Ship operators need a means to analyse their fleet’s performance in order to maintain their competitiveness. Important decisions such as voyage planning and scheduled maintenance are all based on the ship’s performance. However, there seems to be no reliable means to estimate a ship’s performance. The most widely used measurements to estimate a ship’s performance are fuel consumptions, but even with weather filtering, it is difficult to identify how much fuel consumption is due to environmental forces such as wind and waves, and how much is due to the ship’s performance itself.

In 2016, ISO19030 was developed in order to prescribe practical methods for measuring changes in ship specific hull and propeller performance [1]. The emphasis was laid on practical methods, and many analysis techniques, such as calculating added resistance due to waves, were not included in the standard, as there was no practical means of collecting the data necessary to apply such techniques at the time. Instead, filtering and averaging were introduced instead to deal with environmental effects that were not accounted for in the standard. The result was a practical method that measures a ship’s performance with the limitations that it can only be used to compare ship specific changes relatively and unsuitable for comparing different ships, with rather coarse time resolution (the averaging period should be at least 3 month or more and typically a year). However, ship operators need a means by which they can analyse their fleet’s performance for each leg of the journey and often day by day. Decisions to increase a ship’s efficiency are made day to day and three months averaged value of a ship’s performance is far too late and coarse to be useful in a decision-making process.

There have been some previous research on applying ISO19030 methods for the analysis of a ship’s performance. Researches outlined in [2]-[5] are notable examples. All authors commonly express concerns regarding filtering performance, and regret that the ISO19030 cannot provide means of comparing performance between ships or within shorter periods.

There has been other research that has tried to analyse a ship’s performance using different approaches. Notably, J. Park et al. [6] used machine learning to predict a decrease in performance due to fouling. D. Gundermann et al. [7] used a simple regression model to obtain speed and power relationship for cases where no reference data are available. M. Jeon et al. [8] used a meta-regression model to predict fuel consumption. However, all these attempts did not consider a physical model of the wind and waves, and were only based on previous data; therefore, they cannot reflect changes in the operational profile or environment well.

This paper implements a software to analyse a ship’s performance. To meet a ship operator’s requirements, a new analysis method based on ISO15016:2015 is proposed [9]. ISO15016:2015 is the amended standard of the ship speed trial analysis method. Its accuracy objective of 2% in shaft power and 0.1knots in speed was verified from many sources during its development process as it is being used in contracts and mandatory regulations such as the reference speed in EEDI calculations [10]. ISO15016:2015 was also considered during ISO19030 development and many parts of ISO19030, such as wind correction, were directly referenced. The results of its implementation and application to two ships, a 176K bulk carrier and 8600TEU container, are described herein.

A software implementing the analysis method described in section 2 called SPA (Ship Performance Analysis Software) was developed. SPA implements all functions necessary to handle a large amount of data, which is collected during the operation of the ship. Particular care is given to its practical application and the key functions are listed as below:

• ① Fast processing of a large amount of data using an internal database.
• ② Define custom filters to be applied as per the user’s requirements.
• ③ Use of templates for ship definition, data column definition and filter definition, which will enable a user to reuse well-defined templates for subsequent analysis.
• ④ Custom select which measurements to use for speed, power, heading and environmental correction.
• ⑤ Review the results of applying filters, and the analysis results, in a table format inside the software.
• ⑥ Show the speed power graph inside the software for a quick review of the analysis results.
• ⑦ Output detailed analysis results in csv format for further analysis when necessary.

The main window of SPA is shown in Figure 2. The imported data, the results of applying filters, and the final analysis results can all be reviewed within the software as shown in Figure 3. It is also possible to quickly review the analysis results with a speed power graph as shown in Figure 4.

Figure 2: 
Main window of SPA

Figure 3: 
Results view window of SPA

Figure 4: 
Results chart view window of SPA

The data used in the SPA software have already been collected by the ship operators. Table 2 summarises the input data for the SPA software. The output from the SPA software after analysis includes full calculation results in .csv format, which can be used later by the user. Table 3 is a list of the output data.

This section describes the application results of the SPA software to the operational data recorded onboard a 176K bulk carrier and an 8600TEU container carrier. Each analysis was performed over one leg of a journey, where the same loading conditions were maintained, and the duration was approximately two to three weeks for the bulk carrier, and several days to a week for the container carrier.

Figure 5 and Figure 6 show the analysis results of a 176K bulk carrier in ballast and laden loading conditions, respectively. The solid line in the charts represents the model test results. Since ISO19030 results are single values, which represent an averaged decrease in speed from the reference data that are usually the model test results, to make an intuitive visual representation of the results, a dashed line is drawn by shifting the model test curve along the direction of the speed (x) axis by the amount of the averaged Performance Value. In all analysis results, the measured power is corrected to form a line similar to the model test curve and the distance from the model test results to the corrected values, represented by red dots, are how much the performance of the ship has degraded from an as-new condition due to hull fouling, etc.

Figure 5: 
176K bulk carrier ballast condition results

Figure 6: 
176K bulk carrier laden condition results

Figure 7 shows the difference in performance before and after dry-docking. The points represent the analysis results of the method proposed in this paper and the two types of dashed line represent the analysis results using ISO19030. Since ISO19030 relies on averaging out uncorrected environmental effects, at least three months of data is required. Therefore, only the method proposed in this paper can accurately show difference in performance before and after dry-docking with only one leg of a journey before and after. It is clearly shown in Figure 6 and Figure 7 where the performance analysis results of the method proposed in this paper significantly differ from those of ISO19030. Please note that ISO19030 analysis results in Figure 6 are not missing but overlap the model test results.

Figure 7: 
176K bulk carrier ballast drydocking effects

Figure 8 and Figure 9 shows similar results for the cases of an 8600TEU containers with 80% and 90% displacement conditions. In Figure 8, the analysis results using the proposed method and the results using ISO19030 are similar, however this is not the case in Figure 9. Thus, there is no guarantee that the proposed method will always give accurate and better analysis results than ISO19030. However the proposed results are more consistent and realistic than ISO19030, where the performance in some cases being only 0.03% different from the model test results (for the 176K bulk carrier laden condition results) or as low as -20% in speed in comparison with the model test results (for the 8600TEU container carrier 90% loading condition results).

Figure 8: 
8600TEU container carrier 80% displacements results

Figure 9: 
8600TEU container carrier 90% displacements results

Figure 10 shows the dry-docking effects of an 8600TEU container carrier, and the proposed method shows more realistic results than ISO19030, which shows dry-docking has almost no effects.

Figure 10: 
8600TEU container carrier drydocking effects

In this study, a new speed power analysis method based on ISO15016:2015 was proposed. The proposed method first evaluated the added or decreased resistance due to the environmental changes including wind, waves, and water density. Then, the measured shaft power was corrected using the direct power method, and finally, corrected to reference displacement for comparison. The proposed method was implemented in the SPA software, and applied to a 176K bulk carrier and an 8600TEU container carrier. The results show that the proposed method has the following advantages:

• ① Offers more information than just a value (averaged Performance Value) as in ISO19030.
• ② Able to provide a more realistic representation of performance than ISO19030, with only a small amount of data, typically with one leg of a journey.
• ③ Only uses data already collected by ship operators, so that no new data measurements are necessary.

The proposed method and the SPA software that was implemented in this study can be applied to estimate the powering performance of the operating ships used to verify the effects of energy saving devices and low friction paint, which will help decrease fuel consumption and reduce greenhouse gas emissions.