Virginia Tech Research Center Arlington, Virginia, USA. PPT slides will be available at

Transcription

1 SMART BUILDINGS & INFRASTRUCTURES Invited Talk Professor Saifur Rahman Virginia Tech Advanced Research Institute Virginia, USA IIT Delhi New Delhi, India, 02 June 2016 Virginia Tech Research Center Arlington, Virginia, USA PPT slides will be available at

2 Intelligent Interconnected Grid Intelligent Load Demand or price-driven control of appliances Sensors Detect outages, fluctuations, and disturbances Distribution Network Interconnected micro grids Microgrid Distributed Arch. Local monitoring and control Bulk Power Plant Wind Power Park Smart Inverters and Storage Minimize voltage and power fluctuations Saifur Rahman Control Room Functions Balance electricity supply/demand across the grid Smart Cities Smart cities address urban challenges such as pollution, energy efficiency, security, parking, traffic, transportation, and others by utilizing advanced technologies in data gathering and communications interconnectivity via the Internet. Source: Internet of Things Philippines Inc., December

3 Key Areas for Smart Cities Systems are connected together by ICT to transmit and process data in the smart city concept. 5 From Smart City to Smart Buildings 6 Smart City Energy Smart Grid Smart Buildings Supported by ICT and distributed networks of intelligent sensors, data centers/clouds Smart city: Complex system of interconnected infrastructures and services Energy: Smart electric power grids, smart gas networks, smart water systems Smart grid: Bi-directional flows of energy, remote control/automation of power, integrated distributed energy Smart buildings: Intelligent building automation systems, smart devices, productive users, grid integration 3

4 Smart Buildings A smart building connects the building automation system with building operations, such as HVAC, lighting, water supply, sensor network and fire emergency. Implementing smart building solutions can provide: up to 30% savings of water usage up to 40% savings of energy usage reduce building maintenance costs by percent Source: Smart Building Market To Grow 30% by 2020, k et-to-gr ow -30-by-2020, December Smart Building & IoT Evolution < Building Control HVAC control Lighting control Building Automation Building management Building control Building Performance Energy management Remote monitoring Remote control Smart Building Intelligent buildings Green buildings Grid integration Internet of Thing (IoT) Connected Devices 0.2 billions 6.92 billions 50 billions 4

5 Virginia Tech (VT) Solution For Smart Buildings - BEMOSS BEMOSS is a Building Energy Management Open Source Software (BEMOSS) solution that is engineered to improve sensing and control of equipment in small- and medium-sized commercial buildings. BEMOSS BEMOSS monitoring and control: Three major loads in buildings HVAC Lighting loads Plug loads BEMOSS value: Improves energy efficiency and facilitates demand response implementation in buildings. 9 Key BEMOSS Features BEMOSS #1 #2 #3 #4 #5 #6 Open source & open architecture Plug & play Interoperability Cost effectiveness Scalability & reliability Security 10 5

6 BEMOSS Interoperability Communication Technologies q Ethernet (IEEE 802.3) q Serial Interface (RS-485) q ZigBee (IEEE ) q WiFi (IEEE ) RS-485 Data Exchange Protocols q BACnet (IP and MS/TP) q Modbus (RTU and TCP) q Web (e.g., XML, JSON, RSS/Atom) q ZigBee API WEB q Smart Energy (SE) q OpenADR (Open Automated Demand Response) Smart Energy Profile (SEP) BEMOSS Plug & Play CT50 (WiFi) CT80 (ZigBee SE) ICM (WiFi) Nest (WiFi) HVAC Load Controllers 12 BEMOSS automatically discovers new load controllers deployed in a building CT30 (WiFi) VAV controller (Modbus) RTU (Modbus) Philips Hue (WiFi) Lighting load Light switch (WiFi) controller (BACnet) Lighting Load Controllers Step-dimmed ballast (ZigBee) Sensors/ PowerMeters Power meter (Modbus) Power meter (BACnet/M odbus) Light sensor (BACnet) Occupancy sensor (BACnet) Plug load controller (BACnet) Smart plug (ZigBee) Smart plug (WiFi) Plug Load Controllers 12 6

7 BEMOSS on Various Embedded Devices Beaglebone Pandaboard Odroid C3 CPU: 1GHz ARM Cortex-A8 CPU: 1.2GHz Dual core ARM Cortex-A9 CPU: 1.7GHz Quad core ARM Cortex RAM: 512MB RAM: 1GB RAM: 2GB Price: $55 Price: $220 Price: $69 Raspberry Pi Raspberry Pi 2 CPU: 700 MHz ARM1176JZF-S RAM: 256MB Price: $35 CPU: 900 MHz Quad core ARM Cortex A7 RAM: 1GB Price: $35 13 BEMOSS: Solutions for Small Buildings (e.g., multi-tenant office bldg.) BEMOSS HVAC Controllers Plug Load Controllers 14 ZigBee mesh Lighting Load Controllers ZigBee mesh 14 7

8 BEMOSS Scalability: Solutions for Larger Buildings BEMOSS Core Floor 2 Floor 1 BEMOSS Zone 2 BEMOSS Zone 1 15 BEMOSS Reliability and Redundancy BEMOSS Core Suppose BEMOSS Zone 2 fails Floor 2 Floor 1 BEMOSS Zone 2 BEMOSS Zone

9 6/3/16 BEMOSS Reliability and Redundancy BEMOSS Core BEMOSS core will take care of devices in Zone 2 Floor 2 Floor 1 BEMOSS Zone Prince Street, Alexandria, VA

10 6/3/ South Taylor St., Arlington, VA Turner Street, Blacksburg, VA Retailed building: ~50,000 sqft Peak demand: ~160kW Electricity consumption: 4665MWh/month 20 10

11 Classroom Wireless Monitoring Power meter Environmental sensor (CO2, noise, temperature, humidity) BEMOSS core Thermostat Motion sensor Plug load controller 21 Smart AC Control 22 11

12 Smart Appliance Control Remotely accessible WiFi Plug Load Controller 23 Ambient Light Sensor Based Control Office space Lighting 24 12

13 Remote Access Dashboard 25 Thermostat Monitoring 26 13

14 Power Meter Monitoring 27 Occupancy/Motion Sensor 28 14

15 VT BEMOSS - Building Energy Management Open Source Solutions Live Dashboard Home >> Live Dashboard History and Reporting Setting Real-time kw usage HVAC 6.5 kw Lighting loads Plug loads Total Building Consumption 4.3 kw 2.4 kw Real-time weather Daily energy consumption (kwh) Nov 7, 12:00pm Total: 10.2kW HVAC Lighting loads Plug loads Others Today s kwh performance 9.6% 350kWh Partly Cloudy 67 F Chance of Rain 10% Midnight 6am noon Average daily kwh usage 720kWh 29 BEMOSS Potential Applications Integration of machine learning algorithms to get better understanding of power consumption in buildings Integration of algorithms to manage a large amount of data collected from load controllers/sensors Integration of algorithms to allow management of multiple buildings in a transaction-based energy network

16 Thank You Professor Saifur Rahman Virginia Tech Advanced Research Institute Virginia, USA ( 16