a network architecture for localized electrical energy
TRANSCRIPT
A Network Architecture for Localized Electrical Energy Reduction,
Generation and SharingDavid Culler
Randy Katz, Seth Sanders, and TeamUniversity of California, Berkeley
“Energy permits things to exist; information, to behave purposefully.” W. Ware, 1997
A Network Architecture for Localized Electrical Energy Reduction,
Generation and SharingDavid Culler
Randy Katz, Seth Sanders, and TeamUniversity of California, Berkeley
“Energy permits things to exist; information, to behave purposefully.” W. Ware, 1997
A new kind of Roadmap for CS
2
Qua
ds
19731949 2005 2030 2050
A new kind of Roadmap for CS
2
Qua
ds
19731949 2005 2030 2050
A new kind of Roadmap for CS
2
Qua
ds
19731949 2005 2030 2050
1990
A new kind of Roadmap for CS
2
Qua
ds
19731949 2005 2030 2050
1990
A new kind of Roadmap for CS
2
Qua
ds
19731949 2005 2030 2050
199050%
A new kind of Roadmap for CS
2
Qua
ds
19731949 2005 2030 2050
199050%
80%
A new kind of Roadmap for CS
2
Qua
ds
19731949 2005 2030 2050
199050%
80%
Towards an “Aware” Energy Infrastructure
3
Baseline + Dispatchable Tiers Oblivious Loads
Towards an “Aware” Energy Infrastructure
3
Baseline + Dispatchable Tiers Oblivious Loads
TransmissionGeneration DemandDistribution
Towards an “Aware” Energy Infrastructure
3
Baseline + Dispatchable Tiers Oblivious Loads
Non-Dispatchable Sources
TransmissionGeneration DemandDistribution
Towards an “Aware” Energy Infrastructure
3
Baseline + Dispatchable Tiers Oblivious Loads
Non-Dispatchable Sources Aware Interactive
Loads
TransmissionGeneration DemandDistribution
Towards an “Aware” Energy Infrastructure
3
Baseline + Dispatchable Tiers Oblivious Loads
Non-Dispatchable Sources
Communication
Aware Interactive Loads
TransmissionGeneration DemandDistribution
Towards an “Aware” Energy Infrastructure
3
Baseline + Dispatchable Tiers Oblivious Loads
Communication
Non-Dispatchable Sources
Communication
Aware Interactive Loads
TransmissionGeneration DemandDistribution
Towards an “Aware” Energy Infrastructure
3
Baseline + Dispatchable Tiers Oblivious Loads
Communication
Non-Dispatchable Sources
Communication
Aware Interactive Loads
TransmissionGeneration DemandDistribution
Where to Start?
4
Renewable energy consumption
Electricity source
Where to Start?
• Buildings– 72% of electrical consumption (US), – 40-50% of total consumption, – 42% of GHG footprint– US commercial building consumption
doubled 1980-2000, 1.5x more by 2025 [NREL]
4
Renewable energy consumption
Electricity source
Where to Start?
• Buildings– 72% of electrical consumption (US), – 40-50% of total consumption, – 42% of GHG footprint– US commercial building consumption
doubled 1980-2000, 1.5x more by 2025 [NREL]
• Where Coal is used
4
Renewable energy consumption
Electricity source
Where to Start?
• Buildings– 72% of electrical consumption (US), – 40-50% of total consumption, – 42% of GHG footprint– US commercial building consumption
doubled 1980-2000, 1.5x more by 2025 [NREL]
• Where Coal is used• Prime target of opportunity for
renewable supplies
4
Renewable energy consumption
Electricity source
Load-following Supply
5
Load-following Supply
5
Load-following Supply (really !)
6Source: EEX spot prices.
Growing proportion of renewables leads to higher price volatility. October 2008 to March 2010:
-500.02Daily minimum price (indicated in red when negative)Daily maximum price
494.26
€/M
Wh
Merit-orderaveragerange
Start from Scratch?
7
Start from Scratch?
• No!
7
Grid Exists
8
Conventional Electric Grid
GenerationTransmission
DistributionLoad
Internet Exists
9
Conventional Electric Grid
GenerationTransmission
DistributionLoad
Conventional Internet
Intelligent Energy Network as Overlay on Both
10
Conventional Electric Grid
GenerationTransmission
DistributionLoad
Conventional Internet
Intelligent Energy Network as Overlay on Both
10
Conventional Electric Grid
GenerationTransmission
DistributionLoad
Intelligent Energy Network
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
Conventional Internet
Intelligent Energy Network as Overlay on Both
10
Conventional Electric Grid
GenerationTransmission
DistributionLoad
Intelligent Energy Network
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
Conventional Internet
Intelligent Energy Network as Overlay on Both
10
Conventional Electric Grid
GenerationTransmission
DistributionLoad
Intelligent Energy Network
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
Conventional Internet
11
Intelligent Power Switch
• Scalable Building Block– Network + Power– Monitor, Predict, Communicate, Control
• Decouple: Buffering vs Scheduling
Intelligent Power Switch
(IPS)
Energy Network
Power-Comm Interface
EnergyStorage
PowerGeneration
Host Load
energy flows
information flows
A Cooperative Grid
12
comm
powerInternet
Grid
AHU
Chill
CT
A Cooperative Grid
12
comm
powerInternet
Grid
IPS
IPS
IPS
AHU
Chill
CT
A Cooperative Grid
12
comm
powerInternet
Grid
IPS
IPS
IPS
AHU
Chill
CT
IPS
A Cooperative Grid
12
comm
power
now
Load profile
w
Internet
Grid
IPS
IPS
IPS
AHU
Chill
CT
IPS
A Cooperative Grid
12
comm
power
now
Load profile
w$now
Price profile
w
Internet
Grid
IPS
IPS
IPS
AHU
Chill
CT
IPS
A Cooperative Grid
12
comm
power
now
Load profile
w$now
Price profile
w
now
Actual load
w
Internet
Grid
IPS
IPS
IPS
AHU
Chill
CT
IPS
A Cooperative Grid
12
comm
power
now
Load profile
w$now
Price profile
w
now
Actual load
w
Internet
Grid
IPS
IPS
IPS
AHU
Chill
CT
IPS
IPS
Bldg Energy Network
A Cooperative Grid
12
comm
power
now
Load profile
w$now
Price profile
w
now
Actual load
w
IPS
IPS
IPSInternet
Grid
IPS
IPS
IPS
AHU
Chill
CT
IPS
IPS
Bldg Energy Network
A Cooperative Grid
12
comm
power
now
Load profile
w$now
Price profile
w
now
Actual load
w
IPS
IPS
IPSInternet
GridData center
IPS
IPS
M/R Energy
Net
IPS
IPS
IPS
AHU
Chill
CT
IPS
IPS
Bldg Energy Network
A Cooperative Grid
12
comm
power
now
Load profile
w$now
Price profile
w
now
Actual load
w
IPS
IPS
IPSInternet
GridData center
IPS
IPS
M/R Energy
Net
IPS
IPS
IPS
AHU
Chill
CT
IPS
IPS
Bldg Energy Network Power proportional services
Stages of Energy Effectiveness
13
Stages of Energy Effectiveness
• Waste elimination– Do Nothing Well !!!
13
Stages of Energy Effectiveness
• Waste elimination– Do Nothing Well !!!
• Power Proportionality – Power : Performance (utilization)– Partial Load - from nothing to peakl
13
Stages of Energy Effectiveness
• Waste elimination– Do Nothing Well !!!
• Power Proportionality – Power : Performance (utilization)– Partial Load - from nothing to peakl
• Sculpting– Identify the energy slack and utilize it
13
Stages of Energy Effectiveness
• Waste elimination– Do Nothing Well !!!
• Power Proportionality – Power : Performance (utilization)– Partial Load - from nothing to peakl
• Sculpting– Identify the energy slack and utilize it
• Negotiated Grid / Load / Human Interaction– Plan, Forecast, Negotiate, Manage
13
14
Our Buildings
14
Envir
onm
enta
l
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
• Wasteful
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
• Wasteful• <20 % Power Prop.
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
• Wasteful• <20 % Power Prop.
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
• Wasteful• <20 % Power Prop.• Predictable
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
• Wasteful• <20 % Power Prop.• Predictable
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
• Wasteful• <20 % Power Prop.• Predictable• Sculptable ?
Our Buildings
14
Envir
onm
enta
lOpe
ratio
nal
• Wasteful• <20 % Power Prop.• Predictable• Sculptable ?
Our Buildings
Do nothing poorly
15
Envir
onm
enta
lOpe
ratio
nal
Our Buildings
15
Envir
onm
enta
lOpe
ratio
nal
Lighting
Our Buildings
15
Envir
onm
enta
lOpe
ratio
nal
Lighting
Our Buildings
HVAC
15
Envir
onm
enta
lOpe
ratio
nal
Lighting
Our Buildings
HVACIT and Plug Load
15
Envir
onm
enta
lOpe
ratio
nal
Lighting
Servers
PDUs, CRACs
Our Buildings
HVACIT and Plug Load
15
Envir
onm
enta
lOpe
ratio
nal
Lighting
Servers
PDUs, CRACs
Our Buildings
HVACIT and Plug Load
Design
15
Envir
onm
enta
lOpe
ratio
nal
Lighting
Servers
PDUs, CRACs
Our Buildings
HVACIT and Plug Load
DesignUse
15
Envir
onm
enta
lOpe
ratio
nal
Lighting
Servers
PDUs, CRACs
Our Buildings
HVACIT and Plug Load
DesignUse
Waste Less
16
Waste Less
16
Waste Less
16
The Data tells the story…
• Monitor Based Commissioning– Eliminate simultaneous heat/cool– AC91 on schedule
17
The Data tells the story…
• Monitor Based Commissioning– Eliminate simultaneous heat/cool– AC91 on schedule
17
?
The Data tells the story…
• Monitor Based Commissioning– Eliminate simultaneous heat/cool– AC91 on schedule
17
?
IT does nothing poorly too
18
“The Case for Energy-Proportional Computing,” Luiz André Barroso,Urs Hölzle, IEEE Computer December 2007 – study of 5,000 servers
IT does nothing poorly too
18
“The Case for Energy-Proportional Computing,” Luiz André Barroso,Urs Hölzle, IEEE Computer December 2007 – study of 5,000 servers
IT does nothing poorly too
18
“The Case for Energy-Proportional Computing,” Luiz André Barroso,Urs Hölzle, IEEE Computer December 2007 – study of 5,000 servers
ave
IT does nothing poorly too• Designed for peak
18
“The Case for Energy-Proportional Computing,” Luiz André Barroso,Urs Hölzle, IEEE Computer December 2007 – study of 5,000 servers
ave
IT does nothing poorly too• Designed for peak• 10-50% Power Prop.
18
“The Case for Energy-Proportional Computing,” Luiz André Barroso,Urs Hölzle, IEEE Computer December 2007 – study of 5,000 servers
ave
IT does nothing poorly too• Designed for peak• 10-50% Power Prop.• As bad as buildings
18
“The Case for Energy-Proportional Computing,” Luiz André Barroso,Urs Hölzle, IEEE Computer December 2007 – study of 5,000 servers
ave
Where does the Power go?
• The Glue, not the processor
Core i7
Atom 333
Westmere
Power Proportional IT Cloud
20
Capacity
IPSRequests
Power
ForecastsAvailability
Power Proportional IT Cloud
20
Capacity
IPSRequests
Power
ForecastsAvailability
Energy “Slack”
21
Thermostatically Controlled Load
Energy “Slack”
21
Thermostatically Controlled Load
IPS
Energy “Slack”
21
Thermostatically Controlled Load
IPS
Energy “Slack”
21
Thermostatically Controlled Load
IPS
Energy “Slack”
21
Thermostatically Controlled Load Set Point
IPS
Energy “Slack”
21
Thermostatically Controlled Load Set Point
IPS
Energy “Slack”
21
Thermostatically Controlled Load Set Point
IPS
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Supply-Following Loads
22
Cyber / Physical Buildings
BMS
Cyber Physical BuildingLi
ght
Tran
spor
tProcess Loads
Occupant Demand
Legacy Instrumentation & Control Interfaces
Pervasive Sensing
Activity/UsageStreams
BIM BITS
PIB
Activity Models
Multi-Objective Model-Driven
Control
Building IntegratedOperating System
External
HVA
C
Elec
tric
al
Faul
t, A
ttack
, Ano
mal
y D
etec
t &M
anag
emen
t
Control Plan and Schedule
Physical Models
Human-B
uilding
Interfac
e
Layered Arch. for Physical Info
24
Physical Building Systems SensorsComms Links
Logical Meta-Data Networks
ModelPhysical Information
Events Repositories
Analysis
Presentation
SimulationRecommissioning
Diagnosis
Portals User FeedbackOADR
Forecast
Narrow Waist ?
Electrical
Weather
GeographicalWater
EnvironmentalStructuralActuator
Occupancy
sMAP
Modeling
VisualizationContinuous
CommissioningControl
PersonalFeedback
DebuggingStorage LocationAuthentication
Actuation
App
licat
ions
Phys
ical
Info
rmat
ion
sMAP restful web services
/ # list resource under URI root [GET] /data # list sense points under resource data [GET] / [sense_point] # select a sense points [GET] /meter # meters provide this service [GET] /[channel] # a particular channel [GET] /reading # meter reading [GET] /format # calibration and units [GET/POST] /parameter # sampling parameter [GET/POST] /profile # history of readings [GET]/reporting # create and query periodic reports [GET/POST]
POST requests supply JSON objects as arguments: POST: http://meter1.cs.berkeley.edu/reporting/create { "ReportResource" : "/data/325/meter/*/reading", "ReportDeliveryLocation" : "http://webs.cs.berkeley.edu/recv.php", "Period" : 0, "Minimum" : 50, "Maximum" : 100 }
Internet
Cell phonesMAP Gateway
sMAP
Modbus RS-485
sMAP
sMAP Gateway
EBHTTP / IPv6 / 6LowPANWireless Mesh Network
sMAP
sMAP
sMAP
sMAP
Edge Router
Temperature/PAR/TSR
Vibration / Humidity
AC plug meter
Light switch
Den
t ci
rcui
t m
eter
Prox
y Se
rver
EBH
TTP
Tran
slat
ion
California ISO
sMAP Gateway
sMAP Resources Applications
Google PowerMeter
WeathersMAP
Every Building
Database
IP everywhere / Real World Web
Internet
Cell phonesMAP Gateway
sMAP
Modbus RS-485
sMAP
sMAP Gateway
EBHTTP / IPv6 / 6LowPANWireless Mesh Network
sMAP
sMAP
sMAP
sMAP
Edge Router
Temperature/PAR/TSR
Vibration / Humidity
AC plug meter
Light switch
Den
t ci
rcui
t m
eter
Prox
y Se
rver
EBH
TTP
Tran
slat
ion
California ISO
sMAP Gateway
sMAP Resources Applications
Google PowerMeter
WeathersMAP
Every Building
Database
IP everywhere / Real World Web
In an Intelligent Energy Network
28
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
In an Intelligent Energy Network
28
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
In an Intelligent Energy Network
28
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
• Monitor, Model, Mitigate• Deep instrumentation• Waste elimination• Efficient Operation
• Shifting, Scheduling, Adaptation
In an Intelligent Energy Network
28
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
• Monitor, Model, Mitigate• Deep instrumentation• Waste elimination• Efficient Operation
• Shifting, Scheduling, Adaptation
In an Intelligent Energy Network
28
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
• Monitor, Model, Mitigate• Deep instrumentation• Waste elimination• Efficient Operation
• Shifting, Scheduling, Adaptation
In an Intelligent Energy Network
28
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
• Monitor, Model, Mitigate• Deep instrumentation• Waste elimination• Efficient Operation
• Shifting, Scheduling, Adaptation
• Forecasting• Tracking• Market
In an Intelligent Energy Network
28
Load IPS
Source IPS
energy subnet
Intelligent Power Switch
• Monitor, Model, Mitigate• Deep instrumentation• Waste elimination• Efficient Operation
• Shifting, Scheduling, Adaptation
• Forecasting• Tracking• Market
• Availability• Pricing• Planning
Path Forward
29
Des
ign
Use
Path Forward
29
Des
ign
Use
Path Forward
29
Des
ign
Use
Path Forward
29
Des
ign
Use
Path Forward
• Any Power Consumer can …– Measure its Performance and Consumption,– Communicate, and have Intelligence and Control
29
Des
ign
Use
Path Forward
• Any Power Consumer can …– Measure its Performance and Consumption,– Communicate, and have Intelligence and Control
• Eliminate waste
29
Des
ign
Use
Path Forward
• Any Power Consumer can …– Measure its Performance and Consumption,– Communicate, and have Intelligence and Control
• Eliminate waste• Schedule Consumption
29
Des
ign
Use
Path Forward
• Any Power Consumer can …– Measure its Performance and Consumption,– Communicate, and have Intelligence and Control
• Eliminate waste• Schedule Consumption• Adapt Quality of Service
29
Des
ign
Use
Thanks
30
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
Power Proportionality
• Measure of scaling down to partial load31
California Energy Mix
• Renewables Portfolio Standard (RPS)• 36 of 50 states have set goals• Near-term goals range from 10% to 25%
• California RPS• 2008: 10%• 2010: 20%• 2020: 33%
Database of State Incentives for Renewables and Efficiency. http://www.dsireusa.org4/29
California Energy Mix
• Renewables Portfolio Standard (RPS)• 36 of 50 states have set goals• Near-term goals range from 10% to 25%
• California RPS• 2008: 10%• 2010: 20%• 2020: 33%
Database of State Incentives for Renewables and Efficiency. http://www.dsireusa.org4/29
Building Grid
33
Building Grid
33
Building Grid
• large complex loads
33
Building Grid
• large complex loads
33
Building Grid
• large complex loads• Monitor, Model, Mitigate
33
Building Grid
• large complex loads• Monitor, Model, Mitigate
33
Conventional Electric Grid
Generation
Transmission
Distribution
Load
Conventional Internet
Building Grid
• large complex loads• Monitor, Model, Mitigate• In concert with an intelligent
grid
33
Conventional Electric Grid
Generation
Transmission
Distribution
Load
Conventional Internet
Building Grid
• large complex loads• Monitor, Model, Mitigate• In concert with an intelligent
grid
33
Conventional Electric Grid
Generation
Transmission
Distribution
Load
Conventional Internet
Building Information
34
Building Information
34
Building Information
34
Load Tree
Building Information
34
Load TreeClimate Plant
Building Information
34
Load TreeClimate Plant
Building Information
34
Load TreeClimate Plant
Building Information
34
Load TreeClimate Plant
Operations and Environment
Building Information
34
CT: mains power
monitoring
panel level power
monitoring
ACme: plug load energy monitor and controller
Load TreeClimate Plant
Operations and Environment
Building Information
34
CT: mains power
monitoring
panel level power
monitoring
ACme: plug load energy monitor and controller
TemperatureHumidity
Vibration
Pressure
Load TreeClimate Plant
Operations and Environment