SHIP EMISSION ESTIMATION FOR IZMIR AND MERSIN INTERNATIONAL PORTS – TURKEY

Approximately 90% of the world trade is done by sea trade. The world maritime trade fleet is also increasing every year. Since ships use fossil fuels, they contribute to global climate change. It is very important for the environment and human health to calculate emissions in the port areas where ships do cargo operations. In this study, emissions analysis of the two important ports of Turkey’s is performed with using the bottom-up calculation method. In addition, hotelling periods of the vessels are monitored for both ports for one year period. As a result, total Nitrogen Oxide (NOx), Sulphur Dioxide (SO2), Carbon Dioxide (CO2), Volatile Organic Compounds (VOC), Particulate Matter (PM) and Carbon Monoxide (CO) emissions of İzmir Port are calculated as 900 tons/year, 589 tons/year, 45320.5 tons/year, 49.7 tons/year, 77.7 tons/year and 36.9 tons/year, respectively. Moreover, total NOx, SO2, CO2, VOC, PM and CO emissions of Mersin Port are calculated as 1998 tons/year, 1339 tons/year, 102330 tons/year, 114.5 tons/year, 178.5 tons/year and 82.5 tons/year, respectively. The amount of emissions of Mersin port, which has a higher number of movements, was higher than that of İzmir port.


INTRODUCTION
In recent years, emissions from ships have increased due to increased ship activities, intensification of port traffic, and preference of high-powered and heavy-fueled engines. Due to the effects of emissions on global warming and air pollution, monitoring and reduction of shipping emissions have become a necessity. Therefore, International Maritime Organization (IMO) imposes sanctions on ships within the scope of emission reduction preventions. Especially port cities are exposed to serious air pollution due to exhaust emissions from ships berthing to the port. Recently, new regulations have been established by the IMO and the European Union (EU) regarding to the reduction of emissions from ships. Within the framework of these regulations, monitoring of shipping emissions has become more important. In this study, Nitrogen Oxide (NOx), Sulphur Dioxide (SO2), 185 fuel-based calculation with activity-based calculation and stated that activity-based calculations are more accurate than fuel-based calculation due to using real AIS data. In the calculations of Olukanni and Esu [9], Chang et al's [10] study in 2013 was used, it was concluded that the most CO2 released during anchorage and the entry of breakwaters, and 5% of the total emissions at the port were released. In their study, Ledoux et al. [11] have measured exhaust emissions at the port of Calais in Northern France for a period of 3 months and analyzed the internal effects of emissions. They concluded that the effect of maritime transport at the port of Calais on average concentrations is 51% for SO2, 35% for NO, 15% for NO2 and 2% for PM10. Chu-Van et al. [12] made measurements on a dry cargo ship at the ports of Newcastle and Gladstone. Separate measurements were taken from the main engine and exhaust outlets and emissions were calculated at berth, manoeuvring and cruising at sea position. CO, CO2, SO2, NOx, Hydrocarbon (HC), PM emissions were examined, the results were compared with other studies in the literature and the difference in the results obtained was emphasized. Cheng and Haibo [13] examined air pollutants sourced from auxiliary engines of ships in Chinese ports, compared ships that using shore power and ships that do not use shore power in port. In their studies examining emissions from electricity generation in China, they concluded that the annual SO2, NOx and CO2 emissions would decrease by 16 thousand tons, 128 thousand tons and 1435 thousand tons, respectively, by using shore power in ships at ports. Sun et al. [14] calculated the exhaust emissions of ships arriving to China's Quindao port in 5 operating modes (cruising, preparing, decelerating, manoeuvring, and hotelling). As a result, they compared the values of HC, CO2, SO2 and PM with the three ports in the Emission Control Area (ECA) region and emphasized the necessity of taking precautions. In their study, Chen et al. [15] estimated the exhaust emission values of 8690 vessels arriving at Tiajin port during 2014 and compared them with emissions from other ports in China. They compared their own emission estimation method with other methods. Deniz and Durmuşoğlu [16] forecasted exhaust gas emissions from ships in the Sea of Marmara and the Turkish Straits, and compared the results with the other transportation way in Turkey. The authors considered some parameters such as the main engine types, fuel types, operation types, navigation times and speeds of vessels in calculation methods. Schembari et al. [17] examined the change in the concentrations of SO2, NOx and BC in different European ports by a European Union rule that entered into force in 2010, with measurements made at a measurement station placed on a cruise ship called Costa Pacifica. As a result, they emphasized the effect of low sulfur fuel use on air pollution and environmental factors. Consequently, they emphasized the effect of low sulfur fuel usage on air pollution and environmental factors. In an analysis by Zetterdahl et al. [18] on a ship in two stages, the ship used heavy fuel oil in the first stage and low sulfur fuel in the second stage. As a result, SO2 emissions and PM values were found to be decreased in low sulfur fuel usage. In their study, Tichavska et al. [19] calculated and compared the exhaust emissions from ships in ports which were subject to 3 different regulatory applications (EU, Sulphur Emission Control Area (SECA) and non EU / non SECA) for 12 months, according to Ship Traffic Emission Assessment Model (STEAM) and AIS data. In their study, Piris et al. [20] measured the reduction in CO2 emissions of merchant ships for the first time as a result of replacing traditional mooring systems with new automatic mooring systems. Using the ENTEC method [21], they have concluded that CO2 emissions are reduced by 76.78% compared to traditional mooring systems. Bilgili and Çelebi [22] investigated the airborne emissions of nine bulk carrier ships by means of block coefficient and deadweight characteristics during predesign as a result, emissions were calculated with three different methods and some equations was obtained with regards to regression analysis. Ünlügençoğlu [2] developed the web-based data collection software which monitor emissions from ships and performed for Ambarlı Port, Turkey.
In this study, Izmir and Mersin ports were monitored for a period of one year between 2017-2018, NOx, SO2, CO2, VOC, PM and CO emissions from ships arriving at the ports were calculated numerically in terms of cruising, manoeuvring and hotelling positions. In the calculations, the actual data of the ships (power & rpm of main engine, power of auxiliary engine, year of construction, gross tonnage, flag, number of arrivals, cruising, manoeuvring and hotelling) were used. The calculated emissions were classified according to ship types. The ages of the ships arriving at the port in the same period and the average hotelling period in the port were determined. A comparison was made between İzmir and Mersin port and other existing studies.

MATERIAL AND METHOD
İzmir Port is located on the west coast of the Aegean Sea in the Alsancak area of the Gulf of Izmir. The port has a large agricultural and industrial hinterland. İzmir Port, which is the agricultural and industrial port of the Aegean region, also plays a vital role in the export of the country. The port serves all kinds of cargo and is connected to the railway and highway network. Izmir Port serves mainly cargo and cruise passenger ships. 500,000 passengers, 1 million TEU containers and 10 million tons of cargo are handled annually. The Port employs 800 people and has an annual average turnover of 100 million usd [23]. Figure 1 shows the top view of the port of Izmir. In the study, the distance between A-B was accepted cruising, the distance between B-C was accepted manoeuvring and the C-point was accepted as a hotelling (port). Distance A-B is 16.7 km, the distance between B-C is 5.5 km. Point B is the point of taking of the pilot on the ship who helps berthing and departure. The start of the maneuver was accepted as the moment the pilot was taken. The cruising speed of the ships is accepted as 18 knots and the manoeuvring speed is 3 knots. These points, distances and speeds were determined according to the information obtained from the oceangoing captains who worked at the relevant port.

Figure 1. İzmir Port
Mersin International Port, which is 112 hectares in size, is not only one of the most important ports of Turkey, at the same time is one of the most important ports in the eastern Mediterranean with advantages of capacity, geographical location and wide hinterland. The port is connected to the domestic lines and the countries of the Middle East by highways, airways and railways and is ideally located for transit cargoes to Central Asia and the Middle East. The port serves container, general cargo, roro, dry bulk cargo, liquid bulk cargo ships. Mersin International Port, which has 21 berths, provides loading/unloading services to nearly 30 ships at the same time. The port handles over 30 million tons of cargo per year [24]. Figure 2 shows the top view of the port of Mersin. In the study, the distances were accepted as same as İzmir Port.   Table 1 shows the classification of 776 different ships arriving at Izmir port in the date range according to their flags.

METHODOLOGY
There are two methods in the literature in order to calculate emissions from ships numerically. One method is bottom-up methodology based on the activity and movement of ships. In this method, calculations are done by formulas based on main engine auxiliary engine powers, rpm, hotelling time, ect. The other method is the top-down method based on a ship's consumption of fuel oil For estimating emissions this method uses the fuel consumption quantities and corresponding emission factors of the fuel [25].
Bottom-up method is used in this study, the emission factors of the ships are calculated by applying emission factors according to the 3 operating modes of ships (cruising, manoeuvring and hotelling). Emission factors were used depending on the operating modes, rpm of the main engine and the type of fuel used [21]. In order to find the total exhaust gas emissions emitted from the ships arriving at the port, the amounts of the exhaust gases emitted by the vessels during their operation modes of cruising, manoeuvring and hotelling were calculated. The calculations were made with the help of real data such as main engine power and rpm, auxiliary engine power, types of fuel consumed, cruising, manoeuvring and hotelling without using any estimation methodology in the literature. 2068 movements of 776 different ships that arrived to Izmir Port and 4215 movements of 1202 different ships that arrived to Mersin Port were followed in a period of one year, between 01.09.2017 and 01.09.2018. NOx, SO2, CO2, VOC, PM and CO emissions from main and auxiliary engines were calculated for both ports.
The formula given in Equation 1 was used to calculate the exhaust gas emissions generated by the ships arriving to İzmir and Mersin Ports in cruising, manoeuvring and hotelling modes [21].
Where; E, emissions (g) generated in the related operating mode (cruising, manoeuvring and hotelling); T (h), time spent in the related operating mode; ME, main engine power (kW); The LFME represents the main engine load factor (%) in the related operating mode; AE, the power of auxiliary engine (kW); LFAE is the load factor (%) of the auxiliary engine in the related operating mode; EF refers to emission factors (g/kWh) for each type of ship, depending on the type of fuel used in the related operation mode and main engine type.
The main and auxiliary engine power of the vessels required for the related calculation method and the hotelling period of the vessels in the port were determined by monitoring 772 different ships arriving to İzmir Port and 1202 different ships arriving to Mersin Port on the specified dates.
The cruising time for İzmir and Mersin Ports was calculated with dividing the cruising distance made at berthing and departure (2 x 16.7 km) by cruising speed (18 knots, 33.336 km) and found 1 hour (h). The maneuvering time for İzmir and Mersin Ports was calculated with dividing the maneuvering distance made at berthing and departure (2 x 5.5 km) by the maneuver speed (3 knots, 5.556 km) and found 2 h. Hotelling times were obtained by monitoring the ports.
In addition, it was assumed that 2 auxiliary engines were loaded in cruising and manoeuvring and one auxiliary engine in port [2]. Confirmation was received from the captains of the oceangoing ships that worked for the relevant ports for these acceptances. Exhaust gas emissions from each ship were calculated with the developed web-based "Emission Calculation Software" [1] [2].
The engine types according to the main engine rpm are shown in Table 3. In the calculations, the fuel type used for the main engine was Heavy Fuel Oil (HFO) and the fuel type used for the auxiliary engine was Marine Diesel Oil (MDO). Emission factors were determined according to the operating modes and main engine rpm. According to the type of fuel, type of main engine and ship operating modes the emission factors used in the calculation of main and auxiliary engine emissions in different operating modes for each type of ship are shown in Table 3, Table 4 and Table 5.
The emission factors in the calculation of main engine emissions for each ship type are shown in Table  4 according to the type of fuel, main engine and cruising mode of the vessel. In the table, NOx1 refers to the emission factor value for ships built before 2000, NOx2 refers to the emission factor value for ships built after 2000. For each type of ship, the emission factors used in the calculation of emissions from the auxiliary engine are shown in Table 5, depending on the type of fuel used in cruising, manoeuvring and hotelling. For each type of ship, the emission factors of main engine are shown in Table 6, depending on the type of engine and fuel used in manoeuvring and hotelling modes.  Table 7 shows the main and auxiliary engine loads in the different operating modes used in the module in the calculations made in the web-based software [20]. Main engine loads were 80% at cruising and 30% at manoeuvring. Loads of the auxiliary engines were 50% at cruising, 80% at manoeuvring and 50% at hotelling.  Table 8 and total emissions from Mersin Port are shown in Table 9 according to the operating modes of ships. Because of the number of ship movements is higher in Mersin port, all emission values are higher than İzmir port. When the total emission amounts are analyzed, it is seen that the most emitted emission is CO2.  [2] and are shown in Table 10 according to ship operation modes. Since the main engine is only used in cruising and manoeuvring, the table shows only cruise and maneuver modes. In Mersin port, where the average age of ships and number of movements is high, total emission amounts and also unit emission (ton.y -1 / movements) amounts are higher.  [2] and are shown in Table 11 according to ship operation modes. When the average hotelling time and auxiliary engine powers are examined, it is found that the vessels staying in Mersin Port are longer in the harbor and the auxiliary generator powers are higher. Therefore, emissions from auxiliary machinery are higher in Mersin Port.  Table 12  and Table 13. As can be seen from the tables, the emission values from container ships are considerably higher than the other types of vessels in the measured date range. Emitted emissions were mostly from container ships. General cargo, which is the most ship type arriving to the ports, is lower than the containers because the average main engine and generator power is much lower on general cargo ships.

CO2 Emissions according to the ship flags
As a result of the CO2 emission calculation, the flags in the first three places in the unit CO2 emission rankings are Liberia, Marshall Islands and Malta. As shown in Table 14, the number of arrivals of these flagged ships is not in the top three. Therefore, flags such are Liberia, Marshall Islands and Malta should pay more attention to exhaust emissions. Total amounts of ship emissions ports (in ton.y -1 ), in different studies Total amounts of ship emissions ports (NOx, SO2, CO2, VOC, PM and CO) with number of ship calls and inventory year in different studies are shown in Table 15.