DC-Based Open Energy System
Mario Tokoro
Sony Computer Science Laboratories, Inc.
A Sustainable, Dependable, and Affordable Solution
for Next-Generation Electrical Power Infrastructures
©2015 Sony Computer Science Laboratories, Inc.
Power
Server Power
Server Client
Client Client
Power Packet
DC Bus Line
Table-top Experiments, 2005
in Sony CSL
Ghana Expedition, 2010
Okinawa, 2012 Sony CSL - OIST Joint project
©2015 Sony Computer Science Laboratories, Inc.
Bangladesh, 2014
Okinawa, since Dec. 2014
Supported by Okinawa Prefecture Government
©2015 Sony Computer Science Laboratories, Inc.
Our Fundamental Thought Sustainable
Dependable Affordable
Do not pass on liability to next generation
Keep risk manageable
Accessible to everyone on earth
Conventional Power Systems
©2015 Sony Computer Science Laboratories, Inc.
Do not pass on liability to next generation
Keep risk manageable
Accessible to everyone on earth
Renewable Energy Sources, which are
Distributed, Intermittent, and Unstable, with Batteries
Local Consumption of the Locally Produced by Self-Supporting as much, in the form of
Bottom-Up, Autonomous, and Interconnected System can start Small yet Expandable in the form of Bottom-Up, Autonomous, and
Interconnected
Our Approach Sustainable
Dependable
Affordable
Open Energy Systems
©2015 Sony Computer Science Laboratories, Inc.
Conventional systems
• Distributed energy sources
• Local consumption of the locally produced
• Bottom-up and flexible configuration of distributed autonomous systems
• Stock based, asynchronous load/supply balancing
• Exchange network
• Centralized energy source
• Transmission for distant consumption
• Top-down configuration with central control
• Flow based, synchronous load/supply balancing
• Distribution network
Open Energy Systems
How do we achieve OES?
DCOES Technologies
©2015 Sony Computer Science Laboratories, Inc.
Renewable Energy Sources
http://commons.wikimedia.org/wiki/File:Machikawa_power_
station_1.jpg#/media/File:Machikawa_power_station_1.jpg
©OIST
Batteries / Energy Storage
©2015 Sony Computer Science Laboratories, Inc.
SONY Ferro-Phosphoric Acid Li-Ion Batteries
• Full use of capacity
• Long life under tough conditions
• Inherent safety:
no thermal runaway
• Fast Charging time:
1 hour for >90% capacity
• Eco-friendly:
no rare metals
SONY Ferro-Phosphoric Acid Li-Ion Batteries
©2015 Sony Computer Science Laboratories, Inc.
How to balance the PV generation
and battery capacity?
Batteries cannot take whole energy produced by PV
©2015 Sony Computer Science Laboratories, Inc.
Variety in usage pattern
Image ©2014 DigitalGlobe, Map data © 2014 Google, ZENRIN
Energy Exchange Technology
Exchange energy among batteries In the form of DC
In order to maximize the use of PV panels and batteries
through complementing difference in usage patterns
©2015 Sony Computer Science Laboratories, Inc.
• Reactance loss of power is not negligible
• Interconnection of
networks is difficult due to sync.
• Efficient DC/DC converters are now available
• Interconnection of
networks is easy (no sync.
is necessary)
AC for transformers DC
for batteries
Phase Voltage
Frequency
Voltage
Base Theory
Constant Voltage Source keeps
the grid voltage at 350V Constant Current Sources set desired current
One Voltage Source and n Current Sources with Durable and Flexible Distributed Control
©2015 Sony Computer Science Laboratories, Inc.
m-to-n Energy Exchange
DC/DC converter can have 3 modes:
• Waiting (stop)
Constant Voltage mode (CV)
DC/DC #0 #1
DC/DC #2
DC/DC #3
DC/DC #4
DC/DC #5
DC/DC
Set the Grid to 350V by CV mode (#0) Deal 1: Send energy from #1 -> #3 Deal 2: Send energy from #4 -> #2, #5
VG=350V
CC
2A 2A
CC
IG=2A IG=2A
CV CC CC CC
4A
2A 2A
IG=4A IG=2A IG=2A
2A 2A 2A
350V Bus line
Energy Exchange Policies
• Baseline policy
– Capacity available for giving/receiving – Request for consumption
• Advanced policies
– Prediction based on past usage pattern
– Weather forecast for generation and consumption – Dynamic pricing reflecting demand and supply
©2015 Sony Computer Science Laboratories, Inc.
Scalable Architecture
Power Exchange Network
When necessary
When necessary
Most exchanges happen at the lowest level!
©2015 Sony Computer Science Laboratories, Inc.
DCOES@OIST20 System Structure (1)
• Configuration of each house
48VDC
100VAC
Energy System Controller
Charger PV
Bi-directional DC/DC
Inverter and AC switch
350VDC
Energy Server System
Battery
Communication Line
DCOES@OIST20 System Structure (2)
• 19 houses connected
Power Line
house 1 house 2 house 19
switch switch switch
switch switch switch
switch switch switch
...
350VDC
©2015 Sony Computer Science Laboratories, Inc.
Energy Server System
• SONY 48V Li-Ion battery modules
• 350V Grid
• Energy exchange module, DCDC
• DC to AC inverter for appliances
• AC backup by utility company
Main Controller PV Charger BMU Battery 1.2kWh Battery 1.2kWh
DC-DC Conv.
Inverter EMU
PV Charger
Battery 1.2kWh
Battery 1.2kWh
350V DC Power Lines (Privately Owned)
©2015 Sony Computer Science Laboratories, Inc.
©2015 Sony CSL
Example Daily Operation
Performance Evaluation
©2015 Sony Computer Science Laboratories, Inc.
Performance Evaluation
Distributed stand-alone
Top-down by a single entity
• Fixed in size with high initial costs
• Single failure may cause total system outage
• Efficient for a predefined users and Bottom-up by individuals
• Flexible in size and variety with incremental costs
• Failures do not cause total system outage
Centralized
vs
DCOES: Distributed System with Energy Exchange
DCOES
• Flexible in size and variety with incremental costs
• Failures do not cause total system outages
• High efficiency and flexibility in size and variety
©2015 Sony Computer Science Laboratories, Inc.
Variety in usage pattern (Winter)
Q3 Median Q1 outlier
x
Variety in usage pattern (Spring)
Q3 Median Q1 outlier
x
©2015 Sony Computer Science Laboratories, Inc.
Performance Result and Estimation
(Current OIST configuration with baseline policy)
Stand alone DCOES Centralized
Summer: Simulation (19 houses) 2015/7/16- 8/15
Winter: Real OIST Data (19 houses)
2015/1/1- 1/31 Spring: Real OIST Data (19 houses) 2015/4/16- 5/16
34%
60% 63%
38%
67% 71%
42%
74% 77%
0%
20%
40%
60%
80%
100%
Winter Spring Summer
Stand Alone DCOES
Centralized
Performance Result and Estimation
(Solar x2, Battery x2 with baseline policy)
Summer: Simulation (19 houses) 2015/7/16- 8/15
Winter: Real OIST Data (19 houses)
2015/1/1- 1/31 Spring: Real OIST Data (19 houses) 2015/4/16- 5/16
57%
87% 90%
67%
94% 96%
73%
96% 99%
0%
20%
40%
60%
80%
100%
Winter Spring Summer
Stand Alone DCOES
Centralized
©2015 Sony Computer Science Laboratories, Inc.
Self-sufficiency ratio = (Energy used - Energy bought)/ Energy used
Next Plans
DCOES/LV for Houses in a Community
©2015 Sony Computer Science Laboratories, Inc.
350V DC Line PV: 3-5kWp
Batt.: 5kWh
DCOES/MV for Larger Communities
350V DC Line PV: 3-5kWp
Batt.: 5kWh
DCOES/
LV DCOES/
DCOES/ LV LV
DCOES/ DCOES/ DCOES/
LV DCOES/
LV
DCOES/
+ 1KV Line
DCOES/MV
DCOES/MV for Eco Campuses
1KV DC Line PV: 100kWp Batt.: 500kWh
©2015 Sony Computer Science Laboratories, Inc.
DCOES/HV for Cities
DCOES
DC based Open Energy System
For sustainable mankind and development
©2015 Sony Computer Science Laboratories, Inc.