Distributed Load Algorithms
LBNL Demand Response
Automated Server
Internet
OpenADR Client
Weather
data
Siemens Smart Energy Box
Energy Simulation
BMS Adapter
3rd Party Plug-in
Jay Taneja
Nathan Murthy
UC Berkeley
DIADR Mid-Project Demonstration, April 27, 2011
APOGEE BAS
WattStopper
Distributed Load
Control Gateway
Air
handlers/fans
Chillers
1
Distributed DR Algorithms
• Goal: Testing and evaluation of distributed DR
strategies
• Dense deployment of metering devices on
appliances and office equipment, with actuation
by the energy gateway
– Thermostatically-controlled loads (e.g. refrigerators,
space heaters, etc.)
– Battery-powered loads (e.g. laptop computers,
desktop computers with UPS units, etc.)
– Lighting (e.g. overhead lights, lamps, etc.)
– Other office equipment (e.g. printers, routers, etc.)
2
Smart Office (464 SDH)
3
Sensor Data Management: sMAP
• Interface for gathering and storing heterogeneous,
unsynchronized physical data
• Includes data from zone lights and two types of plug meters
http://green.millennium.berkeley.edu:8080/media/graph/demo.html
4
Skipping Refrigerator Cycles
5
Devices with Onboard Batteries
• Case Study: Laptops
• Collected traces to build empirical model of
charge and discharge behavior
• Power delivered is a function of battery capacity
• Developing metrics to design laptop charge
schedule during DR period
• Mix of known state (power consumption, maybe
battery capacity) and unknown state (mobility,
computation load)
6
power
battery
capacity
7
Curtailment of Battery Charging in a DR Event
• Assume N laptops with uniform distributed
capacity states
• Assume laptops leave and enter zone both at a
Poisson rate with λ=1
• Define duration of DR Event
• Throughout DR event, set curtailment ratio c (%
of baseline load) and select laptops to charge
• Choose c to minimize projected peak power for
remainder of DR event
8
9
Charging Curtailment Simulation Results
• 30% curtailment possible
• Choice of curtailment ratio is crucial to how
load management throughout DR event
• Aggressive initial curtailment may offset peak
load reduction towards end of DR event
• Aggregate distributed load in a zone can be
shaped using device energy storage
10
Desktop Power Management
• Desktop + UPS is similar
to laptop
• Collaboration with Dhaani Systems
– Using network appliance to manage state (and
power) of Windows machines
– Machines put to sleep remotely when not in use
– During DR event, aggressiveness can be increased
11
12
12
Lighting
• Lighting zones on SDH 4
– Actuate using Wattstopper via BACnet
– High (50W) and low-power (25W) ballasts in each zone
13
Other Loads
• Printers
– High peak-to-idle ratio (> 75:1)
– Idea: DR-aware print queue
• Avoid concurrent printing (and
resulting high peak load)
• Modify existing print server
14
Next Steps
• Application of techniques to similar loads
• Integrated management of heterogeneous
loads
• Occupant light control
15
Questions?
Acknowledgement
This material is based upon work supported by the Department of Energy under Award Number DE-EE0003847
Disclaimer
This report was prepared as an account of work sponsored by an agency of the United States
Government. Neither the United States Government nor any agency thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its
use would not infringe privately owned rights. Reference herein to any specific commercial product, process,
or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United States Government or any agency thereof. The
views and opinions of authors expressed herein do not necessarily state or reflect those of the United States
Government or any agency thereof.
16