Improving Engine Robustness for Diesel Electric Vessels


At The Motorship Propulsion & Emissions Conference in May, Ricardo UK presented a tool for evaluating how a multi-engine diesel electric vessel’s reliability could be boosted by advanced engine control, with or without an energy storage device.  Here we present the paper, authored by Jonathan Brown, James Manuelyan, Pascal Revereault, in edited form.


Figure 1: Schematic of ship diesel-electric propulsion system

A trend towards utilisation of diesel generators with electric propulsion systems and a move to highly performing high-speed diesel engines is recognised in many vessel segments. Diesel generators (DGs) operate as single speed, variable load units designed for fast transient load response to accommodate demand for power.

The control strategy for DG powertrain units is varied depending upon specification and application.  In many cases it is a simple algorithm without significant optimisation of engine control.  To assess the benefit of a more sophisticated control system either with or without an ESD (ESD), a rapid evaluation tool has been developed that can be tailored to each specific application.

There are usually a number of DG sets on a vessel operating at between 20-100% load. Below a minimum load threshold the lowest loaded DG sets will often be switched off. Although switching off redundant engine units provides significant fuel savings, there is evidence of negative impact on robustness and reliability of the units.  Issues that may be aggravated by repeated on-off cycling include:

  • Tribological wear issues (bearings, pistons & rings, liners, valvetrain)
  • Resonant vibration induced failures (ancillaries, pipework, fuel rails/systems, wiring)
  • Drivetrain reliability (coupling, alternator)
  • Gas leakage – gaskets & cracking (manifolds, pipework)
  • Sensor failure

The number of stop/start cycles may be as high as eight or 10 a day during highly variable duty cycles such as naval exercises or harbour activities.  Data provided under project work at Ricardo supports these figures.

As the propulsive power is provided electrically, by including an ESD it is possible to store energy during periods of lower utilisation rather than switching off units and conversely use stored power to bridge temporary additional demand when additional engines would usually be switched on.  A key to this is sizing the ESD for input/output rates necessary to support standard operation of the ship.  It is then possible to optimise the operating points of the engines within specific constraints to minimise fuel consumption, while providing the potential for a ‘quiet mode’, sailing without engines in environmentally sensitive areas including ports.

A tool has been developed to understand the potential benefits that may be realised with the utilisation of an ESD to improve the reliability and maintenance demands of a vessel.  A set of simulations have been conducted of a representative vessel over different duty cycles.  The simulation has been set up in a two part process:

  1. Baseline optimisation of the engine switching control to minimise on-off cycles within the existing ship architecture (no energy storage)
  2. Optimisation of the propulsive system with an additional ESD

The study has been conducted looking at the constraints on the engine operation based on both minimising the start-stop cycles and the fuel consumption.

Ship specification and operation cycles

In this case a naval vessel has been used as the basis for the study, although this methodology and the provisional benefits will be similarly applicable to many commercial ships, particularly those with a high level of variability within the duty cycle, but also where sensitivity to local environment is necessary, such as work boats, harbour craft (including tug boats), patrol vessels, cruise liners and supply vessels.

For this work the ship is considered to be fitted with four 2.2MW V16 diesel generators, feeding power to the electrical propulsion motors and ship power systems.  A schematic of the ship is provided in Figure 1.

The particular duty cycle that the ship will undergo will have a significant impact on the number of engine stop/start cycles and the range of efficient engine operation during these cycles.  In the case of this assessment three separate duty cycles have been selected:

Cycle 1: A typical naval operational duty cycle on exercise;

Cycle 2: Typical profile of vessel for re-supply;

Cycle 3: A profile for light patrol.


Figure 2: Example ship operating cycles used for the simulation activities

The power demand against time of these three cycles are provided in Figure 2.  These cycles have been developed based on samples from real duty cycle data for naval and work boat applications.  The start and end of the cycle have been maintained for each variant to provide a consistent approach whilst providing a level of variance to cover a number of different potential applications.

The power demand includes propulsion, hotel and systems power requirements, including naval systems such as defensive countermeasures during operations and exercises.  There are some periods of rapidly fluctuating power requirements in each cycle representing transient manoeuvers in port or re-supply at sea.

Description of mathematical model

The model developed to simulate engine and hybrid control consists of three sections.

The main simulation control provides the overall command of the simulation, linking the base engine and hybrid control systems and managing the power demand into these sub-controllers.

The base engine control program is the foundation of the model, controlling engine power output and switching behaviour depending on the current power demand of the cycle and historic data.  It consists of a base engine control map that links an input power demand to the number of engines that are required and their respective power outputs.  This subsystem also contains several other modes it can operate in to help minimise engine stop/start events in different scenarios.

The hybrid control system consists of an optimiser that chooses the charge or discharge rate of the ESD (ESD) to minimise fuel consumption and help prevent engine stop/start events from occurring for given input conditions.


Figure 3: The power management system (PMS) engine control model

Figure 3 shows the layout of the complete model.  This provides a three-layer type control system allowing optimisation of the individual sub-controllers (base engine & hybrid system) underneath an umbrella optimiser in the main control.

For each cycle, three simulations have been performed:

  • A baseline simulation with no hybrid control
  • An “ideal case” simulation where the brake specific fuel consumption (BSFC) map is set so that the engines are always operating at peak BSFC regardless of power output.  This provides a theoretical upper limit to how much fuel consumption can be improved by using a hybrid system
  • A hybrid simulation where hybrid control is active


Table 1 summarises the reductions in engine stop/start events.  The simple switching algorithm refers to the base engine control only using the baseline control map.  The optimised switching algorithm includes the extra modes that the base engine control algorithm can use in particular conditions to help prevent engine stop/starts.

The greatest benefit to performance for both engine stop/start reduction and fuel consumption improvement (see Table 2) was obtained in Cycle 3.  Engine power outputs during the cycle can be seen on the top graph for the baseline line test and bottom graph for the hybrid test.  At around 90 minutes into the cycle, in the baseline test, a third engine is switched on to cater for the increased power demand.  The engine is switched off shortly after due to a large reduction in power demand.  During this period in the hybrid test, the third engine is never switched on as the extra power demand is covered by the hybrid system.The use of the hybrid system also delays the starting of engine two at the start of the cycle.

Table 1: Number of stop/start events saved for each cycle with and without an ESD hybrid system

Number of Stop/Start Events

Percentage Improvement
Cycle Simple Switching Algorithm Optimised Switching Algorithm (baseline test) Optimised Switching Algorithm + Hybrid – ESD Optimised Switching Algorithm Optimised Switching Algorithm + Hybrid – ESD
1 18 14 14 22% 22%
2 6 6 6 0% 0%
3 8 6 4 25%


Total fuel consumption values at the end of the cycle for the different tests were:

  • Baseline test:  4,659kg
  • Ideal case test:  4,475kg
  • Hybrid test:  4,614kg

From the baseline to the hybrid test, there is a 45kg saving in fuel, a 0.97% decrease in consumption.  The theoretical upper limit for fuel saving with hybrid control for this cycle is 184kg, a 3.95% decrease in consumption.  Plotting the difference in fuel consumption from the baseline model to the hybrid and ideal case tests, it is seen that the hybrid system saves fuel during the transient section of the cycle where it can charge and discharge frequently. Combined with the changes to the engine power output already mentioned, it is seen that the hybrid system removes a number of the oscillations of the engine output from 40-60 minutes allowing the engines to run a more stable load closer to the peak BSFC load.  This saves 10kg of fuel an hour in the cycle.   

The impact of both the stop/start and fuel consumption reduction with the standard calibration is not as good in Cycles 1 and 2, with half the targeted reduction in stop/start events and minimum change to fuel consumption in Cycle 1.  Cycle 2 shows no reduction in stop/start events but a better fuel consumption reduction of 0.56%.  Of interest is that the ESD does not provide any further reduction of stop/start cycles over the initial hybrid base engine control in this case.  With development of the tool calibration and further ESD sizing optimisation, the results for Cycles 1 and 2 could be further improved.

Table 2: Fuel consumption over cycles from baseline engine optimised condition with ESD hybrid system

Total Fuel consumption over Cycle
Cycle Baseline (kg) Ideal (kg) Hybrid (kg) % reduction (Ideal) % reduction (Hybrid)
1 6256 5872 6250 6.53 0.09
2 5518 5211 5487 5.88 0.56
3 4659 4475 4614 3.95 0.93


From the simulation model developed for this study it is feasible that, with an additional ESD coupled into a diesel-electric propulsion system with a number of diesel generator units, it should be possible to significantly reduce the number of stop/start engine cycles over a standard simple switching algorithm for certain operating duty cycles, in conjunction with a reduction in fuel consumption.

The evaluation tool developed allows for exploration of different control strategies for optimisation of existing and new plant for on-board power in diesel-electric propulsion vessels.  Results from the tool indicate that further savings may be achieved through utilisation of an ESD as part of the hybrid control system.

The models at this stage have not been setup for an optimum calibration for each duty cycle, as the aim of this initial study was to provide evidence to support the application of a toolset to rapidly evaluate the potential benefit of a hybrid system to improve overall powertrain robustness and reliability.  To provide optimum benefit a detailed record of each vessel’s duty cycle would be required at which point a bespoke optimised calibration of the control strategy would be developed.

Did you subscribe for our daily newsletter?

It’s Free! Click here to Subscribe!

Source: The Motorship


This site uses Akismet to reduce spam. Learn how your comment data is processed.