25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 1 www.fernwaerme.de
FERNWÄRME-FORSCHUNGSINSTITUT
GENETIC ALGORITHM TECHNIQUE TO OPTIMIZE THE CONFIGURATION OF HEAT STORAGE
IN DH NETWORK
Amru Rizal Razani M.Sc.
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 2
Introduction
www.fernwaerme.de
Heat storage integrated in a DH system*
*source: Martin & Thornley, Tyndall Centre for Climate Change Research
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 3
Outline
www.fernwaerme.de
1. DH Network Configuration, Topology & Modeling
(Graph Model)
2. Calculation of heat costumer load profile
3. Determining heat storage layout and volumes
4. Generating cost function
5. Optimization with Genetic Algorithm
6. Result and Discussion
7. Summary and outlook
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 4
DH Network Configuration, Topology
& Modeling (Graph Model)
www.fernwaerme.de
Network Modeling
(3 Scenarios)
1. Centralized
2. Semi decentralized
3. Full decentralized
Network parameter
dimensioning (length,
pipe diameter,
consumer heat load)
Network simulation
(thermohydraulic, heat
loss)
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 5
DH Network Configuration, Topology,
& Modeling (Graph Model)
www.fernwaerme.de
1. Centralized
2. Semi decentralized
3. Full decentralized
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 6
Calculation of Heat Consumer Load Profile
www.fernwaerme.de
• text
Calculating
Consumer Standard
Load Profile
Data processing for
every hour during a
year using outer
temperature data,
building type, week
factor and hour factor
Reference: BGW
-
100.000,00
200.000,00
300.000,00
400.000,00
500.000,00
600.000,00
700.000,00
800.000,00
School Office SME Appartment Public Resto
Yearly energy consumption [kW]
0
200
400
600
800
1000
1200
Lo
ad
[kW
]
Hour
Yearly heat energy load in the network
Max: 950,55 kW
Min: 26,00 kW
Mean: 290,96 kW
Qsp = m Cp ΔT
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 7
Determining heat storage layout and volumes
www.fernwaerme.de
Determining heat storage
layout in the network &
volume for every scenario,
In daily basis according to
Load profile,
Calculate integral
∫ Q dt of the curve->
numerical integration
Using Excel table
0,00100,00200,00300,00400,00500,00600,00700,00800,00900,00
1000,00
Load [
kW
]
Time
Heat consumption line in a winter type day
Max: 950,5 kW
Min: 563,7 kW
Mean: 739,0 kW
Qsp= Msp Cpsp dTsp=Vsp ρsp cpsp (T1sp-T2sp )
Qsp[J] : heat capacity of the storage
Msp [kg] : mass of the storage
Cpsp[J/kgK] : specific heat of storage media (water)
dTsp[K] : temperature difference in storage
Vsp[m³] : volume of the storage
ρsp[kg/m³] : density of storage media
T1sp[°C] :temperature of loaded storage
T1sp[°C] : temperature of unloaded storage
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 8
Generating Cost function, interpolation
www.fernwaerme.de
• text Generating non linear
cost function
(Network, CHP, Storage)
F(d), F(Vol), F(kW),
interpolation
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 9
Optimization with Genetic Algorithm
www.fernwaerme.de
Optimization steps using Genetic
Algorithm
Objective function → cost
function (non linear)
Constraints:
- Capacity of CHP
- Capacity of heat storage
Upper bound
Lower bound
Summary of GA method
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 10
Result and Discussion
www.fernwaerme.de
Results & Analysis
o Network type 1 has the lowest
investition cost but require
more heat energy
o The energy efficiency reached
by network type 3, but it
needs higher initial
investment.
o Network of type 2 has
medium efficiency of cost and
energy, depends on the
installation volume of the heat
storage and its location as
well.
o Network type 3 requires
higher cost, but it offers
higher heat supply security
0,00
5,00
10,00
15,00
20,00
25,00
30,00
35,00
0
20
40
60
80
100
120
140
160
180
200
Type 1 Type 2 Type 3
Comparison of the calculated DH Network Variations
CHP Load [kW] Cost [T€] Heat Storage [m³]
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 11
Summary & Outlook
• Detail calculation of storage volume by utilizing learning algorithm
• Control on heat customer side (demand side management)
increases efficiency of the heat production
• Combination of heat sources (solar cell, geothermal, etc.) in the
network is possible using the same method
• Other layout combinations (close loop, more heat sources) should
be investigated as well.
• Comparing the result with operational data of the heat production
plant with optimization, data integration and performance testing.
www.fernwaerme.de
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 12
Reference
1. BGW Praxisinformation P2007/13 Gastransport/Betriebswirtschaft,
Abwicklung von Standardlastprofilen
2. ASUE (Arbeitsgemeinschaft für Sparsamen und
Umweltfreundlichen Energieverbrauch e.V.), BHKW-Kenndaten
2005
3. http://kfserver.kaiserstadt.de/ (Kostenfunktions-Server)
4. Optimization in Scilab, The Scilab Consortium, July 2010.
5. Phetteplace, Gary, Optimal Design of Piping Systems for District
Heating, August 1995.
www.fernwaerme.de
25 – 26 August 2015 Smart Energy System and 4th Gen DH, Copenhagen Folie 13
Thank you very much
Questions?
www.fernwaerme.de