Report for Prolongation of scholarship

PhD Student Mario Krstovic

0
The main aim of this report is to show:

a) To present the research methodology.

b) Demonstrate the work done and make a clear path for finalisation of the PhD thesis.
Provide the flow chart.
c) Give a short description of modelling, optimisation and data analysis.
All those, parts of this report will be separate chapters in the final version of PhD thesis.
d) Present the basic methodology for the Case Studies. Explain the method not Case Studies

P.S

The novelty explanation point and Case Studies are the theme for a much larger report
contained in my presentations so if there is additional need for the evidence of my novelty I
can send the presentation to the commission. This report explains the developed
methodology new knowledge which principally solves the Case Studies.

1
Aim and Objectives

The aim of this PhD is to define and optimise the set of design and operational parameters
throughout investigated ship exploitation lifetime targeting reduction of fuel consumption and
environmental footprint.
One of main drawbacks in the Marine Engineering and Naval Architecture is slow evolution of rules
resulting in slow implementation of new technologies and low amount of research invested in
development of new types of highly efficient systems. Until now main research is done in
optimisation of existing systems having for consequence low potential of maximal possible
improvement.

The main purpose of this PhD more closely is to apply available knowledge from energy/exergy
modelling, optimisation of ship systems to create new understanding of ship system
interconnections, increase the ship system efficiency and reduce exhaust gas emissions.

Multiple objectives are set for development of optimal design and operational parameters.

Synthesis of optimal design solutions for particular operation lifetime of particular

ship
Best set of operating parameters for every possible operating condition.
Proposal for implementation of new technologies.

To be able to accomplish the previously defined aim and research objectives, range of research steps
should be satisfied:

1. The secondary research to identify the existing methodology gaps and proposals from the
researchers in the field.
1.1. Identification of research gaps.
1.2. Identification of smart solutions from other Industries.
1.3. Analysis of useful methodologies for creation of PhD methodology.
2. Preliminary analysis of ship systems performance and operation profiles to identify the
research direction.
2.1. Analysis of ship energy and exergy flows for development of preliminary modelling
methodology. The modelling is conducted in a way that the precision of systems
depends on the particular system importance and speed needed. Results of analysis are
regarded in ship systems: energy/exergy efficiency, importance ranking and
development of modelling strategy.
2.2. Study of ship operating profile to determine variation of energy and exergy demand in
particular time exploitation periods and definition of constrains which are need to be
satisfied by investigated ship system components.
3. Modelling of ship subsystems and combining them to represent ship system as a whole.
3.1. Development of particular ship energy producers systems and their combination to
represent ship exploitation behaviour.
3.2. Modelling of specific operating demands for different case studies and optimisation
objectives.
3.3. Modelling of Waste Heat Recovery systems necessary for optimisation purposes.

2
 3.4. Verification of modelled systems to ensure that physicality of processes is satisfied.
4. Development of different case studies.
4.1. Creation of different input parameter sets (operational and design) to satisfy operation
needs.
4.2. Development of different system combinations to satisfy the optimisation speed and the
degrees of freedom needed for particular optimisation case study. The systems should
be made to satisfy the requirements of optimisation.
5. Utilisation of different optimisation techniques to find best operational and design measures
for efficiency improvement.
5.1. Development of constrained optimisation algorithm to satisfy the requirements of
optimisation.

Modelling Approach:

The modelling is most labour intensive part of the PhD combining already existing subsystem
modelling methodologies in a way needed to satisfy the optimisation requirements. The whole
process can be divided in few major parts:

Modelling of existing ship subsystems and system which exit represent the measure of ship
efficiency (in preliminary study) or improvement in a way that exit values (objective function
and constrained function) are practically the exit of the system and enter point to the
optimizing algorithm. The main purpose of main ship system is to give information necessary
for study of:

a) Energy/exergy flows. This study is providing information regarding available energy

losses and potential of energy exploitation called exergy as function of different input
parameters (design; load conditions; ambient conditions);
b) Provide simplified representation of ship system interconnections and available energy
and exergy improvement potential between ship subsystems. System interconnections
theoretically are easy to understand but main problem is how to understand critically
the effect which subsystems have one each other and at the whole exit of system.
c) Possess necessary degrees of freedom regarding input decision variables. The system in
different case studies is providing different inputs regarding the objective optimisation.

Modelling two types of operational parameters to satisfy optimisation objective :

a) Operational parameters describing statistical representation of ship lifetime operation
taking in to account exploitation complexities: operation profiles (loaded, ballast, in
port); ambient conditions.
b) Operation parameters representing range of possible situations to find best set of
decision variables needed.

3
 Customisation of optimisation algorithm. The optimisation code is used from already
developed functions in Matlab or in special cases is coded depending on optimisation
complexity of specific Case Study. The main new knowledge regarding optimisation is in:

a) Determination of right decision variables;

b) Model types should be capable of satisfying optimisation requirements. A lot of
information is obtained from previous researches on this topic but every research has it
unique and particular methodology so the literature review can only be taken as starting
point in determination of final models needed. The models need to satisfy requirements
regarding input/output parameters depending on important parameters (decision
variables), research objective and to satisfy the speed requirements. Existing models are
coupled with existing optimisation algorithm, one iteration execution time is multiplied
with determined necessary amount of iterations and total time is calculated. If total
calculation time is unreasonably high, the regression study is conducted on a particular
high time consuming system parts with condition that regression method is capable of
producing results which mimic real ones.

c) Optimisation algorithm is modified to satisfy different case studies but the operation in
principle is kept the same. What is more, the number of iterations is calculated
according to a number of iterations needed, meaning that the number of iterations is
increased until there is not significant difference between results in objective functions.
The results of the study are used on other similar studies.

Generic explanation of optimisation model:

Optimisation model is consisted of:

a) Model describing Operational parameters.

b) Thermodynamic, electric, propulsion, heating model describing the physicality of
existing ship system.
c) Waste Heat Recovery or some other Energy Recovery model.
d) Optimisation algorithm with constraints.
e) The whole system in combined in one unit with different connections and
interconnections in different ship models.

4
 Optimisation Algorithm

Operational Constraints Optimisation

Profile (Loading
Code
Conditions)

Existing System Physical Model

WHR; ERT

Parameter calculation Model

Decision Variables Objective Function Connections C. S. Info.

Picture (1) Principal representation of one Case Study model developed in Matlab for optimisation
purpose. All systems and subsystems are developed as Matlab functions. The fluid properties are
taken from COOLPROP database which is highly verified and is common in this research field.

C. S. Case study

5
Operational Profile Loading conditions model

Case Study

Operation Profile Model Counter

Lading Conditions
Amb. Cond Propulsion EL. Load Heating Load
Load

Picture (2 ) Principal scheme of loading condition

Depending on the type of the optimised system the loading conditions will vary. The operational
profile is taken from available literature or is representation of realistic experience values. The
model is developed as a black box model taking in to account the relations between outputs as it is
shown at the picture. This relation is kept buy counter part of loading conditions on which the
subsystem parts of Operation Profile model are calibrated.

- Case study block provides the information related to the type of the optimisation: design or
operational and information about configuration of optimised system.
- Ambient conditions block provides the information related to the temperature of air and sea
water as the information of relative humidity of the air. Also in ambient conditions are
defined two types of air temperature at the engine intake depending on the configuration
type.
- Propulsion provides the information related to the RPM and Torque depending to the
loading state. There are defined three states related to the loaded state and are: loaded,
ballast and in port.
- Electric load is defined buy electric balance load and is expressed as a percent of load of
auxiliary engines.
- Heating load is defined by heating balance; there will be one separate part of optimisation
which will represent the optimal heating profile for a particular voyage in that case the cargo
heating will be defined more in detail.

6
There are three types of loading conditions models depending on optimisation objectives. Loading
conditions are combination of different parameters: ambient parameters, propulsion, heating and
electric load parameters and the loading conditions input needs to satisfy the combination of those
parameters for particular type of optimisation. The proper estimation of loading parameters is
essential for proper optimisation since one of main objectives of this PhD is to determine proper
combination of design and operational parameters to satisfy the highest energy/exergy efficiency of
ship system. The loading conditions need to represent real cases and depending on the optimised
system, type of optimisation and available data they can be divided in to three types of loading
conditions:

1. Time dependent load conditions. Loading conditions developed to simulate real-time load
conditions with time step between the points.

() = (); = ;

This type of loading conditions is used if there is actual operating data measured in time of
operation. This type of operating parameters would practically give highest precision of optimisation
since in this case the loading conditions represent most accurately real case scenario. The best
possible option would be to be able to have the data from several voyages and then statistically
calculate the input time dependent curves with highest statistical weight.

2. Load conditions dependent on fractions of operation states used in design procedure of ship
systems. Those types of loading conditions actually have lover precision since in this case the
available input are loading conditions of Main engine in ballast and loaded conditions. In this case
the heating and electric loads need to be assumed from (electric and heating) balance and there is
not accurate data related the change of those parameters in time as function of Main engine load,
time of voyage and ambient conditions. The ambient conditions also in this case are not totally
defined since there is not data for them so in this case the ambient conditions are taken as some
average values for particular times of year and route. The precision of this model is lover since the
hating load main engine exhaust gas (mass flow and temperature) are dependent on ambient
conditions. The heating load is also a function of a total time of voyage.

3. Load conditions dependent of possible loading scenarios for development of optimal important
parameter combinations. These loading conditions actually represent the optimal values of
operational parameters for different loading conditions combination of loafing conditions and
ambient conditions. The optimal values of those parameters could be used in exploitation when
similar exploitation conditions are met.

7
Generic explanation ship system model (short description, detailed description will be provided in
the Modelling paragraph):

The ship system is consisted of ship subsystems which are further consisted of subsystem parts. Ship
subsystems are defined the same as ship system.

(i)-may
(i)-loads; have or not (i)-loads;
Amb. have input Amb. Decision
Cond. Decision Cond. Variables
Properties. Variables Properties.

(i) Subsystem (i) System

Energy Losses Important
Work Work Losses
demand Parameters
Energy D.
Parameter Calculation Model

Picture (3) Generic representation of ship system and subsystem

The system is regarded in the same manner as subsystems taking in to account that same types of
inputs and outputs are for the previous and former. The systems and subsystems are consisted of
following parts: box, inputs, outputs, constraints; interconnections.

System or subsystem box practically represents the mathematical model of physical phenomena
happening. In optimisation terminology it represents the transfer function. The type of the models
varies regarding the case study; precision needed and object of optimisation.

System Inputs/Outputs/Interconnections:

The system inputs/outputs/interconnections can be divided in to several categories depending on

the angle of view:

a) Ambient conditions (are input variables defining the weather exploitation conditions there is
several types of ambient condition inputs, most important are temperature of water and air)
b) Loads: Inputs/Interconnections. Loads can be divided in to: heat loads ( heat loads represent
hot water or steam needed for: heating the cargo, fuel, and crew demand on board the ship;
Propulsion load needed to propeller the ship through the water; Electric loads needed for
different processes on board the ship(lighting; pumping; compressed air, navigation etc.

8
 Waste heat/energy subsystems: the waste heat subsystems are subsystems coming after the
main energy production components such as: main engine; auxiliary engine; oil fired boiler
components; The waste heat /energy system needs always to be positioned after the energy
producer which is producing the same type of energy, since load of the energy producer
depends on energy load of system and energy produced by waste heat/energy system.
Interconnections:
Ship system interconnections represent the effect which one ship subsystem has on another
through satisfying energy demand. Ship interconnections in system modelling define the
ranking of subsystems in ship system. Independent systems should be modelled first; the
systems which produce the energy which is already produced by some WER system are to be
modelled letter since the actual load in those systems is actually defined as a difference
between energy load and energy produced by some energy recovery technique.
Principal elementary First type of interconnections example of system containing two types
of energy (mechanical load (propulsion load) and electric load) with all inputs/outputs,
subsystems and interconnections is shown on Picture 4.

EL1 EL1 EL3 msf F 2

Gen. Set.
- WH4
msf F (1,2)
ELPr
WH (1, 2, 3, 4)
ERT WH3
EL(Pr,3)
PL PL Main WH1 WH2
Engine msf F 1

Inputs Interconnections Outputs

Picture (4) Example of system interconnections containing two types of energy (mechanical
load (propulsion load) and electric load) with all inputs/outputs, subsystems and
interconnections

9
Second type of system:

EL msf F2 msf F (1,2)

EL EL1 Gen.
- Set. WH4 WH (1, 2, 3, 4)
ELPr1 msf F 1
PL1 Main EL (Pr1, Pr2, 1)
PL PL WH1 WH3
PTI Engine ERT
WH2 PL1
ELPr2

Inputs Interconnections Outputs

Picture (5) shows principal procedure for interconnecting two subsystems (Propulsion (mechanic)
and electric energy), where system contains feedback line. ERT is producing the electricity which can
go back to the main engine shaft or go back to electric network to reduce the electricity produced
generator set. In this case study the codding process of ship system needs to be iterative since there
is return feedback torque by PTI (power take in) to the main engine. When the energy is put back to
the propeller shaft the amount of waste heat is reduced and iterations should be applied until the
change in the in the waste heat is acceptable. The number of iterations should be enough to satisfy
the required precision but not higher since in contrary the ship system model speed is reduced.

Parameter calculation model

Important
Parameters
Work Energy D.
Losses

Parameter calculation model

Objective Function

Picture (6) Generic representation of Parameter calculation model

10
Input of Parameter Calculation model can be divided in to three principal parts.
First part represents work done by system subsystems depending on the optimisation objective the
work done can be classified as: electric, mechanical, propulsion and heating load.
Second group of inputs is defined as Important Parameters and Energy Demand; it practically
represent all the connections and interconnections from the thermodynamic system entering to the
model explained above, practically represent the mass flow of fuel needed for execution of needed
work. The losses can be described in few categories, exhaust gases, cooling fluids losses and are
defined with mass flow, temperature and fluid properties.
The objective function depends on the type and can be divided in to two types: integral and regular.
a) Integral objective function is used in design optimisation since the end result objective
function defines practically the area under the changeable function in time. If the objective
function is for example the fuel consumption the optimisation point is to actually find the set
of design parameters where the integral version will be minimal.

f0 f
1 fend

Dt

t0 t1 tend

Picture (7) Generic explanation of integral objective function

t end- can represent different values depending on the type of optimisation and it actually
does not need to represent the time it can be defined as some value of 1 or 100 percent,
It usually represents the lifetime exploitation of the ship.

=
0

f- Represents the objective function which is actually the integral under the curve of states in
particular pars of time or some other type of variable.

b) Regular objective functions represent the actual numerical scalar value of important value
which needs to be minimised or maximised depending on optimisation objective.

The objective functions are defined in detail with formulas in some of my previous work but
some of most important are:

Total fuel consumed;

Exergy of system and subsystems;

11
 Percent of exergy of some subsystem compared to the whole system.
Energy efficiency of system.

Percent of energy of subsystem in whole system.

Exergy/ Energy destroyed.

Optimisation:

From secondary and primary preliminary research the optimisation is performed through case
studies. The idea is first to conduct easy problems which practically represent the learning process
unit the final optimisation case studies are performed. First is performed the operational
optimisation meaning that optimal mas flows end pressures are optimised , secondly for different
design solutions should be done through optimal sizing of elements in the optimised systems and
subsystems.

The optimisation is a process which consists of the following steps:

1. Building the simulation of the system.

2. The system needs to be validated with the literature review and common logic.
3. A sensitivity analysis should be performed so that there is obtained the effect of varied
decision variables on the whole system efficiency.
4. The special set of loading conditions should be defined.
5. The optimised algorithm should be developed able of performing the optimisation.
6. The results should be compared with existing results and given estimation of improvement.
7. The different case studies should be compared and obtained best solutions.

One of the objectives of this research is to take into account operational parameters and
also to take in to account combination of systems capable of producing different types of
energy.

In power point through my presentations I explained more than one time the methodologies
and why I believe that my optimisations will be effective if it is needed I can submit them
again.

12
 Achieves List:

Last Year achievements

1 Learned coding
2 ORC Pitch Point Model
3 Oil Fired Boiler Model
4 Report for Funding
5 Main engine
6 Auxiliary Engine
7 Loading Conditions M
8 System Interconnections
9 Heating Balance Model still in process
10 Auxiliary Boiler
11 Calculation Model
12 Heat transfer are single pressure Model
13 Parts for ORC and other waste heat models still need to be combined

I am able to submit the evidence for my last year achievements.

Now needs to be completed

1 Combine ORC components in parallel with optimisation
2 Develop optimisation code
3 Perform Optimisation code on the study cases
Verification of models
Sensitivity study
Results comparative study
4 Writing

13
 Gantt chart: presenting work I have done until now

Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17
Lerning Matlab
ORC Pitch point
Oil Fired Boiler
Report for Funding
Main Engine
Auxiliary Engine
Loading ConditionM
Heating Balance M With this model I had problem practically lost some time
HB CFD Model
Auxiliary boiler
Calculation model
Ship System Model
Waste heat Are Model type 1
ORC; Sat Steam Model Electricity

14
 Gantt chart: presenting work I need to do:

Jul-17 Aug-17 Sep-17 Oct-17 Nov-17 Dec-17 Jan-18 Feb-18 Mar-18 Apr-18 May-18 Jun-18
Flat Finding
Report Writing
Case studies
Optmisation algorithm development
CS1
CS2
CS3
CS4
CS5
CS6
CS7
CS8
Writing and comparative study

Optimsation algorithm development is needed for development of optmisation tools needed for performing the optmisation
CS1 - will be the most easy operational analysis of a 1 heat exchanger heat exchanger Sat Steam Exhaust Gas Economiser
CS2- the complexity will be added more heat exchangers design optmisation infludence of loading conditions ambiend conditions
CS3- Waste Heat recovey model taking the energy of exahaust gasses operational and design optmisation for given operating conditions
sizing of heat exchangers and study of optimsart installed turbo generators
CS4- ORC system with combined energy sources (exhaust gasses, cooling and lybricating liquids, scavange air) design and operational optmisation
CS5 - Combination of systems producing saturated steam and electricity and finding best design solution for the ship lifetime
CS6 -Applying new technology from other industries application of condensating heat exchangers and proposal for their inmplementation in shipping industry
CS7 - Exploitation of different cooling fuids and exploitation of increased esyste meffcicieny
CS8 - Fidning the optmial Crude ooil heating program as a function of time of voyage and ambient conditions.
I have documented all this study cases and the reasons why I need to do them the explanations for them are rellatively long but I could submit it.

15