Content uploaded by Arda Karasu
Author content
All content in this area was uploaded by Arda Karasu on Sep 28, 2023
Content may be subject to copyright.
Climate-resilience examination of retrofitting measures based on future climate data
and energy supply with waste heat
Arda Karasu1, Maximilian Friebe2, Claus Steffan1, Martin Kriegel2
1Technische Universität Berlin, Department of Building Technology and Design, Berlin, Germany
2 Technische Universität Berlin, Department of Energy, Comfort and Health in Buildings, Berlin,
Germany
Abstract
There are two major uncertainties affecting the decision-
making process for retrofitting measures based on
building simulations. These are climatic conditions and
energy resources in the future.
Most designers or simulators produce results for the
present situation using statistical weather data and current
economic values. This way is objective and avoids
speculation. Nevertheless, it automatically leads to
uncertain outputs regarding the operational period of a
building. If the life span of a building is considered, the
realistic approach should be simulation outputs with a
margin related to the future climate.
Offering planners and authorities a margin of outputs
instead of one single number about energy is novel. It
combines the objectiveness of statistical approach and
conceivable eventuality in advance to secure the quality
assurance during the operational period of an building.
This approach is demonstrated through a case study of a
building undergoing retrofitting. The results indicate
planning security through simulations with TRY for 20-
25 years. However, it is important to consider simulations
with future weather data beyond that timeframe.
Additionally, the energy coverage based on renewable
energies may vary significantly due to different weather
scenarios, ranging from 33% to 93% when utilizing waste
heat, depending on the heating system.
Highlights
• Considering future weather data by simulations
• Enhanced quality for energy simulation outputs
• Energy supply with waste heat considering different
climate scenarios
Introduction
This paper focuses on addressing two uncertain factors,
namely future climatic conditions, and their effect on en-
ergy demand and waste heat utilization, during the plan-
ning of an energy concept. It analyses the effects for the
first time in a real case retrofit object in Germany.
It is a clear fact that predicted changes in climate will im-
pact on the building performance (de Wilde &Coley)
where carbon emission and sustainability targets call for
more efficient buildings (Hao et.al.). Building simulations
are hereby established tools since many decades.
The buildings are typically simulated using current
weather data or test reference years, and renovation
concepts are developed accordingly. However, the
measures developed must have the same effect on the
comfort of the building during its life cycle. Nevertheless,
it is certain that due to climate change, different weather
conditions can be expected. Therefore, it is important to
consider future weather conditions when developing such
measures. This growing need is emphasized by Herrera
et.al. to analyse the resilience of building design to cli-
mate change (Herrera et.al.).
Many publications illustrate the fact that when a proce-
dure of verification is applied, for instance related to the
summer heat protection, the calculations based on the fu-
ture weather can often lead to discrepancies compared to
the current procedure (Hoffmann, et.al., Sanchez, et.al.).
In this context, a building located at the Technical Uni-
versity Berlin, which should be retrofitted following the
climate protection agreement (CPE) between Berlin Sen-
ate and University, has been examined in a research pro-
ject on the HCBC-University Campus Berlin-Charlotten-
burg (Münch, et.al.) to determine future climatic changes.
Fifty buildings on the campus of the TU Berlin have been
simulated and different retrofitting scenarios have been
developed. These vary between façade retrofitting and an
urban energy network based on waste heat. All these
buildings on the campus are connected to a district heating
network, that is mainly supplied with fossil fuels cur-
rently. Due to multiple data centres, the campus has a high
waste heat potential that is presently unused. Waste heat
from data centers is particularly well suited for heating
buildings, as it has an almost constant heat output
throughout the year (Wahlroos et al., 2017). Several stud-
ies and projects focusing waste heat utilization are avail-
able. Wahlroos et al. (2018) are presenting an overview of
waste heat utilization in Nordic countries.
In general, waste heat can be utilized either directly to
heat a building or fed into a district heating network. Stan-
ica et al. (2022) developed and analyzed a cascading dis-
trict heating network based on waste heat utilization for
the northern campus of the TU Berlin. This concept offers
lower levelized costs of heating than the current state.
In contrast, the present paper investigates a direct utiliza-
tion of the data center waste heat for a particular building.
Methodology
An in-depth analysis has been run for a particular building
in the campus. It is modelled with DesignBuilder and sim-
ulated with different climate data sets created by
Meteonorm. These data sets are based on representative
concentration pathways of 2.6, 4.5, and 8.5. Furthermore,
the data set of the German Weather Service for 2035 has
been used for comparison. The parameters such as heating
demand, cooling demand, total energy balance, and the
usage of renewable energies have been analysed both sep-
arately and in correlation with each other. In addition, the
utilization of waste heat in this building is investigated for
different heating systems. The respective energy savings
are calculated and analysed using Python.
Case study object
The analysed object was built approximately 40 years ago
as an example of eco-modernity and will undergo renova-
tions in the coming years. It features a high window to
wall ratio (WWR) as it was constructed based on the
greenhouse concept to maximize solar gains. The building
primarily serves as an educational facility and includes of-
fices, seminar rooms, and lecture halls.
Table 1 Building characteristics of the analysed object
Built
1976-86
Number of story proper
9
Net floor area
25.296 m2
Window to wall ration
(above ground)
62,72% (high rise part)
48,21% (including lecture halls)
Specific heat consump-
tion
99 kWh/m2a (35 kWh/m2a after
retrofit
Table 2 Thermal characteristics of the analysed object
Building element
Current
U-Value
W/m2K
U-Value
after potential
retrofit W/m2K
Facade
0,88
0,24
Roof
0,37
0,21
Window
1,75
0,78
Ground floor
0,72
0,29
Figure 1 A typical facade part of the analysed object
Weather data
There are numerous established climate scenarios devel-
oped by various scientists (van Vuuren, et.al.). To avoid
inherent uncertainties, different climate scenarios should
be considered in the analysis. First, Berlin GM was
selected as the location with corrected global radiation,
and data was generated for RCP 2.6, 4.5, and 8.5 for each
scenario. The greenhouse gas concentration ranges here
from 400-1370 ppm CO2-eq and the radiation forcing
ranges from 2.6-8.5 W/m2. The differences between these
scenarios are based on variables such as population
growth, gross domestic product, energy factor (energy
mix), etc. The inherit difference can be seen exemplarily
in Figure 2 for the temperature. Overall, data sets for the
eight decades between 2030-2100 (climate models
CMIP5) were generated, resulting in 24 data sets for the
simulations.
Besides, the DWD (German Weather Service) developed
extra climate data based on 24 regional climate models for
considering long-term requirements of HVAC systems.
Climate models with relevance for a time horizon from
2031 to 2060 form the basis for the data sets of the refer-
ence period 2031-2060. Detailed information about the
future TRY of DWD has been described in the manual of
location accurate TRYs in Germany (5). This data set has
been used as control data set as well.
Figure 2: Comparison of current air temperature of Berlin
with that of 2050 under different scenarios.
Simulation criteria
The object of investigation was modeled and simulated
using DesignBuilder. For the calculations, reference val-
ues of 21°C for heating and 26°C for cool-
ing were used. Both the heating and cooling energy de-
mands were considered after the potential renovation. In
the analysis, final and primary energy were compared
within the respective year and with each other, depending
on the climate scenario and year. The current German fed-
eral uniform primary energy factor of 1.8 for electricity
(mixed source) has been used to have comparable results
as the university uses green energy for electricity cur-
rently. Otherwise, the results would not comparable using
zero for the PE factor. The local PE factor for district heat-
ing is 0.44 and has been calculated by the local supplier.
0
2
4
6
8
10
12
14
16
18
20
22
24
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Air Temperature
Actual RCP2.6 RCP4.5 RCP8.5
In addition to energy calculations, the thermal comfort
during summer was evaluated, using the over temperature
degree hours (OTDH) for the whole building. The OTHD
were calculated using Eq. (1), with the operative temper-
ature and the reference cooling temperature.
.
In the second part of the analysis, environmental energy
sources like solar power and waste heat were considered
as in the building energy concept. Potential changes in the
primary energy factor were not considered for the analy-
sis, to avoid speculation.
Only the standard floors of the building were analysed, as
they were most affected by energy losses based on a pre-
vious analysis. For comparison, the entire building was
simulated with twelve different weather files in a sam-
pling manner, and the results were compared as can be
seen in Figure 3. Depending on the year and scenario, the
differences in cooling demand ranged from 2.9% to 9.9%,
with the largest differences occurring in the milder sce-
nario RCP 2.6. The heating energy demand was on aver-
age 11% higher than in the simulations of standard floors.
The differences between the simulations of standard
floors and the entire building can be explained by the dif-
ferent building types and façade types of the building
parts. As the analysis focuses on future comfort condi-
tions, the analyses were continued on standard floors. En-
ergy concept with waste heat
The building is currently supplied by the district heating
network. Since the campus has a high waste heat potential
(Stanica et al. (2022)), the usage of waste heat was con-
sidered to substitute energy from district heating (see Fig-
ure 4). Additionally, the energy concept of the retrofitted
building consists of PV-installation on roof and façade,
with an annual yield between 205-338 MWh depending
on the installation variation.
A data center in the proximity of the analysed building
offers a high waste heat potential of 2.3 GWh/a with a
temperature of 45 °C. The waste heat load is constant
throughout the year. After comparing the efficiency of
free cooling operation above 0 °C ambient temperature to
waste heat utilization (Münch, et.al.), it was determined
that free cooling would be prioritized. Consequently,
waste heat will be only available when the ambient tem-
peratures is above 0 °C. This waste heat can be utilized to
heat the building when heating supply temperatures are
lower than the waste heat temperature. As a result, part of
the buildings heating demand can be met using waste
heat.
Figure 4: Schematic of the energy concept of the retrofitted
building
The retrofitting of the building enables the use of low tem-
perature heating systems, such as capillary tube mats. For
the analysed building, four different heating systems with
data center
District heating plant
auxiliary source
PV
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
RCP
2.6
2020
RCP
4.5
2030
RCP
8.5
2030
RCP2.6
2040 RCP
4.5
2040
RCP
9.5
2040
RCP
2.6
2050
RCP
4.5
2050
RCP
4.5
2060
RCP
4.5
2070
RCP
8.5
2080
RCP
2.6
2100
RCP
8.5
2100
Energy demand kWh/m2
Cooling demand complete building Cooling demand partially Heating demand complete building Heating demand partially
Figure 3 Comparison between partial and complete simulation models
varying design temperatures were compared (see Table
3).
Table 3Investigated heating systems and their design tempera-
tures
Heating system
radiators
70 °C/55 °C
radiators
60 °C/45 °C
floor heating
40 °C/30 °C
capillary tube mats
35 °C/31 °C
For each heating system a dynamic supply temperature
profile was created using a heating curve. Subsequently
the cover ratio of waste heat was calculated. The impact
of climate change on waste heat utilization was investi-
gated by comparing the TRY (2017) weather data for a
typical year with a scenario based on the most extreme
RCP 8.5 projection of the year 2100.
Results
The initial situation is as follows:
As exemplified in Figure 5, changes in over temperature
hour values in the study object between 2020-2100 are
shown. These operational degree-temperatures between
the years 2020-2100 remain within the range of 80-90%
compatibility during the summer period (during the oper-
ational phase).
The 80% compatibility means that the comfortable indoor
temperature can be 2.5°C higher, while it is 3.5°C at 90%.
Even with the moderate RCP 2.6 scenario, where global
warming is supposed to stay below 2°C, the increasing
trend of excess temperature hours is visible.
Compared to 2020, for example, the indoor temperature
can exceed the comfortable values by 2.5°C for up to 60
hours in 2070. This means that the indoor spaces would
not meet the comfort criteria for more than seven working
days and the cooling demand would increase.
Figure 5: Changes of the over temperature degree hours
between 2020-2100 in the analysed object
Consideration of cooling energy demand:
Even under the mildest scenario, the cooling demand will
increase by 35% between 2020-2100. For the moderate
scenario, the increase is about 56% and under RCP 8.5,
the cooling demand in 2100 will be 2.7 times higher
compared to 2020. The final energy demand varies be-
tween 77,000-220,000 kWh depending on the year and
scenario. When considering primary energy demand, it
could reach 200,000 kWh to achieve climate neutrality in
2050. This, in turn, would mean an increase of 43% when
comparing the current demand with that of 2050 (RCP
8.5).
Figure 6 changes in the specific cooling demand, blue:
final energy, red: primary energy
Consideration of heating energy demand:
Compared to cooling, the trend for heating energy de-
mand is going in the opposite direction. Depending on the
scenario and year, a reduction of the heating energy de-
mand by half can be expected, with the reduction, in the
end, energy demand reaching up to around 20% by the
target year 2050. In terms of primary energy, the entire
building would consume around 6% less energy, with
only 3% for the regular floors.
Figure 7 changes in the specific heating demand, blue:
final energy, red: primary energy
The ratio between cooling and heating:
More interesting is the development of the total energy
demand for heating and cooling together and their rela-
tion. As shown in Figure 8, the current cooling demand is
only about 19% of the current heating demand for end en-
ergy and over 75% for primary energy. In the RCP 8.5
520
530
540
550
560
570
580
590
2020 2030 2040 2050 2060 2070 2080 2090 2100
over temperature hours
year
0,00
10,00
20,00
30,00
40,00
50,00
60,00
2020
2040
2060
2080
2100
2030
2050
2070
2090
2020
2040
2060
2080
2100
RCP 2.6 RCP 4.5 RCP 8.5
Energy demand kWh/m2
0
10
20
30
40
50
60
70
2020
2040
2060
2080
2100
2030
2050
2070
2090
2020
2040
2060
2080
2100
RCP 2.6 RCP 4.5 RCP 8.5
heating energy demand kWh/m2
scenario, the end energy demand for cooling and heating
could be equal in the year 2100, which would mean that
the primary energy demand for cooling could be 3.5 times
that of the energy demand for heating. In other words, the
share of primary energy demand for heating in the future
would be between 9.7-17.2% and that of energy demand
would be between 29.1-45.4%.
The consideration of the overall energy balance for
heating and cooling:
Specifically, substantial changes in cooling and heating
energy demand can be expected. However, when consid-
ering the overall balance, a decrease in total energy con-
sumption for cooling and heating could be expected in 11
out of 30 scenarios. In 18 scenarios, it could be increased,
by 11.1% in primary energy and 5.8% in final energy, re-
spectively.
When considering the year 2050 as a target, only in the
RCP 4.5 scenario, primary energy consumption is ex-
pected to increase by only 1.2%. Partial reductions of up
to 6.4% in end energy consumption and 1.7% in primary
energy consumption can be expected in 2050. Neverthe-
less, if all years up to 2050 are considered, the total end
energy consumption of the building could decrease be-
tween 0.8-7.7%, while the total primary energy consump-
tion could increase between 0.5-7.7%.
Scenarios based on clean energy harvesting:
As shown in Figure 10, the cover ratio of waste heat de-
pends on the installed heating system and the weather data
used. Low temperature systems such as capillary tube
mats or floor heating achieve a degree of coverage of 78
% for the TRY scenario and 93 % for the RCP 8.5 sce-
nario. The supply temperature of these systems remains
below the waste heat temperature throughout the year.
Therefore, waste heat can be utilized to supply these sys-
tems at all times when free cooling is not in operation.
In contrast, the investigated radiator systems have a sup-
ply temperature above 45 °C for part of the year.
Consequently, the degree of coverage is lower. Generally,
due to the milder temperatures in winter, the RCP 8.5 sce-
nario exhibits a higher cover ratio with waste heat.
Figure 9: Degree of coverage of the heating demand
with waste heat for different heating systems for different
climate scenarios
A full coverage of the heating demand with waste heat can
be reached for capillary tube mats or floor heating by us-
ing a thermal energy storage.
The analysis considers the following parameters and as-
sumptions:
• The PV system on the roof has a degradation rate of
0.1% per year, and the performance of the façade sys-
tem stays constant.
• The radiative forcing under different RCP scenarios
has been considered in the energy yield calculation.
• The demand for individual years between 2020-2050
has been interpolated based on simulated values for
the years 2020, 2030, 2040, and 2050.
Depending on the RCP scenario, it is possible to save an
amount of end energy between 32.5-35.3 GWh within 30
years until 2050 by utilizing waste heat, which means
050 100
Radiators 70/55
Radiators 60/45
Floor Heating
Capillary Tube Mats
heating system
coverage ratio in %
RCP 8.5 TRY
0%
50%
100%
150%
200%
250%
300%
350%
2020
2030
2040
2050
2060
2070
2080
2090
2100
2020
2030
2040
2050
2060
2070
2080
2090
2100
2020
2030
2040
2050
2060
2070
2080
2090
2100
RCP 2.6 RCP 4.5 RCP 8.5
ratio of Heating demand / Cooling demand
Figure 8 the ratio between cooling and heating demand according to the scenarios, blue: final energy, red: primary energy
complete self-sufficiency, if a thermal storage is a part of
the system. Otherwise, the maximum cover ratio of the
heating demand is 93%. In addition, the planned PV sys-
tems can cover the cooling demand between 70.3-74.7%
within 30 years, which means an end energy savings of
6.1-6.4 GWh.
Comparison of weather data:
The simulation model was also simulated with the Test
Reference Year (TRY) 2035 Normal Year and TRY 2035
hot summer year. These are based on regional climate
models of the DWD for the period 2021-2050, which pre-
pare weather data for Potsdam. The dataset that is based
on the summer has 26 hot days and thus 16 days more than
the normal year. For summer days, the difference is 11
days and for tropical nights, it is 2 days (a total of 3 for
the hot summer year).
The simulation results after these datasets were compared
with the results after RCP scenarios. As the RCP scenar-
ios supply values only for decades, the average value be-
tween 2021-2050 was taken as the reference value.
For the heating demand, the difference between the results
of simulations with DWD and Meteonorm weather data is
up to 10.5%. For the cooling demand, the difference is up
to 22.5%.
Figure 10 Varied final energy demands based on climate
scenarios, blue: cooling, orange: heating
Consideration of the building's total energy demand:
As described in the section "Comparison of weather data",
the intensity and duration of sunlight hours may vary de-
pending on the scenario. This could have an impact on
daylight use or electricity consumption for lighting, for
example.
The trend of the remaining energy demand is also decreas-
ing. Depending on the climate scenario, there will be a
reduction of 1.2-2% in total energy consumption, mainly
due to increased use of daylight. This affects an energy
amount of approximately 42-65T kWh/year.
Figure 11 Total energy consumption of the building ac-
cording to the scenarios
Conclusion
This research yielded four major findings for the similar
buildings in Germany, which could support decision mak-
ers:
• First, analysis of future weather data confirms that en-
ergy concepts based on simulations with TRY would
perform well within their standard HVAC lifespan of
20-25 years.
• Second, when considering the lifespan of a building
envelope, the period after 25-30 years of operation
should be critically assessed in terms of quality assur-
ance for the energy supply systems.
• Third, as the ratio between cooling and heating energy
demand may change, the supply systems and building
design should be designed to be adaptable.
• The utilization of waste heat can significantly reduce
the consumption of conventional heat sources for
space heating. As future climate scenarios indicate
milder winters, compared to typical TRY data, the
cove ratio of waste heat utilization will increase. This
makes waste heat a suitable heat source for the entire
lifespan of the building.
Due to the different scenarios, the results show relative
differences in heating and cooling demand. However, the
tendency is clear for both, decrease of the heating demand
and increase of the cooling demand. Despite these tenden-
cies, no major changes are expected in the overall energy
balance for the analysed building, as the reduction in heat-
ing demand would compensate for the increase in cooling
demand. It should be noted that the building currently
does not have a cooling system, and according to the re-
sults, the comfort during the summer period is not satis-
factory and is negatively affected by rising temperatures.
By the target year 2050, the developed measures should
function well. Furthermore, the planned measures can
easily function until 2070 due to the moderate climate sce-
nario. After a critical phase between 2070-2080, the
measures should at least work in the analysed period.
0
200
400
600
800
1000
1200
1400
energy demand in MW/a
Cooling Heating
2,6
2,7
2,8
2,9
3
3,1
3,2
3,3
3,4
3,5
2020
2040
2060
2080
2100
2030
2050
2070
2090
2020
2040
2060
2080
2100
RCP 2.6 RCP 4.5 RCP 8.5
Net site energy consumption in GWh/a
However, concerning the worst-case scenario, alternative
measures for active and passive cooling of the building,
or a development plan for the next renovation phase,
should be considered.
Finally, the total energy balance indicates that renewable
energy sources should be used. Utilizing waste heat, it is
possible to achieve up to 93 % savings of heating energy,
depending on the implemented heating system and the cli-
mate scenario. By implementing a thermal energy storage,
a full coverage of the heating demand is possible. Never-
theless, achieving climate neutrality concerning the in-
creased cooling demand will be more difficult for the
building in the future.
In conclusion, it is suggested to consider not only a typical
TRY weather data running building simulations. Further-
more, the regulations for both summer and winter heat
protection should be adapted using future weather data as
well.
Acknowledgements
This project has received funding from the Federal Min-
istry of Economic Affairs and Energy under grant agree-
ment No 03ET1354A. We would like to acknowledge the
contribution of all project partners.
References
De Wilde P., Coley, D., The implications of a changing
climate for buildings, Building and Environment
Volume 55, September 2012, Pages 1-7
Hao., L, Herrera-Avellanosa, D., Del Pero, C, Troi, A.,
What are the implications of climate change for
retrofitted historic buildings? A literature review,
Sustainability 2020, 12(18), 7557
Herrera, M., Natarajan, S., Coley, D.A., Kershaw, T.,
Ramallo-Gonzales, A.P., Eames, M., Fosas, D.,
Wood, M. (2017). A review of current and future
weather data for building simulation. Building Serv.
Eng. Res. Technol. 2017, Vol. 38 (5), 602-627.
Hoffman, S., Keybari, A.G. (2021), Untersuchungen zum
Sommerlichen Wärmeschutz- Teil 1 Vergleich der
Nachweisverfahren unter Berücksichtigung zu-
künftiger Klimadaten. Bauphysik Vol. 43 Issue 1, Feb.
2021, 27-35.
Sanchez, L., Karasu, A., Bachmann, Kriegel, M. (2022).
Simulation-based vulnerability analysis of summer
heat protection calculations in Germany. Proceedings
from Bausim 2022, Weimar (DE), 20-22 Sept. 2022.
CPE Climate Protection Agreement, Link
Münch, B., Karasu, A., Brandt, D., Hantouch, Y.,
Kononenko, N., Küster-Inderfurth, A., Stanica, D.,
Steffan, C., Kriegel, M, Nytsch-Geusen, C., Ziegler,
F. (2018). Eneff: HCBCHochschulCampus Berlin –
CharlottenburgDemonstration eines innovativen
Wärmeenergiemanagements für ein Bestandsquar-
tier. Berlin (DE)
Wahlroos, M., Pärssinen, M., Manner, J., Syri, S. (2017).
Utilizing data center waste heat in district heating –
Impacts on energy efficiency and prospects for low-
temperature district heating networks. Energy, Vol.
140, Part 1, 1228-1238.
Wahlroos, M., Pärssinen, M., Rinne, S., Syri, S., Manner,
J. (2018). Future views on waste heat utilization –
Case of data centers in Northern Europe. Renewable
and Sustainable Energy Reviews, Vol. 82, Part 2,
1749-1764.
Stanica, D.; Kononenko, N.; Bachmann, M.; Schumann,
F.; Friebe, M.; Kriegel, M. (2022). A Holistic
Energetic Transformation Concept for the Heating
and Cooling Supply of a University Campus.
Proceedings from EuroSun 2022, Kassel (DE), 25 –
29 Sept. 2022.
van Vuuren et al (2011) The Representative Concen-
tration Pathways: An Overview. Springer Verlag,
Climatic Change, 109 (1-2), 5-31
BBR and DWD, (2017). Handbook: Test reference years
of Germany for moderate, extreme and future weather
data (in German). Offenbach (DE) Link
