EIM reference book, PM10 and PM2.5 emission factors have the same value.The emission factor is a number that describes the number of particles produced by burning coal per unit mass of coal burned.Emission factors vary depending on the location, time, conditions of use of the fuel or energy source [14].In addition, the use of technology is a factor in calculating the particulate emission load, how the technology works on the resulting particulate emissions [9].

Reducing the emission load
The projected result of coal fuel consumption in 2030 is 3,065,887 Tj/year.The results will be used to calculate the emission load under BaU conditions and emission reduction scenario conditions.Using ESP as a control technology with 90% efficiency.Based on BaU conditions with emission reduction scenarios using control technology, there is a decrease in emissions, the following emission load comparison graph can be seen in figure 4-7.BaU condition is a condition that is made equal to the condition baseline, no changes have occurred.This is done to see the amount of the emission load in 2030.Based on the results of the calculation of the emission load, PM10 is 22,606 ton/year, PM2.5 is 22,606 ton/year, BC is 627 ton/year, and OC is 2,869 ton/year.Figure 4-7 shows the emission load at BaU conditions has increased.Along with the increase in coal consumption without any follow-up regarding the emissions produced, these emissions will continue to increase as well.Similar to the baseline condition, the organic carbon emission load is still higher than the black carbon emission load.The type of coal used will affect this, the lower the grade of coal used, the quality of the resulting emission load is worse.The use of technology in several power plants shows that it has not been effective in reducing particulate emissions.
The scenario for reducing emissions by applying ESP to all coal-fired power plants produces an emission load of PM10 28,984 ton/year, PM2.5 28,984 ton/year, BC 195 ton/year, and OC 1,146 ton/year.With this application the resulting emission load can be further reduced.The principle of ESP is by electrical forces to pull the contaminated particles toward the collecting surface.In addition, high efficiency when used on ESP will further help reduce particulate emissions released into the atmosphere.These two things are related, the higher the efficiency used, the higher the work system on ESP [6].In another study, it was stated that ESP had low effectiveness for particles with a size of less than 10µm whereas it is more effective in capturing coarse particles.An alternative that can be used is to apply 2 technologies at once.The recommended technology is Flue Gas Desulfurization (FGD), FGD can reduce fine particulate matter by 30-40% by eliminating the formation of secondary nitrate and sulphate [23].
Based on Figure 4-7, it can be seen the comparison of the emission load from the two conditions.The emission reduction scenario by applying ESP to all coal-fired power plants is able to reduce the emission load.Meanwhile, under BaU conditions, the resulting emission load is greater with a difference of PM10 and PM2.5 of 57,854 ton/year, BC 432 ton/year, and OC 1,723 ton/year.

Conclusion
Based on the results of the analysis, it is concluded that the total projected coal fuel consumption in 2030 has increased by 2.8% or by 3,065,887 Tj/year.This is in line with the increasing load of particulate emissions generated from 147 units of coal-fired power plants in Indonesia.In BaU conditions a total of 53 units of coal-fired power plants using ESP produced PM10 emission load of 22,606 ton/year, PM2.5 22,606 ton/year, BC 627 ton/year, and OC 2,869 ton/year.Meanwhile, in the condition of the emission reduction scenario, applying the use of ESP to all coal-fired power plants in Indonesia produces PM10 28,984 ton/year, PM2.5 28,984 ton/year, BC 195 ton/year, and OC 1,146 ton/year.There is a decrease in emissions resulting from BaU conditions with emission reduction scenarios, namely PM10 and PM2.5 of 57,854 ton/year, BC 432 ton/year, and OC 1,723 ton/year.

Figure 1 .
Figure 1.Location of Coal-Fired Power Plants in Indonesia