Hybrid Systems for Diesel Powered Ships

Hybrid Systems for Diesel

Powered Ships

Hybrid topologies for slow speed ships

Ship Voyage simulator for performance analysis

Dr. Eleftherios Dedes

(Research performed at the University of Southampton, UK)

ΕΝΩΣΙΣ ΕΛΛΗΝΩΝ ΕΦΟΠΛΙΣΤΩΝ

Presentation Contents

• The emission problem

• Methods to determine fuel consumption/ shipping emissions

• Industry methods to reduce fuel consumption

• Investigation on Hybrid Power Systems

• Description of Ship Voyage Simulator

• ECMS non-linear Optimisation

• Results and Discussion

Typical exhaust pollution

(production of smoke) due to

transient engine loading when

fast ferries getting up to

service speed

Currently…

3

CO2SOx

NOxPM

The emission problem

• Comparison of Shipping with most pollutant Countries:

Source: IMO

FC x 3.1144

Methods to determine fuel consumption/

shipping emissionsTop down method Vs bottom up approach

Emission Factors (Endresen et al., 2003; 2007), Corbett and Coehler (2003)

5

Pollutant

type

Average Power

based factor [g/kWh]

Fuel based factor

[tonnes/day]

PM10 1.5 -

PM2.5 1.2 -

DPM 1.5 -

NOx 17 0.087

SOx 10.5 (for 2.7% S) 0.02 * % S

CO 1.4 0.0074

HC 0.6 -

CO2 620 3.17

N2O 0.031 -

CH4 0.006 0.0003

• Relationship between NOx and SFOC

Methods to determine fuel consumption/

shipping emissions

Therefore…• Research window:

• Emissions environment and fuel consumption

Emission inventories, doubt and unreliable, IMO formulas introduce

large error, strict environmental agenda by 2020

• Profitability of Shipping companies Fuel consumption reduction,

imperative for future sustainability

• Other mature methods to reduce fuel consumption

• Concept visualisation and Data collection:

• Energy requirements (actual operations), statistics

• System sizing (power, density)

• Storage media selection, installation requirements

• Modelling, Simulation and Optimisation

7

Industry methods to reduce fuel consumption

• Operational Measures

• Trim optimisation

• Optimal routing, Just in Time arrival

• Course keeping

• Weather routing

• Technical Measures

• Propeller inflow, outflow optimisation

• Optimal Rudder shape

• Hull lines

• Fuel optimised/ De-rated/ Electronic Diesel Engines

Industry methods to reduce fuel consumption

…>100% Total Savings!!

… Perpetuum Mobili “Αεικίνητον”

Ship-OwnerTechnical Manager Supt. Engineer

Concept Visualisation– Hybrid with batteries• All Energy saving measures except Main Engine retrofits target in

reducing power demand

• Hybrid decouples propeller demand from the efficient operation of the

Main Engine is fully compatible with current market options/retrofits

• Applies Energy Management strategy by:

• Operating prime movers near the most efficient thermodynamic points

• Given the scenario may utilise stored energy or merge electric loads to

main propulsion utilises benefits of All Electric Ships

• Manoeuvrability, ultra slow steaming capability

• Redundancy

• Lower maintenance needs for T/Cs, fuel pumps, complete O/H

• Limited operation of Engines in transient loads

• Limit energy conversions by inserting strict performance parameters

• Allow easier operation of ships sailing to ECAs

• Has a scalable power output and scalable Energy capacity

10

Concept Visualisation– Hybrid with batteries

11

Investigation on Hybrid Power systems• Data Collection from Greek Shipping Companies

• Energy and Power requirements Statistical Analysis

• Selection of Energy storage system

• Structural analysis and hydrostatic loss/ payload reduction

• Preliminary Financial viability assessment

• Development of Ship voyage simulator

• Setting up the Power handling problem in Matlab suite

• Non Linear Optimisation

12

13

Energy/ Power requirements statistical Analysis

• Determination of “off optimum” Engine operation point

14

Selection of Energy storage system

Type of Vessel: Energy [MWh]: Power [MW]:

Handysize 8 1

HandyMax 8 1

Panamax 15 2

Post –

Panamax

5 2

Capesize 4 1

Type Wh/kg Cost [$/kWh]

Lead Acid 35 90

Vanadium - Bromine 50 300

Silver Cadmium 70 -

Zinc - Bromine 70 -

Sodium/nickel chloride 115 110

Lithium Ion 150 600

Sodium Nickel Chloride Battery

• Extreme tests were undertaken in EUCAR organisation

• Impact

• Penetration tests

• Complete immersion in water

• Spike penetration and spraying with water

• Vibration testing

• 30 minute gasoline fire test

15

Sodium Nickel Chloride Battery

• This produces nickel containing re-melt alloy used in the

stainless steel industry.

• The ceramic and salt contained in the cells collects in the slag

and is compatible with their process.

• The slag is sold as a replacement for limestone used in road

construction – nothing goes to the landfill.

• One of the weak links in any recycle process is the collection of

the spent units at the end of life.

16

17

Structural analysis• Void spaces, Double bottoms, bosun store, poop deck

compartments, Top side tanks sufficient free volume

• Weight distribution, Battery Rule compliance, Risk parameters

Ship Type HandySize HandyMax Panamax Post - Panamax Capesize

Required Energy [MWh] 8 8 15 5 4

Required Battery Volume m3

Sodium Nickel Chloride 190Wh/L 42 42 79 26 21

Vanadium Redox Flow 30Wh/L 267 267 500 167 133

Engine Room Volume[m3] 3800 4530 4900 5150 9600

Free volume in current engine room:

35% of total volume

1300 1580 1650* 1760* 3350

Added Volume due to electric components: 1040m3

Additional Engine Volume: 2x100.4m3 + 4x59.30m

3 = 438m

3

Ship

Type

Handy

Size

Handy

Max

Pana

max

Post-

Panam

ax

Capesi

ze

Required Energy [MWh]

8 8 15 5 4

Required Battery weight [tonnes]

Sodium Nickel Chloride 150Wh/kg

70 70 130 43 35

Vanadium Redox Flow 50Wh/kg

160 160 300 100 80

Final Added weight to the vessel (propulsion system

+ storage)

323 323 384 297 288

414 414 554 354 334

Increase in Lightweight [%]

4.1% 3.4% 3.2% 2.0% 1.2%

5.2% 4.3% 4.7% 2.4% 1.4%

18

Hydrostatic impact - payload reductionCondition Cargo

[tonnes]

Cargo Loss

%

Ballast

[tonnes]

LCB

[m]

LCG

[m]

MTC

[tm]

Trim

[m]

Normal Ballast

Departure

0 0.00 23414.3 117.126 117.189 1040.2 2.08

Normal Ballast Arrival 0 0.00 26061.1 114.757 114.855 1068.1 3.045

Heavy Ballast Departure 16487.1 0.00 23411.7 115.991 116.059 1226.5 2.667

Heavy Ballast Arrival 16487.1 0.00 24476.8 116.88 116.682 1191.4 2.265

Grain Departure SF65 60188.4 0.00 2250.7 116.664 116.702 1346.4 1.674

Grain Arrival SF65 60188.4 0.00 2250.7 118.866 118.879 1298.4 0.514

69990 tonnes DWT

cargo Departure

65152.1 0.54 0 115.448 115.501 1364.5 2.323

69990 tonnes DWT

cargo Arrival

65152.1 0.54 0 117.517 117.548 1331.3 1.205

Homogenous Design

Departure

67858.4 0.52 0 115.656 115.702 1364.6 2.09

Homogenous Design

Arrival

67858.4 0.52 0 117.668 117.693 1336.3 0.983

Grain Departure SF42 87866.1 0.40 0 116.837 116.841 1392 0.335

Grain Arrival SF42 87866.1 0.00 939.4 117.483 117.482 1384.9 0

19

• Capacity determined by vessel type and statistical analysis

• Constraints of battery manufacturer e.g max 16 parallel units can

exist

• Nominal voltage of 557V

• 0.285 MWh energy density per bank.

• Possibility to connect multiple banks

• Highest operational efficiency at ultra low currents

NBB

=Energy

req.

0.285

demand

nominaldischarge/ Charge

16BB

P

VI

N

maxnom BB batC N Q

Efficient Sizing of the Battery Banks

Ship Voyage Simulator

• Matlab/ Simulink environment

• Scalable and modular approach (Blocks and signal flows)

• Each block represents machinery, weather, propeller, ship model

• Hull Resistance (Holtrop – Mennen, Hollenbach methods)

• Added Resistance (Aertssen, Kwon methods)

• Wind Resistance (Isherwood, Blendermann methods)

• Wageningen Series open water performance method

• Battery models, Kinetic Energy approach (Manwell, McGowan)

• Experimental Data for Sodium Nickel Chloride

• Shop Test Data and sea trials data for the propulsion machinery

• Wind and wave generator models based on average reported data

20


• Simulator is separated into 2 sub-models in order to reduce the

complexity (Power profile generation, Optimisation)

• Time domain

• Air Emissions are quasi static phenomena No transients

• Generation of power demand using the voyage simulator:

• Random wind and wave generation (Hs, T, μ, Vwind, μwind)

based on captain observations (mean values reported)

• Mean value added resistance model

• RAO block if sea-keeping results are available

• Series 60 mean added resistance value for Tp, Hs and μ

21

ECMS non-linear optimisation

• SQP method, update of the Langrangian

• Converges rapidly to the solutions (compared to other methods)

• Single objective (minimisation of Fuel Consumption)

• Equivalent Cost Minimisation Strategy (ECMS) applications:

• A layouts (Hybrid Auxiliary loads, AES)

• B layout (Hybrid – Conventional system) (propulsion loads only)

• C layout (Fully integrated Hybrid system (Propulsion and

auxiliary loads. Propulsion system coupled with auxiliaries)

23

ECMS non-linear optimisation

• Successful use in automotive industry

• Converges in automotive applications with results of DP

• No a priori knowledge of engine loading or vessel speed

need for time step optimisation

• Unconstrained or constrained

• Principles of ECMS algorithm

• Battery usage has an equivalent fuel cost

• Battery charging has an equivalent fuel saving in the future

• The λ is determined during simulation and is describing the

percentage difference of the SoCt from the SoCref

• Determines how quickly the charging will occur

• SoC reference is predefined setting for SoC (e.g. 35% for

safety while for automobiles is ~65%)24

Optimisation Algorithms

• Optimisation Scenarios (A1,B, C)

25

MAIN 2-STROKE DIESEL ENGINE

AUXILIARY GENSET

AUXILIARY GENSET

AUXILIARY GENSET

ELECTRIC

MACHINE

PTO - PTI SYSTEMPOWER

CONVERTER

CONVERTER

TRANSFORMER

BATTERY

BANKS

GEAR BOX

TO AUXILIARY LOADS

TO PROPULSION LOADS

A B

C

RECTIFIER/

INVERTERTRANSFORMER


• Power layout (D-A2) “All Electric Ship”

26

AUXILIARY GENSET

AUXILIARY GENSET

AUXILIARY GENSET

PROPULSION

MOTOR

POWER

CONVERTER

CONVERTER

TRANSFORMER

BATTERY

BANKS

GEAR BOX

TO AUXILIARY LOADS

TO PROPULSION LOADS

RECTIFIER/

INVERTERTRANSFORMER

PROPULSION

MOTOR

POWER

CONVERTER

CONVERTER

TRANSFORMER

AUXILIARY GENSET

AUXILIARY GENSET

AUXILIARY GENSET

HYBRID

CONTROLLER

• Multiple algorithm starts

• Find all the local minima

• Converge to the global minimum

• Efficiency static values

27

Component Description Necessary in layout Efficiency

Battery Converter and Transformer All layouts 98%

Transmission losses All layouts 99.5%

Gearbox efficiency A1, B, C 98%

Motor Converter Transformer B, C 99%

Electric Machine Power Converter A2, B, C 96%

Electric generator All layouts ~ 96%


• Efficiency curves (polynomials based on real data fit)

28


• Battery Efficiency curves (Based on laboratory measurements)

29


Equivalent Cost Minimisation Strategy

• Layout C (Discharging/Charging respectively):

• Objective Functions (Motor/ Generator)

30

1 1 /

46

m-dis. 4 4 /

2

min M/E

/ , . 5 9 5 9

min 10

M E

i i i i A E sim

i

T F inv conv m

d x x MCR

f g x x x x MCR t

SFOC

x x w x x

1 2 1 2 /

56

gen-dis. /

3

min M/E

/ , . 6 6

min 10

M E

i i A E sim

i

T F inv conv m

d x x x x MCR

f g x x MCR t

SFOC

x w x

1 2 1 2 /

56

gen,charg. /

3

/ , . 6 6 min M/E

min 10

M E

i i A E sim

i

T F inv conv m

d x x x x MCR

f g x x MCR t

x w x SFOC

644 / , min

1

min 10n

i i T F inv sim

i BB bat

xf g x x x w SFOC t

N V

Results

• Ship Voyage Simulator

• Hybrid Power System ECMS optimisation outputs

31

Results


• Results for Estimation of Power/ fuel consumption

32

Results



33

Laden Voyage Ballast Voyage

Vessel 1 Vessel 2 Vessel 3 Vessel 1 Vessel 2 Vessel 3

Simulated

Fuel

consumption

616.11 614.07 1302.70t 978.45t 276.75t390.22t

498.66t

Measured

fuel

consumption

653.40 642.20 1359.10t 1206.80t 354.60t438.20t

594.40t

Fuel

difference-37.29 -28.13 -56.35t -228.35 -77.85t

-47.98t

-95.74t

Percentage

difference-5.71% -4.38% -4.14% -18.92% -21.95%

-10.95%

-16.11%

Ballast voyage Re-analysis of ballast voyage

Vessel 1 Vessel 2 Vessel 3 Vessel 1 Vessel 2 Vessel 3

Simulated

Fuel

consumption

978.45t 276.75t390.22t

498.66t1159.28t 307.43t

461.50t

600.44t

Measured

fuel

consumption

1206.80t 354.60t438.20t

594.40t1206.80t 354.60t

438.20t

594.40t

Fuel

difference-228.35 -77.85t

-47.98t

-95.74t-47.52 -47.17

23.30t

6.04t

Percentage

difference-18.92% -21.95%

-10.95%

-16.11%-3.93% -13.30%

5.32%

1.02%

Results


• Further improvements in Estimation of Power/ fuel

consumption

• Metrological Models, statistical models for wind/waves

• Not suitable for swell effects

34

Results



• Performance analysis

35

Results



• Performance analysis

36

Results

• Determination of power profile for Hybrid scenarios

• On board data collection for auxiliary loads

37

Results

• Layout A1 (Power split for 14.4MWh)

• Battery output, 1 D/G output instead of 2 initially running

38

• Layout A1 (Power split for 2MWh Vs Conventional system)

• Battery output significantly increases the degree of Hybridisation

39

Results

Results

• Layout D-A1

• Feasible solution with non negligible savings that reach up to

6% of auxiliary fuel bill excluding additional savings due to

single D/G operation

• Savings depend on installed power

• quadratic relationship which concaves downwards

• Battery efficiency, charging effect reduces the energy

efficiency

• For 2MWh installed energy

• No significant losses in DWT

• 0.50% of the total A/E fuel bill, excluding again further

savings due to number of engines in operation

40

Results

• Layout D-B

• Taking into account the conversion losses, the Hybrid

solution is not feasible (electromechanical conversions)

• If improving sub component efficiencies, excluding the

battery:

• Daily reduction given the examined load profile is 0.095%

• Layout D-C

• Propulsion fuel consumption savings reaching up to 4.44%

• Cost effective to absorb power with high conversion

losses from the A/E than operate the M/E in less efficient

load

• Is directly dependent on the SFOC curve shape and

steepness 41

Constraints alternation and sensitivity analysis

Battery

Degradation

Battery capacity

2MWh 4MWh 7MWh 10MWh 14.4MWh

Baseline 0.53% 2.62% 4.88% 5.52% 5.70%

1% 0.19% 1.66% 2.98% 3.59% 3.79%

2% - 0.77% 1.34% 1.61% 1.81%

3% - 0.23% 0.28% 0.25% 0.30%

4% - - - - -

5% - - - - -

10% - - - - -

Case

Capacity

24h vector, 2h sample

rate, tref = 72h


rate, tref = 48h


rate, tref = 24h

2MWh 0.53% 0.52% 0.45%

4MWh 2.62% 2.94% 0.60%

7MWh 4.88% 4.91% 1.90%

10MWh 5.52% 5.52% 5.52%

14.4MW

h5.70% 5.70% 5.70%

Conclusions

• Hybrid potential is feasible for Auxiliary loads

• Battery usage is not suitable for application for direct Main

Propulsion due to conversion losses

• Sensitivity analysis show potential and viability

• Coupling the auxiliary generators and the main propulsion

including batteries is a promising alternative that yields to daily

savings of approximately 4.4% (dependent on loading)

• The battery system is working in ultra low currents maintains

high efficiency

• Degree of Hybridisation: ~50% discharging, ~40% charging,

~10% no action (auxiliary loads)

• Feasible system in term of construction

43

Conclusions

• The DWT is not significantly affected

• For A1 scenario, DWT effect is negligible

• For A2 scenario (AES), the Hybrid system is feasible due to the

absence of conversion losses

• However, this system was not simulated due to absence of

Engine data

• Ship Simulator is a useful tool for propulsion assessment and

performance monitoring

• The scalable and the extendable blocks allow great flexibility in

increasing the system’s complexity and the simulator’s precision

for future applications

44

Further details (including reference list)

• E. K. Dedes, D. A. Hudson, and S. R. Turnock. Design of hybrid diesel-electric

energy storage systems to maximize overall ship propulsive efficiency. In

Practical design of ships and other floating structure symposium, 2010.

PRADS’10, pages 703–713. COPPE UFRJ, 2010.

• E.K. Dedes, D.A. Hudson, and S.R. Turnock. Assessing the potential of hybrid

energy technology to reduce exhaust emissions from global shipping. Energy

Policy, 2012; 40: 204-218

• E. K. Dedes, D. A. Hudson, and S. R. Turnock. Technical feasibility of Hybrid

Powering systems to reduce exhaust emissions of bulk carriers. IJME

transactions of RINA, 2013.

• E. K. Dedes, S. R. Turnock, D. A Hudson. A modified activity based approach for

accurate estimation of fuel consumption from global shipping. International

Journal of Transportation Research part D. Under review, 2014.

• E. K. Dedes, S. R. Turnock, D. A Hudson. Diesel Hybrid systems for increase of

fuel efficiency and reduction of exhaust emissions from ocean going ships. To be

submitted to Journal of Energy, 2015.

Hybrid Systems for Diesel

Powered Ships

Thank you for your attention!

Dr. Eleftherios Dedes

Further questions: [email protected]

mailto:[email protected]

Hybrid Systems for Diesel Powered Ships

Documents

Transcript of Hybrid Systems for Diesel Powered Ships