Real- Time Building Energy Modeling and Fault Detection and - - PowerPoint PPT Presentation

ME 4343 HVAC Design

Real- Time Building Energy Modeling and Fault Detection and Diagnostics for a DoD Building

Bing Dong1, Zheng O’Neill2

1 University of T exas, San Antonio, TX, USA 2 University of Alabama, AL, USA The work was done at the United Technologies Research Center

Introduction

• Motivation

Source: NBI report 2008 Energy Performance of LEED For New Construction Buildings

Introduction

• HVAC systems consume >20% more energy than design intent

–

Equipment performance degradation, and interact with other systems.

–

Existing control and information systems do not make visible system level energy consumption.

• Need for a scalable building energy management system that

includes whole building energy diagnostics and visualization

–

Better HVAC operational controls and energy diagnostics

–

Raises the visibility of energy performance to help decision making

Building Facts

• Each 150K sf2 Barrack

– Compartments, classrooms and

cafeteria/galley

• Cooling

– T

wo absorption chiller: 450 ton

– Chilled water loop with fjxed-speed

primary pump

• Heating

– Steam from the base wide central

heating plant

– steam to water heat exchanger

• 5 AHUs for each building
• More than 200 VAV boxes with reheat

coil

• A distributed Direct Digital Control

System (DDC)

4

7114 7113

T echnology Approaches

5

Core Layer: BIM-based Database BIM to BEM Real-time Data Acquisition Application Layer: Real-time energy simulation, visualization and diagnostics

• Overview of the Integrated Infrastructure

T echnology Approaches

• Integrated Energy Modeling Approach

T echnology Approaches

7

• BIM to BEM automatic code generation

Traditional Approach

Building 7114 Architectural Model Building 7114 Mechanical Model BEM (Thermal Network Model)

One Week

T echnology Approaches

8

• BIM to BEM automatic code generation

Automatic data extract

IFC

BIM Database Automatic data extract

BEM Input files

Building 7114 Architectural Model BEM (Thermal Network Model) Building 7114 Mechanical Model

Our Approach Traditional Approach

Building 7114 Architectural Model Building 7114 Mechanical Model BEM (Thermal Network Model)

One Week < 5 minutes!! gbXML

T echnology Approaches

9

• Real-time Data Acquisition

Simens EMS

Our DAQ

sleeping area cafeteria classroom Outside view

Naval Station Great Lakes (Bldg 7114)

Extend BCVTB BACnet actors:

1)

BACnet reader utility: Automatically generate a.xml confjguration fjle and a .csv point description fjle based on the fjle created by Simens EMS

2) StoreBACnetDatatoBIMDatabase:

Based on the .csv fjle, automatically create SQL statements based on the raw data received from EMS

3)

DatabaseManager Establish the connection between BCVTB and BIM-based database

Building Control Virtual Test Bed (BCVTB)

10

Results

• Real-time Energy Performance Visualization

Building Hierarchy Interface Time-Series Energy Flows Interface Energy Statistics Pie Chart Interface

Results

11

• Real-time Energy Simulation

07/06 07/07 07/08 07/09 07/10 07/11 500 1000 BLDG7114 Water Side Load (kW Simulated Measured 07/07 07/08 07/09 07/10 07/11

• 0.1
• 0.05

0.05 0.1 Instant Error

Building 7114 AHU3 secondary and primary system diagram Building 7114 Real-Time Simulation Results from 07/06/2011 to 07/11/2011.

Results

OAT OAT AHU energy AHU energy OAD OAD Airfmow Airfmow Damper Damper Valve Valve

AHU network

Reference ROM Building Operation data

Train Inference

Energy Impact

07/21 07/26 07/31 55 60 65 70 75 80 85 90 95 100 Times Temperature (F) / Damper Position (%) MAT OA Damper DAT DATS OAT

Operation data

OA damper 100% DAT setpoint cannot be maintained

07/17 07/24 07/31 0.4 0.6 0.8 1

OA Damper Position

07/17 07/24 07/31 500 1000

Anomaly Score Actual Expected

Building 7114

Building 7114 Energy Diagnostics: Economizer fault identifjed and corrected

Economizer faults: Enthalpy calculation in control sequences is wrong Faults was corrected on Aug 3rd , 2011. Measured chilled water energy consumption shows 18% savings were achieved

Conclusion

• This study has demonstrated an integrated infrastructure which integrates

design information, database and real-time data acquisition in a real building to support energy modeling, visualization and FDD.

13

Observations and Lessons learned:

• Manually mapping BMS points of each HVAC component.
• The designed control logic in the HVAC control system is usually difgerent from

what is actually implemented locally. Communication with fjeld people is necessary to get an accurate baseline model.

• Acknowledgements:

– DoD ESTCP program manager: Dr. Jim Galvin – UTRC: Dong Luo, Madhusudana, Shashanka ,Sunil Ahuja, T

revor Bailey

– Naval Station Great Lakes

• Energy manager: Peter Behrens
• Mechanical Engineer: Kirk Brandys
• Facility team
• Questions?

14

Thank you!