Defining the Zone of Biological Darkness: A el Area-Based Circadian Daylight Metric for Building Design and Evaluation DIVA DAY 2016 Toronto Kyle Konis, Ph.D, AIA School of Architecture University of Southern California
The Nomadic Laboratory Lab Field Greater precision and control Less control, greater level of reality
O.M.G.
Optimizing the Daylight Ecosystem in Buildings: A workshop on Circadian Lighting, il 21-22, 2015 Mandate light levels for circadian stimulation (0.1CS minimum and 0.3 CS for morning daylight in selected areas) for all new projects and renovations Organizers Kevin Kampschroer, GSA Judith Heerwagen, GSA k Rea, LRC ~20 Invited Participants From Carnegie Mellon Rensselaer, LRC ZGF Architects SERA Architects Property MGMT firms Manufacturers Other government agencies
History of Lighting Research
Health Impacts Lack of sufficient exposure of the retina to bright light can disrupt the circadian system Disruption has been associated with a host of maladies including: sleep disorders, fatigue, performance problems, hormone and metabolic disorders
Emerging Opportunities
Daylighting Design Performance Criteria for Alzheimer Care Facilities Towards Evidence-based Practices for Improved Health
?
History of Lighting Research
Well... Critique Well Building Standard Not strictly a daylighting metric No references/citations to underlying science (why?) Only requires 75% of occupied area to comply 4-hour/day metric discounts need to assess bright light stimulus for AM circadian resetting (6:00 10:00AM) Many latitudes / climates would never comply (if daylight) Problematic when view direction is flexible or unknown
Body Clock Must Be Synchronized With Environment Average period is ~24.2h Larks (23.6h) vs. Owls (25.0h) Entrainment to 24h period requires daily resetting Light is the most powerful time cue for resetting
Timing Adjusting the circadian pacemaker is like pushing a child on a swing the effect depends on the timing of the push Adapted from Rajaratnam & Arendt, 2001
Delay wakeup / sleep onset Timing Exposure at night Adapted from Rajaratnam & Arendt, 2001
Timing Advance wakeup (and sleep onset) Exposure in AM... Adapted from Rajaratnam & Arendt, 2001
NOTE: Circadian resetting period is critical That period is 6:00 to 10:00 AM Perhaps exclude the 6:00 7:00 AM hour depending on program.
NOTE: Cumulative occurrence masks spatial and temporal occurrence of BD zones
Intensity Table: Circadian stimulus effect from various common light sources (lux) Not all lux are created equal Circadian system takes longer to respond than visual system According to Rea et al. model. Lucas has another model Models are becoming more complex as knowledge grows
Field-based Measurement / Verification Procedures
Spectral power (W/m^2) Biological Darkness 0.005 0.004 0% stimulus 27 lux C (l) 0.003 V (l) 0.002 0.001 0.000 350 400 450 500 550 600 650 700 750 Measured SPD C (l) V (l) Wavelength (nm)
Spectral power (W/m^2) Daylight Example 0.035 0.030 0.025 47.7% stimulus 262 lux C (l) 0.020 V (l) 0.015 0.010 0.005 0.000 350 400 450 500 550 600 650 700 750 Measured SPD C (l) V (l) Wavelength (nm)
Evaluating Real Spaces 35 25 15 5
52% N 42% 67% 5 E 53% Note: real daylit spaces are almost always a combination of daylight and electrical lighting
15% 46% N 55% 15 E 29%
35 25 15 5
From Horiz. To Vertical Sensors
Test case, portable school classroom Located in SF climate We do regular old annual hourly CBDM, but Lux * EML.coef = EML Note that the simulations actually have 4 windows...
EML Coefficient Derived from Weighting of Direct and Diff. Sky Illuminance (EPW) (Better than nothing, but M. Inanici has a better method.) Lux * EML.coef = EML San Francisco EPW file
Explain what the daily Circadian Stimulus (CS) criteria is... If >= 250 EML = 1 else = 0 7:00 8:00 AM, ch 20, San Francisco
A grid point location can be scored based on all, or subset of vectors If >= 250 EML = 1 else = 0 8:00 9:00 AM, ch 20, San Francisco
Locations that exceed the EML threshold for all 3 hours are scored as Circadian Effective If >= 250 EML = 1 else = 0 9:00 10:00 AM, ch 20, San Francisco
PROBLEMS HERE!!! Daily CE Area Simple interpretation of daily spatial outcomes... What is this? 40% Annual Mean CE Area
There is a need to assess personalized exposure over time (How much time?)
This is for one grid point location, (using the all-vectors approach)... Daily Compliance Outcomes (gr id point 203) CS 0 Need an indicator for the frequency of CE 47.9% of Year (Stimulus[7 10AM] >= 250 EML) Stimulus is present for only 47% of year, but stimulus is available over ~9 months
Circadian Frequency CS 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7 of 7days 100% 6 of 7 85% 5 of 7 71% 4 of 7 57% 3 of 7 43% 2 of 7 29% 1 of 7 14% 0
Here is what happens if you require regular daily stimulus over the past 6 days to score the analysis day as [1] Daily Compliance Outcomes (gr id point 203) % of Trailing 7 Days When CE Criteria Was Met CE.freq (0 100%) 6-day trailing mean 100% 47.9% of Year (Stimulus[7 10AM] >= 250 EML) 7.7% of Year (CE.freq. >= 100%) Stimulus is present for 47% of year, but stimulus is daily for only 7.7% of year
Now relax the CE.freq requirement to CE achieved for at least 6 of past 7 days Daily Compliance Outcomes (gr id point 203) % of Trailing 7 Days When CE Criteria Was Met 85% 47.9% of Year (Stimulus[7 10AM] >= 250 EML) 24.7% of Year (CE.freq. >= 85%) Annual performance outcome for point 203
Now relax CE.freq to 5 of past 7 days (i.e. >70%) Daily Compliance Outcomes (gr id point 203) % of Trailing 7 Days When CE Criteria Was Met This red circle is the CE.freq acceptability threshold Good 70% 47.9% of Year (Stimulus[7 10AM] >= 250 EML) 40.5% of Year (CE.freq. >= 70%)
. And on and on (e.g. 4 of 7) Daily Compliance Outcomes (gr id point 203) % of Trailing 7 Days When CE Criteria Was Met 57% 47.9% of Year (Stimulus[7 10AM] >= 250 EML) 55.1% of Year (CE.freq. >= 57%)
?
7 Days per week 6 Days per week 5 Days per week 4 Days per week 3 Days per week 2 Days per week 1 Day per week 0
Good Enough? 7 Days per week 6 Days per week 5 Days per week 4 Days per week 3 Days per week 2 Days per week 1 Day per week 0 (Remember this is assessed per grid point location)
So how do we annualize it? Building Geometry + Climate 280 Grid points, 365 days per year
7 CE Area (CE.freq >= 7d/wk) Daily Variation in CE Area for San Francisco Climate (7:00 10:00 AM period) CE Area (CE.freq >= 6d/wk) 6/7 5/7 4/7 CE Area (CE.freq >= 5d/wk) CE Area (CE.freq >= 4d/wk) We can report mean annual CE area 20.5% 29.9% 38.1% 3/7 2/7 1/7 CE Area (CE.freq >= 3d/wk) CE Area (CE.freq >= 2d/wk) CE Area (CE.freq >= 1d/wk) 45.5% 50.4% 53.7% 56.2%
Daily Variation in CE Area for Helsinki Climate (7:00 10:00 AM period) CE Area (CE.freq >= 7d/wk) CE Area (CE.freq >= 6d/wk) CE Area (CE.freq >= 5d/wk) CE Area (CE.freq >= 4d/wk) 18.9% 23.3% 26.6% 29% CE Area (CE.freq >= 3d/wk) CE Area (CE.freq >= 2d/wk) CE Area (CE.freq >= 1d/wk) 31.2% 33.4% 35.7%
So how do we spatialize it? 280 Grid points, 365 days per year
SF Climate
Flattened (Stacked) Outcomes... 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% San Francisco Helsinki
So how do we spatialize AND annualize it?
We have to specific a minimum acceptable CE.freq... CS achieved Good Enough? 7 Days per week 6 Days per week 5 Days per week 4 Days per week 3 Days per week 2 Days per week 1 Day per week 0 (Remember this is assessed per grid point location)
Daily Variation in CE Area for San Francisco Climate (7:00 10:00 AM period) CE Area (CE.freq >= 7d/wk) CE Area (CE.freq >= 6d/wk) CE Area (CE.freq >= 5d/wk) CE Area (CE.freq >= 4d/wk) 20.5% 29.9% 38.1% 45.5% CE Area (CE.freq >= 3d/wk) CE Area (CE.freq >= 2d/wk) CE Area (CE.freq >= 1d/wk) 50.4% 53.7% 56.2% For each grid point location, quantify the percentage of the year where the CE.freq is >= 5 d/wk
SF Climate Circadian Resetting Period (7:00 10:00AM) (Best vector approach) Annual mean CE area is 38%
SF Climate Circadian Resetting Period (7:00 10:00AM) (Fixed vector approach)
Conclusions There are no static zones Dynamic nature of zones can be identified and seasonal patterns can be assessed While the CE zone can be mapped, the boundaries are probably only applicable over short intervals (e.g. month) Annual scores can be used to make relative comparisons between designs for the same site / climate Parameters such as daily CE criteria (EML thresh. and N exposure hours) can be easily adjusted Even with limited scientific information, we can make a judgment that locations with greater CE.freq will be more beneficial for human circadian entrainment See paper in peer-review (Building and Environment) Konis, K. Defining the Zone of Biological Darkness: A el Area-Based Circadian Daylight Metric for Building Design and Evaluation
THANK YOU Kyle Konis, USC kkonis@usc.edu