1
Lawrence Berkeley National Laboratory
Eleanor Lee, Staff ScientistBuilding Technologies Department
Lawrence Berkeley National Laboratory
Stephen Selkowitz, Joseph Klems, Robert Clear, DariushArasteh, Mike Rubin, Phil Haves, Michael Wetter, Jacob
Jonssen, Andrew McNeil, Tianzhen Hong, Christian Kohler, Robin Mitchell, Mehry Yazdanian, Kyle Konis, Windows
and Daylighting Group, Lawrence Berkeley National Laboratory
High Performance Building Facade Solutions
2
Lawrence Berkeley National Laboratory
Broad Context
• Great diversity; creative architectural and engineering solutions• U.S. History
– 37 years since the 1973 OAPEC crisis– Lots of forward progress but lots of steps backwards too: the political
environment supporting energy-efficiency has wavered • Today: The Perfect Storm
– Urgency: Climate change is happening (& finally acknowledged)– ARRA funding supports energy-efficiency measures, green jobs,
workforce training– Opportunity to make significant advances toward solving the GHG
problem by increasing building energy-efficiency, getting buildings off the grid, and improving the everday comfort and satisfaction of building inhabitants
3
Lawrence Berkeley National Laboratory
Inaction: Why?
• Doubt– How long will this Perfect Storm last? – Do we think we can make a difference on GHG?– Too hard, complex, and risky: do we really have to go
with such complicated solutions to achieve our goals?
• Reality– How do I convince the client?– How do I justify the added cost, especially in a
constrained economic environment?
4
Lawrence Berkeley National Laboratory
Context for Facades R&D: Prior Work (past 3+ decades)
• Criteria: Broadly applicable (<10% incremental increase in cost over conventional 2-5 year payback based on energy cost savings – ROI must compete with Wall Street)
• Objective: Develop low-cost, energy-efficient façade technologies for broad market deployment in commercial and residential buildings
– Near-term: Simple, turn-key, replacement technologies (50-70% of the market)
- Thermally-improved framing systems with reduced infiltration- Spectrally selective, low-emittance coatings on glass
– Long-term: Intelligent, integrated façade-lighting-HVAC systems (<1% of the market)
- Switchable, durable electrochromic windows- Local control solutions with minimal sensors, self-
commissioning, little to no O&M costs
5
Lawrence Berkeley National Laboratory
Context of Today (next 15-20+ years)
• Criteria: Broad deployment with life-cycle cost/ global cost approach to accelerate market adoption societal impacts + occupant productivity, health, and satisfaction with their environment
• Net Zero Energy Buildings Approach– Dynamic envelope-lighting-HVAC solutions– Global, optimized control with an intelligent, integrated envelope-
lighting-HVAC system working real-time with BMCS/ campus/ city/ region context
– Smart sensor networked grid (OEM $0.10/point $0.11/point?)– Facility + occupant-based control (user autonomy when appropriate)– Active components
- Switchable glazings- Motorized interior and exterior shading- Active daylighting systems- Active controls for natural or mixed mode ventilation and load
shifting/ thermal mass/ radiant heating & cooling strategies
6
Lawrence Berkeley National Laboratory
Potential Solutions (split between KISS + responsive)
• Hi-R facades– Aerogel– Vacuum glazings– Superwindows (R6+)
• Solar control– Ceramic fritted glass– Exterior metal scrims, fixed overhangs, fins– Automated exterior shades
• Daylight redirection/ lighting quality• Automated, intelligent facades
– Real-time HVAC-Lighting-Façade systems optimization – Daylight + view ↔ SHG + visual comfort/ performance– Low-energy cooling strategies
- Night-time ventilation + chilled ceilings- Double-envelope facades
– Demand response• Photovoltaic-integrated facades for on-site renewable energy
7
Lawrence Berkeley National Laboratory
0% savings: Code baseline: Low-SHGC windows and with manually-operated shades and lighting
8
Lawrence Berkeley National Laboratory
20% savings: Automated interior shades + DALI dimmable lighting systems
The New York Times Headquarters
Daylight “efficient” shading
9
Lawrence Berkeley National Laboratory
Annual Source Energy Use
10
Lawrence Berkeley National LaboratorySouth zone, source kBtu/sf-floor-yr
11
Lawrence Berkeley National LaboratorySouth zone, source kBtu/sf-floor-yr
12
Lawrence Berkeley National Laboratory
http://lowenergyfacades.lbl.gov
13
Lawrence Berkeley National Laboratory
Comfen output
OverviewPerformance Relative Cost Benchmarks Design Advisor
Thermal Comfort Daylight Penetration
Energy Façade Gain
14
Lawrence Berkeley National LaboratoryKyle Konis, UCB/ LBNL
How do we reduce lighting energy use with daylight when there is glare?
15
Lawrence Berkeley National Laboratory
Low-e split blind (fixed relationship between upper & lower slats
Upper: horiz/ Lower: +30°Upper: -45 ° / Lower: 0° Upper: +60° / Lower: +closed
16
Lawrence Berkeley National Laboratory
Split reflective blind
17
Lawrence Berkeley National Laboratory
Diffuser above, blind below
18
Lawrence Berkeley National LaboratoryAutomated roller shade
19
Lawrence Berkeley National Laboratory
Automated daylight blind: concave-up slats with mirrored coating in upper zone and light grey
finish in lower zone
20
Lawrence Berkeley National Laboratory
Sunlight-redirecting blinds
21
Lawrence Berkeley National Laboratory
Avg Annual Lighting Energy Use vs Visual Discomfort: Dynamic systems do better
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
auto-split-mir-VB1
ref-VB
split-opt-VB
diff-VB
auto-VB
split-VB
auto-RS
ref-RS
full power Percent ofdaywindow >2000cd/sqm
LPD (W/sf)
0.57 W/sf0.34 W/sf
0.31 W/sf
10% of day = 1.2 hours
22
Lawrence Berkeley National Laboratory
DGI Summary for 6 Test Conditions (Window View)
Auto-split-mir-VB Auto-VB Split-opt-VB Split-VB Diffuse-VB Auto-RS
Paired Reference
Test Condition
DGI of 16:20 = Just Acceptable, 20:24 = Just Uncomfortable, 24:28 = Just Intolerable, 28:36 = Intolerable
23
Lawrence Berkeley National Laboratory
07.23 08.21 10.08 11.09 11.30
Daily Composites…
Solstice to solstice composite, N = 14 days (clear)Test ConditionE
Diagnostics: Glare Source Composites
Glare Source Dx Dy Dz Steradians Luminance (cd/m^2) Indirect Illuminance (lux)1 0.88 0.30 -0.37 0.27 2771 11432 0.92 -0.37 -0.13 0.45 17993 0.18 -0.96 -0.20 0.01 14324 0.82 0.54 -0.21 0.00 35825 0.96 0.18 0.20 0.14 29406 0.92 0.38 0.11 0.02 31807 0.89 0.45 -0.01 0.02 25068 0.85 0.44 0.30 0.02 20709 0.83 0.51 0.21 0.02 276210 0.85 0.53 -0.01 0.02 246311 0.79 0.61 -0.08 0.02 237012 0.77 0.60 0.23 0.02 279113 0.76 0.64 0.08 0.02 270814 0.75 0.47 0.47 0.02 122615 0.84 -0.06 0.55 0.00 1295
RADIANCE Findglare output x 144 images
24
Lawrence Berkeley National Laboratory
Complex Fenestration Systems (CFS)
South NorthSource: St. Gobain/ Eckelt DLS COOLSHADE HR 32/9
25
Lawrence Berkeley National Laboratory
Definition: BRDFs and BTDFsbi-directional reflectance and transmittance functions
Image from Marilyne Andersen, 2004
26
Lawrence Berkeley National LaboratoryWindows Testbed Facility, LBNL
Accuracy of spatial distribution of flux
27
Lawrence Berkeley National Laboratory
Basis Angle Projections
Klems basis projection 2° basis projection
28
Lawrence Berkeley National Laboratory
Phase II: Exterior shading systems (June 2008- June 2009)
29
Lawrence Berkeley National Laboratory
30
Lawrence Berkeley National Laboratory
Exterior Venetian Blinds
Automated exterior blind Static 3-zone blind with bent slats
31
Lawrence Berkeley National Laboratory
3-zone exterior blindinterior and exterior automated roller shades
71%: peak 18 5 W/ft2-windowReference: Interior venetian blind blocking direct sun
32
Lawrence Berkeley National Laboratory
Visual Discomfort
3-zone: Visual discomfort due to high window luminance reduced from 34% of day down to 6% of day
Automated roller shade: 2% of day
33
Lawrence Berkeley National Laboratory
34
Lawrence Berkeley National Laboratory
30-50% savings: Integrated facades and low-energy cooling
Nord LB, Hannover
35
Lawrence Berkeley National Laboratory
Definition of a smart building skin
Operable façade components: Motors or actuators for shading devices, light-redirecting elements, operable windows, or switchable glass coatings
Interior or exterior sensors that measure relevant quantities that are used by the controller
Control algorithms: Accepts input from sensors or computations then determines how to position the operable façade components
36
Lawrence Berkeley National Laboratory
If Glare and Solar Control is so important, Why not integrate this function into the glazing?
Switchable electrochromicglazings(dimmable via 0-3VDC)
37
Lawrence Berkeley National Laboratory
Human subjects study in LBNL windows test facility
(switchable electrochromicwindows)
38
Lawrence Berkeley National Laboratory
39
Lawrence Berkeley National Laboratory
The New York Times Headquarters, New York, NY
• In 2003, The New York Times approached LBNL regarding feasibility of using automated shade and dimmable daylightingcontrol systems in their new headquarters building (1.2 Msf)
• Questions:– Cost versus potential
benefits? – Features and performance of
commercially-available systems?
40
Lawrence Berkeley National Laboratory
NYT Key Concerns: Optimizing Trade-offs
• Floor to ceiling, clear, low-iron, spectrally-selective low-e glass– Tv=0.75, SHGC=0.39, U-factor=0.28 Btu/h-ft2-°F– Horizontal ceramic tubes,18-inches off exterior face of
window
• Control system trade-offs with automated shading: – Direct sun control: thermal comfort with UFAD system +
visual comfort– Absence of glare (disability, discomfort, luminance ratios,
computer display visibility)– Reduced peak solar load (depending on fabric choice and
control)
– Maximize daylight (lighting energy savings, peak demand reduction)
– Maximize interior brightness, combat gloom– Maximize view out (meaningful stimuli, reduction of stress)– Preserve design aesthetic of a “transparent” façade
41
Lawrence Berkeley National Laboratory
Radiance model: floor plan view
42
Lawrence Berkeley National Laboratory
Approach: Test Energy/Comfort Performance in a Full-Scale Mockup
• Furniture, daylighting, employee feedback and constructability: ~450 m2, 4500 sf mockup
• Core Concerns:– Window glare (Tv=0.75)– Daylight harvesting potential
• Northwest – Southwest corner of a typical floor
• Investigate diverse technological solutions by multiple vendors
• Real sun and sky conditions in nearby climate zone, 6-month monitored period: winter to summer solstice
North
AB
43
Lawrence Berkeley National Laboratory
Mock-up InstrumentationIIluminance sensors
Workstation instrumentation at occupant position with LCD screen
– Luminance and illuminancesensors
– Webcams (10-min time lapse throughout the day)
• Quantified lighting energy savings, visual comfort, and control system behavior
• Surveyed occupant comfort and satisfaction
44
Lawrence Berkeley National Laboratory
45
Lawrence Berkeley National Laboratory9:00 AM
WestSouth
46
Lawrence Berkeley National Laboratory1:00 PM
47
Lawrence Berkeley National Laboratory4:00 PM
48
Lawrence Berkeley National Laboratory5:00 PM
49
Lawrence Berkeley National Laboratory
State-of-the-shelf example• The New York Times Headquarters (specified in 2004)
– Automated shading- Motorized, automated roller shades for direct sun, daylight, glare, and view
control with manual override, scheduling, urban context, demand response- Centralized control +16 networked sensors/ floor + 8 roof-mounted
sensors, real-time, scheduling, model-based/ feedback heterogeneous control
– Dimmable lighting- Dimmable, reconfigurable zones, occupancy, scheduling, tuning, demand
response- First large-procurement of “DALI-enabled” ballasts: each fixture is
addressable & reconfigurable- Distributed control + 16 open-loop sensors/floor + occupancy sensors
– +10-20% incremental cost: competitive bidding, component engineering, turn-key commissioning tools
50
Lawrence Berkeley National Laboratory
Direct sun• 3%-open
shade fabric– direct
source glare
– ~black out shades needed
• But– More
opaque fabric will block more daylight
51
Lawrence Berkeley National Laboratory
52
Lawrence Berkeley National Laboratory
Sky Glare can be a Problem as well as Direct Sun
East view South viewMeasured data
(cd/m2)12/15, 10:00AM
53
Lawrence Berkeley National Laboratory
Shades down (partial on south)
East view West view
Measured data (cd/m2)12/15, 10:00AM
54
Lawrence Berkeley National Laboratory
Luminance Measurements at Typical Workstations
1:3
1:3
1:3 remote
1:3
1:3
1:3
1:3
1:3 remote
55
Lawrence Berkeley National Laboratory
Window luminance: west, vernal equinox, config 3 or 4 sunny, Eglo.avg = 35-53 klux
For 10 days over the 6-month (solstice-to-solstice) monitored period, the west window luminance was greater than 2000 cd/m2 for 10-54 minutes of the day.
56
Lawrence Berkeley National Laboratory
Measured Average Daily Lighting Energy Use Savings
As of May 2008: Avg LPD = 0.35 W/sf
0%
20%
40%
60%
80%
100%
0 2 4 6 8 10 12 14distance from window (m)
daily
ligh
ting
ener
gy s
avin
gs (%
)
Area A west Area B west Area B south
B
B
A
57
Lawrence Berkeley National Laboratory
Simplify wiring and cabling to reduce install costs
photosensor
58
Lawrence Berkeley National Laboratory
Roller shade wiring diagram
59
Lawrence Berkeley National Laboratory
5. Include commissioning and verification in the specifications
Shading systems Daylighting controls
Field commissioning tools: Goal – meet specs before occupancy
60
Lawrence Berkeley National Laboratory
3. Function: How does it do it?
NYT site
61
Lawrence Berkeley National Laboratory
LBNL shade commissioning cart at The New York Times Headquarters Building
62
Lawrence Berkeley National Laboratory
Step-by-step How-to Guide
63
Lawrence Berkeley National Laboratory
Auto mode:
Overridemode:
64
Lawrence Berkeley National Laboratory
Pilot Demonstration
• Conference Room Setting in Washington DC (West-facing)
• 41x50-inch EC windows, SHGC=0.40-0.08, Tv=0.52-0.03, U=0.347 Btu/h-°F-ft2
• Electrochromic Window Controls:– On-off window tinting– Seasonal solar control when unoccupied (tinted
during summer, bleached during winter)– Lower windows tinted based on 20,000 lux threshold
vertical illuminance– Upper windows tinted based on 30,000 lux threshold– Manual override
• Lighting controls:– DALI dimmable ballasts– Architectural scenes, occupancy, daylight controls
• Monitoring underway
65
Lawrence Berkeley National Laboratory
New San Francisco Federal Building: A Naturally Ventilated Office Tower• Shallow plan for daylighting and cross-
flow ventilation• Open plan perimeter offices have no
mechanical cooling or ventilation• Exposed 8-inch ceiling slab acts as
thermal flywheel• Automatically-controlled windows provide
fresh air and night flush• Occupant-operated windows provide local
control • Work supported by the California Energy
Commission, the General Services Administration and the Federal Energy Management Program.
• Images courtesy of Morphosis and Arup
66
Lawrence Berkeley National LaboratoryImages courtesy of Morphosis and Arup
67
Lawrence Berkeley National Laboratory
Images courtesy of Morphosis and Arup
Images courtesy of Morphosis and Arup
68
Lawrence Berkeley National Laboratory
Building controls virtual testbed (BCVTB)
69
Lawrence Berkeley National Laboratory
Control System Testing Using Design Simulations
• Test control system using design simulation:– Real-time EnergyPlus Hardware interface Control
hardware from the building– Does the control program produce the expected
performance?
Virtual Building Real Control SystemAlgorithmsHardwareA/D
D/A +
SPARK + EnergyPlus
EnergyPlus & SPARK
70
Lawrence Berkeley National Laboratory
All software modules reusable without code modification
Controls algorithm(MATLAB/Simulink)
Master Implementation (Ptolemy II)Real-time output (Ptolemy II)
Building temperatures and airflow(EnergyPlus)
Natural Ventilation in SF Fed. Bldg.
71
Lawrence Berkeley National Laboratory
Façade as HVAC + Lighting System
72
Lawrence Berkeley National Laboratory
http://lowenergyfacades.lbl.gov
73
Lawrence Berkeley National Laboratory
Conclusions
• High-performance facade solutions require a cross-disciplinary approach to address:
– Direct sun control– Glare control– Lighting quality– View– Daylight admission for lighting energy savings– Solar heat gain/ cooling load control for HVAC– Daylighting control system performance
• Industry is working toward more cost-effective solutions with features that address practical building owner and occupant needs
• Further work is needed to make these solutions routine, cost-effective, and reliable for all applications.