key: cord-0983029-jjz67ke0
authors: Mujal-Colilles, Anna; Guarasa, Javier Nieto; Fonollosa, Jordi; Llull, Toni; Castells-Sanabra, Marcella
title: COVID-19 impact on maritime traffic and corresponding pollutant emissions. The case of the Port of Barcelona
date: 2022-02-23
journal: J Environ Manage
DOI: 10.1016/j.jenvman.2022.114787
sha: 7a555ded94ddeaaf657c68495a9314dab7d950d0
doc_id: 983029
cord_uid: jjz67ke0

The impact of the SARS-CoV pandemic has gone well beyond health concerns, reaching the maritime industry. The study on the environmental impact of shipping industry during COVID-19 pandemic can provide useful insights to propose new management policies regarding shipping operations, both in-port and on the route. We present a case study centred in the Port of Barcelona covering a 30 nautical miles range in the period March to July 2020, during which different levels of restrictions and stringent lockdown measures were enforced. In this paper, we assess the impact of COVID-19 on maritime traffic and its related emissions in port cities using real-time Automatic Identification System (AIS) data. Interestingly, results show that the decline in maritime traffic is not correlated with a decrease in maritime emissions due to changes in vessel operation. During lockdown (March to June 2020), we observed a 27.9% reduction in the number of port calls compared to the pre-lockdown scenario, whereas pollutant emissions show a moderate decrease (1.8% for CO(2)), no significant reduction (SO(2) and PM) or a slight increase (1.3% for NO(x)). This can be directly assigned to changes in vessel operation mode, i.e. vessels switched from Underway to At Anchor or Moored status, during which auxiliary engines are used at higher loads.

The novel Coronavirus disease was first reported symptomatically somewhere between 30 with ∆ 1,2 being the time difference between consecutive waypoints (hours), 1,2 the 157 instantaneous specific fuel consumption (g/kWh) and 1,2 the instantaneous engine power 158 between consecutive waypoints (kW). Sub-indexes : Engine type, i.e. main engine or 159 auxiliary engine; : Fuel type, i.e. LSHFO, MGO or LNG; and : Vessel emission mode, i.e. 160 cruising, anchored, manoeuvring or hoteling. Low sulfur heavy fuel oil (LSHFO), marine gasoil 161 J o u r n a l P r e -p r o o f maneuvering and hoteling modes, MGO and LNG were preferred over LSHFO. 163

The methodology to estimate SFC was based on the parabolic curve developed by (Jalkanen 164 et al. 2012) , in which the instantaneous specific fuel consumption, 1,2 , is computed through 165 a base and a relative value, as in Eq. ( 3 ): 166

where is the base specific fuel consumption (g/kWh) obtained from the IHS Markit 167 database and is the relative specific fuel consumption (g/kWh) based on engine load 168 and computed by Eq. ( 4 ) ( with being the generic vessel power (kW), the power to speed constant (kW·s/m), and 176 the vessel speed (m/s). The power to speed constant is a parameter that needs to be 177 calibrated. To do so, the actual installed power is considered slightly higher than the actual 178 required power. Then, the maximum engine load factor, , is the ratio between the service 179 and the maximum installed power (J. Table 1 lists the daily number of ships in the 30 nm range. Ship activity in the area over the 245 three periods of analysis shows a drop of 8.4% during lockdown and a significant recovery of 246 5.3% during the post-lockdown. Maritime activity around the port area clearly increases during 247 post-lockdown, which is in accordance with the recovery in the number of port calls inside port 248 premises observed in Figure 1 . 249 J o u r n a l P r e -p r o o f AIS status data revealed the specific actions of the vessels. Figure 3 shows that most of the 252 vessels were Underway (55.6%) while the rest were Moored (34.9%) or At Anchor (9.2%). 253

The number of vessels Not Under Command is merely residual (~0.3%), and is therefore, not 254 considered in the research. As seen in Figure 3 March and throughout April, slightly recovered by May and reached a higher, more common 288 value by July. 289 Table 2a shows the daily-mean values averaged over the different sub periods. It indicates a 290 significant decrease in unique passenger vessels (~-27.3%) during lockdown with a more 291 significant decrease in calls (~-60.5%). This is mainly due to the fact that passenger vessels 292 maintaining the activity were only regular ferry lines whereas cruise ships remained docked in 293 the port area, and therefore they only account for a single call. This is confirmed by the 294 increase in passenger vessel average speed also in Table 2b . Cargo vessel average speed 295 is lower being less vessels in the area but the differences between pre-lockdown and lockdown 296 sub-periods in port calls is insignificant. By contrast, tanker vessel speed increased slightly 297 whereas the number of vessels grew up to 4.9% with a minor decrease of less than 1% in 298 J o u r n a l P r e -p r o o f tanker vessel calls. Table 2b presents the variations of daily means in calls, vessels and speed 299 averaged over lockdown with respect to daily means averaged over pre-lockdown days. 300 it is worth noting that there exists a slight fall in tanker and passenger vessel emissions related 366 to the fact that some of these vessels are powered by LNG, with residual sulphur content 367 leading to an overall reduced contribution of these types of vessel to total SO2 emissions. 368 magnitude of peaks in Figure6-SM because NOx emissions are not fuel but engine related. 370

This type of emissions is heavily dependent on engine characteristics and revolutions. This is 371 the main reason why cruising-related emissions are higher than in previous cases, contributing 372 75.8% compared to 14.6% for hoteling, 6.1% for anchored, and 3.5% for manoeuvring mode. 373

These differences are due to the engine configuration introduced in the model, as preference 374 was given to auxiliary engines with higher revolutions and lower emission factors during 375 hoteling compared to main engines operating at their fullest during cruising. The same applies 376 to cargo and tanker vessels. These have greater contributions to NOx emissions than the 377 more modern passenger vessels which, in compliance with Tier II requirements, typically 378 operate with medium-speed engines with lower NOx emission factors than their counterpart. 379

In relation to PM emissions, variability between days was mostly noticed during the most 380 restrictive days, Figure 6 -SM. Since PM emissions are both fuel and engine related, very 381 similar mode trends to those of NOx emissions were found. 382 Our results show that, during lockdown the number of vessels dropped by 8.4% compared to 401 the pre-lockdown scenario, with a rise in tanker vessels of 4.9%, see Table 1 . Moreover these 402 vessels remained within the area for a longer time, which can be demonstrated by the fact that 403 i) the number of vessels At Anchor, drifting for orders or Moored in the port of Barcelona 404 increased, see Figure 3 and Table 2 Table 3-SM, and  406 iii) the number of vessels returning to their homeport increased as global economy was being 407 shut down. It is worth noticing that at the beginning of March fuel prices fell by 65%, reaching 408 negative values by April 20, 2020, (OPEC 2020). Reduced global fuel demand and 409 impossibility to completely stop fuel extraction resulted in larger vessels turning into floating 410 oil storage units drifting at sea waiting for oil buyers. Therefore, the reduction in speed is 411 consistent with the global situation. However, in the present study results show that fuel 412 consumption and emissions did not decrease with the same trend as the mean ship velocity 413

COVID-19 Lockdown Effects on Air Quality by NO2 in the Cities of 500 Barcelona and Madrid (Spain

Existing Emission Calculation Methods Applied to Monitoring, Reporting 503 and Verification (Mrv) on Board'

An AIS-Based 505 Approach to Calculate Atmospheric Emissions from the UK Fishing Fleet

Shipping and COVID-19: Protecting Seafarers as Frontline 508 Workers

Modelling of Ship Engine Exhaust 510 Emissions in Ports and Extensive Coastal Waters Based on Terrestrial AIS Data -An 511

Green Port Initiatives for a More 513 Sustainable Port-City Interaction: The Case Study of Barcelona

Study on the Impact of Cruise Ships Calling at Barcelona in the City Air Quality 517 Bachelor's Thesis

Sea-Web Database

Revised Guidelines for The Onboard Operational Use of Shipborne Automatic 520

Identification Systems (AIS)'. International Maritime Organisation

International Convention for Prevention of Pollution 522 from Ships (MARPOL)

A Modelling System for the Exhaust Emissions of Marine Traffic and Its 524 Application in the Baltic Sea Area. Suplement 1: The Procedure for Calculating the SO x 525 Emission Factor from Fuel Sulphur Content Is given below . The Units Are given in 526 Parenthesis

A Modelling 528 System for the Exhaust Emissions of Marine Traffic and Its Application in the Baltic Sea 529 Area

A Modelling 531 System for the Exhaust Emissions of Marine Traffic and Its Application in the Baltic Sea 532 Area

Extension 534 of an Assessment Model of Ship Traffic Exhaust Emissions for Particulate Matter and 535 Carbon Monoxide

Global Assessment 537 of Shipping Emissions in 2015 on a High Spatial and Temporal Resolution

Ship 540 Engines and Air Pollutants: Emission during Fuel Change-over and Dispersion over 541 Coastal Areas'

Utilization of Vessel Automatic 544 Identification System (AIS) to Estimate the Emission of Air Pollutant from Merchant 545 Vessels in the Port of Kaohsiung

COVID-19 Impact on Global Maritime 548 Mobility'

Disruptions and Resilience 550 in Global Container Shipping and Ports: The COVID-19 Pandemic versus the 2008-2009 551 Financial Crisis

Observatori Fabra. n.d. 'Monthly Summaries

Organization of Petroleum Exporter Countries

Prediction of the Impact on Air Quality of the Cities Receiving Cruise 557 Tourism: The Case of the Port of

Estimation of Vessel Emissions 559 Inventory in Qingdao Port Based on Big Data Analysis

Changes in Air Quality during the 562 Lockdown in Barcelona (Spain) One Month into the SARS-CoV-2 Epidemic

and port calls. In fact, the number of port calls is significantly reduced during lockdown (-414 27.9%, Table 2a ), summed to the reduction in the number of ships underway (-16%, Table 2 -415 SM). At the same time, the mean velocity is slightly reduced (-4.4% Table 2a) , also with a 416 reduction of the total fuel consumption (-4%, Table 3b ). 417Differences between port calls and ships in the area (especially cargo and tanker vessels, 418 Table 2b ) are due to the maritime traffic parallel to the coast that does not call at the Port of 419Barcelona. However, in general, the increase in the number of vessels was partly related to a 420 higher number of vessels with static or quasi-static positions, i.e. Moored and At Anchor during 421 lockdown, see Figure 3a . In fact, it is not that more vessels arrived but rather that vessels did 422 not leave, increasing their time in port or in the anchorage area. The average distribution of fuel consumption per ship type is of the order of that of emissions 435 per ship type for all four pollutants, see Table 3a . This is, cargo vessels consume ~40% of the 436 total fuel and produce the same amount of pollutants with very few differences between them. 437Likewise, tanker vessels consume ~15% of fuel and are the cause of the same trend of 438 emissions. Finally, passenger vessels consume ~45% of fuel and produce the same 439 percentage of pollutants. 440The scenario presented in Figure 6 is clearly related to the fact that passenger vessels operate 441 at high load constantly and sail at very high speeds. Their shipboard services have a high-442 power demand, which explains their higher installed power. It is then clear that, despite the 443 lower number of this type of vessels, they are a major pollutant in the area, and their 444 contribution was related to the increase in average speed. There is a strong correlation 445 between vessel operation mode and overall contribution to pollution, as plotted in Figure 8 difference is due to the way vessels operated, i.e. average speeds reduced by 4.4% and 463 increased number At Anchor and Moored vessels. It is not that more vessels were reported, 464 but that the ones that were already in the 30 nm range stayed over for longer periods. Cargo 465 and tanker vessels managed to weather the situation by adjusting capacity to real-time 466 demand and showed early signs of recovery as of July 31, 2020. However, passenger vessels 467 succumbed badly to travel restrictions due to the pandemic, and although ferry traffic was 468 resumed when reopening was allowed, ongoing uncertainties related to a still spreading virus 469 do not forecast a smooth second semester for the business. J o u r n a l P r e -p r o o f HIGHLIGHTS -The pollutant emissions evolution derived from maritime activity is computed from real AIS data during the COVID-19 restrictions in 2020. -A significant reduction in the ship calls in the port was observed, but only a mild reduction in the maritime traffic was found in the vicinity of the port -Passenger vessels were responsible for more than 40% of fuel consumption and emissions, but represented only 17.2% of all vessels in the region of interest. -Ships underway dominated the total emissions during lockdown, although the number of vessels underway decreased during this period.J o u r n a l P r e -p r o o f

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:J o u r n a l P r e -p r o o f