Conference PaperPDF Available

Campus Energy Model: Using a Semi-Automated Workflow to Build Spatially Resolved Campus Building Energy Models for Climate Change and Net-Zero Scenario Evaluation

Authors:

Abstract and Figures

To meet carbon reduction goals, municipalities, universities, and other large organizations need reliable and adaptable models that provide detailed building performance metrics to efficiently manage future energy demand. At the urban scale, models are often restricted by access to usage and geometry data and can only accurately predict aggregate energy demand based on historic data or statistical models. This paper seeks to present a novel workflow that uses institutional GIS datasets to produce calibrated multi-zone energy models for future scenario assessment and to inform building retrofitting options. The authors hope that the introduced workflow leads to a wider adoption of the involved tools to support the environmental strategies of others.
Content may be subject to copyright.
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1720
CAMPUS ENERGY MODEL: USING A SEMI-AUTOMATED WORKFLOW TO BUILD
SPATIALLY RESOLVED CAMPUS BUILDING ENERGY MODELS FOR CLIMATE
CHANGE AND NET-ZERO SCENARIO EVALUATION
Thomas Suesser, Timur Dogan
tds78@cornell.edu, tkdogan@cornell.edu
Environmental Systems Lab, Cornell, Ithaca, New York, USA
Abstract
To meet carbon reduction goals, municipalities,
universities, and other large organizations need reliable
and adaptable models that provide detailed building
performance metrics to efficiently manage future energy
demand. At the urban scale, models are often restricted by
access to usage and geometry data and can only accurately
predict aggregate energy demand based on historic data or
statistical models. This paper seeks to present a novel
workflow that uses institutional GIS datasets to produce
calibrated multi-zone energy models for future scenario
assessment and to inform building retrofitting options.
The authors hope that the introduced workflow leads to a
wider adoption of the involved tools to support the
environmental strategies of others.
Introduction
In recent years, climate change and rapid urbanization
have each become inevitabilities. By 2050, two-thirds of
the world’s population will live in urban areas (United
Nations, Department of Economic and Social Affairs, and
Population Division 2012), the same year by which every
nation on earth pledged to obtain zero carbon emissions
in the 2016 Paris Climate Agreement (Chan and Eddy
2015). At the same time, traditional association of cities
as dirty pollution centers have also given way to the
theory that high density urban centers hold the key to
feasible sustainable development. Thus, the need for
sustainable development in urban areas continues to
increase. Since it is usually undesirable to simply
demolish and rebuild in dense urban areas, especially in
developed countries, building retrofitting on massive
scales must occur to reach carbon reduction targets set by
nations, states, and municipalities. Thus, the ability to
analyze the energy use of existing buildings in climate
change and retrofitting scenarios has become increasingly
important.
In order to achieve reliable predictions of future energy
use, it is necessary to first develop detailed models
representative of the existing building geometry which
can be simplified and integrated into energy simulations.
Advances in the field of building simulation have made it
easier to integrate building geometry (Dogan and Reinhart
2017), material definitions, load profiles, and other
properties (Cerezo, Dogan, and Reinhart 2014). However,
processes that yield detailed, accurate, and usable
geometry are often still “error prone and tedious”
(Schlueter and Thesseling 2009). A streamlined, semi-
automated process of integration of building geometry is
yet to be developed. This process must also allow for the
revision of physical building characteristics in order to
simulate and inform retrofitting scenarios. For the same
reason, this model must also validated against historical
energy usage data. One must be able to understand the
effect of multiple parameters on the data, not simply
“fudging” the model parameters in order to achieve
calibration with existing data, but follow a documented
mathematical process. Problems with such a technique
have been documented, such as their inaccuracy and time
and computational intensiveness (Coakley, Raftery, and
Keane 2014). In energy modeling during the building
design phase, more emphasis can be placed on the final
results of the simulation. However, in modeling existing
buildings it is necessary to place greater emphasis on
behavior in time and thus to measure calibration success
based on both aggregate demand and hourly demand.
An accurate and comprehensive modeling tool is needed
in order to obtain this degree of resolution. Energy-Plus
simulation has been touted for its ability to simulate multi-
zone airflow and integrate extensive HVAC systems,
making it one of the most popular simulation programs in
the field of building simulation (Coakley, Raftery, and
Keane 2014; Nguyen, Reiter, and Rigo 2014).
Current building simulation efforts are often beholden to
their access to both geometry and energy data, placing
limitations on calibration accuracy. The author’s home
institution provides an ideal case study for a variety of
factors including the good availability of GIS, building,
and measured high-resolution energy consumption data in
hourly resolution. Even with reliable usage data, success
is usually measured by error in aggregated yearly or
monthly measured vs. simulated energy use (Reinhart and
Cerezo Davila 2016).
This paper seeks to calibrate hourly demand curves while
maintaining high geometric resolution in the models so
that retrofitting recommendations can be given.
Producing geometrically detailed energy models usually
is time and cost-prohibitive as it involves manually
splitting buildings into multiple zones and providing zone
descriptions for the resulting enclosed volumes. This
paper presents an automated workflow that utilizes
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1721
institutional GIS data-sets to generate multi-zone building
energy models for scenario evaluation and planning aid.
Therefore, a variety of different workflows ranging from
GIS parsing, geometric computing and modeling, data
matching, energy model setup, simulation and
optimization tools need to be brought together. Coupling
these complex systems into one easily manipulated and
reliable model remains challenging and hence this paper
seeks to present a feasible and efficient workflow to create
calibrated models of existing buildings on a university
campus scale. The paper implements several automata to
generate relevant input data and seamlessly connects a set
of existing tools. The workflow aims to produce models
that are accurate and detailed enough to be used to inform
design and planning decisions including future retrofitting
projects and climate change scenarios in pursuance of the
institutional carbon neutrality targets. or this study three
buildings have been analyzed. Other campus buildings
will follow in subsequent studies.
Simulation
General Approach
The process of specifying building properties to achieve
calibration was carried out in three main parts. First, the
building geometries were generated from GIS shape files
and split into thermal zones for each floor. Second,
building construction, usage schedules, load profiles, and
ventilation characteristics were specified for each zone.
Lastly, an automated simulation workflow using
EnergyPlus (Crawley et al. 2000) and the Archsim plug-
in (Dogan 2016) for Rhino (Robert McNeel & Associates
2016b) and Grasshopper (Robert McNeel & Associates
2016a) assigned these parameters to the zones as specified
using the Goat optimization solver component with local,
linear approximations (COBYLA) (Flory 2016) and ran
an simulations based on historical weather data measured
with an weather station located on the rooftop of one of
the analyzed buildings.
The geometries were specified from institutional GIS data
and constructed using an automated process. GIS data was
loaded into Rhinoceros and the building footprint
polygons were parsed from the GIS data. A custom script
using a modified Douglas-Peuker algorithm was used to
automatically remove redundant points and to simplify
overly fine discretization of curvature to keep the overall
geometric complexity manageable for BEM tools. Special
attention was given to corner points if they are adjacent to
a neighboring building. Such points were locked in place
and were not subject to simplification. Since adjacency
detection in later steps relies on congruent edges,
overlapping lines were split until congruent. Further, the
footprint area is a relevant property. In this case,
geometric footprint area of the shapefile did not always
match more accurate data entries from institutional
sources that specified either footprint or overall floor area.
Iterative offsetting was used to manipulate the original
shape to match geometric and data derived foot print area.
The processed footprints were then automatically broken
into thermal zones according to the recommendations
given in ASHRAE 90.1 Appendix G (ASHRAE 2013)
using straight skeleton based automatic zoning. (Dogan,
Reinhart, and Michalatos 2016). The zones were then
extruded to form thermal blocks for each floor of the
building. Figure 1 shows the buildings and thermal zone
geometry. The thermal zones were then paired with data
templates that describe materiality, loads and
conditioning settings of each zone. Templates are specific
to each building and include detailed zone descriptions for
core and perimeter regions located in the basement,
ground, intermediate and roof floors. For the template
assignment process, zone adjacencies and overall location
in the building such as in the basement, ground level/first
floor, in-between floor or roof zones are detected
automatically using a custom script that traverses the zone
and face graph.
An Excel spreadsheet was used to gather and compile the
input parameter templates for each zone and then
dynamically linked to the energy model generation
workflow in Rhino/Grasshopper. The spreadsheet format
for model input data aggregation facilitated sharing model
assumptions with others such as facility management as
well as updating batches of input data for multiple
buildings. Further, entries could be updated with a tablet
or smartphone during field assessment.
Window to wall ratios were measured from photographs
taken of building façades and constructions as shown in
Figure 2. The glazing system specification as well as the
Figure 1: Building geometries colored by floor type and thermal zone divisions
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1722
constructions for roof, façade and basement were
specified from field assessment. When construction
assemblies where unknown, visual inspection and heat
flux measurements using the GSKIN U-Value Kit were
performed. Figure 3 shows a long-term study of a wall in
Building A. U-Values at night time where averaged and
used as basis to estimate materials and thicknesses of a
construction. Figures 4-7 show the most important
constructions used in this study. Table1 provides the U-
Values. In addition to the contextual shading from GIS
based building volumes, other significant obstructions
such as large trees or overhangs were added to the model
based on field assessment observations.
Figure 2: Building A façade shown with superimposed
polygons for window to wall ratio calculation
Figure 3: Heat flux measurement to determine wall U-
Values for Building A
Electrical equipment usage schedules were extracted from
metered electricity demand at building level. Occupancy
schedules were based on architectural norm assumptions
(Merkblatt 2006). However, when the electricity demand
indicated regularly occurring peaks in the morning and
lows in the evening - the start and end time of the
occupancy schedules were adjusted accordingly.
Once all inputs were gathered, a Grasshopper based
workflow using Archsim and Energy Plus was used to
batch simulate all models. The models, along with a
custom weather file constructed from historical weather
data measured on a neighboring building, were inputted
into two separate EnergyPlus simulations for heating and
cooling load calibration. One simulation ran for a week in
September and the other for a two-week period in
December.
The rather short calibration periods were selected due to
limited data availability of both the metered energy
demand and weather data. However, the periods are
somewhat representative since they cover the warmer and
more humid season in the summer as well as the winter
period with both an occupied week as well as the winter
holiday season.
Figure 4: Building A and B roof construction.
Figure 5: Building A and B façade construction
Figure 6: Building C roof construction
Figure 7: Building C facade construction
Table 1: Building properties.
Building Name
Bldg A
Bldg B
Bldg C
Floor Area (m2)
6605
4298
2514
Windows
Single
Single
Single
Facade (W/m²K)
1.05
1.05
1.85
Roof (W/m²K)
0.34
0.34
3.2
Power density
(W/m²)
9.33
10.82
46.30
People (p/m2)
0.199
0.072
0.200
Infiltration (ACH)
0.809
1.20
0.738
Internal Mass ratio
(m2/zone area)
0.150
0.028
0.191
-20
-10
0
10
20
30
40
24.9 25.9 26.9 27.9 28.9
Heat Flux [W/m²]
T2 [°C]
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1723
Calibration
The goal of the calibration step is to validate the produced
models so that they accurately predict existing conditions.
The locally measured historic weather data and metered
demand curves were used for this. Calibration success
was measured by comparing simulated cooling and
heating energy with measured usage for the same periods
in September and December 2015, respectively. A hybrid
function of the root mean squared error of the hourly
heating and cooling series and the percent error of the
aggregate heating and cooling demand, as shown in
Equations 1 and 2, was used as fitness function. The
overall percentage difference for the total heating and
cooling load was weighted slightly higher as the accuracy
in hourly resolution.
!"# $%& ' ()*+!,-)*+!./ - $%0 ' 1233
(1)
1233 # 4 ,56789: ;,567:<8
,56789:
-.==>89:;.==>:<8
.==>89:
(2)
The most challenging part of the calibration step is to
identify parameters that are certain and reliable and those
that are uncertain. The authors could specify some
parameters with exact certainty or near certainty, such as
building constructions, internal mass exposed surface area
ratio, window to wall ratios, window glazing materials,
contextual shading and shading systems as well as
electrical equipment usage. Other parameters were less
obvious, hard to measure on site and therefore uncertain.
These could be specified within a reasonable range, but
no further. Sensitivity analysis was then used to narrow
down the focus of the calibration. The authors could
identify two parameters – infiltration rate and people
density - that had significant impact on the simulation
results. Based on the authors’ experience, reasonable
ranges have been assumed to be 0.1 to 3 (ACH), 0.02 to
0.2 (p/m2) for infiltration rate, people density. Less
influential parameters were set to the median value within
the range that was deemed realistic. Linear approximation
(COBYLA) was used to optimize the most influential
parameters over a reasonable range. This minimized error
as specified in Formula 1 and 2. Figures 8 - 11 show
calibrated usage curves in hourly resolution for both
heating and cooling demand for buildings A - C. Table 2
juxtaposes the simulated and metered total heating and
cooling demand for the calibration periods based on
which the percentage errors ranging from 7% to -15.4%
in Table 3 are computed. Metered heating data for
Building C and cooling data for Building B were not
available.
The calibration did yield similar infiltration rates (given
in Table 1). This is plausible given that all three buildings
have similar size, materiality and a relatively poor
construction standard. The estimated people density is
however very significantly higher in Building A and C
compared to the estimate in Building B. A and C are both
used by the Architecture School and densely populated
and usually used around the clock.
Table 2: Total heating and cooling demand metered and
simulated for the summer (09/23/15-09/30/15) and
winder (12/17/15-12/31/15) calibration periods
Bldg A
Bldg B
Bldg C
Metered Heating
Total (kWh/m2)
17.795749
14.163468
N/A
Simulated Heating
Total (kWh/m2)
16.50065
14.155996
N/A
Metered Cooling
Total (kWh/m2)
2.415758
N/A
2.71354
Simulated Cooling
Total (kWh/m2)
2.122319
N/A
2.296996
Table 3: Percentage error in predicted heating and
cooling demand compared to metered data for the
summer (09/23/15-09/30/15) and winder (12/17/15-
12/31/15) calibration periods
Bldg A
Bldg B
Bldg C
Total heating
percentage error
7%
0.053%
N/A
Total cooling
percentage error
-7.3%
N/A
-15.4%
Scenario Assessment
With three buildings calibrated to a satisfactory accuracy
level in temporal thermal behavior and absolute demand,
the models were then used to quantify energy impacts of
climate change predictions and various retrofitting
scenario. Simulations for one entire year using a Typical
Meteorological Year (TMY) weather file from the nearest
airport were performed.
Climate Change
As the world’s climate continues to change and large
entities like universities aim for sustainability and carbon
neutrality in the long term. To achieve such a goal, it is
important to predict how buildings will respond to
changing weather patterns in the future. Using the
CCWorldWeather Gen tool by Jentsch, Bahaj, and James
(2008), climate files for the years 2020, 2050, and 2080
were generated by morphing a TMY from the nearest
airport. Each building is then analyzed in current and
future conditions.
Figures 10 12 show how the three buildings respond to
changing climate. As seen in Figure 10, with no change in
any parameters besides climate, Building A experienced
very little overall change in energy demand. A 4%
reduction in total demand by 2080 can be observed.
However, as the climate warms the model predicts that a
greater share of the energy demand will fall on cooling
loads. In 2015 cooling accounted for 8.8% of the energy
demand (the sum of heating, cooling, and electrical
equipment) for Building A. By 2080, this number has
grown to 23.8%.
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1724
Figure 8: Building A measured vs. simulated hourly heating energy
Figure 9: Building A measured vs. simulated hourly cooling energy
Figure 10: Building B measured vs. simulated hourly heating energy
Figure 11: Building C measured vs. simulated hourly cooling energy
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1725
Figure 10: Building A predicted energy demand based
on climate change.
Figure 11: Building B climate change response
Figure 12: Building C climate change response
This represents an increase by 386%. This shift towards
cooling energy will place a signifiantly higher burden on
the cooling infrastructure of the authors’ home institution.
This trend continues for Building B and Building C, as
seen in Figures 11 and 12, with cooling’s share of total
energy demand jumping from for 6.6% to 21.3% and
25.8% to 41.4% from 2015 to 2080, respectively.
However, neither Building B’s nor Building C’s total
energy demand remains constant. In the case Building B,
total energy demand falls by 11.5% by 2080, while
Building C’s total energy demand grows by 6.2%.
While heating demand shrinks significantly in all
buildings, it remains the dominant load in both Buildings
A and B. This adds to the justification of investment in
sustainable heating sources even in the face of gradually
rising temperatures.
The degree of variance in results between the buildings
reveals the need for separate building simulations, as each
building’s unique characteristics make it respond
differently to climate change.
Electrical Equipment Demand Reduction
One parameter likely to change with advances in
technology is electrical equipment energy demand. In this
model, lighting is grouped into electrical equipment, as
the institution’s facilities management does not meter
electricity separated by end-use. Changing building light
bulbs from incandescent to compact fluorescent lights
(CFL) or LED bulbs can save a tremendous amount of
energy, and is a practice gradually being adopted
worldwide. In large academic buildings, however,
upgrading lightbulbs is only one aspect of potential
electrical equipment energy demand reduction. Switching
office computers from desktops with separate monitors to
more efficient laptops is one such measure. A typical
desktop computer runs at between 80W and 250W,
whereas laptops can be charged around 45W.
Equipment energy demand also adds significant heat to
buildings. In the following study, the authors assume that
electrical equipment energy demand will decrease by 10%
by 2020, 25% by 2050, and 50% by 2080. These scenarios
are analyzed for each building in conjunction with climate
change. The results are given in Figures 13, 14, and 15.
As expected, reducing electrical equipment energy has a
significant impact on the total building energy demand
and therefore is beneficial for any institution’s bottom line
and carbon neutrality efforts. Further, the reduction in
electricity consumption lessens the increase in summer
cooling loads due to climate change. The reduction in
internal gains, however, does not lead to an increase in
heating demand due to the rising exterior temperatures. In
Building C the reduction stabilizes the heating and
cooling loads. While these results are somewhat expected,
quantifying the effect is still valuable. For example, this
information could be used for targeted plug load reduction
in buildings with low capacity heating or cooling systems
to ensure a longer building lifetime if system upgrades are
difficult or too expensive.
271 271 271 271
271 235 209 162
53 73 97 135
-2.6% -3.2% -4.5%
0
100
200
300
400
500
600
2015
2020
2050
2080
Projected Energy USe (kWh/m2/a)
Equipment & Lighting Heating Cooling
83 83 83 83
243 222 199
160
23 32 45
66
-3.4% -6.6%
-11.5%
0
50
100
150
200
250
300
350
2015
2020
2050
2080
Projected Energy USe (kWh/m2/a)
Equipment & Lighting Heating Cooling
118 118 118 118
56 47 40 28
61 72 84 103
0.0% 0.8% 3.0% 6.2%
0
50
100
150
200
250
2015
2020
2050
2080
Projected Energy USe (kWh/m2/a)
Equipment & Lighting Heating Cooling
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1726
Figure 13: Building A site-electrical reduction response
Figure 14: Building B site-electrical reduction response
Figure 15: Building C site-electrical reduction response
Window Retrofitting
Another investment that can save significant amounts of
energy with limited invasion of the existing building is
window retrofitting. All three of the buildings discussed
in this paper were built over one hundred years ago and
contain clear, single pane casement windows. These trap
very little heat, and cause unwanted infiltration at their
numerous seems and joints, making them highly
inefficient during both warm and cool months.
For this study three retrofitting options are analyzed. The
first, referred hereafter as DoublePaneLoE2 is a double
paned window with a low emissivity coating inner surface
of the outer pane. The second, referred to hereafter as
DoublePaneLoE3, is a double paned window with the low
emissivity coating on the interior surface. The third,
referred to hereafter as TriplePaneLoE, is a triple pane
window with the low emissivity coating on layers e2 and
e5, which are the interior of the outermost pane and the
outermost surface of the innermost pane. Table 2 details
the thermal properties of each of these windows, as well
as the existing single clear-paned windows, where Tvis is
the Visible Light Transmission factor, Uval is the U-
Value, and Shgf is the Solar Heat Gain Coefficient.
Window Type
Tvis
Uval
Shgf
SinglePaneClr
0.913
5.894
0.905
DoublePaneLoE2
0.444
1.493
0.373
DoublePaneLoE3
0.769
1.507
0.649
TriplePaneLoE
0.661
0.785
0.764
Table 2: Window Properties
Heating and cooling energy demand for each building was
calculated for each window option in current and future
climate scenarios. The metered electrical equipment
energy was left unchanged but omitted for reporting as the
retrofits have no effect on this value (no daylight
simulations for dimming of electric lighting was
conducted). The results are given as bar charts showing
overall energy impact from 2015 to 2080 (Figure 16-18)
and as linearly interpolated time series (Figures 19-21).
While the results of the climate change and electrical
equipment energy demand reduction scenarios differ for
each building, all three buildings have similar responses
to the window retrofitting scenarios. The introduction of
a double pane glass leads to a significant drop in cooling
and heating energy consumption in all three buildings.
Building A and B yield a 33% and 55% reduction
respectively, whereas Building C offers only a 17.2%
savings potential. The triple pane option does not provide
significant benefits over the double pane glass despite its
better U-Value. This is a result of the remaining poor
façade and roof constructions. Thus, investment should be
directed most immediately to Building B for the greatest
return on investment. Still all three buildings would
achieve significant energy demand reductions if all
windows were replaced by the double pane windows with
low SHGCs.
271 244 203
135
271
249
241
223
53
64
73
83
-6.4%
-13.2%
-25.8%
0
100
200
300
400
500
600
2015
2020
2050
2080
Projected Energy USe (kWh/m2/a)
Equipment & Lighting Heating Cooling
83 75 63 42
243 227
210
181
23
30
38
50
-5.2%
-11.4%
-21.9%
0
50
100
150
200
250
300
350
2015
2020
2050
2080
Projected Energy USe (kWh/m2/a)
Equipment & Lighting Heating Cooling
118 106 88
59
56
43
43
41
61
66
66
66
0.0%
-8.4%
-15.9%
-29.3%
0
50
100
150
200
250
2015
2020
2050
2080
Projected Energy USe (kWh/m2/a)
Equipment & Lighting Heating Cooling
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1727
Figure 16: Energy impact of window retrofitting options
for Building A from 2015-2080.
Figure 17: Current conditions and retrofitting options
for Building B in future climate scenarios.
Figure 18: Current conditions and retrofitting options
for Building C in future climate scenarios.
Figure 19: Performance of window retrofitting options
for Building A over time.
Figure 20: Performance of window Retrofitting Options
for Building B in future climate scenarios
Figure 21: Performance of window Retrofitting Options
for Building C in future climate scenarios
Figures 19 - 21, highlight the change in performance of
each window type with climate change.
This result demonstrates the importance of analyzing
retrofitting options with climate change factored into a
building energy model. However, at current estimates of
the quickness and severity of climate change, weather
morphing should be used to run future scenarios, as the
performance of certain designs may vary with these
13.5
6.7 6.4 6.0
6.3
6.6 7.0 7.3
-33.1% -32.3% -32.6%
0
2
4
6
8
10
12
14
16
18
20
Sp
DpLoe2
DpLoe3
TpLoe
Energy Use from 2015-2080 (MWh/m2)
Heating Cooling
12.9
4.7 4.6 4.3
3.0
2.3 2.6 2.8
-55.6% -54.8% -55.1%
0
2
4
6
8
10
12
14
16
Sp
DpLoe2
DpLoe3
TpLoe
Energy Use from 2015-2080 (MWh/m2)
Heating Cooling
2.6
1.3 1.2 1.1
5.5
5.4 5.6 5.8
-17.2% -15.2% -13.9%
0
1
2
3
4
5
6
7
8
Sp
DpLoe2
DpLoe3
TpLoe
Energy Use from 2015-2080 (MWh/m2)
Heating Cooling
198
200
202
204
206
208
210
212
214
2015 2025 2035 2045 2055 2065 2075
Projected Energy Use (kWh/m2/a)
DP LoE2
DP LoE3
TP LoE
102
104
106
108
110
112
114
116
118
2015 2025 2035 2045 2055 2065 2075
Projected Energy Use (kWh/m2/a)
DP LoE2
DP LoE3
TP LoE
90
95
100
105
110
115
120
2015 2025 2035 2045 2055 2065 2075
Projected Energy Use (kWh/m2/a)
DP LoE2
DP LoE3
TP LoE
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1728
climate shifts, prompting forward thinking designers and
engineers to make different design decisions.
Conclusion
This paper has demonstrated a feasible and efficient
workflow to create calibrated models of existing buildings
that are accurate and detailed enough to be used to inform
decisions for building retrofitting in various climate
scenarios in pursuance of carbon neutrality targets. An
automated process to generate building energy models
from institutional datasets has been outlined.
The usefulness of such the presented workflow was
demonstrated in several scenario evaluations including
climate change, electrical equipment energy demand
reduction and window retrofitting. The information
obtained can be used to select between options, to inform
policy and for targeted investment. This process of
building simulation and its results have underscored the
importance of conducting analyses of separate buildings
even at campus or neighborhood scale. While most of the
results support the intuition of an experienced energy
modeler, there is great value in the ability to quantify
energy impacts at high spatial and temporal resolution.
Unless sufficient funds are available for large scale
overhauls, efforts in pursuit of carbon neutrality, are most
likely to be carried out in small steps. To maximize the
impact of these gradual investments, fast methods for the
generation and calibration of individual building energy
models are needed. Hence, it is the authors’ intention to
conduct future studies of this type on more campus
buildings and to see the results of these studies used to
inform real change in pursuit of the institution’s carbon
neutrality goal.
Acknowledgements
The authors would like to thank the Cornell University
David R. Atkinson Center for a Sustainable Future for
funding this research
References
ASHRAE. 2013. “90.1 Energy Standard for Buildings
Except Low-Rise Residential Buildings.” In
ASHRAE.
Cerezo, Carlos, Timur Dogan, and Christoph Reinhart.
2014. “Towards Standardized Building Properties
Template Files for Early Design Energy Model
Generation.” In Proceedings of the ASHRAE/IBPSA-
USA Building Simulation Conference, 2532.
Chan, Coral Davenport, Justin Gillis, Sewell, and Melissa
Eddy. 2015. “Inside the Paris Climate Deal.” The
New York Times, December 12.
http://www.nytimes.com/interactive/2015/12/12/wor
ld/paris-climate-change-deal-explainer.html.
Coakley, Daniel, Paul Raftery, and Marcus Keane. 2014.
“A Review of Methods to Match Building Energy
Simulation Models to Measured Data.” Renewable
and Sustainable Energy Reviews 37 (September):
12341. doi:10.1016/j.rser.2014.05.007.
Crawley, Drury B, Curtis O Pedersen, Linda K Lawrie,
and Frederick C Winkelmann. 2000. “EnergyPlus:
Energy Simulation Program.” ASHRAE Journal 42
(4): 49.
Dogan, Timur. 2016. Archsim. Accessed May 15.
www.archsim.com.
Dogan, Timur, and Christoph Reinhart. 2017.
“Shoeboxer: An Algorithm for Abstracted Rapid
Multi-Zone Urban Building Energy Model
Generation and Simulation.” Energy and Buildings
140: 140153.
Dogan, Timur, Christoph Reinhart, and Panagiotis
Michalatos. 2016. “Autozoner: An Algorithm for
Automatic Thermal Zoning of Buildings with
Unknown Interior Space Definitions.Journal of
Building Performance Simulation 9 (2): 176189.
Flory, Simon. 2016. Goat Optimization Add-on Solver.
Accessed November 22.
http://www.grasshopper3d.com/group/goat.
Jentsch, Mark F., AbuBakr S. Bahaj, and Patrick A. B.
James. 2008. “Climate Change Future Proofing of
buildingsGeneration and Assessment of Building
Simulation Weather Files.” Energy and Buildings 40
(12): 214868. doi:10.1016/j.enbuild.2008.06.005.
Merkblatt, SIA. 2006. “2024: Standard-
Nutzungsbedingungen Für Die Energie-Und
Gebäudetechnik.” Zürich: Swiss Society of Engineers
and Architects.
Nguyen, Anh-Tuan, Sigrid Reiter, and Philippe Rigo.
2014. “A Review on Simulation-Based Optimization
Methods Applied to Building Performance
Analysis.” Applied Energy 113 (January): 104358.
doi:10.1016/j.apenergy.2013.08.061.
Reinhart, Christoph F., and Carlos Cerezo Davila. 2016.
“Urban Building Energy Modeling A Review of a
Nascent Field.” Building and Environment 97
(February): 196202.
doi:10.1016/j.buildenv.2015.12.001.
Robert McNeel & Associates. 2016a. Grasshopper.
http://www.grasshopper3d.com/.
———. 2016b. Rhinoceros (version 5).
https://www.rhino3d.com/.
Schlueter, Arno, and Frank Thesseling. 2009. “Building
Information Model Based Energy/Exergy
Performance Assessment in Early Design Stages.”
Automation in Construction 18 (2): 15363.
doi:10.1016/j.autcon.2008.07.003.
United Nations, Department of Economic and Social
Affairs, and Population Division. 2014. World
Urbanization Prospects: The 2014 Revision:
Highlights.
PREPRINT Proceedings of the 15th IBPSA Conference
San Francisco, CA, USA, Aug. 7-9, 2017
1729
... Elbeltagi et al. [11] used Grasshopper and EnergyPlus to predict and visualize energy consumption in buildings. Similar software combination predicted aggregate energy demand using GIS data and calibrated multi-zone energy models [12]. Kensek [13] used visual programming for energy and shading analysis. ...
Article
Full-text available
Arch bridges are essential components of transportation infrastructure. Their attractive geometry is based on a multitude of geometric parameters, which makes them a challenging design task. Therefore, arch bridges' optimization should be aided by modern computational techniques and algorithms. This study presents an automated optimization process of steel through arch bridges. We merged visual programming, an accessible text programming alternative, with a genetic algorithm to establish an automated framework. We used Dynamo, an open-source civil engineering visual programming language (VPL), to develop a model generation script. Our finite element method (FEM) package enriched the basic VPL functions; it allowed geometry modeling and static strength analysis inside one parametric environment. Linked genetic algorithm replaced the designer in iterative, time-consuming optimization tasks, automating the process. The algorithm adjusted construction's geometric parameters to provide solutions optimized for the typical objective: minimizing the material consumption while still fulfilling strength requirements. We evaluated the procedure with optimization of selected reference construction. The system dealt with cases of increasing complexity, adjusting cross-section dimensions, static scheme parameters, and material properties. The paper describes practical aspects of implementing and utilizing the visual programming-genetic algorithm solution, which can also be adapted for other structures, additional objectives , and constraints.
... Calibrated energy models would allow designers and engineers to explore more bespoke and effective options for buildings to meet energy performance goals rather than simply implementing general energy conservation measures. Furthermore, the relatively long life spans of buildings warrant that modelers take into account future climate scenarios as decisions based on today's typical weather data may not be ideal in the future (Suesser and Dogan 2017). ...
Conference Paper
Full-text available
As negative effects of climate change become increasingly prevalent, carbon emission reduction has become the need of the hour. To meet carbon reduction goals, municipalities, universities and organizations with large real estate portfolios need reliable and adaptable models that provide detailed building performance metrics to efficiently manage future energy demand. However, creating calibrated energy models at the multi-building scale is often a time-consuming task. This paper, hence, investigates methodologies to partially automate the buildup and calibration of energy models for the authors' home institution. This paper presents a workflow that uses institutional GIS datasets, metered energy use and quick surveys as inputs to generate multi-zone EnergyPlus building energy models that are then calibrated using parameter screening and optimization. The calibrated models are used to assess energy performance under the projected climate in the future and evaluate retrofitting scenarios. Accuracy and applicability of the methodology are demonstrated for one campus building and results show that models can be generated with feasible effort reflect satisfactory accuracy for annual, monthly and daily resolutions.
Article
Full-text available
Today, saving energy is very important. Plenty of research has tried to minimise energy losses. The first step for maintaining energy is to find a way to reduce consumption and save money, which means less cost, fuel and pollution. Hence, there is a need to look for the necessary infrastructure to reduce the use of and distribute optimal energy. In this paper, the Taguchi method for energy modelling (EM) optimisation is proposed to predict the best location for heating and cooling appliances. Here, all information about one of the units of the Toos Arman Star Apartment Hotel located in Mashhad, Iran, was obtained. Based on fundamental analysis, the major factors are building coordinates (X, Y, and Z). According to the major factors, orthogonal array L25 is engaged for Taguchi experiments. The authors tried to maximise the satisfaction of building residents by defining the fitness function approaching to thermal comfort point of 25°C. Based on L25 of Taguchi experiments, the best successful result is analysed to select the level of major factors. The result of Taguchi optimisation was introduced into Comsol Multiphysics software and EM was performed using the determined critical points again. The experimental results show the successful location for cooling and heating appliances compared to the initial design of room 1.
Article
Full-text available
This paper reviews Building Energy Modeling (BEM) development using classification and review of articles for the recent years (2015-2018) to explore how much BEM methodologies have been used Building Information Modeling (BIM) database during this period. Nowadays, BEM and supervision are the most important challenges of smart cities, and BIM has a good and useful source of information for buildings. Recent innovations have provided the opportunity for construction industry to invest in state-of-the-art technologies and adopt new processes. Among these new methodologies is BIM, which has developed for the construction industry over the last two decades. Integrating graphical and non-graphical information, it enables construction industry stakeholders to work collaboratively for efficient project delivery throughout the life cycle of construction projects. The use of BEM in construction industry has come about to respond to the global call for energy conservation and sustainability. BEM tools and processes can be used to simulate energy performance, evaluate energy needs and optimize architectural design. This article reviews and classifies BEM applications into eight different categories such as prediction, estimation, consumption, optimal design, evaluation, efficiency, management, and optimization. In addition to examination of BEM in different categories and according to BIM that is a coordination model and include all information in construction industry as well as electrical, mechanical, and architectural information, this review paper wants to know how many articles have used BIM such as a database in this industry.
Thesis
Full-text available
In today's life, it's necessary to pay attention to saving energy. There are a lot of research and studies in this area to minimize the potential for energy losses. Building Information Modeling (BIM) has been considered as a coordinated model for nearly two decades. BIM is a utility software program that integrates the building information required in the building including a three-dimensional architecture, a map of the electrical installations, etc., and, with the intelligence it is predicted, can provide the necessary offers, time and order of execution to the building engineer, and if, for some reason, a part of the map is changed in any case, report the bugs that arise in the next steps. One of the issues that can be solved by BIM as a powerful database is energy issues. The first step for maintaining energy is to find a way to reduce consumption and save money. In fact, saving energy means less cost, less fuel and less pollution. Hence, we need to look for the necessary infrastructure to reduce the use and distribute optimal energy. This thesis aims to develop the methodology of one of the articles (Gerrish et al. method) by proposing a new and innovative design using thermal energy optimization and temperature estimates obtained from the EM software, and predict the best location for heating and cooling appliances. In this study, all the information about one of the units of the Toos Arman Star Apartment Hotel project located on Navvab Safavi Avenue in Mashhad, near Imam Reza Shrine (PBUH), was obtained from this project, and was modeled on BIM software. Then, the key parameters of the building were investigated and all of them were extracted from the BIM software. In the following, EM was investigated using the energy simulation software. Validation was also done by comparing the simulation results and the actual results. For optimization, two optimization methods of genetic algorithm (GA) and Taguchi method was used to determine the optimal location of cooling devices. In GA, we tried to maximize the satisfaction of building residents by defining the fitness function as approaching the thermal comfort point of 25℃. The results of GA and Taguchi optimizations were introduced into the energy simulation software, and again, EM was performed using determined critical points. Since the proposed topic of this thesis is new and so far no action has been taken to determine the best location for the installation of heating and cooling appliances, the EM optimization result of GA was compared with the EM result of two Signal-to-Noise Ratio analysis (SNR) and Analysis of Variance (ANOVA) of the Taguchi method and EM result of the unit studied. In this comparison, the EM results obtained from GA optimization showed the better results than the Taguchi and EM of the unit studied. In GA optimization method, there were two sensitive points that one point showed a better response than the other point. "Please go to this link: http://shahroodut.ac.ir/en/thesis/thesis.php?thid=TK658"
Article
Full-text available
In this paper the authors present an algorithm that abstracts an arbitrarily shaped set of building volumes into a group of simplified ‘shoebox’ building energy models. It is shown that for generic perimeter and core floorplans the algorithm provides a faster but comparably accurate simulation results of annual load profiles vis-à-vis multi-zone thermal models generated according to ASHRAE90.1 Appendix G guidelines. Envisioned applications range from rapid thermal model generation for urban building energy modelling to schematic architectural design. Following a description of the algorithm, its ability to produce load profiles for a mixed-use neighborhood of 121 fully conditioned buildings for a variety of climates is demonstrated. The comparison yields relative mean square errors in simulated annual building energy use intensity of five to 10 percent compared to ASHRAE 90.1 compliant building energy models while reducing simulation times by a factor of 296.
Article
Full-text available
Over the past decades, detailed individual building energy models (BEM) on the one side and regional and country-level building stock models on the other side have become established modes of analysis for building designers and energy policy makers, respectively. More recently, these two toolsets have begun to merge into hybrid methods that are meant to analyze the energy performance of neighborhoods, i.e. several dozens to thousands of buildings. This paper reviews emerging simulation methods and implementation workflows for such bottom-up urban building energy models (UBEM). Simulation input organization, thermal model generation and execution, as well as result validation, are discussed successively and an outlook for future developments is presented.
Conference Paper
Full-text available
Recent advances in the field of building performance simulation (BPS) have not only expanded the quantity of available tools but also widened their audience from the engineering professions to architects and planners. As the user pool expands and team efforts become more common, issues of standardization and exchange of simulation data have become critical for modelers. Although significant efforts have been made in the development of formats for the exchange of geometry and weather data inputs for simulation, or full building energy models (gbXML, IFC), limited attention has been paid to the documentation and exchange of model inputs for building properties, such as material definitions or load profiles. The lack of a widely accepted file format for their comprehensive management becomes especially problematic in the case of rapid energy model generation during early schematic or urban design explorations, because it reduces the time available for design iterations and increases the risk of errors. Supported by a survey of 150 BPS professionals about their current energy modeling workflows, this paper outlines a vision for a new energy modeling data framework based on the use of building properties (BP) templates as a standard input format through the design stages of a project. A proof of concept implementation of a BP template file format and a BP template editor tool are presented along with an example design simulation exercise.
Article
Full-text available
In this paper, we present a general algorithm to automatically convert arbitrary building massing models into multi-zone building energy models (BEM). The algorithm follows current guidelines for thermal zone discretization of BEMs when actual interior space boundaries are yet undefined. Envisioned applications are for rapid model generation during schematic building design as well as for urban massing studies. We present an argument that current recommendations for separating core from perimeter zones effectively follow a straight-skeleton subdivision. Following a step-by-step explanation of the procedure, a number of example building shapes of varying complexity are shown to demonstrate the algorithm's robustness and suitability for automated multi-zone BEM generation. Going forward, it is recommended that the algorithm is adopted by software developers to ensure more consistent thermal model production within the building simulation community.
Article
Full-text available
A new building performance simulation program that combines the best capabilities and features from BLAST and DOE-2 along with new capabilities is presented. Referred to as EnergyPlus, the program comprises completely new code written in Fortran 90.
Article
Simulation packages for predicting building performance in terms of energy and comfort are becoming increasingly important in the planning process. However, current industry standard weather files for building simulation are not suited to the assessment of the potential impacts of a changing climate, in particular summer overheating risks. In addition, no bespoke climate change weather files are readily available that can be loaded directly into environmental simulation software. This paper describes the integration of future UK climate scenarios into the widely used Typical Meteorological Year (TMY2) and EnergyPlus/ESP-r Weather (EPW) file formats and demonstrates the importance of climate change analysis through a case study example. The ‘morphing’ methodology published by the Chartered Institution of Building Services Engineers (CIBSE) is utilised as a baseline for transforming current CIBSE Test Reference Years (TRY) and Design Summer Years (DSY) into climate change weather years. A tool is presented that allows generation of TMY2/EPW files from this ‘morphed’ data and addresses the requirements related to solar irradiation, temperature, humidity and daylighting beyond the parameters provided by CIBSE weather years. Simulations of a case study building highlight the potential impact of climate change on future summer overheating hours inside naturally ventilated buildings.
Article
Due to the rising awareness of climate change and resulting building regulations worldwide, building designers increasingly have to consider the energy performance of their building designs. Currently, performance simulation is mostly executed after the design stage and thus not integrated into design decision-making. In order to evaluate the dependencies of performance criteria on form, material and technical systems, building performance assessment has to be seamlessly integrated into the design process. In this approach, the capability of building information models to store multi-disciplinary information is utilized to access parameters necessary for performance calculations. In addition to the calculation of energy balances, the concept of exergy is used to evaluate the quality of energy sources, resulting in a higher flexibility of measures to optimize a building design. A prototypical tool integrated into a building information modelling software is described, enabling instantaneous energy and exergy calculations and the graphical visualisation of the resulting performance indices.