use of weather and occupancy forecasts for optimal ... · opticontrol project presentation version...
TRANSCRIPT
Use of Weather and Occupancy Forecasts for Optimal Building Climate Control
(OptiControl)
Project Presentation
The OptiControl Team http://www.opticontrol.ethz.ch/
Version 15. Jan. 2009
OptiControl Project Presentation Version 15. Jan. 2009
2
OptiControl – Overview Aims:
Development of methods to exploit weather forecasts and occupancy-related information aiming at
improving the energy efficiency and comfort of buildings;
reducing peak electricity demand.
Expected Results:
• Methods
• Software/tools
• Benefit-cost analyses
• Application to demonstrator
OptiControl Project Presentation Version 15. Jan. 2009
3
OptiControl – Approach
• Case studies (Selected “Applications”, typical buildings, representative sites etc.)
• Extensive use of computer-based modeling & simulation (Controller development, potential assessment, tests etc.)
• Emphasis on Model Predictive Control (MPC)
• Stepwise refinement & simplification of methods and models
• Field tests of the new control approaches in demonstrator object(s)
OptiControl Project Presentation Version 15. Jan. 2009
4
OptiControl – Research Partners
• ETH Systems Ecology Group Modeling and simulation; project management
• ETH Institut für Automatik Control theory; controller development & tools
• EMPA Building Technologies Laboratory Building physics, HVAC & energy systems, modeling, field tests
• MeteoSwiss Meteorological data and predictions (Europe–local)
• Siemens Building Technologies BAC research and product development
OptiControl Project Presentation Version 15. Jan. 2009
5
OptiControl “Applications”
1. Integrated Room Automation (IRA)
Integrated automation of light, blind, heating, cooling and ventilation
(including TABS and floor heating subsystem variants)
In preparation: 2. Cooling by night-time ventilation 3. Energy Recovery 4. Generic Energy Flux Control
OptiControl Project Presentation Version 15. Jan. 2009
6
Case Study Sites
Zürich Basel-Binningen Genève-Cointrin Lugano Modena Marseille-Marignane Clermont-Ferrand Mannheim Hohenpeissenberg Wien Hohe Warte
x
x
x x
x
x x x x
x
x
OptiControl Project Presentation Version 15. Jan. 2009
7
Buildings Definitions
Controllers Library
Weather DB
Occupancy Datasets
Weather Pred. DB
Occupancy Predictions
Simulation results
Modeling & Simulation Environment
Controller model
“Real” building model
Controller
Modeling
Building specif.
& params
Location specif.
& data
Controller specif.
& params
Occup. specif.
& data
Predicted weather
data
Predicted occup. data
MoEDSiPA Statistics,
Perf. Indices, Graphics etc.
Exper. Definition Simulation Post Analysis
OptiControl Project Presentation Version 15. Jan. 2009
8
0 20 40 60 80 100 120
Ener
gy /
disc
omfo
rt c
osts
Controller Assessment Information Levels: 1. “perfect world – we know everything” 2. “real world, no weather forecasts” 3. “real world, with weather forecasts”
Improvement of present-day control strategies
Transition from perfect models to real world
realistic
Potential
(theoretical)
Reference (today’s practice)
Improved non- predictive control
Predictive control
Performance Bound
OptiControl Project Presentation Version 15. Jan. 2009
9
8 building zone types:
Sample Results: Definition of Simulation Experiments
Façade orientation Southwest Thermal insulation level Swiss average Passive house Construction type Heavyweight Lightweight Window area fraction 30% 80 % Internal gains level low high
HVAC System (OptiControl System #01): • Blinds • Electric lighting • Heating: radiators • Cooling: slow ceiling
– mechanical chiller – free cooling with wet tower
OptiControl Project Presentation Version 15. Jan. 2009
10
Sample Results: Control Strategies Considered
• Ref2 State-of-the-art rule based control
• Ref3 Improved rule based control (new) • MPC-CE MPC-Certainty Equivalent control *)
• PB Performance Bound
n = Narrow thermal comfort range
w = Wide thermal comfort range
*) Using “COSMO-7” weather forecasts by MeteoSwiss, preliminary results.
OptiControl Project Presentation Version 15. Jan. 2009
11
(Ref2n - Ref3n)/Ref2n
-10%0%
10%20%30%40%50%60%70%
0 100 200 300
Ref2n [kWh/m2/a]
Ref2n - Ref3n
-200
20406080
100120140
0 100 200 300
Ref2n [kWh/m2/a]
[kW
h/m
2/a
]
Results (1) – Improved Non-Predictive Control
OptiControl Project Presentation Version 15. Jan. 2009
12
Results (2) – Potential of Weather Forecasts
Ref3w - PBw
0
2
4
6
8
10
0 50 100 150 200
Ref3w [kWh/m2/a]
[kW
h/m
2/a
]
(Ref3w - PBw)/Ref3w
0%
10%
20%
30%
40%
0 50 100 150 200
Ref3w [kWh/m2/a]
OptiControl Project Presentation Version 15. Jan. 2009
0
10
20
30
40
50
60
Ref2 Ref3 MPC-CE PB
[kWh/m2/a]
0
20
40
60
80
100
Ref2 Ref3 MPC-CE PB
[kWh/m2/a]
13
Results (3) – Comparison of Control Strategies
0
5
10
15
20
25
30
Ref2 Ref3 MPC-CE PB
[kWh/m2/a]
0
5
10
15
20
25
30
Ref2 Ref3 MPC-CE PB
[kWh/m2/a]
West / Wien / Swiss average / heavy / windows 30%
South / Zurich / Swiss average / heavy / windows 30%
South / Marseille / Swiss average / heavy / windows 30%
South / Zurich / Passive House / Light / windows 80%
OptiControl Project Presentation Version 15. Jan. 2009
14
• First results are promising
• Benefit of weather predictions varies strongly from case to case
• Appropriate tools are important
• Our sophisticated studies can be useful for identifying simple, improved control strategies
Conclusions