5. 20.06.2018 5
The main challenge in microgrids
• Technical & economical
– Flexible Devices
• Integrate user-specific preferences
– Metering elements
• Low cost
• Interoperability
– Connectivity
• Low energy consumption
– Modeling
• individual modelling is a challenge (heterogeneous appliances)
– Control Objective
• Local vs System-level services
• Regulatory
– Regulation rules may ban microgrids or lead to prohibitive costs
– Belgium : changes are being implemented (as in FR, DE,…)
A few challenges to be tackled
6. 20.06.2018 6
Project STORY E-Cloud GAC
Residential X - X
Industrial - X -
Local Energy (RES) X X X
Battery Storage (non-EV) X - TBD
First results (20.06.18) X X -
Dynamic Pricing X TBD
Energy Allocation
Mechanism
- X TBD
Active Grid Management - X X
Story, E-cloud & GAC
Comparison of project scopes
7. 20.06.2018 7
Agenda
• Introduction
• Story - Project overview (Actility)
• Story - Oud-Heverlee Belgium Demo (Actility)
– Results from Story Belgian Business case
– Residential street-level optimization case
• E-Cloud (ORES)
– E-cloud : Industrial microgrid
• GAC (Greenwatch)
– GAC : Residential microgrid
• Q&A
Overview of today’s discussions
8. 20.06.2018 8
Story
• What happens if a
– Large amount of
– Small-scale storage is
– Integrated into the distribution grid
The core question of the story project
9. 20.06.2018 9
Story - Horizon 2020 Framework
• STORY team
– 18 institutions
– 8 countries
– Use of storage, Residential & Industrial level
Overview
11. 20.06.2018 11
Story - Horizon 2020 Framework
The scope of our contribution
• Actility’s contribution
– Centralized optimization of energy use within a residential
microgrid
• Connection + equipment of a dozen houses
• Optimization of flexible appliances
• Poll 1
– What’s the key objective of Residential microgrid ?
12. 20.06.2018 12
Story - Horizon 2020 Framework
Actility – Demand Response Aggregator : who we are
15. The Story Project – OHL Demo
Geographical situation
15
*The living Lab(source : Google)
*
16. The Story Project – OHL Demo
Connecting a Microgrid
16
(source : Google)
Challenge #3 – Connect
• LoRaWan for Meas. & control
• Detailed measurements are
sent via IP due to data rate
limitation (IoT)
Challenge #1 – Flexible Devices
• Multi-vendor Heating Appliances
• Batteries (installed by Story)
• Multi-vendor EVs
Challenge #2 – Measure
• Electric Consumption
• States (Temperature, On/Off)
Challenge #4 – Control
• Model House & Heating appliances
17. The Story Project – OHL Demo
Lessons learned from installations
17
Fuses location is
unkown
Phases can be
swapped between
Main fuse box & e.g.,
Ground fuse box
Houses are living things.
The installation can
change due to other
providers intervention
Connectivity can be
limited in countryside
19. Story OHL Demo
Case Study 1 : Building level optimization
Building scope • Overall MPC* framework
• Simulation Results
• Real-Life Results
*Model Predictive Control
20. Control of Consumption - MPC
Model Predictive Control & Use cases Case 1 : Dynamic pricing
Case 2 : Minimize grid exchange
21. Control of Consumption – Simulation
Results
Appliance annual energy
cost decrease of about 10%
to 20%
Case 2 : Minimize grid exchange
Case 1 : Dynamic pricing App 1 App 2 App 3 App 4 App 1 App 2 App 3 App 4 App 1 App 2 App 3 App 4
House 1 House 1
Maximum Injection levels are
drastically reduced (even
without battery)
Dynamic pricing likely to increase
maximum offtake & injection
1
2
22. Model & connectivity example
Boiler
Temperature
Power consumption
Control
3 Temperature sensors Communication Optimisation & control
Boiler Energy State
Boiler model
23. Real life result (1/2)
Boiler
Bottom Temp
Top Temperature
Energy Content
(dotted : model)
User behaviour influences massively the boiler states
Water withdrawal have
a stochastic nature
24. Real life result (1/2)
Boiler
Energy price
State
Energy is consumed during low price period
Power
𝟓𝟎𝟎𝒌𝑾
~𝟖𝟑 €/𝑴𝑾𝒉
~𝟑𝟑 €/𝑴𝑾𝒉
Energy price
25. Real life result (2/2)
Building 1 : Direct HP power control (Gear control)
-> Thermal inertia is large due to floor heating but introduces delays
-> Interactions with domestic hot water production not easily handled
26. Real life result (2/2)
Uncontrolled Controlled
Building 2 : Indirect power control via Thermostat (Temperature set-point)
-> Much more difficult than direct control
-> Set-point follows a much more variable pattern when externally controlled
Uncontrolled case : large amount of energy at peak
Controlled case – in case energy is
required at peak, the algorithm limits it
as much as possible (short start)
Controlled case – at first, the model
underestimates thermal inertia
27. 20.06.2018 27
Conclusions
• A house is complex
– Inter-operability is a challenge
– Home are living thing (promote self installation ?)
– Neighborhood : phase identification is needed (coordination)
• Control Results
– Positive results in Simulation
• Case 1 : 10-20 % appliance energy cost reduction in Dynamic Pricing,
leading to larger demand peaks or injection levels.
• Case 2 : Injection can be reduced to a higher extend than offtake
– Results Confirmed in real-life control
• Further assessment is undergoing
– Model Calibration (thermal inertia) is key (machine learning ?)
What we take home in this first part of the Demo
31. Vangulick – BE – S6 - 0221
Some of you may think: Microgrids are the Armageddon of the
DSO’s world
Microgrids
32. Vangulick – BE – S6 - 0221
• The central ideas are
– Generation units are made available to a community
of industrial customers
– Local generation that is consumed locally is
considered as auto-consumption
– Energy that is not produced locally enters the
classical market processes. Customers still have their
freedom of negotiation for this part.
– The storage of electrical energy could help to
increase the consumption of local generation within
the E-Cloud
E-Cloud: General concept
33. Vangulick – BE – S6 - 0221
• The E-Cloud is
– an electrical system not separated from the conventional network
– integrating consumers and local generation units aimed at
optimizing energy flows,
– optimizing for both consumers and producers and for the
Community
– this optimum, which varies in a quarter of an hour, requires to
• Combine implementation of information technology and smart grid
• Customer’s willingness to participate in this process and,
• Educate them to better understand their own consumption
E-Cloud: Key characteristics
34. Vangulick – BE – S6 - 0221
E-Cloud: How does it work ?
Total
Generation
(/1/4h)
Allocation
key
C
o
u
c
h
e
p
h
y
s
i
q
u
e
C
o
u
c
h
e
d
a
t
a
35. Vangulick – BE – S6 - 0221
E-Cloud: How does it work ?
DSO ADT(*) Factory A
SME B
Producer
SME C
(*) ADT= Agence de Développement Territorial : Agency rresponsible for the
implementation, promotion and management of business park (Zone Activité
Economique (ZAE))
E-Cloud: Participation Agreement
• Factory A:
• Production part : 40%
• Rental to pay to producer:
40000€/year
• SME B:
• Production part : 35%
• Rental to pay to producer:
35000€/year
• SME C:
• Production part : 25%
• Rental to pay to producer:
25000€/year
• Producer
• Install 1 MW wind turbine & 250
kW PV
36. Vangulick – BE – S6 - 0221
All customers keep their right to negotiate for their non locally
produced energy
E-Cloud: How does it work ?
SME B
SME C
Facory A
Supplier
Zeta
Supplier
Zeta
Supplier
Omega
37. Vangulick – BE – S6 - 0221
• In each of the factories / SMEs, a (new) role is created: the Energy
Manager
– Played by the customer’ staff or by supplier or by third party
– Receives forecasts of its share of generation & its consumption for
the next day (provided by the DSO)
– can adapt his consumption planning
– During the day, can verify its share of generation meet its actual
consumption (provided by the DSO). It can try to maximize the use of
the local generation
E-Cloud: How does it work ?
Active –consumers (Consom-Acteur)
New Work
planning
38. Vangulick – BE – S6 - 0221
• Customer recruitment
– ZAE Tournai :45 Customers = 14 Recruited !
– ZAE Sart-Tilman: Info session done
• Optimization
– Design on going (result begin of June)
– Target performance: 10% decreasing of the energy
bills
• Regulation issues
– Possibility for a demonstrator to have special market
rules: ok Regulator agreed
– In line with « Winter Package » (Energy community)
Conclusion
Where are we so far ?
40. Vangulick – BE – S6 - 0221
• We do believe that E-Cloud is a 4 « win »:
– Win for businesses, which increase their competitiveness.
– Win for all network users because the E-Cloud relieves
network constraints and, potentially, the investment needs
and operational costs for the DSO.
– Win for the Walloon region, which facilitates the
achievement of renewable production and
competitiveness targets.
– Win for producers, who can install generation capacities
in areas less sensitive to NIMBY.
Conclusion
42. 20-06-18 42
• PV monitoring of 6.000 residential clients
• Multi-energy monitoring
• After-sale service
• Experience since 2009
• Offer and demand flexibility
• Remote load optimization
43. Control of PV
inverters
PV predictions
GATEWAY
Control of
residential loads
ICT platform
• Evaluation, prediction,
activation and monitoring
of demand flexibility
• Market interaction• Monitoring
• Smart load control
20-06-18 43
Offer and demand
flexibility
44. • Neighborhood with 50 houses controled
• DSO is partner of the project
• Preliminary results (+10 years)
Increase in self-consumption
✓ Neighbordhood : 53 % → 74 %
✓ Individual : 39 % → 48 %
Decrease in consumption peaks
(individual and neighborhood)
Project implementation
Without
GAC
With
GAC
39 48
14
26
48
26
0%
20%
40%
60%
80%
100%
Self-consumption (+10 years
situation)
20-06-18 44
45. • White goods
• Electrical boiler
• Electrical heater
• Heat pumps
• Batteries?
• Electric cars
Sensors
Loads
Gateway
GAC
platform
Z-Wave 4G
Control of the loads
Market
actors
Internet
• GWio
• Control of smart
plugs, smart
relays and
specific
interfaces
• Aggregation
• Monitoring
• Optimization
• Tariff
incentives
• Flexibility
• Facilitation of
flexibility
activation
• Flexibility
demands
• Congestion checks
20-06-18 45
47. Valorization of flexibility:
Individual: prosumers
• Self-consumption optimization
Collective: neighbordhood
• Self-consumption optimization
lia
C
cCc
C
ser
seri
Transfo
20-06-18 47
Offer and demand flexibility
49. Valorization of flexibility:
Individual: tariff incentives
• Dynamic pricing
• Peak pricing
• Pricing periods
Collective: neighbordhood
• Flexibility to the DSO
• Flexibility market
lia
C
cCc
ser
seri
Transfo
20-06-18 49
Offer and demand flexibility
50. What are we looking for?
• Additional services we could include in our offer
• Centralised and decentralised batteries, blockchain
• Additional playgrounds to test our solutions & more
• ERA-NET project, H2020, …
20-06-18 50