1 Introduction

In the last decade, the world merchant fleet dedicated to international trade has increased. In January 2016, the ship world fleet grew by 3.5% and reached 1.8 billion DWT that consisted of 90,917 vessels including bulk carriers, oil tankers and container carriers. The sea shipping industry is responsible for about 90% of world trade (in tons) that is representing a total international cargo over 9841 millions of tons [26].

Consequently, it produces a growth of fuel consumption and Green House Gas (GHG) emissions at sea. The GHG emission of ship engines have raised the concern of International Maritime Organization (IMO) on the consequences for environment and human health.

In addition, IMO first adopted MARPOL Annex VI in 1997. It limits the main air pollutants in ships exhaust gas, including sulphur oxides SO\(_X\).

Following entry into force of MARPOL Annex VI, the main changes are a progressive reduction in emissions of SO\(_X\), NO\(_X\) and particulate matter (PM), as well as the introduction of emission control areas (ECAs). ECAs are created to further reduce emissions of those air pollutants in designated coastal areas.

Nevertheless, the shipping industry is facing huge challenges. First, main concern of ship owners is to reduce operating cost and maximize incomes, whereas the fuel price has increased significantly over the years. Second, customers such as shippers and freight forwarders are increasingly demanding on-time delivery [12]. Third, ships must fulfil the rules regarding environmental restrictions implemented by the IMO (emissions limitations).

In the case of fuel cost, the fuel consumption of sea vessels depends heavily on the steaming speed. The practice of slow steaming (SS) (speed reduction denoted in this paper as SS) has become more common in cargo fleets especially in liner shipping [2].

The study of SS practice in liner shipping became more frequent in the last years [2, 23, 27]. There are classic studies [19], which deal with the use of SS. Moreover, in the solid bulk and oil market the freight formation model (supply curve) presented by several authors, such as [21], provides for SS to reduce supply and reduce costs. In these cases unlike liner services (container ships), the lines are not regular, and there is no scheduled service, where the number of ships and the frequency set define the speed [17].

There are studies linking emissions reduction and/or SS, on other types of ships other than container ships, but there are indeed many more publications analyzing the problem in container ships [1, 5, 14].

SS is also a market practice in the bulk carriers trading. However, only few recent studies are studying the economic impact of speed reduction and emissions in this specific case.

In addition, delivering on time is a difficult challenge due to port congestion, inefficient port operations, extreme weather conditions, machine breakdowns and other factors [12]. Besides, some industries criticize the SS because it is necessary to build more ships to transport the same quantity of product and achieve targets of delivery time.

Lastly, one positive effect of SS is that it reduces GHG emissions that are proportional to the amount of fuel burned [2].

Recently, they have been significant advances in SS approach not only in the study of economic aspects [7, 15, 16, 18], but also in others areas as resistance [24], shipping time, bunker cost, ability to deliver on time [12] and environmental advantages [2, 6].

In this paper, the influence of SS on one fleet of 13 bulk carrier ships is analysed through simulation. This model uses criteria based on speed, fuel oil consumption, distance travelled, cargo and emissions quantity (CO\(_2\) and SO\(_X\)). Then, the results of a simulation model suggest that SS implementation is a possible solution to turn navigation more profitable in economic and environmental aspects for bulk carriers.

2 Methodology

This section presents the developments of the model including the explanation of the methodology, the database (DB) used in the analysis as well as the criteria selection. Thereafter, the definition and implementation of the models are presented. The main steps of the proposed methodology are shown in Fig. 1.

The goal of this work is to evaluate the potential economic and environmental benefits of new navigation condition: SS and ultra-slow steaming (USS).

The proposed framework consists of a discrete-event simulation (DES) model to represent the voyage process. Both economic and environmental parameters were considered to assess the influence of SS and USS in a fleet of bulk carrier ships. The design of the alternatives is based on the review literature and expertise.

This approach is similar to previous studies proposed by [4]. The previous mentioned study analyses the influence of the SS and USS on one fleet of 15 bulk carriers considering only the fuel consumption, distance, and cargo. In addition to these variables, the present model consider the emissions that is a critical parameter to fulfil current regulations.

Several alternatives were designed for different variable settings. DES was then used to evaluate key performance improvements.

A DES model can reproduce an existing complex system using a sequence of events and provide to the decision-maker a vision on how that system might perform [22].

The model is developed to balance and evaluate the operational decision on speed reduction with the factors on bunker cost, fuel oil consumption, distance travelled, cargo quantity, carbon dioxide emission and sulphur oxide.

2.1 Database and input analysis

The simulation stage was based on a valid process model. The main steps of the proposed methodology are shown in Fig. 1.

Fig. 1
figure 1

Flowchart of the methodology

Full size image

In this study, 13 bulk carrier vessels from a ship fleet of a unique ship owner are considered. Table 1 gives a highlight on ship main features.

Table 1 Ship fleet parameters (13 vessels)
Full size table

The DB represents a period of 2.5 years. That means 6844 records corresponding to 223 voyages (one-way travels).

The information available is obtained in laden and ballast conditions. The simulation was developed to split the ship fleet in three ship categories based on maximum displacement of each one.

Ship category 1 is composed by vessels that have a maximum displacement between 167,963 and 191,668 tons. It represents 54% of the whole fleet. Ship category 2 is composed by vessels that have a maximum displacement between 201,550 and 224,978 tons. It represents 31% of the whole fleet. Finally, the ship category 3 is composed by vessels that have a maximum displacement between 259,711 and 280,313 tons. It represents the 15% of the entire fleet.

The model simulates one-way voyages of the vessels using ARENA software for both ballast and laden conditions. For each sub-model (original, SS or USS), inputs and outputs are detailed in the Table 2.

Table 2 Input and output values for laden and ballast conditions
Full size table

The present study includes the CO\(_2\) and SO\(_X\) emissions in the model. IMO defined the Energy Efficiency Operational Indicator (EEOI). It is an expression of emission efficiency in the form of CO\(_2\) emitted per unit of transport work [9]. The ECO\(_2\) is given by Eq. 1, ECO\(_2\) represents the amount of CO\(_2\) emission released into the atmosphere, where j is the fuel type, FC is the mass of consumed fuel in kg, \(C_{\rm F}\) is the fuel mass to CO\(_2\) mass conversion factor in kg-CO\(_2\)/t-fuel, see Table 3. M is the cargo carried in tons and D is the travel distance in nautical miles. A higher value of this indicator denotes a lower efficiency

$$\begin{aligned} {\rm ECO}_2 = \dfrac{\sum _{j} FC_{j} \times C_{{\rm F}j}}{M \times D} \end{aligned}.$$
(1)
Table 3 Fuel mass to CO\(_{2}\) mass conversion factors \(C_{\rm F}\) in [kg-CO\(_2\)/t-fuel] [10]
Full size table

Today there is commercial software’s used to estimate the EEOI value before the trip. Even though the indicator EEOI is not enough to measure the overall efficiency of ships, it indicates the amount of CO\(_2\) released into the atmosphere.

The \({\rm ESO}_X\) is given by Eq. 2, \({\rm ESO}_X\) represents the amount of SO\(_X\) emission released into the atmosphere per unit of transport work, where j is the fuel type, FC is the mass of consumed fuel in kg, \(C_{\rm S}\) is the fuel mass to SO\(_X\) mass conversion factor in kg-SO\(_X\)/t-fuel, see Eq. 3. M is the cargo carried in tons and D is the travel distance in nautical miles

$$\begin{aligned} {\rm ESO}_X = \dfrac{\sum _{j} FC_{j} \times C_{Sj}}{M \times D} \end{aligned},$$
(2)
$$\begin{aligned} C_{\rm S} = SF \times 20 \times VB \end{aligned},$$
(3)

where SF is the percentage of sulphur present in the fuel and VB is the volume of bunker in Tons of Fuel.

The \({\rm ESO}_X\) is in kg-SO\(_X\)/t-fuel and it depends on the type and sulphur content of the fuel used by the ship [13]. It has to multiply total bunker consumption by the percentage of sulphur present in the fuel and subsequently by a factor of 20 to compute SO\(_2\) emissions. The 20 SO\(_X\) factor is exact and comes from the chemical reaction of sulphur and oxygen to produce SO\(_2\). A higher value of this indicator denotes a lower efficiency.

The SF is calculated based on the actual sulphur content in the fuel, see Eq. 5. The dilution factor is calculated by the average quantity of sulphur content that depends on the type of bunker fuel (IFO or MDO) on-board and the quantity of the fuel in this operation in the port of refuel. The fuel quality can be altered depending on the refuelling port, which influences the quality of the bunker.

It is assumed that the average sulphur content (SF) for IFO and MDO are 2.5 and 0.25%, respectively, if the information about quality of fuel (sulphur content) is not available [3].

$$\begin{aligned} {\rm SF} = {\rm ABQ} \times {\rm ASC}+{\rm RBQ} \times {\rm RSC} \end{aligned},$$
(4)

where ABQ is the actual quantity of bunker on-board in tons, ASC is the percentage of concentration of SO\(_X\) of the bunker on-board, RBQ is the quantity of bunker to be refuelled in tons, RSC is the percentage of the average concentration of SO\(_X\) of bunker to be refuelled.

The models represent 360 days (1 year), and it is running for 200 iterations to verify the model convergence. Semi-random numbers have been altered between each iteration.

Currently in shipping market, the ships frequently sails in part-load condition and in different speed compared with design speed and not always with an efficient speed [8].

Model 1 represents the original (ORI) condition of the system and others models SS strategies. The main parameter to define the model is the speed. Model 2 represents the SS condition of the system where speed is decreased by 2 knots compared with the ORI model. Model 3 represents the USS condition of the system where speed is decreased by 4 knots compared with the ORI model.

The speed reduction values used in this paper are an outcome of the European Research project entitled ULYSSES of the 7th Framework Programme for Research and Technological Development [6]. The objective of this project that has been conduced between 2011 and 2013 was to demonstrate that the efficiency of the world fleet can be increased to a point where the following CO\(_{2}\) targets are met, through a combination of USS and complementary technologies:

  • Before 2020, reducing greenhouse gas emissions by 30% compared to 1990 levels;

  • Beyond 2050, reducing greenhouse gas emissions by 80% compared to 1990 levels.

The ULYSSES project focused on bulk carriers and tankers, as these ship types produce 60% of the CO\(_{2}\) from ocean-going vessels. The main results defined the requirements for USS, including technical, economic, safety and environmental factors.

In this study, the inputs parameters are fixed for laden (LC) and ballast (BC) conditions as well as for original, SS and USS strategies. Total consumption of fuel and average daily emissions parameters are modified due to speed effect. The simulation work flow used is the same for Original, SS and USS as shown in Fig. 2.

In the simulation, the ships are created and initialized according to specific rules. It has been calculated based on the average travel time per year of whole ship fleet and the number of ships.

The average time between arrivals is about one every four days. Three sub-processes have been created to map the three ship categories defined before. Each of them is respecting the assignments sequence shown in Fig. 2.

Fig. 2
figure 2

Workflow of the voyage simulation of the bulk carrier fleet

Full size image

As an illustration, Table 4 shows the distributions used to define the input parameters of ORI model of ship category 1.

Table 4 Input distributions for ship category 1 ORI model for laden (LC) and ballast (BC) conditions
Full size table

The voyages implemented in each sub-process correspond to Eq. 5 where \(T_{\rm V}\) is the voyage time distribution in days, D is the distance distribution in nautical miles, and S is the average daily speed distribution in nautical miles per day.

$$\begin{aligned} T_{\rm V} = \dfrac{D}{S} \end{aligned}$$
(5)

The model estimate the information above mentioned as results, the total of cargo transported, the total of fuel consumed, the total CO\(_2\) emission, and the total SO\(_X\) emission.

The following variables are evaluated for each ship and each iterations: \(L_{\rm T}\) is the cargo transported, \(T_{\rm FC}\) is the total fuel consumed, \(T_{\mathrm{CO}_2}\) is the total CO\(_2\) emission and \(T_{{\mathrm{SO}}_X}\) is the total SO\(_X\) emission.

3 Results and discussion

The influence of SS and USS on the fleet of 13 bulk carrier ships is shown in this section.

The result for the three models is given in Table 5. We observe the amount of cargo transported (in tons) in each of the proposed alternatives (ORI, SS and USS), total consumption of fuel of the fleet (in tons), and the total of emissions (CO\(_2\) and SO\(_X\)) in tons for a fixed period of 1 year.

Table 5 Result descriptions in tons for ORI, SS and USS models
Full size table

The results show that the values of consumption decrease by 49 and 15% in SS and USS, respectively. The values of CO\(_2\) emissions decrease by 60 and 23% in SS and USS, respectively. Finally, the values of SO\(_X\) emissions decrease by 43 and 13% in SS and USS, respectively, see Fig. 3.

Fig. 3
figure 3

Total consumption of fuel in tons right y-axis and total consumption of emissions in tons left y-axis for ORI, SS and USS models

Full size image

Figure 4 shows the average of fuel consumption for each ship and the number of ships that leave the round trip in the simulation. The round trip represents a laden voyage followed by a ballast voyage. That is important to analyse the advantage of SS and USS to each ship category. In all cases, USS strategy is the most advantageous.

Fig. 4
figure 4

Examples of average of fuel consumption (tons) reduction for ship category 1, 2 and 3, including the quantity of ships that are necessary to complete this simulation

Full size image

The major disadvantage of SS strategy for shippers is the longer shipping times. This factor is associated with the cargo transported. In the model proposed the total cargo transported by each ship in SS and USS strategies is compared with the total cargo transported in original model. Consequently, to move the same amount of cargo (in same number of days), it would be necessary to use a larger number of vessels in the fleet.

The difference of the cargo transported between the models shows that the fleet needs one more ship to transport the same quantity of cargo in both proposed strategies (SS and USS), see Table 6. It is worth to mention that the extra waiting time that may suffer the freight owners due to the longer shipping time is not considered in this study. Moreover, the cargo considered in the study is only composed by minerals whereas cargo quality is not affected by extra shipping time. In the case of grains, beans, corns and cereals, this concern might be an issue that should be carefully examined.

Table 6 Cargo transported by the fleet and the percentage of cargo in the three navigation strategies
Full size table

This analysis is general, it considers the increase of ship in a global way. In future works, the economic analysis can be improved including the increment of ships with a separate analysis for each ship category to transport the same quantity of cargo in SS and USS strategies.

According to [2], to determine the sustainability of SS and USS strategies, the cost of added vessels to a service under this strategy as well as the increase in costs for shippers must be considered. Capital and Operational costs vary according to the number of vessels added and their characteristics.

Assuming the fuel price about 257 USD for a metric ton (Ship-Bunker, 2016), cost savings by fuel consumption of SS and USS are assessed, respectively, to 9.3 and 15.3 millions of USD annually as shown in Table 7.

Table 7 Annual costs and fuel consumption saving in USD and projection to the next 10 years
Full size table

Sea costs in shipping are composed of: Capital expenditure (CAPEX), Operational expenditure (OPEX) and Travel cost at sea. In this study, OPEX costs are based on Moore Stephens reports on ship operating costs, see Table 8 [20]. The CAPEX depends heavily on the ship price, which suffered large fluctuations in recent years. The CAPEX estimate (US$/day) considered three price scenarios for the 2006/2016 period, based on Clarksons data: Highest annual average, lower average annual and average for the period, see Table 9. The CAPEX estimate considers a 10% year of discount rate, a 2.62% year of Commercial Interest Reference Rates (CIRR)- 2016 average, 15 years of use-life.

Table 8 OPEX estimation, based on [20]
Full size table
Table 9 CAPEX estimation price, based on Clarksons database
Full size table

Here, the Travel cost at sea (TCS) or running cost varies for ORI, SS and USS strategies. The TCS is composed by: fuel cost, port charges, channel crossing rates, commissions, cleaning holds and tanks, and other relate expenses. Considering that the ships take the same routes with the three ship categories, only fuel costs could be considered variables [11]. Therefore, the variations of fuel consumption are enough to take into account the variation of TCS.

Considering that the CAPEX price of a new bulk carrier vessel (see Table 9) is in average about $17,339 $USD/day, the projection of CAPEX cost in 10 years is about $63.3 $mUSD. The annual costs and consumption saving can be assessed and projected on 10 years. Then, profitability of SS and USS can be evaluated, see Table 7 [25].

Figure  5 shows the effectiveness analysis of SS and USS simulation, in SS strategy. The emissions are reduced to 57 and 22%, the consumption of fuel is reduced to 49 and 15%, and the total cost in the next 10 year is reduced to 99 and 66%, respectively, in comparison with the ORI model.

Fig. 5
figure 5

Effectiveness analysis of SS and USS simulation

Full size image

The convergence is verified for each model, e.g. cargo transported convergence after 200 iterations is shown in Fig. 6. Semi-random numbers have been altered between each iteration. Figure  6 presents in x-axis the number of iteration and in y-axis the value calculated in the simulation of cargo transported in tons. The three strategies are evaluated in 200 iterations each one with semi-random numbers automatically generated by the ARENA software.

Fig. 6
figure 6

Convergence of cargo transported in tons for the SS model

Full size image

This study considers a sensitivity analysis to show how the CAPEX and bunker value are affecting the total cost of the different strategies. Tables 10 and  11 show the results of a sensitivity analysis comparing the total cost for 10 years prediction by bunker price and CAPEX variation. This results indicate that the use of USS is more profitable than SS and ORI. It is noticed that when the fuel price is low or the CAPEX is high the SS strategy may not be any more an efficient decision.

Table 10 Sensitivity analysis comparing the total cost for 10 years prediction by CAPEX variation
Full size table
Table 11 Sensitivity analysis comparing the total cost for 10 years prediction by fuel price variation
Full size table

4 Conclusions

Main results evidenced the reduction of transported cargo by less than 8% for two conditions (SS and USS), while the total fuel consumption decreased by almost 51 and 85%, respectively.

This study prove that the speed reduction in USS strategy through just-in-time-arrival is possible without reducing the capacity of the maritime transport systems, with the increment of one ship unit. This paper shows that SS has reduced emissions by around 22% over 1 year; it fulfil the target of IMO.

Savings in operational costs, considering fuel consumption and emissions (CO\(_2\) and SO\(_X\)) invites us to reflect on the number of extra vessels required to fulfil the cargo transport objective. Due to the need to increase the number of ships to move the same amount of cargo transported in the same time, USS is more profitable than ORI and SS conditions.

An important consideration for USS strategy is the fact that at this slow speed the engine might be not any more efficient. Therefore, it require additional investigations to confirm the viability of this scenario.

The findings of this study bring useful insights about different simulation approaches used as decision support systems in the field of navigation strategies. This study increment the literature about the use of DES focused on SS and USS for bulk carrier ships.

The models provide helpful insights for ship owners and tactical transportation planners.

This simulation contribute to the limited literature that uses SS with DES. This paper explores the use of DES as modelling tools used to support decision-making.

The use of DES can help to simulate scenarios with real historical data, assisting ship owners in making decisions about the number of ships in their fleet and establishing best operating strategies.

It is suggested that further research investigate the acquisition of new ships in the fleet to evaluate if it is necessary to buy one ship of each type or several of smaller ones. It would improve the economic evaluation in a future research.