Towards Efficient Mapping of Greenhouse Gas Emissions: A Case Study of the Port of Tallinn

Abstract

Global, regional and national policies and regulations are providing incentives to reduce greenhouse gas (GHG) emissions in ports, and the first step in this effort is to obtain a detailed overview of the main sources of emissions. The rapid developments in port GHG mapping have led to the need for a case study to assess the effectiveness and practical use of these methodologies and to suggest best practices for ports just starting this mapping process. Here, we present the current state of the art in the assessment of GHG emissions in ports. This analysis enabled us to identify the most promising methodologies to assess GHG emissions in ports in an efficient, reliable and near real-time manner. We then tested the best methodologies and practices that emerged from the review articles to build a GHG assessment system in the Port of Tallinn. Finally, we assess the advantages and disadvantages of current approaches and suggest promising ways forward.

1. Introduction

The acceleration of climate change is one of the most important environmental policy challenges in today’s world. The essence of the problem is that too much greenhouse gases (GHG) are released into the atmosphere due to human activities, which causes temperatures to rise, increases the frequency of extreme weather events and storms and has many other unpredictable consequences. The primary sources of greenhouse gas emissions are transportation, electricity and heat production, industry, commercial and residential buildings, agriculture, land use, land-use change and forestry. The transport sector is responsible for approximately 23% of total energy-related CO₂ emissions. It has been estimated that transport emissions could increase at a faster rate than emissions from other energy end-use sectors if no strong mitigation processes are initiated [1].

Maritime transport plays an important role in the world economy. Although it is one of the most energy-efficient modes of transport, it is nevertheless a large and growing contributor to greenhouse gas emissions. Shipping accounted for almost 3% of global emissions of CO₂ in 2018 [2]. According to a business-as-usual scenario, it is expected that this emission will increase by up to 50% by 2050 [2]. On the other hand, the International Maritime Organization (IMO) aims at halving GHG emissions by 2050. In July 2021, the European Commission set the target to achieve climate neutrality in the EU by 2050, including a target of at least 55% net reduction in greenhouse gas emissions by 2030; these actions also include shipping. Despite this ambitious goal, the decline is slower than expected. Moreover, many seaports still do not report their emissions [3].

International, regional and national policies, as well as regulatory measures, are creating favorable conditions for reducing greenhouse gas (GHG) emissions in port environments. The resulting growth in research focused on reducing emissions is further accelerating the development of effective strategies and technologies, thereby supporting this important sustainability initiative. The first step in this endeavor is to obtain a detailed overview of which are the main sources of emissions. The next step is to identify the most optimal solutions to reduce emissions and then monitor whether these measures have had the desired impact. This process is not trivial as already the mapping of emissions involves a multitude of co-existing sources and interactions with the surrounding environment.

A recent review [3] called for a thorough assessment of port emissions as many maritime ports still do not publicly provide GHG emissions reports. Recent years have seen very rapid developments in this area, and in just a short period of time different ports have developed very diverse solutions for mapping GHG emissions; for extensive reviews of port emission calculation systems see [4,5,6]. Such rapid developments have led to the need for a case study to assess the effectiveness and practical use of these methodologies to suggest best practices for ports that are just starting this mapping process.

In this paper, we first provide the current state of the art of the assessment of GHG emissions at ports. This analysis made it possible to identify rewarding methodologies that enable the assessment of GHG in ports most efficiently, reliably and almost in real time. We then tested the best methodologies and practices that emerged from the review articles to build a greenhouse gas assessment system in the Port of Tallinn. Ultimately, we assess what are the advantages and disadvantages of the current approaches and suggest rewarding ways forward.

2. Current Methods to Assess GHG Emissions at Ports

The amount of greenhouse gases emitted into the atmosphere by human activities is usually converted into carbon dioxide, or CO₂ equivalent (eq) emissions, which is calculated using the relative greenhouse gas effects of different gases, i.e., Global Warming Potential (GWP) [7]. GHG accounting only covers estimates of anthropogenic emissions, taking into account the following gases per 100 years:

• Carbon dioxide (CO₂) released from the combustion of fossil fuels (coal, oil, oil shale, natural gas and peat). The GWP of CO₂ is 1.
• The GWP of methane (CH₄) is 25 times higher than that of carbon dioxide. At the same time, CH₄ emissions are also an order of magnitude lower than those of carbon dioxide.
• The GWP of nitrous oxide (N₂O) is 298 times higher than that of carbon dioxide, but nitrous oxide emissions are also several orders of magnitude lower than CO₂.
• F-gases are emitted when using aerosols (deodorants, various foams), refrigerators and freezing systems, air conditioners, fire extinguishers and chemical cleaners. While emissions of fluorinated gases are low, their potential to cause a greenhouse effect is several orders of magnitude higher than that of carbon dioxide.

Port emission data reporting is quite a new area. [3] analyzed how maritime port emissions are reported as part of sustainability reporting. They studied the world’s top 49 container ports. Less than half of the assessed ports provide publicly available GHG emissions reports.

In recent years, several ports have started to measure their GHG emissions. Ref. [6] provide a comprehensive overview of the methods used to calculate GHG emissions. The review paper highlights two different approaches which are most commonly used when estimating port emissions: the top-down approach and the bottom-up approach. They define the top-down approach as a fuel-based method where fuel sales are converted to fuel consumption and then combined with emission factors to produce emission estimates. The bottom-up approach combines each port activity with its engine energy output. The bottom-up approach usually uses the automatic identification system (AIS) data to estimate ship movements.

In their study, ref. [6] analyzed 32 port emissions studies, of which 29 were bottom-up approaches and 3 used both bottom-up and top-down approaches. They conclude that the general methodologies for estimating ship emissions in ports are based on a combination of data on ship engines, fuel consumption and movements, together with their impact on engine performance and emission factors based on energy consumption. Finally, they propose a common framework based on time in port, emission factors and engine power. Recent papers presenting bottom-up approaches are, for example, ref. [8] by using activity-based methodology, ref. [4] by using VTS data, ref. [9] by summarizing ship activities (navigation, maneuvering and hoteling) and ref. [10] by using ships’ own reports. Top-down approaches are presented, for example, in ref. [11] using activity-based modelling and ref. [12] using spatial projections of ship tracks, voyage and engine specification data.

In another comprehensive review, ref. [4] classified the calculations as top-down, which is the fuel-based approach, and bottom-up, which is the activity-based approach. They argue that the bottom-up approach is generally more accurate and more widely used than the former. For bottom-up approaches, see, for example, [13,14,15]. In recent years, top-down approaches seem to be more popular, see [16,17,18,19,20,21,22,23,24].

In addition, ref. [4] analyzed 50 studies and found that analysis based on engine power was more common than analysis based on fuel consumption. In addition, they argue that AIS data are commonly used to track vessel movements. However, there are some inaccuracies in AIS data and they suggest that a shore-side system of the vessel traffic service (VTS) data should be used instead.

As an example of a bottom-up approach, ref. [25] estimated GHG emissions from shipping activities in container ports. Their method considered engine power, load factor and fuel emission factor. Three ship activities during berthing were analyzed, i.e., maneuvering activities, waiting in the port basin and berthing conditions. Another example is [26], who call their method activity-based modelling. In their model, regional ship exhaust emissions were simply estimated based on the number and size of ships and the type of fuel they used.

As the state of the art, ref. [5] studied European port emissions by allocating a verified amount of emissions reported in the THETIS-MRV database of the European Maritime Safety Agency (EMSA) [27] to individual ports using cargo data. Total emissions are calculated for each type of ship (e.g., gas carrier, container ship, etc.) and then allocated to ports according to how much of the cargo associated with that type of ship (e.g., LNG, containers) is handled in each port. A standard amount of emissions is therefore allocated to each unit of cargo handled. This method may penalize ports that receive goods from nearby areas and is limited to operational emissions from ships in the MRV scope, covering only the last and first legs of voyages to and from the EU and all emissions between EU ports.

Finally, a recent study by ref. [28] raises the issue that environmental performance is much larger than emissions. They propose a conceptual performance assessment tool to evaluate the environmental performance of small seaports. The proposed environmental performance assessment tool has four specific categories: (1) environmental management, (2) responsibility, (3) impact assessment and (4) self-monitoring.

3. Building a Greenhouse Gas Assessment System in the Port of Tallinn

3.1. Mapping the Scene and Choosing the Inventory Approach

The IPCC instructions [7,29] allow emissions to be calculated using three methods: Tier 1, Tier 2 and Tier 3. Tier 1 is the basic method which uses the default value of the specific emission factor of the IPCC methodology in addition to the national basic data. Tier 2 is the method of medium difficulty that uses national basic data and specific emission factors. Tier 3 is the most sophisticated method that requires accurate emission data on different pollution sources. As selected methods and assumptions behind them significantly affect the results of GHG mapping, the port emission guides emphasize that port emissions inventory should have a very systematic and well-reported approach to assessing GHG emissions from all port-related sources, including seagoing vessels, domestic vessels, cargo handling equipment, heavy-duty vehicles, locomotives, heat production, the electrical grid, etc. (e.g., [30]). The GHG emissions mapping will be repeated in subsequent years to plan key decision steps related to port emissions reduction and monitor the impact of the measures implemented. The same guideline also acknowledges that the main objective of GHG mapping at ports is to prepare baseline data for emission reductions, whereas comparing the emissions of different ports is challenging due to large methodological differences.

One of the most important elements of a port GHG emissions inventory is data. It is advisable that data should be collected directly from the sources being inventoried and validated before use. This enables uncertainties associated with different data elements to be assessed. As data collection and validation is the most time-consuming phase of an emissions assessment, our experience at the Port of Tallinn confirmed that this process should be automated as much as possible. Even if an initial investment in the automation activity seems high, it will pay off within only a few years. Specifically, knowledge of the GHG emission mapping process changes very rapidly over time, and if mapping processes are fully or partially automated, it is possible to change different elements of the assessment process with little effort and great flexibility. Moreover, this approach also allows for effective quality control and transparency.

The Port of Tallinn is the largest cargo and passenger port complex in Estonia, which plays an important role in the Estonian transport system and the economy as a whole. The GHG inventory covered the parent company of the Port of Tallinn and the port’s subsidiaries TS Laevad (operates ferries on routes between the Estonian mainland and islands) and TS Shipping (provides icebreaking services in Northern Estonian ports and harbors, and off-shore activities during the summer). As GHG mapping of the Port of Tallinn involves a very large amount of specialized data collection, validation, harmonization and complex analyses and studies, the decision was taken already at the beginning of the mapping process to digitize the emissions mapping as much as possible. This decision is supported by the machine readability of many of the databases used for GHG mapping. The main activities of the mapping and digitalization process were (1) creating a central database containing the tables needed to map GHG emissions and defining relationships between these tables; (2) dynamic linking of various databases, which contain basic information on GHG emissions, to the central database; (3) the development of a web application to allow for (a) the operational collection of baseline GHG emissions data from different operators and (b) the operational modification of emission parameters and the addition of as-yet-undigitized baseline data by an authorized employees of the Port of Tallinn; (4) creating the scripts to calculate GHG emissions and publishing these results in the form of tables and figures on the Port of Tallinn website.

The results of the GHG mapping and process digitalization allow for an operational calculation of GHG emission metrics and for this this information to be published on the Port of Tallinn website. The up-to-date publication of the study and data will enable the customers of the Port of Tallinn, including shipping companies, cargo operators and tenants, to analyze the GHG emissions caused by their activities and seek solutions to reduce their environmental impact. The data can be used by the Estonian Ministry of the Environment and the Environment Agency to monitor the environmental performance of companies, and by environmental experts to analyze environmental impact assessments. The results will also provide a long-term opportunity to monitor the GHG emissions of the Port of Tallinn and its customers operating in the port area and to adjust their activities for the benefit of society and the goal of achieving climate neutrality by 2050.

3.2. Developing Database and Calculation Scripts

3.2.1. Compiling the Existing Databases of the Port of Tallinn into a Central SQL Database

GHG emissions of the Port of Tallinn were assessed according to the following scope areas, based on the ownership or control of the pollution source: Scope 1—direct sources of GHG emissions in the port. These include vessels, vehicles and other equipment, and boiler houses owned by the port; Scope 2—indirect sources of GHG emissions in the port. These sources include purchased electricity and heat for the buildings and infrastructure owned by the Port of Tallinn. Electricity and heat in buildings used by tenants and operators are not included in this scope; Scope 3—other indirect sources of GHG emissions. These sources are related to the activities of tenants and operators and the traffic in the port area and include ships calling at the port, ro-ro cargo, cargo handling equipment, railway locomotives, electricity and heat purchased by tenants and operators (excluding that purchased from the port), means of transport originating from the port area (including taxis, pick-up and drop-off cars, regular buses, tourist buses and vehicles serving the port and ships, etc.), the personal vehicles of port workers and all other sources of emissions from the port area. The calculation of GHG from mobile sources (ships, vehicles and other means of transport and equipment) covered by Scope 3 was limited to activities in the port area.

In order to assess the GHG emissions of the Port of Tallinn, the existing data on direct and indirect sources of emissions in the port were compiled. Input data on heat production and consumption, electricity consumption and vehicle movement into the Port of Tallinn were obtained from different databases managed by the port. In the course of GHG mapping and digitalization, all these databases were dynamically linked to a central MS SQL GHG database, which contains all the tables necessary for mapping GHG emissions and defines the relationships between the different tables to automate the process as much as possible. Then, specific queries were developed to calculate emissions and visualize the results (see the Section 3.2.4 below). This database contains the baseline data related to the GHG emissions mapping of the Port of Tallinn since 2019.

In most cases, it was possible to obtain information directly on the use of different fuels as well as the production and consumption of heat and electricity. However, as not all data were in databases and not all databases are managed by the Port of Tallinn, we sometimes had to take a flexible approach to mapping. For example, it was necessary to use indirect indicators, e.g., the fuel consumption of vehicles owned by the Port of Tallinn was calculated on the basis of the money spent on fuel (accounting data).

A separate web application tool was created (see the Section 3.2.2 below), through which GHG emission information was collected from operators and stored directly in the main tables. All these tables are used by an aggregation script (see the Section 3.2.4 below) which links the underlying data to the calculation rules and writes the result to the master table used by Microsoft Power BI to visualize the mapping results on the Port of Tallinn website.

To obtain the fuel consumption of the vehicles operating in the port area, the Smart Port automatic number plate recognition of the cars was used, and then automated queries were sent to the Estonian Transport Administration to obtain the vehicle-specific emission parameters of the cars that visited the port (for more details, see the Section 3.2.3 below).

In order to increase the mapping accuracy of those sources for which the emission rate is the predominant part of the total emissions, i.e., emissions from visiting vessels, we attempted to make use of existing EU infrastructures. Here, we used data imported from the Automatic Identification System (AIS) of vessel traffic and the European Maritime Safety Agency (EMSA) database on the reporting of the GHG emissions of vessels (see the Section 3.2.3 below). In some cases, VTS data can be a more accurate source of vessel movement information, but in the case of the Port of Tallinn, the VTS reports AIS data. As AIS is more uniform all over the world, we used this source of information.

3.2.2. Collecting Vital Information Using Web Tools

The web application was created to meet three major sub-objectives of the GHG emission mapping: (1) the web application allows GHG-specific emissions baseline data from operators to be operationally collected; (2) the same web application allows authorized employees of the Port of Tallinn to operatively modify the emission parameters and other baseline conditions necessary for the calculation of GHG emissions, to edit the entire existing baseline dataset and to add baseline data that have not yet been digitized; (3) the web application also displays initial thematic aggregated assessments, which allows for the identification of possible input errors as well as harmonizes and validates data. If required, the data tables can also be downloaded in csv, excel and pdf format, copied elsewhere or printed out. Once the administrator has finished validating the data, it is possible to transfer all the data in the application to the appropriate tables in the MS SQL database.

A questionnaire was developed to collect GHG data from cargo operators in an efficient and harmonized manner. The questionnaire was designed with a level of detail that would allow analyses to be carried out in order to make recommendations for reducing GHG emissions. The questionnaire was published online and an invitation was sent to all cargo operators in the Port of Tallinn to participate in the mapping of GHG emissions. Operators were asked for information on heat production and consumption, electricity consumption, the fuel and energy consumption of stationary equipment and vehicle movements which do not leave the port territory. The latter was necessary because the mapping of vehicles based on the car registration number does not include vehicles that only move in and never leave the territory of the port. The Port of Tallinn also acts as an intermediary for heat and electricity sales. In this case, when an operator bought heat and electricity from the Port of Tallinn, they did not have to provide any data and the data were automatically entered into the central database. When the operator submitted the questionnaire to the Port of Tallinn, the answers were locked and further correction by the operator was only possible if the administrator opened the same questionnaire again. A web-based application also allowed immediate validation of the basic data. In the case that an operator entered any incorrect data, the operator could be immediately notified of the problem fields by error messages, but some errors (e.g., amounts of fuel entered with units of measurement) could also be removed automatically by means of quality control.

To access the application username and password are needed. There are two types of users: administrators and operators. Administrators have access to all parts of the application. Each operator logs in to the web application with his own password and they only have access to their own company’s data in order to edit the basic GHG emission data previously entered for their company or to add new data.

3.2.3. Linking Central Database with External Information Sources

Smart Port is the automatic traffic management system of the Port of Tallinn that is used to map the traffic of vehicles in the port area (taxis, buses, passenger escorts, guests, ship and port services). This system stores the registration number of each vehicle that passes through a barrier. To assess which categories of vehicles (e.g., M1, N2) are visiting the port area and which fuel (e.g., petrol, diesel) is used by these vehicles and how much, these registration numbers were used to send automated queries to the Estonian Transport Administration requesting vehicle-specific emission parameters for these vehicles. Existing data were converted into frequency distributions, which were then used to calculate the probabilities of port-based vehicle types for those vehicles for which no information was available from the Estonian Transport Administration. Importantly, the ID number of the barrier defines the expected mileage of each vehicle in the port area. Consequently, the type of vehicle and fuel, the emission parameters (e.g., expected vehicle-specific fuel consumption per 100 km in urban area) and the expected mileage in the port area allow an indirect estimation of the GHG emissions of each vehicle in the port area.

During the mapping process, we also tried an alternative approach to identify the vehicle type of those vehicles that were not registered in the Estonian Transport Administration’s database. A photograph of each vehicle was taken with the aim of classifying the vehicles using image analysis. Unfortunately, the first attempts to classify the vehicle type on the basis of these photographs (an AI algorithm) yielded results with an accuracy of only 60%. Thus, at this stage it is not yet reasonable to apply the image recognition algorithm; instead, we used a statistical approach (i.e., average vehicle emissions) for those individual vehicles for which emission parameters were not available.

Blue Flow is a monitoring system that measures real-time fuel consumption from ferries owned by the port’s subsidiary TS Laevad. The system stores information on the fuel consumption of these ferries and transmits this automatically to the MS SQL database. Specifically, BlueFlow sends a monthly structured report file in *.csv format to the Port of Tallinn’s mail server, where this information is automatically loaded into the central MS SQL database.

AIS collector is an application created and installed on the server of the Port of Tallinn, which reads information of the vessels’ automatic tracking system shared by The Estonian Transport Administration (i.e., the automatic identification system, AIS) and converts this into a digital format that scripts can read and analyze further. These data include vessel name, details on ship navigational status (e.g., underway using engine, moored), location, speed and heading. During conversion, we validated possible incorrect entries in the ship type, status, IMO number (i.e., Ship Identification Number of the International Maritime Organization), time stamp and coordinates fields and erroneous data were removed from the database. Surprisingly, several ship types (e.g., passenger ships and tankers) or different ship dimensions (vessel length, draft) were also given for the same IMO number.

We used IMO number as a common denominator and developed a script to link data collected by the AIS collector with the vessel-specific GHG emission coefficients obtained from the THETIS-MRV database of the European Maritime Safety Agency (EMSA) [27]. The THETIS-MRV reports verified amounts of GHG emissions and other relevant information from large ships calling at EU and EEA ports. The system covers GHG emissions from ships underway and at berth and shares other related information such as energy efficiency, distance travelled, time spent at sea and cargo carried. In the process of this digitization, we used information on the average annual fuel consumption and GHG emission rate reported for the vessel for the distance travelled (tons of fuel consumption and kg GHG emission per nautical mile), the total annual GHG emissions at sea and at berth (tons) and time spent at sea and in port (hours). A script was created that first used AIS data to estimate the distance travelled by each vessel in the port area. AIS data were often inaccurate when calculating a vessel’s time in port. Instead, the port’s own data (FlexPort database) were used. Then, the script used ship-specific emission factors imported from the EMSA database to calculate the GHG emissions of each vessel separately for maneuvering and mooring. Our methodology is similar to that used by [5], but differs in that we calculated the total GHG emissions of each ship visit to the port area (both maneuvering and berthing), independent of cargo load. Moreover, as not all ships fall within the scope of MRV, we have ensured that emissions from these ships are reported using alternative methods (e.g., spent fuel).

3.2.4. Calculating GHG Emissions by Source

In order to calculate GHG emissions from existing data in the MS SQL database, the calculation rules were defined. These rules describe the functional relationships between different tables and include all formulas for calculating GHG emissions. After the GHG emission calculations, the script merges all data into a single table and publishes in a format readable directly by the Microsoft Power BI environment.

For the assessment of GHG related to the activities of the Port of Tallinn, we followed the methodological basis of the Environmental Footprint Assessment Guidelines (https://envir.ee/kliima/toetavad-materjalid/organisatsioonide-khg-jalajalg accessed on 1 December 2022) developed by SEI Tallinn in 2021. The developed calculation methodology is based on widely accepted international guidelines and standards for determining the greenhouse gas footprint, including the Greenhouse Gas Protocol, the IPCC Assessment Report and ISO 14064-1:2018. At the same time, it incorporates specific assumptions and conditions that are unique to Estonia. As a result, this approach harmonizes the methodological basis and data set, making it easier for Estonian organizations to accurately calculate their environmental footprint. Consequently, this is helping organizations to harmonize their GHG footprint calculations.

The following generalized formulae have been used to calculate GHG emissions. The heat values and specific emission factors in the formulae depend on the type of fuel, energy and equipment, as well as the year.

• Vehicles and Mobile Devices:

  Fuel consumption (TJ) = Fuel consumption (in liters) × Heat value (TJ L⁻¹)

  CO₂ emissions (in tons) = Fuel consumption (TJ) × CO₂ specific emission factor (tCO₂ TJ⁻¹)

  CO₂ emissions in CO₂ equivalent (in tons) = CO₂ emissions (in tons) + 25 × CH₄ emissions (in tons) + 298 × N₂O emissions (in tons)

  CH₄ emissions = Fuel consumption (TJ) × CH₄ specific emission factor (t CH₄ TJ⁻¹)

  N₂O emissions = Fuel consumption (TJ) × N₂O specific emission factor (t N₂O TJ⁻¹)
• Electricity Consumption:

  Fuel consumption (TJ) = Annual electricity consumption (MWh) × Heat value (TJ MWh⁻¹)

  CO₂ emissions (in tons) = Fuel consumption (TJ) × CO₂ specific emission factor (tCO₂ TJ⁻¹)

  CO₂ emissions in CO₂ equivalent (in tons) = CO₂ emissions (in tons) × CO₂ equivalent specific emission factor (tCO₂ TJ⁻¹)

Stationary Devices: the GHG emissions of stationary devices were calculated based on whether the device consumed fuel (calculated as vehicles) or electricity (calculated as electricity consumption).

• Heat Production:

  Fuel consumption (TJ) = Fuel consumption (m³) × Heat value (TJ m⁻³)

  CO₂ emissions (in tons) = Fuel consumption (TJ) × CO₂ specific emission factor (tCO₂ TJ⁻¹)

  CO₂ emissions in CO₂ equivalent (in tons) = CO₂ emissions (in tons) + 25 × CH₄ emissions (in tons) + 298 × N₂O emissions (in tons)

  CH₄ emissions = Fuel consumption (TJ) × CH₄ specific emission factor (t CH₄ TJ⁻¹)

  N₂O emissions = Fuel consumption (TJ) × N₂O specific emission factor (t N₂O TJ⁻¹)
• Heat Consumption:

  Fuel consumption (TJ) = Annual heat consumption (MWh) × Heat value (TJ MWh⁻¹)

  CO₂ emissions (in tons) = Fuel consumption (TJ) × CO₂ specific emission factor (tCO₂ TJ⁻¹)

  CO₂ emissions in CO₂ equivalent (in tons) = CO₂ emissions (in tons) × CO₂ equivalent specific emission factor (tCO₂ TJ⁻¹)

According to this methodology, for all mapped sources, generalized factors were compiled that were used to convert measured quantities (e.g., MWh, petrol or diesel consumed) into CO₂ equivalent. However, the specific emission factors outlined in this methodology do not allow for a detailed calculation of emissions from visiting ships and vehicles. As these emission sources account for a significant proportion of the total emissions, it was considered that a more detailed approach was needed (see the Section 3.2.3 for ship- and vehicle-specific GHG emission factors). A flowchart showing how different data sources were linked and used in the common algorithm to produce the GHG emissions report is shown in Figure 1.

The established GHG mapping and assessment methodology was then used to assess the emissions of the Port of Tallinn for the year 2021, selected due to the often-delayed publication of annual emission parameters. Unsurprisingly, the main source of GHG emissions was visiting vessels, which accounted for 55% of the total. Other significant contributors were electricity consumption (28%) and heat generation and consumption (10%). The port’s direct and indirect emissions, categorized as Scope 1 and 2, contributed a minor 1.9% of total emissions (

Table 1. The individual contributions of different sources per scope to total greenhouse gas emissions in the Port of Tallinn in 2021.

Emission Scope	Source of Emission	CO2 Equivalent
Scope 1	Total	1199
	Heat	615
	Machinery	554
	Vessels	30
Scope 2	Total	1208
	Electricity	826
	Heat	382
Scope 3	Total	126,507
	Electricity	35,363
	Heat	7590
	Machinery	12,259
	Vessels	71,295
Grand Total		128,914

).

Our findings suggest that the most effective measure to reduce GHG emissions at the port would be to implement berth power solutions powered by green, zero-emission electricity sources, as the majority of emissions from visiting vessels come from hotel operations and diesel engine use. In addition, there are significant benefits to be gained from converting all other electricity needs to green electricity. If these strategies were implemented together, a potential 83% reduction in the port’s greenhouse gas emissions could be achieved.

4. Advantages and Disadvantages of the Current Approaches and Rewarding Ways Forward

Mapping greenhouse gas (GHG) emissions in port areas is becoming increasingly important as countries strive to reduce their carbon footprint and mitigate climate change. In this paper, we used a hybrid approach to apply, combine and develop several existing methodologies to provide accurate and reliable estimates of GHG emissions in the Port of Tallinn. Mapping can be performed using top-down or bottom-up systems [6]. We used the bottom-up system because the solutions developed were easier to implement and faster to update. Wherever possible, direct inventory-based approaches were used to collect data on activities and equipment in ports and to use these data to estimate GHG emissions. These approaches required a significant amount of data collection and analysis but could provide detailed and reliable estimates of GHG emissions.

As data collection and validation is the most time-consuming phase of an emissions assessment, we tried to automate this process as much as possible by linking many machine-readable databases. A good approach was to use the combined databases of Smart Port and the Estonian Transport Administration for the vehicle GHG mapping. Similarly, the combined AIS and Thetis-MRV data were used to estimate emissions from ships in ports. The use of advanced algorithms enabled the analysis of large amounts of data to provide reliable and near real-time estimates of ship-specific GHG emissions. For ships that were owned by the Port of Tallinn, we were also able to take advantage of on-board continuous emission monitoring systems installed on ships (e.g., Blue Flow) to accurately measure their emissions in real time. Additionally, because we used mostly automatic inputs to update the central GHG mapping database, the amount of manual work required to enter values was low, which is a clear advantage over many other practices.

The mapping of emissions in the Tallinn Port area confirmed the assumption that shipping accounts for a very large share of GHG emissions. Therefore, great care was taken to estimate the GHG emissions from shipping as accurately as possible. The methodology developed is based on the best available information, i.e., using the AIS system to map the movement of vessels in the port area and linking these data to actual GHG emission values reported and verified for these vessels in the Thetis-MRV database.

The advantages of the developed solution over previous practices are the automation and accuracy of the calculations. Emission studies carried out by many other ports generally do not use measured and verified ship-specific emission parameter values [5], but are based on theoretical relationships between the nature of ship engines and GHG emissions. However, these emission rates are often generalized to the type of ship and do not take into account its specificities. For example, the same conversion factor is used for all tankers to calculate GHG emissions during maneuvering or berthing [31,32,33,34]. This situation is due to the fact that the annual obligation to report GHG emissions from ships to the European Commission only came into force in 2019, when data for 2018 were submitted. As a consequence, ship-based GHG emission parameters are only available in the EMSA/THETIS-MRV database for the most recent years.

This hybrid approach provided a comprehensive assessment of GHG emissions in the Port of Tallinn, including emissions from ships, cargo handling equipment and other sources. Advanced methods and algorithms have helped identify areas where emissions are highest and target measures to reduce emissions. As a significant part of the data are collected in near real time, the port can quickly identify and respond to changes in emissions. Efficient data collection and automated calculations have significantly reduced the time and effort required to collect data. This also helps to reduce costs and improve the efficiency of GHG emissions assessments. Importantly, accurate and reliable GHG emissions data are becoming increasingly important for environmental compliance, and the use of these methodologies can help ports comply with regulations and avoid potential penalties.

One of the biggest challenges in mapping GHG emissions has been receiving timely feedback from operators. Therefore, it is reasonable to simplify the questionnaire sent to operators for the following years. As the capabilities of automated mapping continue to grow, any fields that do not concern the given operator and/or that can be measured automatically can be left out of such a dynamic questionnaire. A good example of this is the mapping of all vehicles entering or leaving the port area at the barriers (i.e., the Smart Port system). Surprisingly, many operators were unable to estimate the number of vehicles that visited them each year.

Another major challenge is the standardization of GHG emissions mapping in ports. In the absence of a commonly agreed methodology, many ports only estimate Scope 1 and 2 emissions and do not consider Scope 3 emissions due to the complexity of assessment, making it unreasonable to compare GHG emissions between ports. The use of standardized methodologies can help ensure that GHG emissions assessments are consistent and comparable between ports. This will also facilitate the benchmarking and sharing of best practices between ports, leading to more effective GHG mitigation strategies.

The use of new technologies is likely to be integrated into GHG mapping in port areas. These technologies can provide more accurate and detailed information on GHG emissions in ports, making it easier to identify emission sources and implement mitigation strategies. Overall, it is important to keep pace with the use of the most rewarding and up-to-date methodologies for assessing GHG emissions in ports, as this can provide numerous benefits, including improved accuracy, efficiency, transparency and compliance, all of which can help reduce emissions and improve the sustainability of ports.

Last but not least, the use of these advanced methodologies leads to better collaboration between stakeholders. Importantly, such collaboration, including port authorities, shipping companies and local communities, will be critical to achieving meaningful GHG reductions in port areas. Stakeholders can work together to share data, identify emission hotspots and develop joint mitigation strategies. For example, the expansion of carbon markets provides opportunities for port authorities and shipping companies to generate revenue from GHG reductions. If carried out on a sustainable basis, this may encourage greater investment in GHG mitigation strategies in port areas [35].