Transcript Storage optimization in distribution systems

Storage Optimization in
Distribution Systems
Roger Cremers & Gabriël Bloemhof
KEMA (The Netherlands)
CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
Need for tool to assess storage solutions

Utilities are faced with increasing number of
distributed energy sources. Storage devices can
facilitate the implementation of these sources in
the power system

The implementation of storage devices in power
systems faces the utilities however with a lot a of
questions that need to be answered
CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
Typical questions
Can a storage system alleviate the problems in
my distribution network?
 I have a limited amount of money to buy storage
systems. What systems should I buy?
 I need a certain amount of storage capacity in
my power system. Should I buy only one storage
device or multiple smaller devices?
 Can storage based solutions compete with
classical solutions?

CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
Too many alternatives to analyse…

Optimizing the location, type and size of mobile
storage systems is a combinatorial problem with
(too) many possible solutions
Number of unique solutions
100000000
Possible combinations
10000000
1000000
100000
50 nodes
100 nodes
200 nodes
10000
1000
100
10
1
2
3
4
Desired number of storage systems
CREMERS – NL – RIF Session 4 – Paper 0180
5
Frankfurt (Germany), 6-9 June 2011
PLATOS
New planning tool for optimising the application
of (mobile) storage systems in electrical power
systems
 PLATOS can assist to address relevant issues
involved with storage applications

CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
PLATOS is based on artificial evolution
Create 1000 random solutions
Analyze each solution:
Costs, Benefits, Performance…
Select best 50 solutions
Generate 1000 new solutions on best 50:
Survival, Inheritance, Mutation, Cross-over
Repeat until convergence
CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
Performance indicator

Performance of each unique solution is indicated
by a performance indicator
PI = NPV (annual benefits, annual costs)
CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
User definable solution space

User defines:





What benefits and costs to be taken into account
Number of storage types and sizes to choose from
Desired storage technology
Optimization objective
Network constraints
CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
Graphical and tabular output
Use of single line diagram to present results
graphically
 Creation of Excel file containing all relevant
results

2
2
50 kWh
50 kWh
2
2
50 kWh
50 kWh
CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
2
50 kWh
2
50 kWh
2
50 kWh
Frankfurt (Germany), 6-9 June 2011
Frankfurt (Germany), 6-9 June 2011
Main features of PLATOS

Optimization of storage application in power systems




Performance indicators can be defined by the user







Optimization of location, size and type
Optimization criteria can be changed by the user
Monitoring of optimization process
Definition of points of interest within power system
Both technical and economical performance indicators
Graphical and tabular output
Comparison with classical non storage based solutions
User definable load and generation patterns
Tool can be used for each voltage level
Transparant and uniform comparison of storage and non storage
based solutions to network problems
CREMERS – NL – RIF Session 4 – Paper 0180
Frankfurt (Germany), 6-9 June 2011
Questions
?
CREMERS – NL – RIF Session 4 – Paper 0180