Doctoral Thesis
BUILDINGS AS POTENTIAL URBAN MINES: QUANTITATIVE,
QUALITATIVE AND SPATIAL ANALYSIS FOR VIENNA
submitted in satisfaction of the requirements for the degree of
Doctor of Science
of the Technische Universität Wien, Faculty of Civil Engineering
Dissertation
GEBÄUDE ALS POTENZIELLE URBANE ROHSTOFFLAGER:
QUANTITATIVE, QUALITATIVE UND RÄUMLICHE ANALYSE FÜR
WIEN
ausgeführt zum Zwecke der Erlangung des akademischen Grades eines
Doktors der technischen Wissenschaft
eingereicht an der Technischen Universität Wien Fakultät für Bauingenieurwesen
von
Dipl.-Ing. Fritz Kleemann
Matrikelnummer: 0440532
Wiedner Hauptstraße 40/GS1/9
1040 Wien
Gutachter: Ass.-Prof. Dr. Johann Fellner
Institut für Wassergüte, Ressourcenmanagement und Abfallwirtschaft
Technische Universität Wien
Karlsplatz 13/226, A-1040 Wien
Gutachter: Ass.-Prof. Ramzy Kahhat, PhD
Departamento de Ingeniería,
Pontificia Universidad Catolica del Peru
Avenida Universitaria 1801, San Miguel, Lima, Perú
Wien, August 2016
I
Acknowledgements
First and foremost I would like to thank Johann Fellner for giving me the opportunity to carry out this
research within the Christian Doppler Laboratory for Anthropogenic Resources at the Technische
Universität Wien, and for his constant supervision.
I would further like to thank: Jakob Lederer for his support during the entire research project, and for
not letting the working environment become too deadly serious; Philipp Aschenbrenner, especially for
helping me carrying out the case studies done in this research; Helmut Rechberger, as head of the
Institute for Water Quality, Resource and Waste Management, for his confidence and support; and
Ramzy Kahhat for co-supervising this thesis and giving me the opportunity to conduct my research in
a different set-up, namely the Pontificia Universidad Catolica del Peru.
I thank all my colleagues at the institute for creating a great place to work, and especially my office
mates Resi, Dana and Jakob (again). Hanno, thank you for always spreading a positive atmosphere and
for countless lunchbreaks – it was my pleasure! David, Julia, Verena, and Andrea, thank you for your
collegiality within the CDL-bubble. Oliver Schwab, thank you for a great road trip to Surrey and for
sharing one or another beverage. Oliver Cencic, thank you for input with regard to the uncertainty
analysis. Uli, thanks for the exchange and discussion on the topic!
I thank the whole team connected with the project “Hochbauten als Wertstoffquelle”, which was
carried out in the course of this thesis, as well as all municipal departments involved. Special thanks
go to Claudia Schrenk and Josef Zeininger for coordinating the project and facilitating data exchange.
I thank Martin Denner for his work on the GIS data and Helmut Augustin for his valuable input and
help in identifying data sources. I thank Hubert Lehner and Anna Szczypińska for their work on the
change detection data and for her contribution to the associated publication, and Hannes Kirschner and
Eva Büchelhofer for making the analysis of construction files possible. Franz Oberndorfer initiated the
project “Hochbauten als Wertstoffquelle”, for which I thank him. Financial support came from the
Wiener Linien GmbH & Co KG, which is gratefully acknowledged.
Furthermore, I thank all building owners, project developers and wrecking companies who were
willing to cooperate with us by sharing data or by giving us access to buildings to enable us to collect
the relevant data for this research.
Finally, I thank my family and friends, especially Patricia for enriching my life beyond the university!
III
Abstract
Buildings constitute a major contributor to material use and accumulation in human settlements.
Therefore, they play an important role when moving toward a sustainable use of natural resources.
Knowledge about the material composition of buildings and dynamics in the building stock are
considered a prerequisite to defining effective resource management measures. The thesis aims to
investigate buildings as potential urban mines in the sense that existing material stock can be used as a
future mine for secondary resources. The city of Vienna has been chosen as a case study.
In order to generate data about the composition of buildings in Vienna, specific material intensities for
different building categories are defined. This is done based on different data sources. A practical
method (1st Paper) is presented to characterize the material composition of buildings prior to their
demolition. The characterization method is based on the analysis of available construction documents
and different approaches of on-site investigation. The method is tested in case studies carried out, and
results indicate that the documents are useful to quantify bulk materials (e.g. bricks, concrete,
sand/gravel, iron/steel and timber). On-site investigations are necessary to locate and determine
materials of lower concentration such as metals (e.g. copper and aluminium) or plastics. To enlarge the
sample size of buildings being investigated, construction files of already demolished buildings are
analysed to determine the specific material composition of the buildings. Additionally, new buildings
are investigated based on existing life cycle assessments, accounting documents and construction
plans. The database for specific material intensities for different building categories is complemented
with data from the literature.
In the second part of the thesis (2nd
Paper) material stocks in buildings and their spatial distribution are
analysed. In particular, the building structure is analysed by joining available geographical information
systems (GIS) data from various municipal authorities. The previously generated specific material
intensities for different building categories are subsequently combined with the data on the building
structure. This allows the overall material stock in buildings in Vienna to be calculated as well as the
spatial distribution of materials in the municipal area to be assessed. This research forms the basis for
a resource cadastre, which provides information about gross volume, construction period, utilization,
and material composition for each building in Vienna.
In a further step, the information about the material composition of buildings is combined with data on
the demolition activity in order to estimate quantity and quality of demolition waste generated. The
volume of demolished buildings is calculated based on two different data sources (demolition statistics
and change detection data) and multiplied with the respective material composition. In the 3rd
Paper an
approach is presented that allows demolition statistics to be validated by using data of automatized
change detection of the building stock. Based on this technique, building demolition activities in the
municipal area are detected based on yearly aerial images. Results show that demolition statistics do
not cover all demolition activity in Vienna and, consequently, demolition waste generation figures
solely based on statistical data of demolition activities would underestimate the total waste generation.
The approach used in this study can be useful for validating existing data on demolition waste
generation and demolition statistics or to generate data if no data is available.
V
Kurzfassung
Hinsichtlich Materialverbrauch und Lageraufbau von Materialien in menschlichen Siedlungen spielen
Gebäude eine maßgebliche Rolle. Folglich spielen Gebäude eine wichtige Rolle wenn es um die
nachhaltige Verwendung natürlicher Ressourcen geht. Das Wissen über die Materialzusammensetzung
von Gebäuden und die Dynamiken des Gebäudebestands sind Voraussetzungen um effektive
Maßnahmen im Ressourcenmanagement zu setzen. Ziel der Arbeit ist es das Potenzial von Gebäuden
als urbane Mine zu bewerten (im Sinne eines Lagers für zukünftige Sekundärrohstoffe) zu
untersuchen. Wien dient hierzu als Fallstudie.
Um Daten über die Materialbeschaffenheit von Gebäuden in Wien zu generieren, werden spezifische
Materialintensitäten für unterschiedliche Gebäudekategorien definiert. Dies geschieht basierend auf
unterschiedlichen Informationsquellen. Eine Methode zur Bestimmung der Materialzusammensetzung
von Gebäuden vor dem Abbruch wird im ersten Paper präsentiert. Die Methode basiert auf der
Auswertung von vorhandenen Unterlagen über das Gebäude sowie unterschiedlichen Untersuchungen
vor Ort. Die Methode wird in Fallstudien angewendet und die Ergebnisse zeigen, dass auf Basis der
vorhandenen Unterlagen die Hauptmaterialien (z.B. Ziegel, Beton, Sand/Kies, Eisen/Stahl oder Holz)
quantifizieren werden können. Die Untersuchungen vor Ort sind nötig um Materialien in geringer
Konzentration wie Metalle (z.B. Aluminium und Kupfer) und Kunststoffe abschätzen zu können. Um
die Stichprobe untersuchter Gebäude erhöhen zu können, werden bereits abgebrochene Gebäude
anhand von Konstruktionsplänen hinsichtlich ihrer Materialzusammensetzung analysiert.
Lebenszyklusanalysen, Abrechnungsunterlagen und Konstruktionspläne ermöglichen es neue Gebäude
zu bewerten. Der Datensatz über spezifische Materialintensitäten unterschiedlicher Gebäudekategorien
wird durch Literaturdaten ergänzt.
Im zweiten Teil der Arbeit (zweites Paper) werden das Materiallager in Gebäuden und dessen
räumliche Verteilung analysiert. Die Gebäudestruktur wird durch die Verknüpfung vorhandener
Geoinformationssysteme (Informationen zu Gebäudefläche und -höhe, Nutzung und Bauperiode)
unterschiedlicher Magistratsabteilungen analysiert. Die zuvor generierten spezifischen
Materialintensitäten für unterschiedliche Gebäudekategorien werden mit Daten über die
Gebäudestruktur kombiniert. Diese Methode erlaubt es das gesamte Materiallager in Gebäuden Wiens
sowie dessen räumlich Verteilung im Stadtgebiet zu bewerten. Dies bildet die Grundlage für einen
Ressourcenkataster für Hochbauten welcher Informationen zu Volumen, Bauperiode, Nutzung und
Materialzusammensetzung aller Gebäude enthält.
Die Informationen über die Materialzusammensetzung von Gebäuden werden in einem weiteren
Schritt mit Daten über die Abbruchaktivität verknüpft, um Menge und Qualität von Abbruchabfällen
abschätzen zu können. Das Volumen abgebrochener Gebäude wird, basierend auf zwei
unterschiedlichen Datenquellen (Abbruchstatistik, fernerkundliche Veränderungsdetektion), berechnet
und mit den relevanten spezifischen Materialintensitäten multipliziert. Das dritte Paper präsentiert
einen Ansatz der es erlaubt, anhand von Daten aus der automatischen Erkennung von Veränderungen
im Gebäudebestand, Abbruchstatistiken zu verifizieren. Dabei werden mit Hilfe von Luftbildern
Abbruchaktivitäten im Stadtgebiet identifiziert. Die Ergebnisse zeigen, dass die Abbruchstatistiken
nicht alle Abbrüche in Wien erfassen und Berechnungen der Abfallmengen auf Basis dieser Quelle das
Abfallaufkommen unterschätzen. Der präsentierte Ansatz stellt eine Möglichkeit dar, existierende
Daten zu Abbrüchen und Abbruchabfällen zu validieren bzw. Daten zu generieren, wenn keine
vorhanden sind.
VII
Published articles
1st Paper
A method for determining buildings’ material composition prior to demolition
Fritz Kleemann, Jakob Lederer, Philipp Aschenbrenner, Helmut Rechberger and Johann Fellner
Building Research and Information
DOI: 10.1080/09613218.2014.979029
Contribution: literature research; planning, organisation and implementation of case studies;
collection and evaluation of data; writing
2nd
Paper
GIS-based analysis of Vienna’s material stock in buildings
Fritz Kleemann, Jakob Lederer, Helmut Rechberger, and Johann Fellner
Journal of Industrial Ecology
DOI: 10.1111/jiec.12446
Contribution: literature research; planning and design of the method; collection and evaluation of
data; writing
3rd
Paper
Using change detection data to assess amount and composition of demolition waste from
buildings in Vienna
Fritz Kleemann, Hubert Lehner, Anna Szczypińska, Jakob Lederer, and Johann Fellner
Resource Conservation and Recycling
DOI: 10.1016/j.resconrec.2016.06.010
Contribution: literature research; planning and design of the method; collection and evaluation of
data; writing
IX
Contents
1. Introduction ................................................................................................................................... 1
2. Objectives ....................................................................................................................................... 3
3. Scientific background .................................................................................................................... 5
3.1. Data generation on a building level ......................................................................................... 5
3.2. Modelling materials in the building stock ............................................................................... 6
3.3. Demolition waste generation ................................................................................................... 8
4. Methods ........................................................................................................................................ 11
4.1. Specific material intensities for different building categories ............................................... 11
4.1.1. Case Studies .................................................................................................................. 11
4.1.2. Construction plans of demolished buildings ................................................................. 14
4.1.3. New buildings ................................................................................................................ 15
4.1.4. Literature ....................................................................................................................... 15
4.2. Building structure and total material stock of buildings in Vienna ....................................... 16
4.3. Waste generation through building demolition ..................................................................... 18
4.3.1. Statistics-based analysis of the demolition activity ....................................................... 18
4.3.2. Demolition activity analysis based on remote sensing image matching ....................... 18
5. Results........................................................................................................................................... 21
5.1. Case studies ........................................................................................................................... 21
5.2. Specific material intensities of different building categories ................................................ 27
5.3. Building structure and total material stock of buildings in Vienna ....................................... 29
5.4. Waste generation through building demolition ..................................................................... 32
6. Discussion ..................................................................................................................................... 39
6.1. Uncertainties .......................................................................................................................... 39
7. Conclusion and outlook ............................................................................................................... 45
Literature ............................................................................................................................................. 47
Appendix ................................................................................................................................................ 1
1st Paper ............................................................................................................................................... 1
2nd
Paper .............................................................................................................................................. 3
3rd
Paper ............................................................................................................................................... 5
XI
List of figures
Figure 1: Schematic overview of the information used in this research to quantify the material stock in
buildings and building demolition waste ............................................................................................... 11
Figure 2: Notional sections (displayed on the left) in areas with many installations (right). Based on
Figure 1 in the 1st Paper. ........................................................................................................................ 14
Figure 3: Information flows of different data sources used in this research. Based on Figure 1 in the 2nd
Paper. ..................................................................................................................................................... 17
Figure 4: Schematic illustration of the generation of the height difference modes based on two
DSMs ..................................................................................................................................................... 19
Figure 5: Origin of materials for CS1. Based on Figure 3 and 4 in the 1st Paper. ................................ 26
Figure 6: Location and quantity of valuable metals (steel, aluminium, copper) in CS1. Based on Figure
5 in the 1st Paper. ................................................................................................................................... 27
Figure 7: Gross volume (m³) of different building categories in the city of Vienna in the year 2013.
Based on Figure 2 in the 2nd
Paper. ....................................................................................................... 29
Figure 8: Material composition of the total building stock in the city of Vienna in the year 2013. Based
on Figure 3 in the 2nd
Paper. .................................................................................................................. 30
Figure 9: Spatial distribution of minerals in buildings in Vienna (kg/m² built-up area). Based on Figure
4 in the 2nd
Paper ................................................................................................................................... 31
Figure 10: Demolished building volume (m³ gross volume) in 2013 and 2014 based on statistics of the
municipal building inspection department. Based on Figure 4 of the 3rd
Paper. ................................... 32
Figure 11: Gross volume of demolished buildings (given in m³/a) based on statistics of the building
inspection department (one-year average of 2013 and 2014), and image matching based change
detection (one-year equivalent of data from July 2013–June 2014). Based on Figure 5 of the
3rd
Paper. ................................................................................................................................................ 34
Figure 12: Cumulative curve of demolished building volume (m³ gross volume) of demolition projects
ranked by size ........................................................................................................................................ 35
Figure 13: Standard error of estimated mean value depending on the number of samples
(Cencic 2016) ........................................................................................................................................ 41
XIII
List of tables
Table 1: Buildings investigated as case studies and their main characteristics ..................................... 12
Table 2: Comparison of material composition of the case studies (kg/m³ gross volume) ..................... 25
Table 3: Specific material intensities (kg/m³ GV) of different building categories in Vienna (rounded
to two significant digits). Based on Table 1 in the 2nd
Paper. ............................................................... 28
Table 4: Per capita figures on the materials present in buildings in Vienna (t/cap) (rounded to two
significant digits). Based on Table 2 in the 2nd
Paper............................................................................ 31
Table 5: Quantity and composition of waste from demolished buildings in Vienna (rounded to two
significant digits). Based on Table 2 of the 3rd
Paper. ........................................................................... 36
Table 6: Estimated uncertainties for data sources used to generate specific material intensities for
different building categories. ................................................................................................................. 40
Table 7: Contribution of the different building categories to the overall building volume (given in %)
and number of buildings analyzed with respect to their material composition. Parenthesized number
represents data from literature. .............................................................................................................. 41
Table 8: Estimated uncertainty of data used to analyze the building structure ..................................... 42
1
1. Introduction
In urban environments around the world, large amounts of materials are consumed and accumulate
over time, with buildings and network infrastructures being the major contributors. Building up this
so-called anthropogenic stock requires that primary natural resources be exploited, which subsequently
remain the paramount source for current material demand. The lifetime of both buildings and
infrastructure is long compared to cars or consumer goods, which plays an important role for the
recycling of the materials since today’s waste arises from construction activities lying decades in the
past. At the same time, new materials and technologies are brought into the building stock today,
becoming the waste of the future. Increased awareness about the necessity of sustainable use of
resources and a general concern about the linear use of materials has resulted in calls for stronger
recycling efforts. Global urbanization (United Nations, 2014) and the concomitant demand for
construction material underline the topicality and need for appropriate measures in this context.
On a European level, it is estimated that between 25 and 30% of the overall waste generated can be
attributed to construction and demolition activities (European Comission, 2014). Due to the high
potential to reduce the use of primary resources and landfill space by closing material cycles in the
construction sector, recycling targets for construction and demolition waste (CDW) for all member
states were defined by the European Parliament and the Council of the European Union (2008).
Accordingly, a minimum of 70% (by weight) of non-hazardous construction and demolition waste
(soil and stone excluded) shall be prepared for reuse, be recycled or undergo other material recovery
by the year 2020.
In Austria between 20 and 30% of the total waste generated results from construction and demolition
activities, excluding excavation material (BMLFUW, 2011). This amounts to 800-1000 kg of CDW
per capita (cap) and year (a). About 90% of the material is recycled or backfilled. However, no
detailed information about material flows is available (BMFLUW, 2015). In order to assess to what
extent CDW management can be improved, data on waste generation and waste flows are required.
Contrary to many other waste streams (e.g. packaging waste, WEEE, etc.), CDW management is
mostly realized on a regional level, and only a very small share of the overall CDW stream (e.g.
metals) is subject to the national and international market. The reason is that the mineral fraction,
which represents the bulk of the CDW, is characterized by a rather small trade and transport radius due
to its large quantity and the low value per mass unit if compared to other waste such as metal scrap or
electronic waste. Hence, regional and local data on CDW needs to be generated, verified and brought
2
together to obtain coherent information for CDW management and recycling, which in turn is
necessary to analyse current practices in managing CDW and to detect optimization potentials.
Compared to network infrastructure such as streets, subway lines or tram lines, buildings are very
heterogeneous in their material composition. This has great influence on the recyclability of materials
at the end of a building´s life, and requires precise deconstruction and demolition measures. Therefore,
it is crucial to obtain a better understanding about the material composition of the building stock and
related material flows. A further interesting point of relevance about buildings compared to network
infrastructure is their ownership structure. While network infrastructure is often publicly owned and
managed, the building sector is characterized by more privately owned property. Not only buildings,
but also the stakeholders involved are, therefore, key factors for proper CDW management.
3
2. Objectives
The main research goal of this thesis is to investigate the resource potential of the building stock in
Vienna. Therefore, an estimation of the material composition of the building stock as well as its
dynamics is necessary. The high density of buildings in urban areas potentially allows efficient
management of anthropogenic resources. Here waste materials from e.g. demolition activities occur in
relative proximity to places of material demand e.g. through construction activities. In order to exploit
this potential, resource management strategies require reliable data to plan, implement and monitor
productive measures to support the cyclic use of materials. Although compared to many other
European countries, the availability and quality of waste-related data is good in Austria, the
construction and demolition sectors in particular often remain intransparent and existing data are hard
to verify.
Three main areas of investigation can be differentiated in this research; they are, however, intertwined
with each other.
I) The generation of specific material intensities for different building categories is the basis
for characterizing the overall building stock as well as estimating waste streams from
building demolition. Therefore, a suitable method following a bottom-up approach to
assess the material composition of buildings before their demolition has been developed.
Furthermore, information about the location and quality of the materials within the
buildings is necessary. After the demolition of the buildings investigated, data provided by
contracted wrecking companies can be verified. Additional sources of information about
the material composition of different building types include construction files of already
demolished buildings in Vienna, bills of quantities, and LCA (Life Cycle Assessment)
data of new buildings as well as relevant literature.
II) Based on the specific material intensities of different building categories, and information
about the building structure, the quantification of the overall material stock embedded in
the buildings of Vienna is possible. By combining information of different GIS
(Geographical Information System) data, available through various municipal departments
of the City of Vienna, the building structure is investigated with regard to the share of
different building categories of the overall building stock.
III) The amount and quality of demolition waste (DW) from buildings is estimated by using
the specific material intensities generated in combination with data about the demolished
building volume of the different building categories. The demolished building volume is
determined using statistical data and image matching change detection data respectively,
in combination with GIS data.
4
Based on the above-mentioned factors, the following main research questions are examined:
1) How can the material composition of buildings be determined in an efficient but
comprehensive way?
2) How can specific material intensities for different Viennese building types be generated?
3) What is the building structure in Vienna and how can it be assessed?
4) What is the total material stock in Viennese buildings?
5) What waste occurs through building demolition and is available for recycling?
6) What measures are available to assess and monitor the demolition activity in Vienna?
In accordance with the main areas of investigation, three papers were authored within this thesis. The
first paper of the thesis (A method for determining buildings’ material composition prior to
demolition) tackles questions of how to generate data about the material composition of single
buildings in an efficient way. The second paper (GIS-based analysis of Vienna’s material stock in
buildings) aims at characterizing the material composition of the building stock of Vienna by
analysing the building structure based on existing municipal GIS data. A data set giving information
about gross volume, construction period and utilization of each building in Vienna is generated and
combined with specific material intensities of different building categories. The third paper (Using
change detection data to assess amount and composition of demolition waste from buildings in
Vienna) focuses on analysing the amount and quality of waste generated through building demolition.
The specific material intensities are combined with change detection data derived from yearly aerial
images of the municipal area of Vienna.
5
3. Scientific background
Although there are different approaches to gaining knowledge about the characterization of the
building stock and its dynamics (Kohler and Hassler, 2002), its material composition is still not well
known (Kohler and Yang, 2007). Challenges in this regard lie in generating data about individual
buildings on the one hand, and in projecting this information for selected geographical areas of interest
on the other hand.
3.1. Data generation on a building level
For individual buildings, specific material intensities can be generated through (i) information about
the materials used to construct buildings, by (ii) analysing generated waste streams after demolition, or
by (iii) investigating buildings prior to their demolition. For (i) and (ii) many studies exist and
numerous examples can be given:
Görg (1997) considered legal requirements and norms valid at the time of construction to develop a
model for predicting future demolition wastes, whereas Kohler et al. (1999) investigated buildings in
Germany on a national level in terms of the mass flows and the costs based on macro-economic input-
output data and by assigning specific mass flows to the buildings, building parts and materials.
Gruhler et al. (2002) and Deilmann and Gruhler (2005) defined and characterized different building
types in terms of the material and the energy intensity, and Baccini and Pedraza (2006) developed the
ark-house method to assign different material intensities to buildings of different ages and usage. This
approach was also applied by Lichtensteiger and Baccini (2008) to determine bulk materials in
buildings (minerals, steel, and wood), and by Wittmer and Lichtensteiger (2007) to calculate the
anthropogenic copper stock in buildings.
The studies mentioned greatly contributed to the investigation of building stocks and identified some
of the major challenges when characterizing the built-in materials by applying methods (i) and (ii).
Using information about the materials that have been used to construct buildings (method i), one often
neglects or underestimates renovations and technical upgrades during the use phase, which usually
increase the complexity of built-in materials. For example, Krook et al. (2011) described the
installations that are out of use. They often stay in hibernation and are not recovered during technical
upgrades, although new installations are added, which increases the material content compared with
the initial state. An idea about the relevance of renovation work is provided by Bergsdal et al. (2007)
for the CDW sector in Norway. By contrast, studies that estimate the dynamics of the building stock
based on data from the demolition waste (method ii) face the challenge of under- or overestimating the
content of so-called trace materials (e.g. copper). This problem lies in the difficulties of accurately
sampling high quantities of waste streams of heterogeneous composition and in the quality of the
6
waste stream data as reported by demolition contractors and collected by authorities (Blengini, 2009;
Cochran et al., 2007; McCauley-Bell et al., 1997). In addition, the relevant information of the detailed
location of valuable trace materials within a building is often not provided in studies that present
highly aggregated data sets. Therefore, the authors suggest the application of different methods to
investigate the building stock, particularly for trace materials with a share of less than 1%
(Lichtensteiger and Baccini, 2008).
Considering these challenges, it is assumed that an end-of-life investigation such as, for instance,
performed by Clement et al. (2011) gives the most accurate results regarding building stock
characterization and thus about materials potentially available for later recycling. The thesis at hand,
therefore, developed a method allowing the quantification of the materials actually in stock just before
the buildings are demolished.
3.2. Modelling materials in the building stock
Besides studies about the material composition of individual buildings, the last decade has exhibited
an increasing interest in the existing building stock (Kohler and Hassler, 2002). Studies have
determined the gross volume, the built-in materials, the age as well as renovation intervals of buildings
in order to estimate the demand for heating and cooling of buildings or renovation requirements of old
structures (Dall et al., 2012; Fabbri et al., 2012; Hu et al., 2010a; Kavgic et al., 2010; Pauliuk et al.,
2013; Sandberg et al., 2011; Sandberg and Brattebø, 2012). Within the life-cycle of a building, all
these examples can be allocated to the use phase.
Another way to look at the building stock is by examining the end-of-life phase, particularly by
projecting the quantity and quality of construction and demolition (C&D) wastes and subsequently
considering the building stock as a future anthropogenic resource deposit for secondary raw materials
(Bergsdal et al., 2007; Hashimoto et al., 2009; Sartori et al., 2008). Studies on the building stock focus
on a multitude of issues: different construction materials, building characteristics, a consideration of
different scales (cities, countries), and the use of different data sets (e.g. statistical data, GIS data sets).
In Austria, the built environment has so far only been estimated on a national level, based on available
statistical data by Stark et al. (2003). Their approach combines different statistical data about the use
of construction materials (mineral, organic, metallic). On a city, federal state or cantonal level, no
studies using statistical data have been carried out in Austria, in contrast to other countries. Schneider
and Rubli (2007) and Stäubli and Winzeler (2011) determined the stock of minerals built-in in
infrastructure and buildings in the urban canton of Zürich, Switzerland, by a dynamic model using
historical data sets. Other materials than minerals, e.g. plastics and metals, are not considered in their
model. Materials in lower concentrations like copper are targeted by Bader et al. (2011) for
7
Switzerland and Kral et al. (2014) for the cities of Taipeh (Taiwan) and Vienna (Austria). Both
examples consider not only the material stock present in buildings, but also in infrastructure, electronic
devices, and other items. This wide and different focus leads to a loss in detail when it comes to the
sole investigation of the building stock. Lichtensteiger and Baccini (2008) and Baccini and Pedraza
(2006) explored the national building stock in Switzerland using the ark-house-method mentioned,
where buildings are characterized by utilization and construction period, which would potentially
allow quantities of demolition waste to be predicted. Hu et al. (2010b) and Hu et al. (2010a) developed
a dynamic building model based on statistical data for the urban housing stock in rural and urban
China in general and Beijing in particular, focusing on C&D wastes as well as iron and steel.
Hashimoto et al. (2007) and Hashimoto in a subsequent study (Hashimoto et al., 2009) projected not
only the future demand of construction minerals in Japan, but also the amount of C&D waste minerals
that can serve as potential secondary raw materials for the construction industry. Their approach is
based on input-output analysis and statistical data on stocks (i.e. total floor area). Bergsdal et al.
(2007) projected the amount of C&D waste from construction and renovation and related demolition
activities in the residential sector in Norway for a period of 23 years (1995-2018) based on statistical
data on the building stock (floor area, age) and historical building materials consumption. They
considered the most relevant materials: minerals, plastics, metals, and wood. On a higher scale,
Wiedenhofer et al. (2015) modelled stocks and flows of minerals for residential buildings and
transportation networks in the EU 25 based on statistical data, on floor sizes and building age, on the
one hand, and specific material composition for different residential buildings, on the other hand.
With a few exceptions, most of the studies are based on statistical data sets either focusing on specific
material fractions (i.e. minerals, metals) or on selected building types (i.e. residential buildings, but not
industrial or commercial buildings). The reason for the latter might be the better availability of
statistical data. Even though the use of statistical data sets, as practised in most of these studies, allows
the stock size as well as the gross-generation of C&D wastes (potential secondary raw materials) to be
estimated, it does not allow a detailed localization of both the material stock as well as the generation
of C&D wastes. By employing GIS-data, it is possible to overcome this shortcoming. However, only a
few studies using GIS for the purpose of determining and localizing the building stock as well as the
generation of C&D wastes as potential secondary raw materials exist so far. A spatial and temporal
assessment of construction material in buildings and infrastructure was carried out by Tanikawa and
Hashimoto (2009) in two urban areas of 8 km² and 11 km² in Japan and the UK, using a 4d GIS
database that includes not only spatial (size and location of a building), but also temporal information
(i.e. year of construction of a building). A recent study of Tanikawa et al. (2015) presents a project on
how to map construction material of buildings and infrastructure by using GIS in Japan. Meinel et al.
(2009) worked on the establishment of a database about buildings in Germany based on building data
8
obtained from topographic maps, digital maps and statistical data. The system is intended to be used as
a nationwide monitoring system of settlement and open space development. Marcellus et al. (2012)
utilized GIS to track construction material stocks and flows of LEED certified buildings in
Pennsylvania and Philadelphia (US) in order to foster material recycling efficiencies. For the same
region Marcellus‐Zamora et al. (2015) used GIS to characterize buildings described by land use type.
Gruen et al. (2009) described plans of how 3D building and city models can be generated from images
and semantic sources and how they can be integrated into GIS systems.
All of the studies mentioned show the strengths of applying GIS datasets in the determination of
selected parts and features of the building stock in cities and countries. What some of them point out is
the difficulty of combining different GIS datasets (e.g. one dataset on spatial information and another
dataset on temporal information), on the one hand. On the other hand, the lack of information on the
most important attributes of buildings when it comes to the determination of the stock in buildings
such as, for instance, materials present in low concentrations (like metals and plastics) is also
problematic. This thesis presents an approach for analysing the buildings structure and the material
stock in buildings in Vienna.
3.3. Demolition waste generation
From a waste management perspective, knowledge about waste materials leaving the building stock
through demolition activities is important. Demolition waste (DW) from buildings plays a crucial role
in waste management as the material composition is heterogeneous compared to civil infrastructure
and because in cities buildings rather than civil infrastructure are the main drivers of material
consumption, accumulation, and DW generation (Tanikawa and Hashimoto, 2009).
Research addressing CDW management has resulted in numerous publications in recent years (Yuan
and Shen, 2011), of which some have already been mentioned. The focus of studies ranges from single
buildings to global material flow studies of different materials, which in many cases also consider the
built environment. Studies focusing on CDW generation include the work of Hashimoto et al. (2009)
where material stocks in buildings and infrastructure and occurring waste streams are estimated for
Japan. For the calculation of waste from buildings, the demolished floor space and material per m² of
floor space are used. Tanikawa and Hashimoto (2009) developed the 4d GIS to investigate spatial and
temporal characteristics and changes in the accumulation of material in buildings and infrastructure.
Bergsdal et al. (2007) presents a procedure to project waste flows entering the Norwegian waste
management system. A national model of stocks and flows of buildings and materials as well as
Monte Carlo simulation is, therefore, used. Fatta et al. (2003) estimates quantities of CDW for Greece
based on the number of demolition licenses and certain assumptions about the amount of waste
9
generated per m² of demolished building area. For the region of Lisbon in Portugal, De Melo et al.
(2011) describe how CDW generation can be estimated based on construction activity and waste load
movements of different stakeholders involved in construction activities in the region. Solís-Guzmán et
al. (2009) developed a model based on the investigation of the bill of quantities of 100 residential
projects to estimate the waste volume generated through demolition and to calculate costs arising for
deposition of the waste. In a building scale case study, used materials are identified and resulting
waste volumes calculated. Cheng and Ma (2013) based their CDW estimation on building information
modelling (BIM) in Hong Kong. Their system can extract information about materials in the building
and estimate waste amounts to efficiently plan recycling and reuse and to calculate disposal fees. For a
large scale region in the United States, Cochran and Townsend (2010) used a material flow analysis
approach to estimate CDW based on assumptions about historically used building materials and
average lifetimes of the materials.
While in many of the studies mentioned a lot of effort is put into developing models to simulate
material stocks and flows in the built environment, the basis of the calculation, e.g. material
composition of buildings, often seems not to be considered of equal importance. Moreover, the actual
lifetime of buildings depends on numerous factors (e.g. materials used, construction method, external
factors, heritage protection), making reliable predictions about average lifetimes almost impossible
and hence putting into question the overall applicability of buildings models based on lifetime
assumptions (Kohler and Yang, 2007). In addition, the decoupling of CDW management from its
regional context appears inappropriate as long as the actual management practice remains at a local
level. In order to generate data on the amount and quality of DW occurring in a city, detailed
knowledge of the material composition of buildings, on the one hand, and of the demolition activity,
on the other, is required. As shown in Kleemann et al. (2014), however, the quality (incl.
completeness) of statistical data on the material composition of DW from buildings is often poor. With
regard to data on the demolition activities, only rudimentary documentation is frequent. Consequently,
alternative approaches to generate data on DW generation and composition are required in order to
assess the quality of existing data. Thus, a novel approach was applied in Vienna, aiming at estimating
the amount and composition of DW generated through demolition activities in the building sector.
11
4. Methods
This part of the thesis follows the structure of the main research questions and is therefore divided into
three sections (4.1, 4.2, 4.3). Figure 1 shows a schematic overview of the method of this thesis,
showing that all areas are related and based on one another.
Figure 1: Schematic overview of the information used in this research to quantify the material stock in
buildings and building demolition waste.
4.1. Specific material intensities for different building categories
The material composition of a building depends, among other characteristics, on the utilization and
construction period of a building. Viennese Wilhelminian style houses, for example, are usually
constructed with bricks and are characterized by ceiling heights of between 3.5 and 4.5 meters (m).
Residential buildings from the 1960s and 1970s are mainly constructed from concrete with lower
ceiling heights (<3 m), whereas newer buildings are often more complex with regard to material
composition, and contain higher amounts of plastics and composite materials (e.g., insulation). In
order to generate specific material intensities for different building categories in Vienna, various
approaches to evaluating the material composition of single buildings with different characteristics
were applied. The building categorization itself was conducted based on available information about
the building structure (construction period and utilization) and is explained in chapter 4.2.
4.1.1. Case Studies
To collect information about the material composition of buildings, several buildings in Vienna were
selected and investigated with respect to their material composition. The buildings investigated
feature different characteristics regarding their size, age and utilization and are, therefore, different in
their material composition. All of the investigated buildings are at the end of their life, meaning that
12
they are not in use anymore and about to be demolished. Through that, all adaptions and technological
upgrades during the lifetime of the buildings can be considered. Furthermore, it is possible to apply
methods of destructive investigation to assess the material composition of certain building parts during
the on-site inspection. Table 1 summarizes the main characteristics of the case study buildings.
Table 1: Buildings investigated as case studies and their main characteristics
Case
study
Construction
year
Main construction
material
Gross volume
(GV m³)
Gross floor
area (GFA
m³)
Utilization Demolished
during Project
CS1 1960 Concrete 60,000 18,000 Residential yes
CS2.1 1870 Bricks 13,000 2,800 Commercial yes
CS2.2 1870 Bricks 18,000 3,800 Commercial yes
CS2.3 1870 Bricks 16,000 3,600 Commercial yes
CS2.4 1870 Bricks 15,000 3,400 Commercial yes
CS2.5 1960 Bricks/Concrete 7,200 2,200 Commercial yes
CS2.6 2003 Concrete 11,000 2,500 Commercial yes
CS3 1900-1930 Bricks 21,000 3,900 Industrial yes
CS4 1878 Bricks 40,000 10,000 Residential yes
CS5 1859 Bricks 3,700 1,100 Residential no
CS6 1953 Bricks/Concrete 20,000 7,100 Residential yes
CS7 1976 Concrete 150,000 39,000 Commercial no
CS8 1925 Bricks 1,500 440 Residential yes
CS9 1979 Concrete 97,000 26,000 Commercial no
Available Documents
Available documents are collected from the building owner or the building inspection department to
be analysed with regard to information about the material composition of the building. Documents can
be of different quality and, therefore, also the collectable information varies. Usually construction
plans are the main source of information. For some of the buildings investigated a waste management
concept for demolition sites (Abfallkonzept für Baustellen) is available. However, no information
about estimated amounts of material is included in these documents. The report on the investigation of
pollutants in buildings before demolition (Schadstofferkundung von Gebäuden vor Abbrucharbeiten)
gives valuable information, if available, especially with regard to hazardous substances. In cases
where no information is accessible via the building owner or contractor, the building inspection
department can draw up the construction files, containing construction plans of the respective
buildings under investigation.
The construction plans are the basis for the material quantification of each building. The dimensions
and composition of building components such as foundations, walls, ceilings, partitions, roofs, etc. are
13
measured from the plans and further processed. Based on the volume and density of the respective
materials, the mass is calculated. For composite materials such as reinforced concrete or brick work,
the share of the different materials (steel & concrete or mortar & brick) is estimated. Reinforcement
steel is quantified as a volumetric share of 0.5% of the volume of the reinforced concrete (Bergmeister
et al., 2009) and can therefore also be quantified based on the construction plans. Mortar in the
demolished buildings´ masonry is quantified assuming a volumetric share of 19%, a value which is
based on on-site measurements. The construction plans are further used during the on-site
measurements and selective sampling, to count windows and doors and to assign specific attributes to
certain areas of the buildings. Additionally, the construction plans are compared to the actual building
to record possible errors.
On-site measurements and selective sampling
In most cases only the main building materials (e.g. bricks and mortar, concrete and certain amounts of
wood and steel) can be quantified based on the constructions plans. Of note is that building materials
of low concentration from installations or fittings can usually not be covered through the analysis of
construction plans. To collect data on these materials, on-site investigations are carried out. During the
on-site investigation, data about built-in materials are collected through measurements and selective
sampling, which included the dismantling, weighing and measuring of components such as windows,
doors, partitions, ceiling suspensions, floor and roof constructions, wires, pipes, etc. Depending on the
conditions on-site, the following strategies are applied:
Sampling of representative areas: As most buildings have certain reoccurring structural
characteristics (e.g. dwelling units, floors, façade structures, roof structures), representative
areas are investigated, if appropriate, and the overall material amount is projected based on
these investigations.
Investigation of rising and distribution mains: In many cases sanitary and electrical
installations are centrally located in buildings. To efficiently generate data about the majority
of installations, the number and dimensions of installations as well as the building height
(rising mains) or length (distribution mains) is used to calculate the associated materials.
Investigation of the material composition of typical units: Doors, windows, partitions, or
heaters in a building are often similar in their design and material composition. Therefore,
common building parts are investigated and the overall materials projected by using the
number of units.
Notional sections: In areas with a lot of installations (e.g. central heating room), notional
sections were set length- and crosswise (in some cases also horizontally) through rooms.
Along the sections, all orthogonal running wires, pipes and other installations are counted and
14
dimensions documented. For the length of the installations the dimensions of the particular
room (perpendicular to the cross section) is assumed (Figure 2).
Material inventory of specific areas of buildings: Specific features of a building are
documented and attributed to certain building parts or areas.
Figure 2: Notional sections (displayed on the left) in areas with many installations (right). Based on Figure
1 in the 1st Paper.
Based on the data collected, information about the different built-in materials is aggregated, and the
mass is calculated based on volume, area or number of the particular material and based on data about
its specific density or weight. Relevant information on density of building materials or weight of
installations is either obtained by sampling or from literature about construction methods and design
norms.
In the case that the building investigated is demolished within the duration of the project, results
generated in this study are compared with official reports about the waste amount generated by the
contracted demolition company after the demolition. The reason for this comparison is to have another
potential source of information about the material composition of demolished buildings.
4.1.2. Construction plans of demolished buildings
In order to investigate a larger sample of buildings in terms of their material composition, construction
files of already demolished buildings are used. These construction files of buildings reported to have
been demolished in Vienna in the years 2013 and 2014 are available for research purposes through the
municipal building inspection department (MA37). The information content of the construction files
varies considerably and mainly construction plans are utilized for the analysis of the material
composition. Depending on the quality of the plans, building materials used for wall, ceiling and
sometimes also for roof or floor construction can be determined. Data gaps, mostly regarding materials
of low concentration, are filled by data from the case studies analysed in detail. 40 additional buildings
are investigated using this data source.
15
4.1.3. New buildings
In order to get information about the material composition of buildings recently constructed in Vienna,
different data sets are utilized. Mainly life cycle assessments (LCA) of residential buildings are used,
but also tendering documents, construction plans and accounting documents of new construction
projects are utilized. The LCA data are obtained from the Austrian Institute for Healthy and Ecological
Building (IBO), whereas data from tendering documents, construction plans and accounting
documents are collected in accordance with the respective building owners and contractors. The
inclusion of this data was necessary as only few newer buildings could be covered by analysing the
construction files of demolished buildings. New buildings are of importance because building material
used today will represent the DW generated in the future. Materials such as gypsum, agglutinated
insulation materials and other composite materials are believed to be particularly problematic in that
they cause problems with regard to the recyclability at the end of the use phase.
4.1.4. Literature
In order to put the collected data in perspective to previous studies, different literature from Europe
was reviewed and, if suitable, used for the generation of specific material intensities for the different
building categories in the city of Vienna. The work of Görg (1997), Schulze et al. (1990), and Gruhler
et al. (2002) provide material intensities for multifamily houses in Germany, whereas Rentz et al.
(2001) focus on residential buildings as well as administration buildings. Wedler and Hummel (1947)
focus on debris management after WWII in their work but also give detailed information about the
material composition of characteristic buildings in Germany’s major cities. Baccini and Pedraza
(2006), Bader et al. (2011), and Lichtensteiger and Baccini (2008) emphasize copper in their studies
based on the investigation of single houses, multifamily houses, commercial buildings and industrial
buildings. In the study of Blengini (2009), one residential building is assessed in detail during its
demolition with regard to its material composition. Müller (2013) highlights the wide range of the
overall material intensity of different building types.
After collecting the data about the material composition, the mass (kg) for each material is calculated
and divided by the gross volume (GV) of the respective building in order to determine comparable
material intensities per volume (kg/m³ GV). The gross volume was chosen as a reference value as it is
compatible with the procedure of analysing both the buildings’ structure (chapter 4.2) and the
demolition activity (chapter 4.3). In other literature, the gross volume as well as the gross floor area or
the used area can be found as a reference value.
16
4.2. Building structure and total material stock of buildings in Vienna
To estimate the material stock in Viennese buildings, both the specific material intensities of different
building categories (chapter 4.1), and information about the building structure are used. This part of
the thesis, therefore, aims at generating a dataset with information about size (gross volume),
construction period, and utilization for each building in Vienna. Thus, GIS data containing different
information are spatially joined using Esri software. The following data are used to carry out the
analysis:
1. A map of the municipal department for city surveying, MA 41 (2013), includes information
about the area and the relative height of each building or building part in Vienna. It is based
on terrestrial surveying and therefore provides the most accurate data with regard to
buildings. This dataset is therefore used as target data set, meaning that other information
about buildings (such as construction period, utilization) is added to this data set. The relative
height in the multipurpose map is defined as the distance between ground level and eaves,
meaning that basement floors and the roofs of the buildings are not represented. The specific
material intensities of buildings (chapter 4.1), which are, however, later combined with the
GIS data, are related to the whole building (including basement and roof). To include
basements and roofs in the GIS data, a small survey among experts is carried out to estimate
the average height of basements and roofs (for each building category). These values are
added to the relative height of the buildings. The data set used for this research was obtained
in 2013. According to the MA 41, the data are updated continuously and are, at most, 3 years
old.
2. A data set provided by the municipal department for urban planning and land use, MA 21
(2013b), is spatially joined to the target map (1) to assign information about utilization and
construction period to each building. These data are constantly updated and are maximum 10
years old.
3. Data gaps with regard to utilization and construction period exist for some areas of the city
when using only 1. and 2. To fill these gaps, data from the municipal department for city
development and planning, MA 18 (2013) on the construction age of building blocks as well
as the generalized zoning plan (utilization) of the MA 21 (2013a) are used. These data sets
only contain information on a building block level and are therefore less accurate. However,
for buildings where no other information is available, this approximation on a larger scale
represents the only possibility for categorization and characterization.
4. Another data source containing information about the utilization and construction period of
buildings is the building and flat index of the municipal building inspection department,
17
MA 37 (2013), which is continuously updated. To improve the data quality, especially with
regard to newer buildings (constructed after 2008) this data set is added to the target data set.
Based on the available GIS data on construction period and utilization, 15 building categories have
been defined. They distinguish five different construction periods (before 1918, 1919–1945, 1946–
1976, 1977–1996, and after 1997), which correspond to the available GIS data, and three different
categories for utilization (residential, commercial, and industrial). Existing information about the
utilization has been strongly aggregated from 78 categories originally. This step was taken because
material intensities are not available for all 78 categories, and differences in the material composition
are most likely not significant among all of the categories.
In order to quantify the material stock in buildings, the information about the building structure
generated is combined with specific material intensities (chapter 4.1). Figure 3 summarizes the
information flows of different data sources used for the analysis of the building stock in Vienna and its
material composition (including the generation of specific material intensities for different building
categories).
Figure 3: Information flows of different data sources used in this research. Based on Figure 1 in the 2nd
Paper.
18
4.3. Waste generation through building demolition
From a waste management perspective it is especially interesting to know about the waste flows
leaving the buildings’ stock e.g. through demolition activities. Besides the amount, it is also important
to consider the quality and composition of waste here. This part of the thesis, therefore, focuses on the
development of a method to collect data on building demolition in Vienna.
Two data sources, (i) statistical data and (ii) image matching based change detection data, are utilized
in this context. The demolition activity of different building categories is expressed in demolished
gross volume (m³). The results of analysing the two data sources are compared to identify
inconsistencies. Subsequently, specific material intensities (chapter 4.1) are assigned to the
demolished building volume of the respective building categories (chapter 4.2) to assess the amount
and composition of the demolition waste generated. All data sources are processed in a GIS model.
4.3.1. Statistics-based analysis of the demolition activity
Any demolition activity taking place in Vienna has to be reported through a notification to the
municipal building inspection department (MA 37). Only in heritage protected areas of the city is
permission necessary to carry out a demolition. The notifications include the address of the object
being demolished and usually information about the contracted wrecking company. No information
about any characteristics of the respective building such as size, construction period, utilisation, or
construction type is provided. In order to obtain data about these characteristics, the target data set
generated as described in chapter 4.2 is used. Address points of the municipal department for urban
planning and land use (MA 21) are used in order to locate the notified demolition sites (addresses in
the demolition statistics). Subsequently, orthophotos of different years, which are processed from the
annual aerial image campaign of the municipal department for urban survey (MA 41), are used to
verify whether a demolition took place. Furthermore, it is possible to determine which parts of a
building were subject to demolition. The GIS data allows size, construction period and utilization of
the demolished buildings to be determined.
4.3.2. Demolition activity analysis based on remote sensing image matching
Observations by members of the Institute for Water Quality, Resource and Waste Management of the
Technische Universität Wien indicated that the demolition statistics (chapter 4.3.1) did not consider all
demolitions carried out in Vienna. To verify the statistics based data on demolition activities in the city
of Vienna, data based on so-called image matching change detection is used. The method developed
by the MA 41 allows negative changes in the building stock (demolitions) to be detected by comparing
two digital surface models (DSM). A DSM is a digital height model, which contains the earth´s
surface but also all natural and man-made objects (e.g. buildings, vegetation). The DSM is derived
through image matching, where corresponding points in overlapping images are identified. Yearly
19
orthophotos provided by aerial image campaigns of the city of Vienna are the basis for the image
matching process and for generating the DSMs of different years. To generate the specific data needed
for this project, a height difference model of two DSMs of 2013 and 2014 is calculated by subtracting
the older surface model from the more recent one (Figure 4). Doing so, areas with higher height values
in 2013 than 2014 (e.g. demolished buildings) can be detected. During the process all areas with
values equal or lower than -3 m (height) and higher than 30m² (area) were converted into polygons in
order to create single objects for the further processing in GIS. Small objects, especially cars, are
excluded in this procedure, whereas small garden cottages typical for some areas of the city of Vienna
are still considered. To exclude remaining vegetation in the polygon data, the so-called Normalized
Difference Vegetation Index (NDVI) is used. Subsequently, the final polygons (≥30m²; > 3 m) are
visually checked using the orthophotos of 2013 and 2014 in order to verify whether and to what extent
a demolition took place. In cases in which a demolition of a building can be verified, the GIS Data
(chapter 4.2) is used to identify size, construction period, and utilization of the buildings being
demolished.
Figure 4: Schematic illustration of the generation of the height difference modes based on two DSMs
Finally, the resulting data sets (GV, construction period and utilization of demolished buildings) of
both analyses (statistics; image matching change detection) are combined with specific material
intensities (chapter 4.1) of the respective building categories (chapter 4.2) to draw conclusions on the
amount and composition of DW from buildings.
21
5. Results
Through the application of the methods described it is possible to generate data on the material
composition of different building types and to derive specific material intensities of different building
categories. The building categories are defined based on the available GIS data, which is used to
determine the building structure of the city of Vienna. The building structure allows the determination
of the overall material stock in buildings in Vienna. By combining data on the demolished building
volume with the specific material intensities, it is further possible to estimate amount and quality of
waste arising from building demolition in Vienna.
5.1. Case studies
A major part of the thesis was the development, planning, and implementation of a method to assess
the material composition of building prior to demolition. The results are provided in the following
subsections.
Applicability of the method
Information from the construction plans allows one to determine the volume of walls and floors. Based
on this information, the bulk materials (concrete, bricks, sand, gravel, etc.) can be quantified using
data from the literature about the density and composition of different building materials. Depending
on the quality of the plans, details about the floor and the roof assemblies are available. In this case,
filling or lining material and materials for impact sound insulation and flooring can be quantified.
Information about the installations (heating, sanitation, electricity) is only sporadically available in the
existing documents. Hence, they cannot be used to quantify the associated amounts of relevant
materials such as copper, steel or insulation material. For reinforced concrete buildings, the
reinforcement steel as a major contributor can be quantified based on the construction plans. Similarly,
steel girders in large brick buildings are usually drawn in the plans.
Documents besides the construction plans can usually only be used as supporting information. The
investigation of pollutants in buildings prior to demolition is not meant to provide comprehensive
information about the total material composition of the building. It mainly focuses on hazardous
substances that must be safely removed prior to the main demolition work. It contains qualitative
rather than quantitative information about materials in the building. The waste management concept
for demolition sites is supposed to contain qualitative and quantitative information about the built-in
materials. However, quantitative data are rare. Based on the case studies investigated, it appears that
the waste management concept and the investigation of pollutants are conducted to satisfy the legal
22
requirements rather than being used to plan the demolition work in advance. In some cases, these
documents are part of the tender offer for the demolition work and are provided only shortly prior to
the commencement of the demolition.
The on-site investigation allows quantification of materials such as steel, copper, aluminium, plastics,
wood and other materials used in low concentration compared to the overall building mass. The
application of the methods of on-site investigation can be summarized as follows:
Sampling of representative areas: This strategy is appropriate for buildings with similar floor
design and which are similar in construction and outfit. In many cases some areas were
similar, which allowed this method to be applied.
Investigation of rising and distribution mains: Each case study is different and shows that the
electrical and sanitary installations often undergo technical upgrades and alterations during the
lifetime of a building. Focussing on rising and distribution mains helps to cover a majority of
installations with manageable effort.
Investigation of the material composition of ‘typical’ installations: windows, doors, heaters
and suspended ceilings are investigated as typical installations in all of the case studies. Some
of the measured data about heaters and suspended ceilings can be used for several case
studies.
Notional sections: For rooms with a high number and multiplicity of installations, notional
sections appear to be the most efficient approach for a structured determination of built-in
materials.
Inventory of particular areas: In case the previously described strategies are not applicable,
information about the structures and fittings are simply noted and attributed to the rooms or
areas of the building.
The case studies showed that it may be impossible to characterize the material composition of
buildings in a comprehensive way solely based on available documents. The bulk materials of the
building structure can be well quantified, but materials in low concentration compared to the overall
building mass are not quantifiable. On the contrary, an on-site investigation allows the quantification
of materials in low concentration but is not suitable to quantify bulk materials because the use of
heavy machinery would be necessary to obtain information about the composition of walls or ceilings.
The effort needed to carry out the document-based and on-site investigations of the buildings is high.
Data collection through the analysis of documents and on-site investigation as well as the evaluation of
the data results in a workload which cannot be realistically maintained for a high number of buildings.
From the authors’ perspective, however, the approach is appropriate to generate information about the
23
specific content of materials of low concentration present in buildings. Despite their low amount in
volume, their impact on the overall material value of buildings (e.g. copper) or their hazard potential
(e.g. asbestos, tar) in the generated demolition waste can be significant.
Material composition of buildings
shows the material composition of the case studies expressed in kg per 1m³ of gross volume of the
building. Empty cells indicate that this material was not present in the respective case study. For each
case study, the material origin and the material mass calculation method were extensively documented
to be able to trace back the results. The results of the case studies indicate a dominant share of mineral
matter, which amounts to between 94% and 98% of the total mass. In comparison, the share of all
other materials present is relatively low. However, their total amounts can be significant depending on
the overall size of the building.
The overall material intensity varies mainly due to the design, geometry and compactness of the
buildings. Relatively low ceiling heights, and small dwelling units usually result in a high overall
material intensity (e.g. CS1). Most of the older buildings have a similarly high material intensity of
between 400 and 440 kg/m³ gross volume. The room height in buildings from this time of up to 4.5 m
is compensated by a wall thickness of approximately 1 m at the basement. CS5 shows the highest
overall material intensity, with thick brick walls and relatively small rooms. This reason and a
relatively small gross volume explain the comparably high overall material intensity. The comparably
new building of CS2.6 has a significantly lower overall material intensity. The ceiling height is
approximately 4 m, and the overall construction height with two floors relatively low, reducing the
required wall thickness at the base. In addition, most partitions inside the building are plasterboard
constructions with concrete columns, which support the ceiling. CS3 consisted mainly of production
halls, which imply a relatively low overall material intensity through its construction style with huge
areas, and ceiling heights of up to 9 m. The relatively low overall material intensity of CS7 can be
explained by the existence of halls and parking lots in the lower and basements floors and its concrete
frame construction.
Unsurprisingly, the share of steel tends to be higher when reinforced concrete is used as the main
building material (CS1, CS2.6, CS7, CS9). However, large buildings, which are mainly made of bricks
(CS2.1, CS2.2, CS2.3), may also be characterized by relatively high steel content, particularly if
massive steel girders are used. In CS3, comparably large quantities of steel were used for the roof
construction. The steel content of CS5 and CS8 is notably low, particularly in comparison with coeval
buildings. In CS5 the ceilings were constructed of massive timber beams instead of steel girders,
which explain the high wood and the low steel content.
24
The amount of aluminium varies considerably from building to building as it strongly depends on the
design of the structural elements and installations. In CS1, aluminium was used as a frame for the
facade and partly for doors and windows; in CS2.6, aluminium was used as part of the roofing. CS7
was equipped with aluminium windows and doors throughout. In CS9, in addition to the aluminium
windows, parts of the façade were shielded with solid aluminium.
In the case studies investigated, copper was mainly used for electrical and sanitary installations. The
specific intensities are notably comparable and in the same order of magnitude for all case studies,
except for CS3. This particular building was uninhabited for approximately two years before it was
demolished. During that time most of the copper disappeared through unknown paths, supposedly
informal recycling collectors.
25
Ta
ble
2:
Co
mp
ari
son
of
ma
teria
l co
mp
osi
tio
n o
f th
e ca
se s
tud
ies
(kg
/m³
gro
ss v
olu
me)
26
Location of different materials
Beside the quantification of the materials in buildings, the location and concentration of materials are
of interest. Figure 5 exemplarily shows the origin of different materials for CS1. The knowledge about
the location of materials in a building, as shown here, can be helpful to plan the recycling processes
effectively and to estimate the effort needed to recover certain materials before the building demolition
(selective deconstruction). The separability largely influences the technical and organizational
processes necessary for material recycling. Regarding valuable materials, Figure 6 further
distinguishes the location of steel, aluminium and copper exemplarily for CS1. Most of the steel is
reinforcement steel, which has to be separated from the concrete using hydraulic crushers. The other
steel stems from different installations in the upper floors and the basement floor. Most of the
aluminium is located in windows and doors in the upper floors, and the facade construction comprises
a significant amount of aluminium. Copper is mostly found in the electric main supply lines in the
basement of the building. A smaller amount is located in electrical installations in the upper floors.
Figure 5: Origin of materials for CS1. Based on Figures 3 and 4 in the 1st Paper.
27
Figure 6: Location and quantity of valuable metals (steel, aluminium, copper) in CS1. Based on Figure 5
in the 1st Paper.
Disposal reports of demolition contractors
The comparison of collected data with official reports of the demolition contractor show different
results. In some cases the values are consistent, especially with regard to the mineral fractions
(brick/mortar, concrete, etc.) as well as wood; in other cases the numbers totally diverge. Besides a
lack of documentation, the main reason for some of the discrepancies is that large amounts of material
are left on the site after demolition. It is common practice in Vienna to backfill demolition material
into the basement floors up to ground level. In these cases only a part of the waste is covered by the
disposal reports, which leads to highly divergent numbers. With regard to materials present in lower
concentrations compared to the overall mass of the building, the numbers are even less consistent. It is
notable that valuable fractions, such as copper and aluminium, are neglected in the documentation as
on-site visits during the demolition work showed that these fractions were well separated but not
mentioned in the reports. Copper is not accounted for in any of the records of the demolition
companies. Therefore, the documentation about waste streams provided by the wrecking companies
cannot be utilized as a reliable source of information about the material composition of demolished
buildings.
5.2. Specific material intensities of different building categories
In order to generate specific material intensities of different building categories, the data of the
material composition of single buildings is comprised in one data set. The data shown in (material
composition of case studies) is expanded by the results of the analysis of construction files of
demolished buildings, new buildings, and the reviewed literature. To make the information
comparable, the material data of all sources is put into relation of 1m³ gross volume of the respective
building. Each building is associated with one of the 15 categories (chapter 4.2). In Table 3, the
specific material intensities of mass (kg) per GV (m³) are shown in an aggregated form. The material
categories (mineral, organic, metal) are further divided into specific materials such as concrete, bricks,
28
glass, asbestos, wood, plastics, bitumen, iron, aluminium, and copper. Detailed information about the
data set on the specific material intensities of different building categories is provided in the
supplementary information of the 2nd Paper.
Table 3: Specific material intensities (kg/m³ GV) of different building categories in Vienna (rounded to
two significant digits). Based on Table 1 in the 2nd
Paper.
Period of
construction Utilization
Mineral
materials
Organic
materials Metals Total
-1918
Residential 390 19 3.1 410
Commercial 430 3.7 4.4 440
Industrial 280 5.8 8.8 300
1919-1945
Residential 410 13 4.8 430
Commercial 340 7.1 6 360
Industrial 320 28 5.8 350
1946-1976
Residential 430 6.5 7.3 450
Commercial 350 7.6 5.7 360
Industrial 340 7.6 13 350
1977-1996
Residential 430 6.7 7.1 460
Commercial 380 1 13 400
Industrial 170 1 15 180
1997-
Residential 380 10 15 410
Commercial 320 5.7 10 340
Industrial 290 5.6 13 310
The material intensities of the building categories are calculated based on a different number of
information sources, meaning that for some building categories specific material intensities have been
derived from many different data sets, whereas for one building category (after 1997, industrial,
representing 0.5% of the total GV) no data were available at all. For this building category a material
mixture of industrial buildings from 1977-1996 and residential and commercial buildings from after
1997 was assumed. This problem is further considered in the discussion (chapter 6.1)
29
5.3. Building structure and total material stock of buildings in Vienna
Building structure
The results of the analysis of the building structure in Vienna are based on a data set of approximately
580,000 dates. Each date represents a building or a part of a building. The splitting up of buildings into
several parts is necessary in cases where parts of one building constitute different heights. According
to the municipal department for economy, labour and statistics, MA 23 (2011), there are
approximately 165,000 buildings in the city of Vienna. The overall volume of buildings in Vienna
amounts to 860 million m³ GV, with a large share being used for residential purposes (540 million m³
GV). Regarding the age of the building stock, it is noticeable that the largest share of buildings was
constructed before 1918 (290 million m³ GV). Due to the fact that the city of Vienna experienced a
boost in construction activity during the second half of the 19th century (which went along with a
quadrupling of the population: 1850: 551.000 inhabitants and 1916: 2,240,000 inhabitants, according
to Statistik Austria (2014), figures refer to the present city border), a significant share (about 25-30%)
of Vienna´s buildings stock was erected between 1850 and 1918. Figure 7 illustrates the distribution of
the GV among the different building categories. The data set obtained from the city of Vienna refers to
the year 2013.
Figure 7: Gross volume (m³) of different building categories in the city of Vienna in the year 2013. Based
on Figure 2 in the 2nd
Paper.
0
100
200
300
400
500
600
Residential Commercial Industrial No
Information
Mil
lion
m³
gv
Before 1918
1919-1945
1946-1976
1977-1996
After 1997
No information
30
Material stock
By combining information about the building structure (volume per building category) with specific
material intensities of different building categories (chapter 5.2), the overall material stock in
buildings in Vienna in the year 2013 was calculated to 380 million metric tons (t). The bulk of the
material (>96%) is mineral, mainly represented by concrete (150 million t), bricks (130 million t) and
mortar (50 million t). Organic materials and metals constitute a very small share of the overall material
stock in buildings, of which wood and steel are the major contributors (Figure 8).
Figure 8: Material composition of the total building stock in the city of Vienna in the year 2013. Based on
Figure 3 in the 2nd
Paper.
Considering the Viennese population of 1.8 million, the material stock in buildings per capita amounts
to about 210 t/cap. As Vienna is an economic centre of Austria with around 250,000 commuters, a part
of the building stock also services people residing outside of Vienna. This applies mainly to
commercial and industrial buildings. Table 4 summarizes per capita figures on the buildings stock in
Vienna.
Spatial distribution of materials
The combination of material information described with GIS data not only allows the estimation of the
overall material stock in buildings but also its spatial distribution. By mapping the material data on a
building level, reference values about the material composition of specific buildings can be generated.
This allows estimations of demolition waste arising during demolition works as well as the forecasting
of the value or costs connected to the disposal or recovery of demolition waste. This combination of
GIS data with specific material intensities further allows the establishment of a resource cadastre,
which illustrates the spatial distribution of materials throughout the city. Figure 9 shows the material
31
intensity of buildings in Vienna expressed as mass per built-up area of the buildings exemplary for
mineral materials.
Table 4: Per capita figures on the materials present in buildings in Vienna (t/cap) (rounded to two
significant digits). Based on Table 2 in the 2nd
Paper.
Mineral 200 Organic 5.5 Metal 3.3
Concrete 83 Wood 4.0 Iron/Steel 3.2
Bricks 70 Various plastics 0.35 Aluminium 0.045
Mortar/plaster 29 Bitumen 0.22 Copper 0.031
Mineral fill 7.8 Carpet 0.19 Lead 0.0023
Slag fill 3.4 Heraklit 0.16 Brass 0.0012
Gravel/sand 2.7 Asphalt 0.13
Natural stone 0.72 PVC 0.10
Foamed clay bricks 0.71 Polystyrene 0.076
Plaster boards/gypsum 0.74 Paper/Cardboard 0.059
Ceramics 0.42 Laminate 0.029
(Cement) asbestos 0.34 Linoleum 0.014
Glass 0.22
Mineral wool 0.21
Mineral wool boards 0.017 Total 210
Figure 9: Spatial distribution of minerals in buildings in Vienna (kg/m² built-up area). Based on Figure 4
in the 2nd
Paper
32
5.4. Waste generation through building demolition
Based on the statistics from the building inspection department (MA 37) on the demolition activity in
2013 and 2014, the orthophotos of the years between 2012 and 2014, the polygons of potentially
demolished buildings based on the height difference model of 2013 and 2014, and the GIS model from
2012 it is possible to calculate the demolished gross volume of different building categories.
Figure 10: Demolished building volume (m³ gross volume) in 2013 and 2014 based on statistics of the
municipal building inspection department. Based on Figure 4 of the 3rd
Paper.
In 2013 around 300 demolition notifications are documented by the MA 37, with a total demolished
gross volume of 1.9 M m³. In 2014 demolition notifications amount to 240, which is equivalent to a
demolished building volume of 1.5 M m³. Figure 10 shows the distribution of demolished building
volume among the building categories for the years 2013 and 2014. About 2.3% (2013) and 8.6%
(2014) of the overall demolished building volume cannot be assigned to a utilization or construction
period category (indicated by no category assignable – N/A). For these buildings it has been assumed
0
2
4
6
8
10
12
2013 2014 2013 2014 2013 2014 2013 2014
Residential Commercial Industrial N/A
x 1
00
,00
0 m
³ g
ross
vo
lum
e/a
Year of statistic and building category
-1918
1919-1945
1946-1976
1977-1996
1997-
N/A
Construction
period
33
that their age and utilization are consistent with the total demolished building volume. For buildings
for which only information about the construction period was available, the utilization pattern has
been assumed to be in accordance with the data available for the respective construction period. An
analogical procedure has been applied to buildings with information on utilization but with no data on
the construction period. Figure 10 shows that the distribution of demolished building volume among
the building categories varies considerably from year to year because very few large demolition
projects strongly influence the overall distribution.
As an alternative method to evaluate the demolition activity in Vienna, change detection data based on
image matching is used. To facilitate the data, each polygon of a potentially demolished building is
checked manually. Although changes smaller than 30m² and less than 3 m in height are excluded
computationally, many polygons do not represent buildings but e.g. train wagons, containers, trucks or
soil movement. These polygons are excluded, resulting in a reduction of polygons from around 2,600
to around 500. As the manual checking is very labour intensive, the computational procedure should
be improved when applying the method on a regular basis. Different options should be tested in this
regard. In the case that the threshold of 30m² is increased, it is necessary to check whether buildings
are neglected by doing so and what influence this might have on the final results. Another possibility
would be to exclude polygons of certain shapes (width-length-ratio) in order to detect e.g. trucks, train
wagons and containers. In Figure 11 the results based on the change detection data are shown in
comparison to the statistic based results (average of 2013 and 2014). As mentioned, the aerial imagery
used for this study was taken between July 2013 and June 2014. To make the results comparable to the
average results of the yearly statistics, the 10.5 month period between the aerial image campaigns used
in the change detection analysis was projected to one year. For the projected 1.5 months it is assumed
that the demolition activity and the shares of building categories being demolished are equal to the
actual change detection results (10.5 month period). This seems appropriate as no trend in demolition
activity could be observed analysing the notification statistics. In contrast to construction activities,
which are strongly influenced by the season, demolition works are also carried out during winter.
34
Figure 11: Gross volume of demolished buildings (given in m³/a) based on statistics of the building
inspection department (one-year average of 2013 and 2014) and image matching based change detection
(one-year equivalent of data from July 2013–June 2014). Based on Figure 5 of the 3rd
Paper.
The results based on the change detection analysis differ from the statistics results, with a total gross
volume of demolished buildings of 2.8 M m³/a and 1.7 M m³/a, respectively. This significant
difference clearly indicates that demolition activities are only partly covered by the statistics of the
building inspection department. This is also supported by the fact that some demolition projects were
observed by the project team and could not be found among the demolition notifications. As the
statistic is based on the demolition notifications, it can be assumed that a considerable number of
0
2
4
6
8
10
12
Sta
tist
ics
201
3/2
01
4
Chan
ge
det
ecti
on
anal
ysi
s
Sta
tist
ics
201
3/2
01
4
Chan
ge
det
ecti
on
anal
ysi
s
Sta
tist
ics
201
3/2
01
4
Chan
ge
det
ecti
on
anal
ysi
s
Sta
tist
ics
201
3/2
01
4
Chan
ge
det
ecti
on
anal
ysi
s
Residential Commercial Industrial N/A
x 1
00
,00
0 m
³gro
ss v
olu
me/
a
Data source and building utilization
-1918
1919-1945
1946-1976
1977-1996
1997-
N/A
Constructio
n period
35
demolition projects are not reported. However, for some building categories (e.g. residential 1977-
1996, commercial 1919-1945, industrial 1977-1996) the statistics give larger demolished building
volumes in comparison to the change detection analysis (Figure 5). This observation is explained by
the fact that the time frame of both approaches is not equal, meaning that demolition projects might
have been reported but have been demolished prior or after the aerial imagery has been taken (July
2013 and June 2014).
Figure 12 underlines the fact that a few big demolition projects have a significant influence on the
total demolished building volume by showing that 6-7% of the biggest demolition projects represent
50% of the total demolished building volume. Furthermore, the results indicate that the demolition of
smaller buildings is underrepresented in statistics.
Figure 12: Cumulative curve of demolished building volume (m³ gross volume) of demolition projects
ranked by size
By combining the GV of demolished buildings with specific material intensities for the respective
building categories, the amount and composition of demolition waste can be estimated. Table 5
summarizes the results of different approaches applied in this study for assessing the demolition
activities in Vienna (statistics versus image matching based change detection). The results of the
image matching based change detection approach imply waste generation from building demolition in
the city of Vienna of about 1.1 M t/a (or 610 kg/cap/a). This represents about 0.3% of the overall
material stock embedded in buildings (210 t/cap, see chapter 5.3). In comparison, the estimated annual
consumption of materials for constructing buildings amounts to about 1600 kg/cap/a (based on data on
new buildings of the municipal building inspection department, MA 37), indicating an annual growth
in the Viennese buildings stock of 2.9 M t/a, which represents about 0.8% of the overall material stock
embedded in buildings.
0
5
10
15
20
25
30
0 100 200 300 400 500
x 1
00
,00
0 m
³ g
ross
vo
lum
e
Number of demolition projects
Statistics 2013
Statistics 2014
Change detection analysis
Data
36
Table 5: Quantity and composition of waste from demolished buildings in Vienna (rounded to two
significant digits). Based on Table 2 of the 3rd
Paper.
Material (tons)
Demolition
activity
(volume) based
on statistics for
2013
Demolition
activity (volume)
based on
statistics for 2014
One year average
demolition
activity (volume)
based on
statistics for 2013
and 2014
Demolition activity
(volume) based on image
matching based change
detection (projected for
a period of one year)
Mineral 750000 540000 640000 1000000
Concrete 310000 240000 270000 500000
Gravel/sand 5000 3000 4000 4700
Bricks 280000 190000 230000 330000
Mortar/plaster 110000 75000 91000 140000
Mineral fill 22000 15000 19000 27000
Slag fill 7300 5100 6200 9000
Expanded clay bricks 2400 1500 2000 2200
Plaster boards/gypsum 3800 2000 2900 5100
Glass 800 540 670 1200
Ceramics 1200 780 1000 1500
Natural stone 8200 7300 7700 11000
Mineral wool 550 290 420 600
Mineral wool boards 150 56 100 140
(Cement) asbestos 820 500 660 780
Organic 21000 13000 17000 27000
Wood 15000 10000 12000 20000
Reed 420 270 340 440
Paper/Cardboard 110 60 86 81
PVC 370 240 310 490
Various plastics 590 370 480 630
Carpet 410 240 330 340
Laminate 63 39 51 57
Linoleum 54 29 42 66
Asphalt 2100 650 1400 2500
Bitumen 1700 840 1300 2400
Polystyrene 180 75 130 190
Metal 13000 13000 13000 22000
Iron/Steel 13000 12000 12000 21000
Aluminium 210 310 260 470
Copper 140 150 150 270
Total 780000 560000 670000 1100000
37
The bulk of the material (96%) is mineral, mainly represented by concrete (44%), bricks (28%) and
mortar (13%). Organic materials and metals constitute a very small share (4%) of the overall material
composition of DW from buildings, of which wood and steel are the major contributors. With regard
to the recyclability of materials, it is important to mention that the compound in which they are used is
crucial. Materials in old buildings tend to be more easily separated (e.g. bricks, mortar, wood) than
materials such as reinforced concrete or various composite materials. Currently, most of the DW from
Vienna is transported and recycled outside the city. Concrete aggregate is usually used in road
construction, and crushed bricks find application as a plant substrate or are utilized as raw material in
the cement industry. Recycling in the sense that e.g. recycled concrete is used in new concrete as
aggregate is rare. Most organic material is thermally treated, and in some cases wood is used in the
wood board industry. Due to their value, metals are recycled at very high rates.
The comparison in Table 5 makes it clear that the statistically based approach neglects a relevant part
of demolition activity taking place in Vienna and related waste streams. As all polygons of potentially
demolished buildings were checked and compared with orthophotos, an overestimation of the
demolition activity via image matching based change detection can be excluded.
39
6. Discussion
This research developed methods to generate data on the material composition of buildings through
different approaches to generating specific material intensities for different building categories in
Vienna. To further develop the database of specific material intensities, it would be beneficial to
increase the number of buildings in Vienna being investigated in terms of their material composition.
Collaboration with other research fields studying buildings but whose primary focus is not necessarily
material composition (e.g., on energetic performance of buildings, heritage buildings, building
information modelling, and so on) is, therefore, desirable.
The analysis of the building structure in Vienna and its associated material stock demonstrates that it is
possible to characterize the material composition of the Viennese building stock by combining
available data sets from the municipal authorities with collected data on the material intensities of
different buildings types. Besides the further development of the database mentioned in terms of
specific material intensities of different building categories, this building stock analysis can be
improved by enhancing the data on the buildings (e.g. adding a city-wide roof model and adding more
precise information about basement floors).
As far as a continuation of building stock estimation is concerned, the research shows to what extent
remote sensing can contribute to monitoring demolition activities in the city of Vienna. In combination
with material data of different building categories it allows the amount and quality of waste occurring
from demolished buildings to be estimated.
6.1. Uncertainties
Uncertainties arise at different stages of the research carried out and are a major point of discussion.
All three areas of investigation are subject to uncertainties at different levels. In the following the most
important issues are discussed.
Possible sources for uncertainties include the data collection during the investigation of buildings and
building parts, outdated construction plans of demolished buildings and the use of literature from other
regions. The document-based investigation, focusing on bulk materials, relies on ostensibly accurate
documents about the buildings, which implies uncertainties itself. The documents used to analyse
buildings which have already been demolished are assumed to be accurate, but might be less so than
expected. Another source of uncertainties is the use of default data, e.g. for the ratio of reinforcement
steel in reinforced concrete or of mortar in masonry. The on-site investigations are believed to give the
most reliable information about materials in low concentrations within a building. However, also here
40
uncertainties exist. The mass of pipes can serve as an example for that: As it is not feasible to check
the wall thickness of each pipe in a building, existing norms for dimensions of pipes are used.
Uncertainties with respect to the material mass per length of installation (pipe) can be assumed to be
rather low (below 5%). However, the uncertainty of the measured installation length, determined by
notional sections, can be estimated to be around 20%. Special components such as elevators, which
can account for a significant share of a material fraction in a building, cannot be dissembled within the
on-site investigation, and thus estimates together with general producer information are used. Hence, a
higher level of uncertainties exists. Estimated uncertainties of data sources for the generation of
specific material intensities are shown in Table 6.
Table 6: Estimated uncertainties for data sources used to generate specific material intensities for
different building categories.
Data source Estimated uncertainty (%)
Inspection and selective sampling 5-20
Documents case studies <5
Documents of demolished buildings (MA 37) 5-20
Data new buildings <5
Data from literature 20-50
The investigation of a higher number of case studies can increase knowledge about the material
composition of different buildings as well as associated uncertainties. The main barrier in this respect
is the high effort needed to carry out the case studies. Therefore, the existing data has been analysed to
suggest a certain number of samples to achieve a defined level of uncertainty. Selected materials are
plotted in Figure 13 to exemplarily show how uncertainties decrease with an increasing number of
samples. If an additional sample features a value outside the “range” of the existing sample set,
uncertainty can also increase again. For copper that means that the uncertainty is more than 50% when
only one sample is analysed. When analysing seven samples, and assuming that only random errors
are present, the uncertainty decreases to approximately 20% (equals 54%√7⁄ ). The material intensity
for bricks in buildings built before 1918 and being commercially used show lower uncertainty levels.
In this case the estimated mean value of one sample already shows an uncertainty of below 10%.
41
Figure 13: Standard error of estimated mean value depending on the number of samples (Cencic 2016)
As mentioned in chapter 5.2, the material intensities of the building categories are calculated based on
a different number of information sources. Table 7shows that for a small share of the building stock
(1.8% of all buildings), no information about construction period and utilization is available.
Fortunately, material intensities of building categories, which constitute most of the overall GV, are
based on a higher number of buildings analysed (Table 7). However, for industrial buildings after
1997 no data were available. For this building category, it was assumed that its composition can be
approximated by a mixture of industrial buildings from 1977 to 1996, and residential and commercial
buildings from after 1997. The contribution of this category (industrial buildings after 1997) to the
overall GV of buildings in Vienna is 0.5% only. Thus, vague estimates about the material intensities
for this category have little impact on the result.
0%
10%
20%
30%
40%
50%
60%
0 5 10 15 20 25 30 35
stan
dar
d e
rro
r o
f es
tim
ated
mea
n v
alue
sample size n
Copper Case Studies Concrete LCA
Plastics LCA Bricks -1918/residential
Bricks -1918/commercial
42
Table 7: Contribution of the different building categories to the overall building volume (given in %) and
number of buildings analyzed with respect to their material composition. Parenthesized number
represents data from literature.
Residential Commercial Industrial No information
unit % of
total GV
No. of
buildings
analysed
% of
total GV
No. of
buildings
analysed
% of
total GV
No. of
buildings
analysed
% of
total GV
No. of
buildings
analysed
Before 1918 24.7 12+(8) 6.9 8+(1) 1.2 2+(1) 1.4 -
1919-1945 5.4 6+(7) 0.9 1+(2) 0.7 1+(1) 0.5 -
1946-1976 16.9 16+(15) 3.8 2+(3) 2.3 0+(1) 1.3 -
1977-1996 9.6 4+(7) 6.5 1+(1) 2.7 0+(1) 1.1 -
After 1997 5.1 12+(1) 3.3 3+(0) 0.5 0+(0) 0.5 -
No information 1.4 - 0.8 - 0.5 - 1.8 -
Apart from the specific material intensities for different building categories, the analysis of the
building structure also underlies uncertainties. Table 8 shows estimated uncertainty ranges of the
different data sources used to analyse the building structure and to subsequently estimate the overall
material stock. Sources for uncertainties mainly refer to the GIS data used. The data gaps for the
construction period and utilization at a building level were filled with data available at a building
block level which are, of course, characterized by higher uncertainties. Data about construction period
and utilization are up to 10 years old, whereas GIS data about newer buildings might not be updated
properly. Estimated uncertainties of material intensities for different building categories refer to
materials such as concrete, bricks/mortar, and constructional wood or steel. Low volume materials, as
used for fittings and installations, tend to generate higher uncertainty.
Table 8: Estimated uncertainty of data used to analyze the building structure
Data source Estimated uncertainty (%)
Area and height of buildings (MA 41) <5
Construction period & utilization (MA21) 5-20
Construction period & in the building block (MA18) 5-20
Utilization in the building block (MA21) 20-50
Construction period & utilization of new buildings (MA 37) 20-50
Because data on the average height of roofs and basement floors for different building categories were
compiled based on the estimation of experts, larger uncertainties exist. Because there is no other
source of information, the approach chosen represents the only possibility to predict the overall GV.
The share of additional volume added by considering roofs and basement floors amounts to almost
30% of the overall GV. Therefore, this share is quite significant for estimating the overall material
43
stock in buildings in Vienna. Details concerning expert estimation of the “additional” building volume
(for roofs and basements) are provided in the supporting information of the 2nd
paper.
As described above, the material data can be mapped in GIS and provide material information on a
building level to estimate the amount and quality of DW. Here it is important to consider that
uncertainties are unequally higher compared to the building stock analysis at a city level. Uncertainties
regarding the demolished building volume mainly result from the quality and topicality of the GIS
data comprised in the building model used (Table 8). However, the change detection data involve
additional uncertainty. New buildings on a demolition site which have been built up to the same height
or higher as the demolished building in the time frame between the two aerial pictures being taken
cannot be detected. It is assumed that this is rarely the case, and if so, only small buildings are
supposedly affected. However, due to this a slight systematic underestimation of the demolished
building volume is possible. Further investigations in this respect are necessary, especially when the
method is continuously applied to monitor the demolition activity. The amount and quality of related
demolition waste underlie uncertainties which result from the specific material intensities of different
building categories. As described, specific material intensities are generated based on different sources
of information, ranging from document analysis and on-site investigation to the use of data from
relevant literature. Through the categorization of buildings, generalization was necessary, which by
nature implies uncertainties.
45
7. Conclusion and outlook
The built environment constitutes a dominant stock and a potential future source for different
secondary raw materials. Therefore, detailed information about the material composition of buildings
is necessary to determine the resource potential of the building stock and urban settlements. The
individuality and peculiarities of urban planning, architecture and material use means that regions,
cities and buildings have different characteristics.
To generate specific intensities for different materials in buildings, various data sources are used. The
investigation of buildings prior to demolition considers changes in buildings through renovations and
technical upgrades during the use phase. Building documents are the basis to determine the bulk
materials within a building, whereas on-site investigations enable more specific information to be
gathered about the location and amount of materials of lower concentration. To allow investigations of
a higher number of buildings, construction plans of demolished buildings are analysed and
supplemented by results from the case studies, especially with regard to materials in lower
concentration. Material information from LCA data are used for the characterization of newer
buildings. Future research should identify more reliable data sources about the material composition of
buildings and develop methods to integrate the material data collection in routine building
investigations (e.g. energy performance certificate, static certificates, investigation of pollutants).
The quantification and localization of the material stock in buildings in Vienna provide the necessary
basis for a continuous monitoring of the Viennese building stock as well as for better knowledge about
the flows and stocks of materials throughout the city. The precision of the material stock analysis
depends both on the quality of data on the building structure and the data on the material intensities of
the different building categories. Future improvements regarding the data on the building structure
mainly concern the size of the buildings and their mapping in GIS. In this respect the municipality of
Vienna is currently developing a roof model for the entire city, which will allow the exact shape and
volume of all roofs to be mapped. However, exact data on basement floors in buildings will not be
available soon and expert judgements, as carried out in this study, will be retained. With regard to the
availability of GIS data, the municipality of Vienna is continuously expanding the number of openly
available data sets.
Combining the results of the building stock analysis with information about the demolition activity in
Vienna allows the amount and composition of wastes from demolition activities to be estimated. The
method proposed in this research can be practically implemented by the building inspection
department to monitor the building stock. Knowledge about demolition activities and resulting waste
46
streams in Vienna allow the amount of materials available for recycling and secondary use to be
predicted. By combining this information with data on planned buildings and construction in the city
(city planning), this projection may help to coordinate the potential use of the recycling material. For
the management of CDW these approaches can provide comparative values to validate statistical data
and to help detect inconsistencies. The ongoing monitoring of the demolition activity in Vienna is
considered important to understand the dynamics of the building stock and should, therefore, be
implemented on a regular basis.
47
Literature
Baccini, P., Pedraza, A., 2006. Bestimmung von Materialgehalten in Gebäuden, Bauwerke als
Ressourcennutzer und Ressourcenspender in der Langfristigen Entwicklung urbaner Systeme. vdf
Hochschulverlag AG, Zürich.
Bader, H.P., Scheidegger, R., Wittmer, D., Lichtensteiger, T., 2011. Copper flows in buildings,
infrastructure and mobiles: a dynamic model and its application to Switzerland. Clean Technologies
and Environmental Policy 13, 87-101.
Bergmeister, K., Wörner, J.D., Fingerloos, F., 2009. Konstruktiver Hochbau - Aktuelle
Massivbaunormen. Ernst und Sohn, Berlin.
Bergsdal, H., Bohne, R.A., Brattebø, H., 2007. Projection of Construction and Demolition Waste in
Norway. Journal of Industrial Ecology 11, 27-39.
Blengini, G.A., 2009. Life cycle of buildings, demolition and recycling potential: A case study in
Turin, Italy. Building and Environment 44, 319-330.
BMFLUW, 2015. Die Bestandsaufnahme der Abfallwirtschaft ind Österreich - Statusbericht 2015.
Bundesministerium für Land- und Forstwirtschaft, Umwelt und Wasserwirtschaft (BMFLUW).
BMLFUW, 2011. Bundes-Abfallwirtschaftsplan 2011. Bundesministerium für Land- und
Forstwirtschaft, Umwelt und Wasserwirtschaft (BMFLUW), Vienna.
Cencic, O, 2016. Personal communication.
Cheng, J.C., Ma, L.Y., 2013. A BIM-based system for demolition and renovation waste estimation and
planning. Waste management 33, 1539-1551.
Clement, D., Hammer, K., Schnöller, J., Daxbeck, H., Brunner, P., 2011. Wert-und Schadstoffe in
Wohngebäuden. Österreichische Wasser-und Abfallwirtschaft 63, 61-69.
Cochran, K., Townsend, T., Reinhart, D., Heck, H., 2007. Estimation of regional building-related
C&D debris generation and composition: Case study for Florida, US. Waste Management 27,
921-931.
Cochran, K.M., Townsend, T.G., 2010. Estimating construction and demolition debris generation
using a materials flow analysis approach. Waste Management 30, 2247-2254.
Dall, G., Galante, A., Torri, M., 2012. A methodology for the energy performance classification of
residential building stock on an urban scale. Energy and Buildings 48, 211-219.
De Melo, A.B., Goncalves, A.F., Martins, I.M., 2011. Construction and demolition waste generation
and management in Lisbon (Portugal). Resources, Conservation and Recycling 55, 1252-1264.
Deilmann, C., Gruhler, K., 2005. Stoff-und Energieflüsse von Gebäuden und Infrastrukturen als
Grundlage für ein vorausschauendes szenariogeleitetes Stoffstrommanagement. Österreichische
Wasser-und Abfallwirtschaft 57, 103-109.
European Comission, 2014. Construction and Demolition Waste (CDW). European Commission
European Parliament, Council of the European Union, 2008. Directive 2008/98/EC of the European
Parliament and of the Council of 19 November 2008 on waste and repealing certain Directives.
48
Fabbri, K., Zuppiroli, M., Ambrogio, K., 2012. Heritage buildings and energy performance: Mapping
with GIS tools. Energy and Buildings 48, 137-145.
Fatta, D., Papadopoulos, A., Avramikos, E., Sgourou, E., Moustakas, K., Kourmoussis, F., Mentzis,
A., Loizidou, M., 2003. Generation and management of construction and demolition waste in
Greece—an existing challenge. Resources, Conservation and Recycling 40, 81-91.
Görg, H., 1997. Entwicklung eines Prognosemodells für Bauabfälle als Baustein von
Stoffstrombetrachtungen zur Kreislaufwirtschaft im Bauwesen. Verein zur Förderung des Instituts
WAR - Wasserversorgung · Abwassertechnik · Abfalltechnik · Umwelt- und Raumplanung der TH
Darmstadt, Darmstadt, Germany.
Gruen, A., Behnisch, M., Kohler, N., 2009. Perspectives in the reality-based generation, n D
modelling, and operation of buildings and building stocks. Building Research & Information 37, 503-
519.
Gruhler, K., Böhm, R., Deilmann, C., Schiller, G., 2002. Sofflich-energetische Gebäudesteckbriefe -
Gebäudevergleiche und Hochrechnungen für Bebauungsstrukturen. Institut für ökologische
Raumentwicklung Dresden, Germany.
Hashimoto, S., Tanikawa, H., Moriguchi, Y., 2007. Where will large amounts of materials
accumulated within the economy go?–a material flow analysis of construction minerals for Japan.
Waste Management 27, 1725-1738.
Hashimoto, S., Tanikawa, H., Moriguchi, Y., 2009. Framework for estimating potential wastes and
secondary resources accumulated within an economy–A case study of construction minerals in Japan.
Waste management 29, 2859-2866.
Hu, M., Bergsdal, H., van der Voet, E., Huppes, G., Müller, D.B., 2010a. Dynamics of urban and rural
housing stocks in China. Building Research & Information 38, 301-317.
Hu, M., Pauliuk, S., Wang, T., Huppes, G., van der Voet, E., Müller, D.B., 2010b. Iron and steel in
Chinese residential buildings: A dynamic analysis. Resources, Conservation and Recycling 54, 591-
600.
Kavgic, M., Mavrogianni, A., Mumovic, D., Summerfield, A., Stevanovic, Z., Djurovic-Petrovic, M.,
2010. A review of bottom-up building stock models for energy consumption in the residential sector.
Building and Environment 45, 1683-1697.
Kleemann, F., Lederer, J., Aschenbrenner, P., Rechberger, H., Fellner, J., 2014. A method for
determining buildings’ material composition prior to demolition. Building Research & Information, 1-
12.
Kohler, N., Hassler, U., 2002. The building stock as a research object. Building Research &
Information 30, 226-236.
Kohler, N., Hassler, U., Paschen, H., 1999. Stoffströme und Kosten in den Bereichen Bauen und
Wohnen. Equete-Komission "Schutz des Menschen und der Umwelt" des 13. Deutschen Bundestages,
Bundeshaus, D-53113 Bonn, Heidelberg, Germany.
Kohler, N., Yang, W., 2007. Long-term management of building stocks. Building Research &
Information 35, 351-362.
Kral, U., Lin, C.Y., Kellner, K., Ma, H.w., Brunner, P.H., 2014. The Copper Balance of Cities. Journal
of industrial ecology 18, 432-444.
49
Krook, J., Carlsson, A., Eklund, M., Frändegård, P., Svensson, N., 2011. Urban mining: hibernating
copper stocks in local power grids. Journal of cleaner production 19, 1052-1056.
Lichtensteiger, T., Baccini, P., 2008. Exploration of urban stocks. detail 5, 16.
MA 18, 2013. GIS data about the construction age of buildings on a building block level, in: 18,
S.u.S.-M. (Ed.). Stadtentwicklung und Stadtplanung - MA 18.
MA 21, 2013a. Generalized zoning plan (GIS), in: Stadtteilplanung und Flächennutzung - MA 21
(Ed.), Wien.
MA 21, 2013b. GIS data about construction period and utilization on a building level, in:
Stadtteilplanung und Flächennutzung - MA 21 (Ed.), Wien.
MA 23, 2011. Gebäude und Wohnungen - Statistiken. Wirtschaft, Arbeit und Statistik (MA 23), Wien.
MA 37, 2013. GIS data about new buildings in Vienna, in: 37, B.-M. (Ed.).
MA 41, 2013. Flächenmehrzweckkarte, in: Stadtvermessung Wien - MA 41 (Ed.).
Marcellus‐Zamora, K.A., Gallagher, P.M., Spatari, S., Tanikawa, H., 2015. Estimating Materials
Stocked by Land‐Use Type in Historic Urban Buildings Using Spatio‐Temporal Analytical Tools.
Journal of Industrial Ecology.
Marcellus, K., Spatari, S., Gallagher, P., 2012. Utilizing GIS as a geospatial tool to inventory LEED
certified buildings and Construction and Demolition (C&D) waste flows in the United States, Life-
Cycle and Sustainability of Civil Infrastructure Systems: Proceedings of the Third International
Symposium on Life-Cycle Civil Engineering, Hofburg Palace, Vienna, Austria, October 3-6, 2012.
CRC Press, p. 373.
McCauley-Bell, P., Reinhart, D., Sfeir, H., Ryan, B., 1997. Municipal Solid Waste Composition
Studies. Practice Periodical of Hazardous, Toxic, and Radioactive Waste Management 1, 158-163.
Meinel, G., Hecht, R., Herold, H., 2009. Analyzing building stock using topographic maps and GIS.
Building Research & Information 37, 468-482.
Müller, A., 2013. Das Ressourcenpotenzial von Bauabfällen (article 1691). Müllhandbuch 3, 1-32.
Pauliuk, S., Sjöstrand, K., Müller, D.B., 2013. Transforming the Norwegian dwelling stock to reach
the 2 degrees Celsius climate target. Journal of Industrial Ecology 17, 542-554.
Rentz, O., Seemann, A., Schultmann, F., 2001. Abbruch von Wohn- und Verwaltungsgebäuden -
Handlungshilfe, in: Baden-Württemberg, L.f.U. (Ed.), Karlsruhe, Germany.
Sandberg, N.H., Bergsdal, H., Brattebø, H., 2011. Historical energy analysis of the Norwegian
dwelling stock. Building Research & Information 39, 1-15.
Sandberg, N.H., Brattebø, H., 2012. Analysis of energy and carbon flows in the future Norwegian
dwelling stock. Building Research & Information 40, 123-139.
Sartori, I., Bergsdal, H., Müller, D.B., Brattebø, H., 2008. Towards modelling of construction,
renovation and demolition activities: Norway's dwelling stock, 1900–2100. Building research &
information 36, 412-425.
Schneider, M., Rubli, S., 2007. Ressourcenmodell mineralischer Baustoffe auf der Ebene Stadt Zürich.
Wertstoff-Börse GmbH, Schlieren.
50
Schulze, H.-J., Walther, P., Schlüter, M., 1990. Gebäudeatlas-Mehrfamilienwohngebäude der
Baujahre 1880-1980. Bauinformation.
Solís-Guzmán, J., Marrero, M., Montes-Delgado, M.V., Ramírez-de-Arellano, A., 2009. A Spanish
model for quantification and management of construction waste. Waste Management 29, 2542-2548.
Stark, W., Vogel-Lahner, T., Frühwirth, W., 2003. Bauwerk Österreich–Endbericht. GUA -
Gesellschaft für umfassende Analysen GmbH, Wien.
Statistik Austria, 2014. Vorausberechnete Bevölkerungsstruktur für Wien 2013-2075 laut
Hauptszenario.
Stäubli, B., Winzeler, R., 2011. Bau und Rückbau im Kanton Zürich: Modell für die Materialflüsse
1900 bis 2008. Müll und Abfall 06/2011, 262-270
Tanikawa, H., Fishman, T., Okuoka, K., Sugimoto, K., 2015. The Weight of Society Over Time and
Space: A Comprehensive Account of the Construction Material Stock of Japan, 1945–2010. Journal of
Industrial Ecology, n/a-n/a.
Tanikawa, H., Hashimoto, S., 2009. Urban stock over time: spatial material stock analysis using 4d-
GIS. Building Research & Information 37, 483-502.
United Nations, 2014. World Urbanization Prospects - The 2014 Revision, New York.
Wedler, B., Hummel, A., 1947. Trümmerverwertung: technische Möglichkeiten und wirtschaftliche
Bedeutung. Wilhelm Ernst & Sohn.
Wiedenhofer, D., Steinberger, J.K., Eisenmenger, N., Haas, W., 2015. Maintenance and Expansion -
Modeling Material Stocks and Flows for Residential Buildings and Transportation Networks in the
EU25. Journal of Industrial Ecology.
Wittmer, D., Lichtensteiger, T., 2007. Development of anthropogenic raw material stocks: a
retrospective approach for prospective scenarios. Minerals and Energy 22, 62-71.
Yuan, H., Shen, L., 2011. Trend of the research on construction and demolition waste management.
Waste management 31, 670-679.
Appendix
1st Paper
A method for determining buildings’ material composition prior to demolition
Fritz Kleemann, Jakob Lederer, Philipp Aschenbrenner, Helmut Rechberger and Johann
Fellner
Building Research and Information
DOI: 10.1080/09613218.2014.979029
RESEARCH PAPER
Amethod fordeterminingbuildings’materialcompositionprior to demolition
Fritz Kleemann, Jakob Lederer, Philipp Aschenbrenner, Helmut Rechberger and Johann Fellner
ChristianDoppler Laboratory for Anthropogenic Resources,ViennaUniversity of Technology,Karlsplatz13/226,ViennaA-1040, Austria
A prerequisite of the efficient recycling of demolition waste and its evaluation in terms of the material specific recycling
rates is information on the composition of the building material stock (as the source of future demolition waste). A
practical method is presented that characterizes the material composition of buildings prior to their demolition. The
characterization method is based on the analysis of available construction documents and different approaches of on-
site investigation. The method is tested in different buildings and the results from four case studies indicate that the
documents are useful to quantify bulk materials (e.g. bricks, concrete, sand/gravel, iron/steel and timber). However,
on-site investigations are necessary to locate and determine the trace materials such as metals (e.g. copper and
aluminium), or different types of plastics. The overall material intensity of the investigated buildings ranges from 270
to 470 kg/m3 gross volume. With ongoing surveys about the composition of different buildings, the collected data
will be used to establish a building-specific database about the amount of materials contained in Vienna’s building stock.
Keywords: buildings, deconstruction, demolition, demolition waste, material composition, material intensity, recycling,
urban mining, waste management
IntroductionIn 2009, almost 6.5 million tonnes of demolition wastewere generated in Austria. In comparison, municipalsolid waste accounted for only 3.9 million tonnes(BMLFUW, 2011). Accordingly, the recycling ofdemolition wastes has great potential to reduce theoverall amount of waste landfilled and save primaryresources. In order to project the future availabilityof recyclable material from demolition activities, it isessential to obtain information about the materialcomposition of the building stock and the buildingsthat will be demolished in future. Although there aredifferent approaches to gain knowledge about thecharacterization of the building stock and its dynamics(Kohler & Hassler, 2002), its material composition isnot well known (Kohler & Yang, 2007). The chal-lenges are in generating data about individual buildingsand extrapolating this information for selected geo-graphical areas of interest. The aim of this study is,therefore, to apply and subsequently evaluate amethod for characterizing buildings with respect totheir material composition. The generated data willhelp to provide specific material intensities for differentbuilding categories (based on age and functions). Thesespecific material intensities should allow future
research to estimate the overall material resources inVienna’s buildings. By combining this informationwith data about demolition and renovation activitiesin Vienna, it will be possible to determine the futureamounts and composition of demolition waste.
Specific values about the material composition ofbuildings can be generated by three methods:(method i) using information about the materialsused to construct the buildings, (method ii) by analys-ing the generated waste streams after demolition, or(method iii) by investigating the buildings prior totheir demolition. There are many studies on the firsttwo approaches, and numerous examples can be pro-vided: Gorg (1997) considered the legal requirementsand the valid norms in the referring period of construc-tion to develop a model to predict future demolitionwastes, whereas Kohler, Hassler, and Paschen (1999)investigated buildings in Germany on a national levelregarding the mass flows and the costs based onmacro-economic input–output data and by assigningspecific mass flows to the buildings, building partsand materials. Gruhler, Bohm, Deilmann, and Schiller(2002) and Deilmann and Gruhler (2005) defined andcharacterized different building types regarding the
BUILDING RESEARCH & INFORMATION 2014
http://dx.doi.org/10.1080/09613218.2014.979029
# 2014 Taylor & Francis
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
material and the energy intensity, and Baccini andPedraza (2006) developed the ARK-House method toassign different material intensities to buildings ofdifferent ages and usage. This approach was alsoapplied by Lichtensteiger and Baccini (2008) to deter-mine the bulk materials in the buildings (minerals,steel and wood) and by Wittmer and Lichtensteiger(2007) to calculate the anthropogenic copper stock inthe buildings. These studies greatly contributed to theinvestigation of building stocks and identified someof the major challenges when characterizing the built-in materials by applying methods i and ii. Usinginformation about the materials that have been usedto construct buildings (method i), one often neglectsor underestimates renovations and technical upgradesduring the use phase, which usually increase the com-plexity of built-in materials. For example, Krook,Carlsson, Eklund, Frandegard, and Svensson (2011)described the installations that are out of use. Theyoften stay in hibernation and are not recoveredduring technical upgrades, although new installationsare added, which increases the material content com-pared with the initial state. An idea about the rel-evance of renovation works is provided by Bergsdal,Bohne, and Brattebø (2007) for the construction anddemolition (C&D) waste sector in Norway. On thecontrary, studies that estimate the dynamics of thebuilding stock based on data from the demolitionwaste (method ii) face the challenge of under- or over-estimating the content of so-called trace materials (e.g.copper). This problem lies in the difficulties of accu-rately sampling high quantities of waste streams ofheterogeneous composition and the quality of thewaste stream data as reported by demolition contrac-tors and collected by authorities (Blengini, 2009;Cochran, Townsend, Reinhart, & Heck, 2007;McCauley-Bell, Reinhart, Sfeir, & Ryan, 1997). Inaddition, the relevant information of the detailedlocation of valuable trace materials within a buildingis often not provided in studies that present highlyaggregated data sets. Therefore, the authors suggestthe application of different methods to investigatethe building stock, particularly for trace materialswith a share of less than 1% (Lichtensteiger &Baccini, 2008).
Considering all these challenges, it is assumed in thepresent study that an end-of-life investigation (e.g.Clement, Hammer, Schnoller, Daxbeck, & Brunner,2011) provides the most accurate building stockcharacterization results and information about thematerials that are potentially available for later recy-cling. This paper describes a method to quantify thematerials that are actually present in buildings immedi-ately prior to their demolition. Four case studies ofmajor demolition projects in Vienna were chosen toexamine the applicability of the presented method.The case studies are used to address the followingresearch questions:
. What is the material composition of selected build-ings in Vienna?
. Where in a building are the materials? In particu-lar, where are the valuable trace materialslocated and how are they embedded?
. How well do the collected data match the officialreports and the information provided by the demo-lition contractors?
Materials andmethodsTo evaluate comprehensively the material compositionand constitution of the buildings prior to their demoli-tion, a stepwise approach based on both the evaluationof available documents and on-site investigation wasestablished and tested.
As a first step, all available documents of the buildingare collected and analysed for the materials used toconstruct and modify the building. The documentsare retrieved from various sources; therefore, thecontent and quality of information may vary consider-ably. In the frame of the present study, the followingdocuments were considered:
. Construction plans provided by the building-owners or the building authorities.
. Investigation of pollutants in buildings beforedemolition (Schadstofferkundung von Gebaudenvor Abbrucharbeiten). This document must be pre-pared by an authorized person and is legallyrequired from the owner of a building to be demol-ished and characterized by a volume of more than5000 m3 (Austrian Standards, 2006).
. Waste management concept for demolition sites(Abfallkonzept fur Baustellen). This document isalso prepared by an authorized person andrequired by the authorities from the owner of abuilding to be demolished with a volume of morethan 5000 m3 (Wiener Landtag, 2012).
In the second step, the data about the built-in materialsare collected through an on-site inspection, measure-ments and selective sampling before the deconstruc-tion/demolition of the building. This step includes thedismantling, weighing and measuring of componentssuch as windows, doors, partitions, ceiling suspensions,floorand roof constructions,wires,pipes, etc.Dependingon the on-site conditions, different strategies are applied:
† Sampling of the representative areas
The built-in materials are determined and pro-jected for areas in the buildings, which have
Kleemann et al.
2
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
similar construction, fittings and repetitive appear-ance (e.g. dwelling units, floors, facades or roofarea). For instance, one flat of a dwelling buildingis investigated in detail, and the total mass ofdifferent materials in all flats can then be calcu-lated based on the floor area of the building.
† Investigation of the rising mains and distributionmains
The rising mains are identified and counted, andthe dimensions of pipes and wires are documented.The length of the mains is subsequently calculatedusing the total height of the building. The distri-bution mains at each floor are measured by count-ing and documenting the dimensions of the wireconnections in the main fuse boxes and calculatingthe length using the average distance to the roomsthat are supplied. The pipes for water, wastewater,heating and air-conditioning are counted, and thedimensions are documented. By measuring the dis-tance from the connection points (sanitary facili-ties, heater) to the rising mains, the length of thepipes can be determined.
† Investigation of the material composition of‘typical’ installations
Units commonly used in one building (e.g. differ-ent types of windows, doors, heaters, partitions,suspended ceilings or raised floors, and roof con-structions) are investigated for their material com-position (e.g. dismantling a window and weighingdifferent materials) and projected by area ornumber.
† Notional sections
For areas in buildings with many installations,notional sections are set length-wise and cross-wise (in some cases, also horizontally) throughthe rooms, as indicated in Figure 1. With these sec-tions, all wires, pipes and other installations thatrun orthogonally are counted, and the dimensionsare documented. For the installation lengths thedimensions of the investigated room (perpendicu-lar to the cross-section) are assumed.
† Inventory of specific areas
The previously described strategies are not alwaysmeaningful when applied. Special components canbe simply documented and related to a room or afloor.
Based on the collected data from documents and siteinspections, information about different built-inmaterials is aggregated, and the mass is calculatedbased on the volume, area or number of the particular
material and on data about its specific density orweight. Relevant information on the density of thebuilding materials or the weight of the installations isobtained either by sampling or from the literature onconstruction methods and design norms.
Finally, the collected data are compared with the dataof waste fractions and quantities, which are reportedby the demolition company to the building owner.These data also serve as a basis to calculate the totalamounts of C&D wastes that are generated andmanaged in Austria (BMLFUW, 2011).
Case studiesTo verify the presented method, four case studies wereperformed in the city of Vienna (Figure 2). Addition-ally, information about the current practices of plan-ning, coordination, time management and the workduring the demolition was gathered from the stake-holders involved in the demolition process. Thechosen case studies allowed the investigation of awide range of building types and represent a sampleof actually demolished buildings in Vienna. The latteris important because certain buildings tend to bereplaced by new buildings, and others are renovated.
In contrast to case studies 1, 3 and 4, which comprisedsingle buildings of different sizes, case study 2 (a publichospital) represented a complex of six buildings. Fourof these buildings with a gross volume of 13 000–17 000 m3 each were built approximately in 1870and made of brick. A smaller building (7200 m3),which was completed in 1960, was constructed withbrick walls and ceilings of reinforced concrete. A newoperating theatre (11 000 m3), which was built in2003, was constructed with reinforced concrete.Table 1 outlines the characteristics of the four casestudies. The buildings of case study 2 are shown separ-ately according to their time of construction. Therein,CS2.1 comprises four buildings from 1870.
Results and discussionApplicability of the presentedmethodInformation from the construction plans allows one todetermine the volume of walls and floors. Based on thisinformation, the bulk materials (concrete, bricks, sand,gravel, etc.) can be quantified using data from the lit-erature about the density and composition of differentbuilding materials (e.g. the ratio of reinforcement steelin concrete and the ratio of mortar in brickwork).Depending on the quality of the plans, details aboutthe floor and the roof assemblies are available. In thiscase, filling or lining material and materials forimpact sound insulation and flooring can be quantified.Because the flooring may have been changed during thelifetime of a building, the provided information in the
Determining thematerial composition of buildings
3
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
construction plan must be crosschecked on-site withthe actual type of flooring. Information about theinstallations (heating, sanitation, electricity) is onlysporadically available in the existing documents.Hence, they cannot be used to quantify the associatedamounts of relevant trace materials such as copper orsteel. A partial estimate of steel can be based on theconstruction plans using the assumed ratio of steelfrom the concrete reinforcement and the girders,
which are marked in the plans. As noted previously,an on-site investigation is needed to find steel andother metals used in a non-structural way.
The report about the investigation of pollutants inbuildings prior to demolition is not meant to providecomprehensive information about the total materialcomposition because it mainly focuses on hazardoussubstances that must be safely removed prior to the
Figure 1 Notional sections (displayed on the left) in areaswith many installations
Figure 2 The sites of the 4 case studies carried out ^ CS1upper left,CS4 lower right (Microsoft 2014, adapted)
Kleemann et al.
4
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
main demolition work. However, it usually contains asection with some general information about the build-ing, which is helpful for planning on-site investi-gations. In some cases, the report is part of the tenderoffer for the demolition work and provided shortlyprior to the commencement of the demolition.It is often viewed as a necessary document to fulfilthe legal requirements instead of a useful source ofinformation to better plan the deconstruction/demolition.
Similarly, the waste management concept for demoli-tion sites is often provided together with the reportabout the investigation of pollutants in buildings andthe contractor’s bid for the demolition work. Basedon the investigated case studies, it appears thatthe waste management concept is conductedto satisfy the legal requirements instead of being usedto plan the demolition work in advance. The wastemanagement concept should contain qualitativeand quantitative information about the built-inmaterials; however, quantitative data are rare ormissing.
The on-site investigation allowed a quantification oftrace materials such as iron, copper, aluminium,plastics, wood, etc. Additional information aboutthe typical nominal sizes of pipes, wires or otherinstallations was derived from design norms andproducers. For most of the work, available construc-tion plans were notably helpful to sketch-ininformation or relate notes to specific areas ofthe building. The lessons that were learned from theperformed case studies can be summarized as follows:
† Sampling of representative areas
This strategy was appropriate for the dwellingunits of case study 1 because all 254 units weresimilar in construction and fit-out. In case studies2–4, some floors were similar which allowed thismethod to be used.
† Investigation of rising mains and distributionmains
Each case study was different and showed that theelectrical and sanitary installations often undergotechnical upgrades and alterations during the life-time of a building.
† Investigation of the material composition of‘typical’ installations
Windows, doors, heaters and suspended ceilingswere investigated as typical installations in allinvestigated case studies. Some of the measureddata about heaters and suspended ceilings couldbe used for several case studies.
† Notional sections
This procedure was used in all the case studies,where the installations were numerous (in particu-lar, in basement floors). For rooms with a highnumber and multiplicity of installations, notionalsections appeared to be the most efficient approachfor a structured determination of built-inmaterials.
† Inventory of particular areas
The previously described strategies were often notapplicable and information about the structuresand fittings were simply noted and attributed tothe rooms or areas of the building.
In conclusion, it may be impossible to characterizecomprehensively the material composition of buildingsbased only on the available documents. The bulkmaterials of the building structure can be well quanti-fied, but the trace materials used for the installationsare not quantifiable. On the contrary, an on-site inves-tigation allows the quantification of trace materials butis not suitable to quantify bulk materials because the
Table 1 General information about the case studies (CS)
CS1 CS2.1 CS2.2 CS2.3 CS3 CS4
Building use Residential Administrationhospital
Hospital Hospital Industrialproduction
Commercialresidential
Constructionmaterial
Reinforcedconcrete
Brickwork Reinforcedconcrete,brickwork
Reinforcedconcrete
Steel,brickwork
Brickwork
Completion 1970 1870 1960 2003 1900 1859
Grossvolume (m3) 60 000 62 000 7200 11000 21000 3700
Gross£oor area (m2) 18 000 13 400 2200 2500 3900 1100
Determining thematerial composition of buildings
5
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
use of heavy machinery would be necessary to obtaininformation about the composition of walls or ceilings.
The effort needed to carry out the document based aswell as the on-site investigation of the buildings ishigh. Data collection through the analysis of docu-ments and on-site investigation as well as the evalu-ation of the data results in a workload that makes itinapplicable for a high number of buildings. To gener-ate the data about the material composition of casestudy 2, which was the most work intensive, about200 man-hours were necessary. From the authors’ per-spective, however, the approach is appropriate to gen-erate information about the specific content of tracematerials present in buildings. Despite their lowamount in volume, this information is crucial as tracematerials impact on the overall material value ofbuildings (e.g. copper) or contain potentialcontaminants (e.g. plastics) in the generated demoli-tion waste.
Material composition of buildingsTable 2 shows the results of the material compositionof different case studies in aggregated form and inrelation to 1 m3 of gross volume of the building.Gross volume was chosen as a reference valuebecause it includes all parts of a building. Withregard to the planned upscaling of the data, thismetric fits well with the available GIS data (geographi-cal information system) about buildings in Vienna. Thebuildings of case study 2 are again analysed separatelyaccording to the time of construction. For each casestudy, the material origin and the material mass calcu-lation method were extensively documented to be ableto trace back the results.
The results of the case studies indicate a dominantshare of mineral matter, which amount to between
94% and 98% of the total mass. In comparison, theshare of all other present materials is relatively low.However, their total amounts can be significantdepending on the overall size of the building. Unsur-prisingly, the share of steel tends to be higher whenreinforced concrete is used as the main buildingmaterial (CS1, CS2.3); however, large buildings,which are mainly made of bricks (CS2.1), may alsobe characterized by relatively high steel content, par-ticularly if massive steel girders were used. In CS3,comparable large quantities of steel were used for theroof construction. The steel content of CS4 is notablylow, particularly in comparison with the coeval build-ings of CS2.1. Here, the ceilings were constructed ofmassive timber beams instead of steel girders, whichexplains the high wood and the low steel content.The amount of aluminium varies considerably frombuilding to building and depends on the design of thestructural elements and installations. In CS1, alu-minium was used for the facade, doors and windows;in CS2.3 and CS4, aluminium was used as part of theroofing; and the aluminium content in the other build-ings was insignificant. In the investigated case studies,copper was mainly used for electrical and sanitaryinstallations. The specific values are notably compar-able and in the same order of magnitude for all casestudies, except for CS3. This particular building wasuntenanted for approximately two years before itwas demolished. During that time, most of thecopper disappeared through unknown paths, suppo-sedly informal recyclable collectors. Polyvinylchloride(PVC) as a flooring material and cement asbestos as acover for roofs and facades were found in some ofthe case studies.
Regarding the overall material intensity, the valuesfor CS1 are relatively high, which can be explainedby the compact geometry of the buildings, the rela-tively low ceiling heights of approximately 3 m,
Table 2 Comparison of material composition of the case studies (kg/m3 gross volume)
CS1:1970 CS2.1: 1890 CS2.2: 1960 CS2.3: 2003 CS3: 1900 CS4:1859
Minerals (bricks, concrete, gravel, sand) 430 420 410 320 260 450
Cement asbestos 1.5 0.04 ^ ^ 0.14 ^
Steel 7.6 5.1 4.6 9.5 5.8 0.97
Aluminium 0.22 0.049 0.057 0.22 0.03 0.16
Copper 0.11 0.15 0.16 0.24 0.0019 0.062
Wood 2.3 4.3 2.2 0.62 3.6 20
PVC 0.52 0.19 0.21 0.18 0.0093 0.20
Various plastics 1.3 0.16 0.35 4.9 0.14 0.46
Others (e.g.mineral wool, bitumen, linoleum) 1.1 0.54 1.2 0.54 0.43 0.13
Total 440 430 420 340 270 470
Kleemann et al.
6
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
and small dwelling units of approximately 40–60 m2. Compared with CS1, the overall materialintensity of CS2.1 is only slightly lower, althoughceiling heights reach approximately 4.5 m. Thisresults from a wall thickness of approximately 1 mat the basement, as is usual for buildings of thisbuilding period. CS2.2 has a slightly lower overallmaterial intensity than CS1. This observation canbe attributed to the use of bricks for wall construc-tions instead of heavier concrete walls and consider-ably larger room sizes. The comparably new buildingof CS2.3 has a significantly lower overall materialintensity. The ceiling height is approximately 4 m,and the overall construction height with two floorsrelatively low, reducing the required wall thicknessat the base. In addition, most partitions insidethe building are plasterboard constructions withconcrete columns, which support the ceiling. CS3consisted mainly of production halls, which imply arelatively low overall material intensity through itsconstruction style with huge areas, and ceilingheights up to 9 m. CS4, from the same era as thebuildings of CS2.1, has the highest overall materialintensity, with thick brick walls and relatively smallrooms. This reason and a relatively small grossvolume explain the comparably high overall materialintensity.
Location of di¡erent materialsBeside the quantification of the materials in buildings,a main aim of the study is to investigate where the
materials are located and concentrated inside the build-ing. Figures 3 and 4 exemplarily show the origin ofdifferent materials for CS1.
The knowledge about the location of materials in abuilding as shown in Figures 3 and 4 is helpful toplan the recycling processes effectively and estimatethe effort needed to recover certain materials beforethe building demolition (selective deconstruction).Regarding the valuable material, Figure 5 furtherdistinguishes the location of steel, aluminium andcopper for CS1.
In addition to the material location, important infor-mation for recycling is the assemblage of certainmaterials. The separability largely influences the tech-nical and organizational processes necessary formaterial recycling. In case study 1, most of the steelis reinforcement steel. This steel has to be separatedfrom the concrete using hydraulic crushers. The othersteel stems from different installations (Figure 5) inthe upper floors and the basement floor. Most of thealuminium is located in windows and doors in theupper floors, and the facade construction comprises asignificant part of aluminium. The copper is mostlybond in the electric supply lines in the basement ofthe building; a smaller amount is located in electricalinstallations in the upper floors. Therefore, to berecycled, it must be separated from the PVC or otherplastics used as cable insulation. In this case study,most of the pipework was made of steel; however,the use of copper for sanitary installations or as
Figure 3 Origin of di¡erent mineral materials for case study1
Determining thematerial composition of buildings
7
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
roofing material can significantly increase the overallamount.
Comparisonwith other studiesTable 3 compares specific material quantities of differ-ent studies with the present ones in aggregated formand in relation to 1 m3 of gross volume of the building.The material intensities calculated by Gorg (1997) andGruhler et al. (2002) apply to multifamily houses. Pub-lished by the federal state of Baden-Wurttemberg,Germany, this study (Rentz, Seemann, & Schultmann,2001) focuses on dwelling and administration build-ings. Baccini and Pedraza (2006) emphasized copperin their study and investigated single houses,
multifamily houses, commercial buildings and indus-trial buildings. In the study of Blengini (2009), one resi-dential building was assessed.
The figures of the present study are consistent with theoverall material intensity of previous studies (Table 3).However, in most cases, lower metal values were cal-culated in this study. The main reason for the signifi-cant difference in the steel figures compared withGruhler et al. (2002) is the unrealistically highcontent of steel in reinforced concrete of 3–5% byvolume, whereas 0.5% by volume was assumed inthis study. For comparison, Gorg (1997) assessed asimilar value of 0.4–0.8 vol.%. The specific figuresfor copper available from the literature (Baccini &
Figure 4 Origin of di¡erent non-mineral materials for case study1
Figure 5 Location and total quantity of valuablemetals (steel, aluminium, copper) in case study1
Kleemann et al.
8
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
Pedraza, 2006; Wittmer & Lichtensteiger, 2007) areconsistent with the data of the present study;however, the copper content reported varies consider-ably. Muller (2013) highlighted the wide range of theoverall material intensity of different building types(100–1200 kg/m3 gross volume) in different studies,which resulted in similarly wide ranges for all materialfractions.
Comparison of compositional data with demolitioncontractor reportsThe initial idea of this study was to validate the col-lected data and the calculated values of the materialcomposition of buildings using the figures of thewaste disposal reports (Entsorgungsnachweise),which are provided by the contracted demolition con-tractor. Currently, the reports of case studies CS1–CS3have been made available by the demolition contrac-tors. Tables 4 and 5 show the comparison of self-col-lected and reported material data of CS2 and CS3,respectively. The reports of CS1 and CS2 havesimilar data quality.
The collected data (Table 4) are notably consistentwith the mineral main fractions (brick/mortar and con-crete) and wood. The amount of steel was underesti-mated using the material characterization conductedprior to demolition. This underestimation can beeither a result of a relatively low assumed share ofreinforcement steel in concrete of 0.5% by volume(Bergmeister, Worner, & Fingerloos, 2009) or aresult of the impurities in the iron scrap fraction thatwas separated. According to the report of the demoli-tion contractor, the only recovered aluminium orig-inates from cables, which could not be detectedduring the on-site investigation. Interestingly, nocopper is accounted for in the records of the demolition
contractor. There are also big differences in theamount of cement asbestos. The demolition contrac-tor reported more than six times the quantity that hasbeen found during on-site investigations. One expla-nation may be that mineral fractions which are con-taminated with cement asbestos have also beenreported as cement asbestos. Gypsum is usually not
Table 4 Comparison of collected data with o⁄cial data of thewaste disposal reports of case study 2 (in tonnes)
Data collectedprior to demolition
O⁄cial data of wastequantities (provided by thedemolition contractor)
Bricks/mortar 24 000 27000 Mineral demolition waste
Concrete 7800 7600 Concrete waste
Steel 490 680 Ferrous scrap metalwith impurities
Aluminium 11 6 Aluminium cable
Copper 13 ^
PVC 15 ^
Wood 290 290 Waste wood
Bitumen 51 70 Roo¢ng paper andtar residues
Cementasbestos
4.5 28 Cement asbestoswaste
Gypsum 250 25 Gypsum
Other plastics 52 85 Wastes from thedemolition site
Others 21 ^
^ 9600 Excavated soil
Total 33 000 45 000 Total
Table 3 Comparison of speci¢c material values of di¡erent studies adapted [kg/m3 gross volume]
Presentstudy
Go« rg(1997)
Rentz et al.(2001)
Gruhler et al.(2002)
Baccini andPedraza (2006)
Blengini(2009)
Minerals (bricks, concrete,gravel, sand)
260^450 n.d. 100^419 n.d. n.d. 387
Cement asbestos 0^1.5 n.d. n.d. n.d. n.d. n.d.
Steel 0.1^8.6 2.08^23.22 2^16a 0^37 n.d. 14.6
Aluminium 0.03^0.55 n.d. n.d. n.d. n.d. 0.013
Copper 0.002^0.5 n.d. n.d. n.d. 0.05^0.24 0.023
Wood 0.6^20 n.d. 2^28 n.d. n.d. 0.44
Various plastics 0.15^5.1 n.d. n.d. n.d. n.d. 0.96
Total 270^470 382^556 127^431 370^480 180^570 404
Note: aIncludes all metals.
Determining thematerial composition of buildings
9
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
extracted and separately collected in Austria, whichresults in significant parts of the gypsum accumulatedin the fraction of mineral demolition waste, whichmay explain the differences of both data sets. Thesoil that was accounted for by the demolitioncontractor was excavated during the preparation ofthe construction site for the new building projectand could not be considered as part of theinvestigation.
The two data sets show some discrepancies, particu-larly regarding the trace materials in small quan-tities. Note that valuable fractions such as copperand aluminium appear neglected in the documen-tation. On-site visits during the demolition work,however, showed that these fractions were wellseparated.
The data in Table 5 demonstrate large discrepanciesbetween the quantities that are reported in the officialwaste disposal report and the material amounts thatare assessed during on-site investigations. Because nosteel was considered in the official waste disposalreport, additional information was requested fromthe demolition contractor, which finally resultedin another set of material data (right column inTable 5). Surprisingly, the data that were directlyobtained from the contractor showed considerableinconsistencies with respect to the official waste dispo-sal report. Compared with the data that were collectedduring the on-site investigation, the differences arenoticeable, particularly regarding the bricks/mortarand steel. There are large differences in the amountof mineral materials that were assessed and reportedbecause the basement floors were filled with parts ofthe mineral material, which was not reported as demo-lition waste. The discrepancies for steel and timbercannot be explained from the authors’ perspective.The differences between the information that wasdirectly obtained from the contractor and the officialreport (provided by the same contractor) show
general problems in the documentation and reporting,which results in unreliable data.
Obtaining information from the demolition contrac-tors was a significant challenge, and in some casesnot possible at all. One reason is that the contractorsreport the waste composition according to the categor-ization in the Austrian List of Waste Ordinance (Bun-desabfallwirtschaftsplan). The codes used in thisordinance do not always give information on thewaste composition. For instance, building debris mayconsist of mortar, concrete and bricks, but the shareof each of them is not required to be determined andreported by the contractor. In some situations, partsof this debris from a demolished building are directlyreutilized as backfill material on the site. The reasonsfor the differences between reported and determineddata for valuable trace materials (e.g. copper) are notyet entirely clear and are subject of furtherinvestigation.
In general, the comparison of the collected and the cal-culated data (prior to demolition) with the officiallyreported data from the demolition contractors indi-cates that the official documents can only be used asa rough estimate of the bulk materials. In most cases,valuable trace materials, such as copper or aluminium,are not reported. The reports of CS3 showed highlyuncertain and missing data even for materials of signifi-cant quantities such as steel. Currently, neither dataabout waste from demolition projects nor data aboutdemolition activities are recorded by authorities.Demolition contractors are only obliged to reportabout the disposed waste streams in highly aggregatedform on an annual basis (e.g. tonnes of demolitionmaterial from all private households in 2012). Thewaste disposal reports serve solely as a proof for thebuilding owners that legal requirements have been ful-filled. Only a few of these reports (selected on arandom basis or due to legal notifications) arechecked by the authority in charge. For a better under-standing of demolition activities in Vienna and associ-ated material flows, records should not only be filed bydemolition contractors and building owners, but bedirectly integrated into waste statistics. Althoughsuch integration would not solve the problem of datainconsistencies it would be a first step towards highertransparency in the C&D waste sector.
Data uncertaintiesGenerating primary data, as it has been done in thisstudy, inevitably raises the question about uncertain-ties in the different approaches. The initial idea to vali-date the collected data was to compare it with datafrom the contractors. However, this was not possibleas the data from the contractors were very inconsistent.Thus, a higher number of case studies need to be inves-tigated to obtain consistent data about the material
Table 5 Case study 3: comparison of collected data with data ofwaste disposal reports and data of the demolition contractor(in tonnes)
Data collectedprior to
demolition
O⁄cial wastedisposalreport
Figures of thedemolitioncontractora
Bricks/mortar 3500 2000 410
Concrete 930 1500 1370
Steel 120 n.d. 53
Wood 75 204 52
Total 4700 3700 1900
Note: aProvided on request.
Kleemann et al.
10
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
composition of different buildings, as well as theexpected uncertainties. Based on the information athand, only qualitative assumptions on the origin ofuncertainties can be given.
The document-based investigation, focusing on bulkmaterials, relies on accurate documents about thebuildings, which implies uncertainties itself. In thisstudy documents were assumed to be accurate. Withregard to bulk materials, inconsistencies comparedwith the data from the demolition contractor arebelieved to be mainly due to material remaining onsite as filling material after the demolition. Thismaterial is disposed of at a later stage and might beaccounted as waste of the new construction projecton the respective site. Another important source ofuncertainties is the use of default data. In case study2 most of the 490 tonnes (t) of steel (collected data)originated from steel girders (210 t) and reinforcedconcrete (100 t). For reinforced concrete a steel shareof 0.5% by volume was assumed. The underestimationcompared with the data from the contractor can resultfrom a wrong assumption regarding the reinforcement.Assuming a volumetric steel share of 0.7% wouldincrease the mass of reinforcement steel by 40% to140 t. Furthermore, the steel girders could only bemeasured at their narrow side during the on-site inves-tigation. Here an average material thickness of 12 mmwas assumed – 13 mm would raise the amount of steelfrom girders by 10% to 230 t. Additionally, someimpurities can be part of the steel fraction and increaseits weight.
The on-site investigations are believed to give the mostreliable information about trace materials within abuilding. However, uncertainties also exist. The massof pipes can serve as an example for that. As it is notmanageable to check the wall thickness of each pipein a building, existing norms for dimensions of pipesare used. Uncertainties with respect to the materialmass per length of installation (pipe) can be assumedto be rather low (below 5%). However, the uncertaintyof the measured installation length, e.g. determined bynotional sections, can be estimated to be around 20%,which results in overall uncertainties of above 20%.Special components such as lifts, which can accountfor a significant share of a material fraction in a build-ing, could not be dissembled, and estimates togetherwith general producer information were used. Herehigher uncertainties exist.
ConclusionsThe built environment constitutes a dominant stockand possible future source of different materials.Therefore, detailed information about the materialcomposition of buildings is necessary to determinethe resource potential of the building stock and
urban settlements. The individuality and peculiaritiesof urban planning, architecture and material usemeans that regions, cities and buildings have differentcharacteristics. The characterization method that wasapplied to four case studies demonstrates these differ-ences at the building level. To generate specific valuesfor different materials in buildings, the method thatwas applied in this study investigates the buildingsprior to demolition. The changes to the buildingsthrough renovations and technical upgrades duringthe use phase were considered. In comparison withanalysing waste streams, this method investigates allmaterials in detail before being potentially mergedwith another fraction. Although the building docu-ments can help to determine the bulk materials (i.e.bricks, concrete, sand or gravel), the on-site investi-gations enable more specific information to be gath-ered about the location and concentration of tracematerials in buildings. A comparison of the collecteddata with the data in official reports, which were pro-vided by the demolition contractors, indicated insuffi-cient reliability of the officially reported figures.Therefore, a comprehensive material characterizationof demolished buildings based on these data is imposs-ible. The implications of the uncertainties in thereported data go even beyond this project, as thereported data are collected, aggregated and publishedin national waste management plans.
Further research should apply the described approachof document-based evaluation and on-site investigationto more buildings and different building types to enlargethe data set about the material composition of build-ings. To allow investigations of a higher number ofbuildings and to generate comparative values for build-ings’ composition, other approaches for data gener-ation should also be considered. Building componentcatalogues, for instance, seem promising in thisregard, especially for newer buildings. In future, com-bining the document-based evaluation with knowledgeand data from on-site investigations (e.g. specificmaterial intensities of trace materials present in differ-ent heating, ventilation, air-conditioning or electronicinstallations) seems appropriate to reduce effortswhen characterizing a larger sample of buildings.
A reliable data set is considered a prerequisite to sub-sequently link the information that is gathered at abuilding level with GIS data. To characterize theentire building stock of Vienna, different GIS datasets about building age and utilization will be com-bined, and depending on the building category (age/utilization) specific material intensities (kg/m3 grossvolume) can be assigned. On-site investigations areplanned for 20 buildings. The document-basedapproach combined with information about tracematerial intensities (obtained from the 20 on-site inves-tigations) will be applied to a sample of about 200demolition projects.
Determining thematerial composition of buildings
11
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
FundingThe presented work is part of a large-scale researchinitiative on anthropogenic resources (ChristianDoppler Laboratory for Anthropogenic Resources)and was funded by the Federal Ministry of Science,Research, and Economy and the National Foundationfor Research, Technology and Development. Industrypartners co-financing the research initiative are AltstoffRecycling Austria AG (ARA), Borealis group, voestal-pine AG, Wien Energie GmbH, Wiener Kommunal-Umweltschutzprojektgesellschaft mbH, and WienerLinien GmbH & Co. KG.
ReferencesAustrian Standards. (2006). Schadstofferkundung von Gebauden
vor Abbrucharbeiten.Baccini, P., & Pedraza, A. (2006). Bestimmung von Materialge-
halten in Gebauden Bauwerke als Ressourcennutzer undRessourcenspender in der Langfristigen Entwicklungurbaner Systeme. Zurich: vdf Hochschulverlag AG.
Bergmeister, K., Worner, J. D., & Fingerloos, F. (2009). Kon-struktiver hochbau – aktuelle massivbaunormen. Berlin:Ernst und Sohn.
Bergsdal, H., Bohne, R. A., & Brattebø, H. (2007). Projection ofconstruction and demolition waste in Norway. Journal ofIndustrial Ecology, 11(3), 27–39. doi:10.1162/jiec.2007.1149
Blengini, G. A. (2009). Life cycle of buildings, demolition andrecycling potential: A case study in Turin, Italy. Buildingand Environment, 44(2), 319–330. http://dx.doi.org/10.1016/j.buildenv.2008.03.007
BMLFUW. (2011). Bundes-Abfallwirtschaftsplan 2011. Vienna:Bundesministerium fur Land- und Forstwirtschaft, Umweltund Wasserwirtschaft (BMFLUW).
Clement, D., Hammer, K., Schnoller, J., Daxbeck, H., &Brunner, P. H. (2011). Wert- und schadstoffe in wohnge-bauden. Osterreichische Wasser- und Abfallwirtschaft,63(2–3), 61–69.
Cochran, K., Townsend, T., Reinhart, D., & Heck, H. (2007).Estimation of regional building-related C&D debris gener-ation and composition: Case study for Florida, US. WasteManagement, 27(7), 921–931. doi:http://dx.doi.org/10.1016/j.wasman.2006.03.023
Deilmann, C., & Gruhler, K. (2005). Stoff-und Energieflusse vonGebauden und Infrastrukturen als Grundlage fur ein vor-
ausschauendes szenariogeleitetes Stoffstrommanagement.Osterreichische Wasser-und Abfallwirtschaft, 57(7–8),103–109.
Gorg, H. (1997). Entwicklung eines Prognosemodells fur Bauab-falle als Baustein von Stoffstrombetrachtungen zur Krei-slaufwirtschaft im Bauwesen. Darmstadt, Germany: Vereinzur Forderung des Instituts WAR – WasserversorgungAbwassertechnik Abfalltechnik Umwelt- und Raumplanungder TH Darmstadt.
Gruhler, K., Bohm, R., Deilmann, C., & Schiller, G. (2002). Sof-flich-energetische Gebaudesteckbriefe – Gebaudevergleicheund Hochrechnungen fur Bebauungsstrukturen. Dresden:Institut fur okologische Raumentwicklung.
Kohler, N., & Hassler, U. (2002). The building stock as aresearch object. Building Research & Information, 30(4),226–236. doi:10.1080/09613210110102238
Kohler, N., Hassler, U., & Paschen, H. (1999). Stoffstrome undKosten in den Bereichen Bauen und Wohnen. Heidelberg:Springer.
Kohler, N., & Yang, W. (2007). Long-term management ofbuilding stocks. Building Research & Information, 35(4),351–362.
Krook, J., Carlsson, A., Eklund, M., Frandegard, P., & Svensson,N. (2011). Urban mining: Hibernating copper stocks in localpower grids. Journal of cleaner production, 19(9), 1052–1056.
Lichtensteiger, T., & Baccini, P. (2008). Exploration of urbanstocks. detail, 5(6), 16.
McCauley-Bell, P., Reinhart, D., Sfeir, H., & Ryan, B. (1997).Municipal solid waste composition studies. PracticePeriodical of Hazardous, Toxic, and Radioactive WasteManagement, 1(4), 158–163. doi:10.1061/(ASCE)1090–025X(1997)1:4(158)
Microsoft. (2014). Bing Maps – Screenshots of Vienna. doi:http://www.bing.com/maps/
Muller, A. (2013). Das ressourcenpotenzial von Bauabfallen(article 1691). Mullhandbuch, 3, 1–32.
Rentz, O., Seemann, A., & Schultmann, F. (2001). Abbruch vonWohn- und Verwaltungsgebauden – Handlungshilfe. Karls-ruhe: LFU – Landesanstalt fur Umweltschutz Baden-Wurttemberg.
Wiener Landtag.. (2012). Gesetz uber die Vermeidung undBehandlung von Abfallen und die Einhebung einer hiefurerforderlichen Abgabe im Gebiete des Landes Wien(Wiener Abfallwirtschaftsgesetz – Wr. AWG).
Wittmer, D., & Lichtensteiger, T. (2007). Development ofanthropogenic raw material stocks: A retrospectiveapproach for prospective scenarios. Minerals and Energy –Raw Materials Report, 22(1–2), 62–71.
Kleemann et al.
12
Dow
nloa
ded
by [
TU
Tec
hnis
che
Uni
vers
itaet
Wie
n] a
t 05:
06 1
5 D
ecem
ber
2014
2nd
Paper
GIS-based analysis of Vienna’s material stock in buildings
Fritz Kleemann, Jakob Lederer, Helmut Rechberger, and Johann Fellner
Journal of Industrial Ecology
DOI: 10.1111/jiec.12446
R E S E A R C H A N D A N A LYS I S
GIS-based Analysis of Vienna’s MaterialStock in BuildingsFritz Kleemann, Jakob Lederer, Helmut Rechberger, and Johann Fellner
Summary
The building stock is not only a huge consumer of resources (for its construction andoperation), but also represents a significant source for the future supply of metallic andmineral resources. This article describes how material stocks in buildings and their spatialdistribution can be analyzed on a city level. In particular, the building structure (buildingsdifferentiated by construction period and utilization) of Vienna is analyzed by joining availablegeographical information systems (GIS) data from various municipal authorities. Specificmaterial intensities for different building categories (differentiated by construction periodand utilization) are generated based on multiple data sources on the material compositionof different building types and combined with the data on the building structure. Utilizingthese methods, the overall material stock in buildings in Vienna was calculated to be380 million metric tonnes (t), which equals 210 t per capita (t/cap). The bulk of thematerial (>96%) is mineral, whereas organic materials (wood, plastics, bitumen, and soon) and metals (iron/steel, copper, aluminum, and so on) constitute a very small share, ofwhich wood (4.0 t/cap) and steel (3.2 t/cap) are the major contributors. Besides the overallmaterial stock, the spatial distribution of materials within the municipal area can be assessed.This research forms the basis for a resource cadaster, which provides information aboutgross volume, construction period, utilization, and material composition for each building inVienna.
Keywords:
building materialbuilding stockgeographic information systems (GIS)industrial ecologyurban metabolismurban mining
Supporting information is linkedto this article on the JIE website
Introduction
The last decade has shown an increasing interest in thebuilding stock (material stock of the built environment) as aresearch object (Kohler and Hassler 2002). Studies have de-termined the gross volume (GV), built-in materials, and age aswell as renovation intervals of buildings in order to estimate thedemand for heating and cooling of buildings or renovation re-quirements of old structures (Dall et al. 2012; Kavgic et al. 2010;Fabbri et al. 2012; Sandberg et al. 2011; Pauliuk et al. 2013; Huet al. 2010a; Sandberg and Brattebø 2012). Within the life cy-cle of a building, all these examples can be allocated to the usephase. Another way to look at the building stock is to focus on
[After initial online publication, the article was corrected to adjust for the inadvertent use of unrepresentative data on lead from buildings containingx-ray rooms. The article was also corrected to adjust for the erroneous use of unverified data on zinc from a German source.]Address correspondence to: Fritz Kleemann, Christian Doppler Laboratory for Anthropogenic Resources, Technische Universitat Wien, Karlsplatz 13/226.2, 1040 Wien,Austria. Email: [email protected], Web: http://iwr.tuwien.ac.at/anthropogene-ressourcen/home.html
© 2016 by Yale UniversityDOI: 10.1111/jiec.12446 Editor managing review: Seiji Hashimoto
Volume 00, Number 0
the end-of-life phase, particularly the projection of the quan-tity and quality of construction and demolition (C&D) wastes,and subsequently the consideration of the building stock as a fu-ture anthropogenic resource deposit for secondary raw materials(Sartori et al. 2008; Bergsdal et al. 2007; Hashimoto et al. 2009).
Studies on the building stock focus on a multitude of issues:different construction materials; building characteristics; con-sideration of different scales (cities, countries); and the use ofdifferent data sets (e.g., statistical data, geographical informa-tion systems [GIS] data sets).
In Austria, the built environment has been estimated on anational level, based on available statistical data by Stark and
www.wileyonlinelibrary.com/journal/jie Journal of Industrial Ecology 1
R E S E A R C H A N D A N A LYS I S
colleagues (2003). Their approach combines different statisti-cal data on the use of construction materials (mineral, organic,and metallic). Obernosterer and colleagues (1998) conducteda top-down analysis of the urban metabolism of Vienna. Fur-ther studies on a city, federal state, or other regional level usingstatistical data have been carried out in different countries.Schneider and Rubli (2007) and Staubli and Winzeler (2011)determined the stock of minerals embedded in infrastructureand buildings in the urban canton of Zurich, Switzerland, apply-ing a dynamic model using historical data sets. Other materialsthan minerals (e.g., plastics and metals) are not considered intheir model. Materials in lower concentrations such as copperare targeted by Bader and colleagues (2011) for Switzerland andKral and colleagues (2014) for the cities of Taipeh (Taiwan) andVienna (Austria). Both examples consider not only the materialstock present in buildings, but also such stock in infrastructure,electronic devices, and other items. This wide and varied focusleads to a loss in detail when it comes to the sole investigationof the building stock. Van Beers and Graedel (2007) analyzedthe spatial distribution of in-use stocks of copper and zinc inAustralia by combining concentrations of the respective metalswith GIS. Lichtensteiger and Baccini (2008) and Baccini andPedraza (2006) explored the national building stock in Switzer-land using their so-called ark-house method, where buildingsare characterized by utilization and construction period, whichwould potentially allow quantities of demolition waste to bepredicted. Hu and colleagues (2010a, 2010b) developed a dy-namic building model based on statistical data for the urbanhousing stock in rural and urban China in general and Beijingin particular, focusing on C&D wastes as well as iron and steel.Hashimoto and colleagues (2007, 2009) projected not only thefuture demand of construction minerals in Japan, but also theamount of C&D waste minerals that can serve as potentialsecondary raw materials for the construction industry. Theirapproach is based on input-output analysis and statistical dataon stocks (i.e., total floor area). Bergsdal and colleagues (2007)projected the amount of C&D waste from construction andrenovation and related demolition activities in the residentialhousing sector in Norway for a period of 23 years (1995–2018)based on statistical data on the building stock (floor area, age)and historical building materials consumption. They consid-ered the most relevant materials: minerals, plastics, metals, andwood. On a larger scale, Wiedenhofer and colleagues (2015)modeled stocks and flows of minerals for residential buildingsand transportation networks in the European Union (EU) 25based on statistical data on floor sizes and building age, on theone hand, and specific material composition for different resi-dential buildings, on the other. With a few exceptions, most ofthe studies are based on statistical data sets either focusing onspecific material fractions (i.e., minerals, metals) or on selectedbuilding types (i.e., residential buildings, but not industrial orcommercial buildings). The reason for the latter might be thebetter availability of statistical data.
Even though the use of statistical data sets, as utilized inmost of these studies, allows the stock size as well as the grossgeneration of C&D wastes (potential secondary raw materi-
als) to be estimated, it does not allow a detailed localizationof both the material stock as well as the generation of C&Dwastes. By employing GIS data, it is possible to overcome thisshortcoming. However, only a few studies using GIS for thepurpose of determining and localizing the building stock as wellas the generation of C&D wastes as potential secondary rawmaterials exist so far. A spatial and temporal assessment of con-struction material in buildings and infrastructure was carriedout by Tanikawa and Hashimoto (2009) in two urban areas of8 square kilometers (km²) and 11 km² in Japan and the UK,using a four-dimensional GIS database that includes not onlyspatial (size and location of a building), but also temporal infor-mation (i.e., year of construction of a building). A recent studyof Tanikawa and colleagues (2015) presents a project of howto map construction material of buildings and infrastructureby using GIS in Japan. Meinel and colleagues (2009) workedon the establishment of a database on buildings in Germanybased on building data obtained from topographic maps, digitalmaps, and statistical data. The system is intended to be usedas a nation-wide monitoring system of settlement and openspace development. Marcellus and colleagues (2012) utilizedGIS to track construction material stocks and flows of LEEDcertified buildings in Pennsylvania and Philadelphia (UnitedStates) in order to foster material recycling efficiencies. Forthe same region, Marcellus-Zamora and colleagues (2015) usedGIS to characterize buildings described by land-use type. Gruenand colleague (2009) described plans of how three-dimensionalbuilding and city models can be generated from figurativeand semantic sources and how they can be integrated intoGIS systems.
All of these studies show the strengths of applying GIS datasets in the determination of selected parts and features of thebuilding stock in cities and countries. What some of them pointout is the difficulty of combining different GIS data sets (e.g.,one data set on spatial information and another data set ontemporal information), on the one hand, as well as the lack ofinformation on the most important attributes of buildings whenit comes to the determination of the stock of, for instance,select quantitative minor materials (e.g., metals, plastics) inbuildings, on the other. This article presents an approach foranalyzing both the buildings structure and the material stockin buildings using the case study of the city of Vienna in Aus-tria. The building structure is analyzed by combining differentGIS data sets, which allows a localization of different build-ing types. Data on the material composition of different build-ing types, including bulk materials (e.g., minerals, metals) andquantitatively minor, but, in terms of their resource and envi-ronmental potential, relevant materials (e.g., plastics, copper,aluminum, and asbestos) are collected through various sourcesand approaches, ranging from the analysis of documents andon-site investigations to the review of relevant literature. Thestudy aims at mapping information about the material stockin buildings in the city of Vienna as a basis for efficient re-source management (e.g., projecting the flow of wastes and sec-ondary raw materials from demolition activities in the buildingsector).
2 Journal of Industrial Ecology
R E S E A R C H A N D A N A LYS I S
Material and Methods
The study at hand combines different approaches to analyzethe building structure in Vienna with regard to size (GV), con-struction period, and utilization (building usage), as well as togenerate a database about the material intensities (kilogramsper cubic meter [kg/m³] GV) of different building categories(construction period [i], utilization [j]). Subsequently, data onbuilding structure and material intensities of different buildingcategories are combined to assess and localize the material stockin all buildings in Vienna (see equations below).
Mm i, j = (GVi, j x MIm i, j
)
M =k, l ,n∑
i =1, j =1, m=1
Mm,i, j
Mm i, j = mass of material m built in in the building categoryi, j [kg]
GVi, j = gross volume of building category i, j [m³]MIm i, j = specific material intensity for material m for the
building category i, j [kg/m³]
Building category:
i = utilizationj = construction period
The total mass Mm i, j of material m in the building categoryi, j equals the product of the specific material intensityMIm i, j and the respective gross volume GVi, j . By adding upall material masses (for each building category and each mate-rial), the total material stock M [kg] is determined.
Building Structure
As described above, in a first step, the building structure inVienna is analyzed aiming at the generation of a city-wide GISdata set that comprises information about GV, constructionperiod, and utilization for each building. The steps conductedto combine the different data sets are explained in the following.
Step 1: The so-called multipurpose map of the municipaldepartment for city surveying (MA 41 [2013]) includes, amongother data, information about the area and the relative heightof each building or building part in Vienna. According to MA41, the data of the multipurpose map are updated continuouslyand, at most, 3 years old. The data set used for this article wasobtained in 2013. It is based on terrestrial surveying and there-fore the most accurate data with regard to buildings. Hence, thismap is used as a target data set, meaning that other informa-tion about buildings (such as construction period, utilization)is added to this existing data set. The relative height in themultipurpose map is defined as the distance between groundlevel and eaves, therefore basement floors and the roofs of thebuildings are not represented.
Because specific material intensities of buildings (expressedas mass [kg] per GV [m³]; see the section on Material Intensities),
which are subsequently combined with the GIS data, are relatedto the whole building (including basement and roof), expertestimates about the average height of basements and roofs (foreach building category) had to be added to the relative heightgiven in the multipurpose map.
Step 2: To assign information about utilization and con-struction period to each building, a data set provided by themunicipal department for urban planning and land use (MA21 2013) is spatially joined to the multipurpose map (step 1).These data are constantly updated and might be, at maximum,10 years old.
Step 3: With respect to utilization and construction period,data gaps exist for some areas of the city (for 18% of the buildingareas no utilization and for 21% of the building areas no con-struction period can be assigner through step 2). To fill thesegaps, data from the municipal department for city developmentand planning (MA 18 [2013]) on the construction age of build-ing blocks as well as the generalized zoning plan (utilization)of the MA 21 (2013) are used. Both data sets contain infor-mation only on a building block level; however, for buildingswhere no other information is available, this approximation ona larger scale represents the only possibility for categorizationand characterization.
Step 4: Another data source on the utilization and con-struction period of buildings represents the building and flatindex of the municipal building inspection department (MA37a [2013a]). The data on new buildings (constructed after2008) are especially robust and thus added to the data set to im-prove its quality. These data, provided by MA 37, are constantlyupdated.
Step 5: Based on the available GIS data on constructionperiod and utilization, building categories have been defined.Altogether, 15 categories, differentiating between constructionperiod and utilization of buildings, are distinguished: fivedifferent construction periods (before 1918, 1919–1945,1946–1976, 1977–1996, and after 1997), which correspondto the available GIS data, and three different categories forutilization (residential, commercial, and industrial). Existinginformation about the utilization has been aggregated as theoriginal data set distinguished between 78 categories. This stepwas necessary because, on the one hand, material intensitiesare not available for all 78 categories and, on the otherhand, differences in composition might not be significantbetween all these categories. Detailed information regardingthe categorization of buildings is provided in the SupportingInformation available on the Journal’s website.
Material Intensities
Given that the material composition varies depending onthe utilization and construction period of buildings, specificmaterial intensities per cubage can be determined and assignedto different building categories. Viennese Wilhelminian stylehouses (Grunderzeithaus), for example, are usually constructedwith bricks and are characterized by ceiling heights of between3.5 and 4.5 meters (m). Residential buildings from the 1960s and1970s are mainly constructed from concrete with lower ceiling
Kleemann et al., GIS-based Analysis of Material Stock in Buildings 3
R E S E A R C H A N D A N A LYS I S
heights (<3 m), whereas newer buildings are often more com-plex with regard to material composition, with higher amountsof plastics and composite materials (e.g., insulation).
A complementary approach—described in the followingsubsections—is used to derive information about the materialintensities of different buildings using various sources. This in-formation is subsequently used to establish a database aboutthe material intensities of the 15 different building categoriesdefined.
In particular, the approach chosen considers (1) 14 casestudies, that is, buildings for which the material compositionhas been analyzed in detail on-site, (2) around 40 buildingswhose composition was estimated based on construction files(plans, verdicts, statics, and structural-physical documents), (3)detailed data of 12 buildings recently constructed (e.g., lifecycle assessment [LCA] inventories, tender documents), and(4) literature in the case that (i) and (ii) did not yield sufficientinformation.
Case StudiesAs described in Kleemann and colleagues (2016), major de-
molition projects in Vienna are investigated and built-in mate-rials are quantified before the demolition of the building basedon an analysis of available documents (construction plans, ex-pert´s reports) and on-site investigation, including selectivesampling. Available documents usually include not only con-struction plans about the existing buildings, but also reports onpollutants within a building, or a waste management conceptalready focusing on the demolition of the building may containvaluable information. During the on-site investigation, data onbuilt-in materials are collected through labor-intensive mea-surements and selective sampling (weighting, and measuringof components such as windows, doors, partitions, ceiling sus-pensions, floor and roof constructions, wires, pipes, and so on).Depending on the conditions on-site, the following strategiesare applied:
� Sampling of representative areas such as dwelling units orfloors
� Investigation of rising and distribution mains to efficientlycollect data on sanitary and electrical installations
� Investigation of the material composition of typical unitssuch as doors, windows, partitions, and heaters, which arecommon in the building
� Notional sections set length and cross-wise (in some casesalso horizontally) through rooms with numerous installa-tions. Along the sections, all orthogonal running wires,pipes, and other installations are counted and dimensionsare documented. For the length of the installations, thedimensions of the particular room (perpendicular to thecross-section) are assumed.
� Material inventory of specific areas of buildings
Based on the data collected, information about the differentbuilt-in materials is aggregated and the mass is calculated basedon volume, area, or number of the particular material as well ason data on its specific density or weight. Relevant information
on the density of building materials or the weight of installationsis either obtained by sampling or from literature on constructionmethods and design norms. The results of the case studies arespecific material intensities of different buildings. Until now,14 major buildings have been investigated in detail.
Construction Files of Demolished BuildingsIn order to increase the sample size (number and GV of
buildings characterized), construction files (mainly construc-tion plans) of buildings reported to be demolished in the cityof Vienna are analyzed, leading to an additional sample of 40buildings (overall GV of 780,000 m³). The documents wereprovided by the MA 37b (2013b). Depending on the qual-ity of the documents, building materials used for wall, ceiling,and sometimes also for roof or floor construction can be de-termined. Data gaps, mostly regarding low-volume materials(e.g., plastics in installations and fittings, copper for electricity,and wood/aluminum/PVC [polyvinyl chloride] for windows),are filled using specific material intensities obtained from the14 case studies analyzed in detail.
Data on New BuildingsIn order to complete the data set on the specific material
intensities with regard to newer buildings, which are rarely de-molished, existing data on the material composition of thesebuildings are utilized. This material data have been derivedfrom LCAs, tendering documents, construction plans, and ac-counting documents. Altogether, data of 12 buildings have beenanalyzed, covering a GV of 170,000 m³.
LiteratureTo put the material intensities generated in context with
other studies carried out, literature data on the material compo-sition of buildings are utilized. In particular, the following stud-ies have been analyzed in detail: Studies addressing the materialcomposition of buildings in Germany have been carried out byAlbrecht and colleagues (1984) and Gruhler and colleagues(2002). Lippok and Korth (2007) provided material intensitiesof different building categories in order to plan upcoming demo-lition work. Focusing on the federal state of Baden-Wurtemberg,a guideline was prepared by Rentz and colleagues (2001) to es-timate material amounts and associated deconstruction costs.Schulze and colleagues (1990) assessed buildings of the formerGerman Democratic Republic, whereas Wedler and Hummel(1947) characterized buildings in a very detailed manner in or-der to reuse debris in German cities after air raids in World WarII. More recent studies were carried out by Baccini and Pedraza(2006) in Switzerland and Blengini (2009) in Italy, the lat-ter of whom conducted a study that closely documented wastestreams of a building during the demolition process. Only dataon buildings that could be built in similar form in Vienna areused. High-rise slabs, as typically found in the former GermanDemocratic Republic, are, for example, excluded given thatsuch buildings do not exist in Vienna.
Depending on the utilization and construction period ofthe building, specific material intensities per GV (kg/m³) are
4 Journal of Industrial Ecology
R E S E A R C H A N D A N A LYS I S
Figure 1 Information flows of different data sources used in this research.
assigned. The average specific material intensities are derivedfrom the data sources previously described. Thereto, all dataobtained are translated into material intensities and assigned tothe respective building category. Subsequently, for each build-ing category, average material intensities are calculated as amedian value (out of the assigned data sets—see the Support-ing Information on the Web). These medians are finally used inconjunction with the GIS data to calculate the overall materialstock of buildings in Vienna. Figure 1 summarizes the differentdata sources used in this research—arrows indicate informa-tion flows. The shading indicates whether the data used wasgenerated specifically for this study or existing data, originallygenerated for different purposes, could be utilized.
Results
Building Structure
The results of the analysis of the buildingstructure of the cityof Vienna are based on a data set with 580,000 dates. Each daterepresents a building or a part of a building. The splitting up ofbuildings into several parts is necessary in cases where parts ofone building are characterized by different heights. Accordingto the MA 23 (2011), there are, in total, around 165,000 build-ings in the city of Vienna. The overall volume of buildings inVienna amounts to 860 million m³ GV (figure 2), with a large
share being used for residential purposes (540 million m³ GV).With regard to the age of the building stock, it is noticeablethat the largest share of buildings was constructed before 1918(290 million m³ GV). Because of the fact that the city of Vi-enna experienced a real boost in construction activity duringthe second half of the nineteenth century (which went alongwith a quadrupling of the population: 1850: 551.000 inhabi-tants and 1916: 2,240,000 inhabitants, according to StatistikAustria [2014], figures refer to the present city border), a signif-icant share (around 25% to 30%) of Vienna´s buildings stockwas erected between 1850 and 1918. Figure 2 illustrates thedistribution of the GV among the different building categories;an illustration showing the share of utilization and constructionperiod separately can be found in the Supporting Informationon the Web (MA 21 [2013]). The data set obtained from thecity of Vienna refers to the year 2013.
Material Intensities of Different Buildings Types
The analysis of all material data collected resulted in specificmaterial intensities for different building categories, expressedin material mass [kg] per GV [m³] (table 1). The materialscategories (mineral, organic, and metal) are further dividedinto specific materials such as concrete, bricks, glass, asbestos,wood, plastics, bitumen, iron, aluminum, or copper. Detailed
Kleemann et al., GIS-based Analysis of Material Stock in Buildings 5
R E S E A R C H A N D A N A LYS I S
Figure 2 Gross volume of different building categories in the year 2013.
Table 1 Specific material intensities (given in kg/m³ GV) of differentbuilding categories in Vienna (rounded to two significant digits)
Period of Mineral Organicconstruction Utilization materials materials Metals Total
Residential 390 19 3.1 410Before 1918 Commercial 430 3.7 4.4 440
Industrial 280 5.8 8.8 300Residential 410 13 4.8 430
1919–1945 Commercial 340 7.1 6 360Industrial 320 28 5.8 350Residential 430 6.5 7.3 450
1946–1976 Commercial 350 7.6 5.7 360Industrial 340 7.6 13 350Residential 430 6.7 7.1 460
1977–1996 Commercial 380 1 13 400Industrial 170 1 15 180Residential 380 10 15 410
After 1997 Commercial 320 5.7 10 340Industrial 290 5.6 13 310
Note: kg/m³ GV = kilograms per cubic meter gross volume.
information about the data set on the specific material intensi-ties of different building categories is provided in the SupportingInformation on the Web.
The material intensities of the single building categories(given in table 1) are based on different numbers of informa-tion sources, meaning that for some building categories specificmaterial intensities have been derived from many different data
sets, whereas for one building category (after 1997, industrial,representing 0.5% of the total GV) no data were available atall. For this building category, a material mixture of industrialbuildings from 1977 to 1996 and residential and commercialbuildings from after 1997 was assumed. This problem is furtherdiscussed in the section on Uncertainties and Data Gaps.
Material Stock
By combining information about the building structure andthe specific material intensities of different building categories,the overall stock in buildings in Vienna in the year 2013 wascalculated to 380 million metric tonnes (t). The bulk of thematerial (>96%) is mineral, mainly represented by concrete(150 million t) and bricks (130 million t) and mortar (50 mil-lion t). Organic materials and metals constitute a very smallshare (4%) of the overall material stock in buildings, of whichwood and steel are the major contributors (figure 3). A figureshowing the share of the different building categories of the to-tal material stock is provided in the Supporting Information onthe Web.
Considering the Viennese population of 1.8 million, the ma-terial stock in buildings per capita amounts to approximately210 t per capita (t/cap). Because Vienna is an economic centerin Austria with around 250,000 commuters, a part of the build-ing stock also services people residing outside Vienna. Thisapplies mainly to commercial and industrial buildings. Table 2summarizes per capita figures on the buildings stock in the cityof Vienna.
6 Journal of Industrial Ecology
R E S E A R C H A N D A N A LYS I S
Figure 3 Material composition of the total building stock in the city of Vienna in the year 2013.
Table 2 Per capita figures on the materials present in buildings in Vienna (t/cap) (rounded to two significant digits)
Mineral 200 Organic 5.5 Metal 3.3Concrete 83 Wood 4.0 Iron/steel 3.2Bricks 70 Various plastics 0.35 Aluminium 0.045Mortar/plaster 29 Bitumen 0.22 Copper 0.031Mineral fill 7.8 Carpet 0.19 Lead 0.0023Slag fill 3.4 Heraklit 0.16 Brass 0.0012Gravel/sand 2.7 Asphalt 0.13Plaster boards/gypsum 0.74 PVC 0.10Natural stone 0.72 Polystyrene 0.076Foamed clay bricks 0.71 Paper/cardboard 0.059Ceramics 0.42 Laminate 0.029(Cement) asbestos 0.34 Linoleum 0.014Glass 0.22Mineral wool 0.21Mineral wool boards 0.017 Total 210
Note: t/cap = tonnes per capita; PVC = polyvinyl chloride.
Obernosterer and colleagues (1998) carried out a top-downanalysis about the urban metabolism of Vienna, which resultedin an estimated per capita stock of 315 t/cap for both build-ings and infrastructure. Iron stocks amount to a total of 3.5t/cap, whereas lead totals 0.23 t/cap. Compared to the presentstudy, the amount of lead seems overestimated. A comprehen-sive dynamic material flow study on aluminum stocks and flowsin Austria was recently published by Buchner and colleagues(2015), resulting in an aluminum stock of 90 to 160 kg/capfor both buildings and infrastructure. The main reason for theconsiderably higher values in comparison to the results of thepresent study (44 kg/cap for buildings only) is the inclusion ofinfrastructure. Further, the building stock in Vienna is obviouslyolder and comprises a lower share of detached houses comparedto the average in Austria, both resulting in a lower aluminumshare.
The question about what material can be expected to bestocked in Vienna’s infrastructure based on other studies car-ried out is difficult. Especially problematic with regard to min-eral materials stocked in infrastructure are the big differencesbetween urban and rural areas when national statistics are usedand the question of what should be included in the calculation(terrain shaping, backfilling, and so on). A detailed study aboutnetwork infrastructure in the city of Vienna based on availableGIS data is currently in progress.
Spatial Distribution of Material Stock
From a resource and waste management perspective, the spa-tial distribution of materials is at least as interesting as know-ing about the overall amount of materials built in. Informa-tion about the spatial distribution of materials on a buildinglevel allows the amount and composition of C&D waste arising
Kleemann et al., GIS-based Analysis of Material Stock in Buildings 7
R E S E A R C H A N D A N A LYS I S
Figure 4 Spatial distribution of material intensity ([a] minerals and [b] wood) in buildings in Vienna (kg/m² built-up area). kg/m2 =kilograms per square meter.
during the demolition of specific buildings to be forecasted andthe value/costs arising with the disposal/recovery of C&D wasteto be estimated. Moreover, projections of secondary raw mate-rials potentially arising from C&D activity can be made.
The combination of GIS data with specific material inten-sities further allows the establishment of a resource cadaster,which illustrates the spatial distribution of materials through-out the city. Figure 4 shows the material intensity of buildings
8 Journal of Industrial Ecology
R E S E A R C H A N D A N A LYS I S
Figure 5 Net accumulations of selected materials in the building stock over time: (a) bricks, concrete, wood, and steel; (b) wood steel.
in Vienna expressed as mass per built-up area of the buildingsfor mineral materials and wood.
Whereas the mineral material distribution mainly dependson the height of the buildings, the distribution of wood alsodepends on the age of the buildings. Figure 5 illustrates thenet accumulation of selected materials in the building stockof Vienna and indicates that wood is mainly present in olderbuildings and that its stock increases at a much lower rate incomparison to iron and steel, which correlates with the in-creased use of concrete. The figure does not take losses throughdemolition, war, or natural disasters into account, but ratherdisplays the current building stock and its net accumulationover time.
Discussion
The results demonstrate that it is possible to characterize thematerial composition of the Viennese building stock by com-bining available data sets from the municipal authorities withcollected data on the material intensities of different buildingstypes.
A lot of effort was put into the collection of data on the ma-terial intensities of buildings from different sources. However,this laborious bottom-up approach was necessary given that thedata availability with respect to material intensities of differentbuildings was very limited. In order to continuously refine thedata on materials with regard to different buildings categories,
Kleemann et al., GIS-based Analysis of Material Stock in Buildings 9
R E S E A R C H A N D A N A LYS I S
Table 3 Estimated uncertainties for different data sources used inthis research
Material intensities for Estimateddifferent building categories uncertainty (%)
Inspection and selective sampling: case studies 5–20Available documents (e.g., construction
plan): case studies<5
Available documents (e.g., constructionplan): MA 37b (2013b)
5–20
Data new buildings <5Data from literature 20–50Building structure (building categories)Area and height of buildings (MA 41 2013) <5Construction period and utilization (MA 21
2013)5–20
Construction period and utilization forbuildings for which information is onlyavailable at building block level (MA 18and MA 21 2013)
20–50
Construction period and utilization of newbuildings (MA 37a [2013a])
5–20
further research is necessary. In this respect, it would be benefi-cial to collaborate with other research fields studying buildings,but not necessarily focusing on their material composition (e.g.,on energetic performance of buildings, heritage buildings, andso on). Given that building information modeling is increas-ingly used when constructing new buildings, it can help to easilyaccess information about the material composition of buildingsand estimate demolition waste (Cheng and Ma 2013).
Uncertainties and Data Gaps
Uncertainties of the material stock estimation arise at dif-ferent stages of data collection and analysis. Table 3 showsestimated uncertainty ranges of the different data sources usedin this research in analogy to figure 1. Possible sources for un-certainties include the data collection during the investigationof buildings and building parts, outdated construction plans of
demolished buildings, use of literature from other regions, orthe used GIS data. The data gaps for the construction period(21% of the area) and utilization (18% of the area) at a build-ing level were filled with data available at a building blocklevel, which are, of course, characterized by higher uncertain-ties. Data about construction period and utilization are up to10 years old, whereas GIS data about newer buildings might notbe updated properly. Estimated uncertainties of material inten-sities for different building categories refer to materials such asconcrete, bricks/mortar, and constructional wood or steel. Low-volume materials, as used for fittings and installations, tend tohave higher uncertainty (see the Supporting Information onthe Web).
Because data on the average height of roofs and basementfloors for different building categories were compiled based onthe estimation of experts, larger uncertainties exist. Becausethere is no other source of information, the approach chosenrepresents the only possibility to predict the overall GV. Theshare of additional volume added by considering roofs and base-ment floors amounts to almost 30% of the overall GV (see theSupporting Information on the Web). Therefore, this part isquite significant for estimating the overall material stock inbuildings in Vienna. Details concerning expert estimation ofthe “additional” building volume (for roofs and basements) areprovided in the Supporting Information on the Web.
As indicated in figure 2, for a small share of the buildings(1.8% of all buildings), no information about construction pe-riod and utilization is available. For these buildings, it has beenassumed that their age and utilization are consistent with thetotal building stock. For buildings for which only informationabout the construction period was available, the utilization pat-tern has been assumed to be in accord with the data availablefor the respective construction period. An analogical procedurehas been applied to buildings with information on utilization,but with no data on the construction period.
Fortunately, averages about material intensities of buildingcategories, which constitute most of the overall GV, are basedon a higher number of buildings analyzed (see table 4). For in-dustrial buildings after 1997, however, no data were available.For this building category, it was assumed that its composition
Table 4 Contribution of the different building categories to the overall building volume (given in %) and number of buildings analyzed withrespect to their material composition
Residential Commercial Industrial No information
% of No. of buildings % of No. of buildings % of No. of buildings % of No. of buildingsUnit total GV analyzed total GV analyzed total GV analyzed total GV analyzed
Before 1918 24.7 12 + (8) 6.9 8 + (1) 1.2 2 + (1) 1.4 —1919–1945 5.4 6 + (7) 0.9 1 + (2) 0.7 1 + (1) 0.5 —1946–1976 16.9 16 + (15) 3.8 2 + (3) 2.3 0 + (1) 1.3 —1977–1996 9.6 4 + (7) 6.5 1 + (1) 2.7 0 + (1) 1.1 —After 1997 5.1 12 + (1) 3.3 3 + (0) 0.5 0 + (0) 0.5 —No information 1.4 — 0.8 — 0.5 — 1.8 —
Note: Parenthesized number represents data from literature. GV = gross volume.
10 Journal of Industrial Ecology
R E S E A R C H A N D A N A LYS I S
Figure 6 Resource cadaster including material information on a building level.
can be approximated by a mixture of industrial buildings from1977 to 1996 and residential and commercial buildings from af-ter 1997. The contribution of this category (industrial buildingsafter 1997) to the overall GV of buildings in Vienna is 0.5%only. Thus, vague estimates about the material intensities forthis category do not impact the overall material stock.
Because information about the building composition of dif-ferent building categories results partly from data collectedspecifically for this study (own investigations) and partly fromsecondary data (e.g., literature), the estimation of the overalluncertainty is complex and thus the subject of discussion in afollow-up article. Generally, data for matrix materials (mineralmaterials) are assumed to be more robust than for materials withsmall shares (metals, organics). Uncertainties associated withprimary data collection on the material composition of build-ings have been evaluated in Kleemann and colleagues (2016).
Conclusion and Outlook
This research has demonstrated an approach to quanti-fying and localizing the material stock in buildings in Vi-enna by combining data on the building structure and mate-rial intensities characteristic for different building categories.The results of this study provide the necessary basis for a re-source cadaster (figure 6), and with it, continuous monitor-
ing of the Viennese building stock as well as better infor-mation about the flows and stocks of materials throughoutthe city.
Combining the results of the study with information aboutthe demolition activity in the city allows the amount and com-position of wastes from demolition activities to be quantified,thus providing vital information on the materials potentiallyavailable for recycling.
The precision of the material stock analysis depends bothon the quality of data on the building structure as well as thedata on the material intensities of the different building cate-gories. Future improvements regarding the data on the buildingstructure mainly concern the size of the buildings and theirmapping in GIS. In this regard, the city of Vienna is currentlydeveloping a roof model for the entire city, which will allow theexact shape and volume of all roofs to be mapped. Exact dataon basement floors in buildings, however, will not be availablesoon and expert judgement, as carried out in this study, will beretained. With regard to the availability of GIS data, the city ofVienna is continuously expanding the number of openly avail-able data sets. In order to evaluate the accuracy of the studyresults, information about the timeliness of the data and alsoabout how they have been collected would be crucial.
To improve and maintain the database on the materialintensities of the different building categories defined, furtherresearch is necessary especially with regard to the material
Kleemann et al., GIS-based Analysis of Material Stock in Buildings 11
R E S E A R C H A N D A N A LYS I S
intensities of the building categories, presently under-represented by data sources (table 4). In a future step, the av-erage material intensities of different categories will be used toevaluate the material composition at the building level.
Acknowledgments
The presented work is part of a large-scale research ini-tiative on anthropogenic resources (Christian Doppler Lab-oratory for Anthropogenic Resources). The financial supportof this research initiative by the Federal Ministry of Science,Research, and Economy and the National Foundation for Re-search, Technology and Development is gratefully acknowl-edged. Industry partners co-financing the research center onanthropogenic resources are Altstoff Recycling Austria AG(ARA), Borealis group, voestalpine AG, Wien Energie GmbH,Wiener Kommunal-Umweltschutzprojektgesellschaft GmbH,and Wiener Linien GmbH & Co KG.
References
Albrecht, R., L. Paker, S. Rehberg, and Y. Reiner. 1984. Umweltent-lastung durch@: okologische Bau-und Siedlungsweisen [Environmen-tal relief through ecological building and settlement patterns].Wiesbaden, Germany: Bauverlag.
Baccini, P. and A. Pedraza. 2006. Bestimmung von Materialgehal-ten in Gebauden [Identification of material contents in build-ings]. In Bauwerke als Ressourcennutzer und Ressourcenspender inder Langfristigen Entwicklung urbaner Systeme [Buildings as resourceconsumer and resource donor]. Zurich, Switzerland: vdf Hochschul-verlag AG.
Bader, H. P., R. Scheidegger, D. Wittmer, and T. Lichtensteiger. 2011.Copper flows in buildings, infrastructure and mobiles: A dynamicmodel and its application to Switzerland. Clean Techologies andEnvironmental Policy 13(1): 87–101.
Bergsdal, H., R. A. Bohne, and H. Brattebø. 2007. Projection of con-struction and demolition waste in Norway. Journal of IndustrialEcology 11(3): 27–39.
Blengini, G. A. 2009. Life cycle of buildings, demolition and recyclingpotential: A case study in Turin, Italy. Building and Environment44(2): 319–330.
Buchner, H., D. Laner, H. Rechberger, and J. Fellner. 2015. Dynamicmaterial flow modeling: An effort to calibrate and validate alu-minum stocks and flows in Austria. Environmental Science & Tech-nology 49(9): 5546–5554.
Cheng, J. C. and L. Y. Ma. 2013. A BIM-based system for demolitionand renovation waste estimation and planning. Waste Manage-ment 33(6): 1539–1551.
Dall, G., A. Galante, and M. Torri. 2012. A methodology for theenergy performance classification of residential building stock onan urban scale. Energy and Buildings 48: 211–219.
Fabbri, K., M. Zuppiroli, and K. Ambrogio. 2012. Heritage buildingsand energy performance: Mapping with GIS tools. Energy andBuildings 48: 137–145.
Gruen, A., M. Behnisch, and N. Kohler. 2009. Perspectives in thereality-based generation, nD modelling, and operation of build-ings and building stocks. Building Research & Information 37(5–6):503–519.
Gruhler, K., R. Bohm, C. Deilmann, and G. Schiller. 2002. Sofflich-energetische Gebaudesteckbriefe-Gebaudevergleiche und Hochrech-nungen fur Bebauungsstrukturen [Material and energetic building char-acterization]. Band 38 vols, IOR-Schriften. Dresden, Germany: In-stitut fur okologische Raumentwicklung.
Hashimoto, S., H. Tanikawa, and Y. Moriguchi. 2007. Where will largeamounts of materials accumulated within the economy go?—Amaterial flow analysis of construction minerals for Japan. WasteManagement 27(12): 1725–1738.
Hashimoto, S., H. Tanikawa, and Y. Moriguchi. 2009. Framework forestimating potential wastes and secondary resources accumulatedwithin an economy—A case study of construction minerals inJapan. Waste Management 29(11): 2859–2866.
Hu, M., H. Bergsdal, E. van der Voet, G. Huppes, and D. B. Muller.2010a. Dynamics of urban and rural housing stocks in China.Building Research & Information 38(3): 301–317.
Hu, M., S. Pauliuk, T. Wang, G. Huppes, E. van der Voet, and D. B.Muller. 2010b. Iron and steel in Chinese residential buildings:A dynamic analysis. Resources, Conservation and Recycling 54(9):591–600.
Kavgic, M., A. Mavrogianni, D. Mumovic, A. Summerfield, Z. Ste-vanovic, and M. Djurovic-Petrovic. 2010. A review of bottom-upbuilding stock models for energy consumption in the residentialsector. Building and Environment 45(7): 1683–1697.
Kleemann, F., J. Lederer, P. Aschenbrenner, H. Rechberger, and J. Fell-ner. 2016. A method for determining buildings’ material compo-sition prior to demolition. Building Research & Information 44(1):51–62.
Kohler, N. and U. Hassler. 2002. The building stock as a researchobject. Building Research & Information 30(4): 226–236.
Kral, U., C. Y. Lin, K. Kellner, H. W. Ma, and P. H. Brunner. 2014.The copper balance of cities. Journal of Industrial Ecology 18(3):432–444.
Lichtensteiger, T. and P. Baccini. 2008. Exploration of urban stocks.Journal of Environmental Economics and Management 18(1): 41–48.
Lippok, J. and D. Korth. 2007. Abbrucharbeiten—Grundlagen,Vorbereitung, Durchfuhrung [Demolition work—Fundamentals,preparation, implementation]. Koln, Germany: Verlagsge-sellschaft Rudolf Muller.
MA 18. 2013. Magistratsabteilung 18—[Municipal department 18—City development and planning]. Vienna: GIS Building Data.
MA 21. 2013. Magistratsabteilung 21 Stadtteilplanung undFlachennutzung [Municipal department 21—Urban planning andland use). Vienna: GIS Building Data.
MA 23. 2011. Magistratsabteilung 23—Wirtschaft, Arbeit undStatistik [Municipal department 23—Economy, labor andstatistics]. Gebaude und Wohnungen—Statistiken [Buildingsand dwellings—Statistics]. www.wien.gv.at/statistik/verkehr-wohnen/wohnen/. Accessed 31 August 2015.
MA 37a. 2013a. Magistratsabteilung 37—[Municipal department 37—Building authority]. Vienna: GIS Building Data.
MA 37b. 2013b. Magistratsabteilung 37—[Municipal department 37—Building authority]. Vienna: Construction files of demolishedbuildings of the years 2012 and 2013.
MA 41. 2013. Magistratsabteilung 41—Stadtvermessung [Municipaldepartment 41—City surveying]. Vienna: GIS Building Data.
Marcellus, K., S. Spatari, and P. Gallagher. 2012. Utilizing GIS as ageospatial tool to inventory LEED certified buildings and Con-struction and Demolition (C&D) waste flows in the UnitedStates. Paper presented at Life-Cycle and Sustainability of CivilInfrastructure Systems: Proceedings of the Third International
12 Journal of Industrial Ecology
R E S E A R C H A N D A N A LYS I S
Symposium on Life-Cycle Civil Engineering, Hofburg Palace, 3–6 October, Vienna.
Marcellus-Zamora, K. A., P. M. Gallagher, S. Spatari, and H. Tanikawa.2015. Estimating materials stocked by land-use type in historicurban buildings using spatio-temporal analytical tools. Journal ofIndustrial Ecology. First published online 11 September 2015. DOI:10.1111/jiec.12327.
Meinel, G., R. Hecht, and H. Herold. 2009. Analyzing building stockusing topographic maps and GIS. Building Research & Information37(5–6): 468–482.
Obernosterer, R., P. Brunner, H. Daxbeck, T. Gagan, E. Glenck, C.Hendriks, L. Morf, R. Paumann, and I. Reiner. 1998. Materialsaccounting as a tool for decision making in environmental pol-icy: MAcTEmPo case study report. Urban metabolism, the cityof Vienna. Vienna: Institute for Water Quality and Waste Man-agement, Vienna University.
Pauliuk, S., K. Sjostrand, and D. B. Muller. 2013. Transforming theNorwegian dwelling stock to reach the 2 degrees Celsius climatetarget. Journal of Industrial Ecology 17(4): 542–554.
Rentz, O., A. Seemann, and F. Schultmann. 2001. Abbruch vonWohn- und Verwaltungsgebauden—Handlungshilfe [Demolitionof residential and administration buildings—Instruction], editedby L. f. U. Baden-Wurttemberg. Karlsruhe, Germany. Lan-desanstalt fur Umweltschutz Baden-Wurttemberg.
Sandberg, N. H. and H. Brattebø. 2012. Analysis of energy and carbonflows in the future Norwegian dwelling stock. Building Research &Information 40(2): 123–139.
Sandberg, N. H., H. Bergsdal, and H. Brattebø. 2011. Historical energyanalysis of the Norwegian dwelling stock. Building Research &Information 39(1): 1–15.
Sartori, I., H. Bergsdal, D. B. Muller, and H. Brattebø. 2008. Towardsmodelling of construction, renovation and demolition activities:Norway’s dwelling stock, 1900–2100. Building Research & Infor-mation 36(5): 412–425.
Schneider, M. and S. Rubli. 2007. Ressourcenmodell mineralischerBaustoffe auf der Ebene Stadt Zurich [Resource model of mineralbuilding material in Zurich]. Schlieren, Switzerland: Wertstoff-Borse GmbH.
Schulze, H.-J., P. Walther, and M. Schluter. 1990. Gebaudeatlas-Mehrfamilienwohngebaude der Baujahre 1880–1980 [Buildingatlas—Multi family houses built 1880–1980]. Berlin, DeutscheDemokratische Republik. Bau Bauakademie der DDR.
Stark, W., T. Vogel-Lahner, and W. Fruhwirth. 2003. BauwerkOsterreich—Endbericht [Built environment Austria—Final report].Vienna: GUA (Gesellschaft fur umfassende Analysen GmbH).
Statistik Austria. 2014. Statistisches Jahrbuch der Stadt Wien 2014 [Sta-tistical yearbook of the city of Vienna 2014]. Vienna: StatistikAustria.
Staubli, B. and R. Winzeler. 2011. Bau und Ruckbau im Kanton Zurich:Modell fur die Materialflusse 1900 bis 2008 [Construction anddeconstruction in the canton Zurich: Model for material flows].Mull und Abfall 6: 262–270.
Tanikawa, H. and S. Hashimoto. 2009. Urban stock over time: Spatialmaterial stock analysis using 4d-GIS. Building Research & Infor-mation 37(5–6): 483–502.
Tanikawa, H., T. Fishman, K. Okuoka, and K. Sugimoto. 2015. Theweight of society over time and space: A comprehensive accountof the construction material stock of Japan, 1945–2010. Journalof Industrial Ecology 19(5): 778–779.
Van Beers, D. and T. Graedel. 2007. Spatial characterisation of multi-level in-use copper and zinc stocks in Australia. Journal of CleanerProduction 15(8): 849–861.
Wedler, B. and A. Hummel. 1947. Trummerverwertung: Technis-che Moglichkeiten und wirtschaftliche Bedeutung [Debris utiliza-tion – technical possibilities and economical relevance]. Berlin:Wilhelm Ernst & Sohn.
Wiedenhofer, D., J. K. Steinberger, N. Eisenmenger, and W. Haas.2015. Maintenance and expansion: Modeling material stocks andflows for residential buildings and transportation networks in theEU25. Journal of Industrial Ecology. 19(4): 538–551.
About the Authors
Fritz Kleemann is a Ph.D. candidate and Jakob Lederer isa postdoctoral researcher at the Christian Doppler Laboratoryfor Anthropogenic Resources, Institute for Water Quality, Re-source and Waste Management at TU Wien (Vienna, Aus-tria). Helmut Rechberger is a professor at and head of theResearch Center of Waste and Resource Management, Insti-tute for Water Quality, Resource and Waste Management atTU Wien. Johann Fellner is an associate professor for WasteManagement and head of the Christian Doppler Laboratory forAnthropogenic Resources at TU Wien.
Supporting Information
Supporting information is linked to this article on the JIE website:
Supporting Information S1: This supporting information includes 12 tables and two figures with more details on buildingcategorization, building structure, and material intensities.
Kleemann et al., GIS-based Analysis of Material Stock in Buildings 13
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-1
Kleemann, F., J. Lederer, H. Rechberger, and J. Fellner. 2016. GIS-based analysis of Vienna’s material stock in buildings. Journal of Industrial Ecology.
This supporting information includes 12 tables and 2 figures with more details on building
categorization, building structure, and material intensities.
[After initial online publication, the article was corrected to adjust for the inadvertent use of
unrepresentative data on lead from buildings containing x-ray rooms. The article was also corrected
to adjust for the erroneous use of unverified data on zinc from a German source.]
Content Building categorization ............................................................................................................................ 1
Building Structure .................................................................................................................................... 5
Material intensities of different building categories ............................................................................... 6
Expert’s estimation of the average height of basement floors and roofs ............................................ 14
Number of data sources for different materials of a building .............................................................. 15
References ............................................................................................................................................. 16
Building categorization The differentiation between construction periods of buildings is chosen based on information
of the available data set. Some information is aggregated because the level of detail is not
meaningful with regard to available material intensities, or represents only a very small share
of the overall building stock. Table S1 shows the age categories of all datasets and the
construction periods used for this research.
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-2
Table S1 Age data of the different data sets (left) and construction periods used in this
research (right)
Existing information about the utilization of buildings is strongly aggregated to three main
categories (residential, commercial, industrial) as the original data set distinguished between
78 categories. Table S2 shows the different utilization categories and their attribution to the
three main categories. Some utilization categories are excluded because the available material
intensities do not fit these special buildings – this share of building volume is treated as if no
information is available.
Age data of all source Used construction periods
Before 1683
Before 1918
1683 – 1740
1741 – 1780
1781 – 1848
1849 – 1859
1849 – 1918
1860 – 1883
1884 – 1918
Until 1848
1919 – 1945 1919 – 1945
1946 – 1967 1946 – 1967
1977 – 1996 1977 – 1996
After 1996
After 1997
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-3
Table S2 Utilization categories of all data sets (left) and utilization categories used in this
research
Utilization categories of all data sets Used utilization categories
Retirement home
Residential
Summerhouse
Apprentice home
Home for nurses
Student home
Business apartment
Residential building
Residential use (other)
Academic secondary school (general)
Commercial
Ambulance
Pharmacy
Professional school
Higher-level vocational schools
Library
Office / operations management
Center for adult education
Retail goods of medium / long-term needs
Retail - moving consumer good
Medical specialist
Leisure, entertainment complex
Gastronomy
Financial institution, Bank
Commercial garage
Wholesale
Cooperative middle school
Hotel, hostel
Youth center
Day-care center
Cinema
Monastery
Hospital
Museum
Public administration
Nursing home
Police station
Polytechnic
Post
Medical practitioner
Ambulance station
Special school
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-4
Utilization categories of all data sets Used utilization categories
Other commercial use
Other personal services
Other technical services
Social counseling centers
Sports facility
Telecommunication
Theater
University, college
Event venues
Elementary school
Freight handling
Federal armed forces
Industry
Firefighters
Garage
Commerce and industry
Agricultural building
Shed
Storage buildings
Traffic requirement for business
Traffic requirement for freight
Traffic requirement for passenger transport
Water supply
Wine cellar (Heuriger)
Not specified
No information
No specification
Superstructure
Other building
No information
Abandoned
Sewerage
Excluded (No information)
Mining facility
Energy supply
Cemetery equipment
Greenhouse
Church, place of assembly
Power station
Waste
Gas station
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-5
Building Structure Figure S1 illustrates the share of utilization and construction period of gross volume of
buildings separately.
Figure S1 Share of gross volume of buildings. Utilization (left) and construction period
(right)
Figure S2 shows the share of the different building categories of the total material stock in
analogy to Figure 2 in the main manuscript. Generally, this graph shows a lot of similarities to
the mentioned figure in the main manuscript. One notable difference is the comparatively
higher contribution of old buildings to the building volume than to the total stock.
Figure S2 Share of the different building categories to the total material stock in 2013
0
50
100
150
200
250
Residential Commercial Industrial No information
Mill
ion
to
ns
[t]
Before 1918
1919-1945
1946-1976
1977-1996
After 1997
No information
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-6
Ma
teri
al
inte
nsi
tie
s o
f d
iffe
ren
t b
uil
din
g c
ate
go
rie
s T
he
mat
eria
l co
mposi
tion o
f dif
fere
nt
buil
din
g t
ypes
and
gen
erat
ed s
pec
ific
mat
eria
l in
tensi
ties
for
buil
din
g c
ateg
ori
es a
re l
iste
d i
n t
he
foll
ow
ing
table
s (T
able
S3-
Tab
le S
10)
Tab
le S
3 M
ate
rial
inte
nsi
ty f
or
resi
den
tial
bu
ild
ings
bu
ilt
bef
ore
1918
Cat
ego
ry
Re
fere
nce
Cas
e s
tud
y
OB
G
Co
nst
ruct
io
n f
ile
s (n
r.
3)
Co
nst
ruct
io
n f
ile
s (n
r.
11)
Co
nst
ruct
io
n f
ile
s (n
r.
29)
Co
nst
ruct
io
n f
ile
s (n
r.
30)
Co
nst
ruct
io
n f
ile
s (n
r.
42)
Co
nst
ruct
io
n f
ile
s (n
r.
45)
Co
nst
ruct
io
n f
ile
s (n
r.
50)
Co
nst
ruct
io
n f
ile
s (n
r.
51)
Co
nst
ruct
io
n f
ile
s (n
r.
59)
Co
nst
ruct
io
n f
ile
s (n
r.
60)
Co
nst
ruct
io
n f
ile
s (n
r.
65)
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
We
dle
r an
d
Hu
mm
el,
1947
We
dle
r an
d
Hu
mm
el,
1947
We
dle
r an
d
Hu
mm
el,
1947
We
dle
r an
d
Hu
mm
el,
1947
Re
nz
et
al.,
2001
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an
valu
e
Me
dia
n
Stan
dar
d
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l44
9.31
446.
8428
5.90
302.
0335
8.09
408.
9536
5.96
425.
8338
9.76
491.
1753
6.56
444.
7439
2.19
416.
6129
6.74
265.
6927
0.27
294.
8733
9.00
405.
9537
9.32
390.
9776
.81
Co
ncr
ete
15.6
810
0.76
18.3
946
.17
31.9
620
.60
33.7
820
.60
20.6
041
.83
71.7
420
.60
22.7
933
.22
0.97
5.81
4.71
0.18
25.0
053
.65
29.4
521
.70
24.5
6
Gra
vel/
san
d
Bri
cks
303.
7523
7.96
203.
4016
7.75
218.
6826
6.71
218.
7027
8.79
249.
5431
2.20
328.
7329
6.92
212.
0122
5.41
173.
8015
1.94
153.
5017
5.59
214.
0011
1.70
225.
0521
8.69
59.6
5
Mo
rtar
/pla
ste
r97
.97
74.3
963
.28
54.2
173
.15
87.7
876
.38
92.4
685
.71
102.
9610
2.08
93.4
310
9.99
111.
7396
.08
88.1
088
.51
92.1
010
0.00
113.
2090
.17
92.2
815
.70
Min
era
l fil
l32
.88
32.8
832
.88
32.8
832
.88
32.8
832
.88
32.8
832
.88
32.8
844
.86
44.3
017
.60
3.62
3.20
19.6
528
.87
32.8
812
.01
Slag
fil
l31
.28
0.89
14.1
513
.80
0.58
12.1
413
.80
12.5
8
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s
Pla
ste
r b
oar
ds/
gyp
sum
0.24
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.16
0.15
0.03
Gla
ss0.
270.
210.
160.
370.
250.
300.
230.
450.
380.
660.
310.
250.
180.
200.
300.
230.
300.
260.
12
Ce
ram
iCas
e s
tud
y0.
130.
470.
470.
470.
470.
470.
470.
470.
470.
470.
470.
477.
231.
886.
256.
551.
700.
472.
50
Nat
ura
l sto
ne
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
127.
4010
.63
0.02
36.7
7
Min
era
l wo
ol
0.00
0.02
0.03
0.02
0.13
0.02
0.02
0.02
0.02
0.02
0.02
0.02
2.53
1.95
0.34
0.02
0.81
Min
era
l wo
ol b
oar
ds
(Ce
me
nt)
asb
est
os
0.40
0.04
3.34
0.16
0.99
0.28
1.58
Org
anic
20.9
223
.52
21.7
612
.14
8.02
13.9
06.
9621
.67
16.4
110
.51
19.1
425
.03
15.5
614
.01
21.7
725
.89
26.3
326
.45
8.00
17.9
017
.79
18.5
26.
44
Wo
od
20.1
923
.11
21.3
811
.74
7.61
13.4
96.
5621
.27
15.9
810
.11
18.4
624
.63
15.5
614
.01
21.5
825
.71
26.2
026
.15
8.00
17.9
017
.48
18.1
86.
42
He
rakl
it0.
270.
270.
27
Pap
er/
Car
db
oar
d
PV
C0.
200.
200.
20
Var
iou
s p
last
iCas
e s
tud
y0.
420.
340.
340.
340.
340.
340.
340.
340.
370.
340.
340.
340.
350.
340.
02
Car
pe
t
Lam
inat
e
Lin
ole
um
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.00
Asp
hal
t
Bit
um
en
0.10
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.02
0.19
0.18
0.13
0.30
0.08
0.02
0.09
Po
lyst
yre
ne
Me
tal
1.26
2.23
4.86
2.79
2.80
5.82
2.43
2.10
2.54
2.61
4.14
2.15
3.12
7.52
4.86
3.48
7.00
3.63
2.80
1.80
Iro
n/S
tee
l1.
012.
144.
772.
712.
655.
742.
342.
012.
422.
524.
052.
063.
117.
524.
853.
476.
103.
502.
711.
75
Alu
min
ium
0.14
0.02
0.02
0.02
0.08
0.02
0.02
0.02
0.05
0.02
0.02
0.02
0.85
0.10
0.02
0.23
Co
pp
er
0.05
0.06
0.06
0.06
0.06
0.06
0.06
0.06
0.06
0.06
0.06
0.06
0.05
0.10
0.06
0.06
0.01
Lead
0.05
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Bra
ss0.
010.
010.
010.
010.
010.
010.
00
Tota
l47
1.49
472.
5931
2.52
316.
9736
8.91
428.
6837
5.36
449.
6040
8.71
504.
3055
9.84
471.
9240
7.74
430.
6232
1.62
299.
1030
1.46
324.
8035
4.00
423.
8540
0.20
408.
2374
.98
Be
fore
191
8 -
Re
sid
en
tial
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-7
Tab
le S
4 M
ate
rial
inte
nsi
ties
for
com
mer
cial
an
d i
nd
ust
rial
bu
ild
ings
bu
ilt
bef
ore
1918
Cat
ego
ry
Re
fere
nce
Cas
e s
tud
y
KES
P1
Cas
e s
tud
y
KES
P2
Cas
e s
tud
y
KES
P3
Cas
e s
tud
y
KES
VW
Cas
e s
tud
y
GZL
Co
nst
ruct
io
n f
ile
s (n
r.
18)
Co
nst
ruct
io
n f
ile
s (n
r.
20)
Co
nst
ruct
io
n f
ile
s (n
r.
27)
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Cas
e s
tud
y
WP
O
Co
nst
ruct
io
n f
ile
s (n
r.
54)
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l43
4.18
401.
4443
0.42
429.
3640
4.69
550.
1250
7.97
416.
0026
0.00
426.
0242
9.36
79.7
925
5.24
426.
0627
8.40
319.
9027
8.40
92.6
6
Co
ncr
ete
21.4
619
.87
22.8
519
.73
44.2
119
7.38
32.1
820
.60
88.4
051
.85
22.8
558
.90
44.9
048
.02
110.
2067
.71
48.0
236
.83
Gra
vel/
san
d
Bri
cks
292.
2426
7.05
286.
1329
4.06
252.
5423
8.84
330.
2327
3.04
52.0
025
4.01
273.
0480
.22
169.
3928
8.55
69.6
017
5.85
169.
3910
9.62
Mo
rtar
/pla
ste
r83
.70
77.1
083
.19
80.7
669
.50
79.5
997
.34
88.4
362
.40
80.2
280
.76
10.1
540
.11
88.2
952
.20
60.2
052
.20
25.0
7
Min
era
l fil
l32
.74
31.3
934
.32
31.9
335
.59
32.8
847
.10
32.8
834
.85
32.8
85.
12
Slag
fil
l
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s0.
600.
600.
600.
380.
380.
38
Pla
ste
r b
oar
ds/
gyp
sum
2.30
4.10
2.46
1.44
1.32
0.17
0.17
0.17
1.52
1.38
1.40
Gla
ss0.
310.
410.
490.
270.
270.
650.
320.
260.
370.
320.
140.
271.
090.
680.
680.
58
Ce
ram
iCas
e s
tud
y0.
440.
970.
610.
930.
730.
470.
470.
470.
630.
540.
220.
090.
09
Nat
ura
l sto
ne
0.10
0.02
0.02
0.02
57.2
011
.47
0.02
25.5
646
.40
46.4
046
.40
Min
era
l wo
ol
0.32
0.34
0.22
0.22
0.20
0.14
0.14
0.14
0.21
0.21
0.08
0.02
0.02
Min
era
l wo
ol b
oar
ds
0.03
0.09
0.05
0.02
0.10
0.06
0.05
0.03
0.05
0.05
0.05
(Ce
me
nt)
asb
est
os
0.03
0.13
0.24
0.13
0.13
0.10
0.14
0.14
0.14
Org
anic
3.67
3.47
3.47
8.20
2.03
9.10
12.7
510
.07
2.60
6.15
3.67
3.90
3.77
16.0
15.
808.
535.
806.
56
Wo
od
3.26
3.04
2.99
7.89
1.48
8.14
12.2
89.
602.
605.
703.
263.
833.
6215
.94
5.80
8.45
5.80
6.57
He
rakl
it0.
070.
070.
070.
070.
070.
00
Pap
er/
Car
db
oar
d
PV
C0.
320.
170.
100.
150.
480.
240.
170.
160.
010.
010.
01
Var
iou
s p
last
iCas
e s
tud
y0.
030.
190.
060.
030.
070.
830.
340.
340.
240.
130.
270.
070.
07
Car
pe
t
Lam
inat
e
Lin
ole
um
0.01
0.24
0.00
0.04
0.04
0.04
0.06
0.04
0.09
Asp
hal
t
Bit
um
en
0.06
0.07
0.07
0.14
0.02
0.02
0.02
0.06
0.06
0.04
0.14
0.14
0.14
Po
lyst
yre
ne
0.00
0.00
0.00
Me
tal
6.02
6.44
6.05
3.19
3.09
5.36
2.53
3.01
4.46
4.27
1.65
5.82
11.2
68.
758.
618.
752.
72
Iro
n/S
tee
l5.
765.
875.
713.
112.
985.
052.
432.
914.
234.
081.
505.
7911
.22
8.70
8.57
8.70
2.72
Alu
min
ium
0.06
0.07
0.06
0.01
0.05
0.24
0.03
0.03
0.07
0.05
0.07
0.03
0.03
0.03
0.03
0.00
Co
pp
er
0.20
0.17
0.10
0.07
0.07
0.06
0.06
0.06
0.10
0.10
0.07
0.05
0.00
0.01
0.05
0.02
0.01
0.03
Lead
0.32
0.18
0.01
0.01
0.01
0.11
0.01
0.14
0.00
0.00
Bra
ss
Tota
l44
3.87
411.
3543
9.94
440.
7640
9.81
564.
5852
3.25
429.
0826
2.60
436.
1443
9.94
83.4
126
4.84
453.
3229
2.95
337.
0429
2.95
101.
68
Be
fore
191
8 -
Co
mm
erc
ial
Be
fore
191
8 -
Ind
ust
rial
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-8
Tab
le S
5 M
ate
rial
inte
nsi
ties
for
resi
den
tial
bu
ild
ings
bu
ilt
bet
wee
n 1
919 a
nd
1945
Cat
ego
ry
Re
fere
nce
Cas
e s
tud
y
BFS
Co
nst
ruct
io
n f
ile
s (n
r.
14)
Co
nst
ruct
io
n f
ile
s (n
r.
15)
Co
nst
ruct
io
n f
ile
s (n
r.
16)
Co
nst
ruct
io
n f
ile
s (n
r.
34)
Co
nst
ruct
io
n f
ile
s (n
r.
48)
We
dle
r an
d
Hu
mm
el,
1947
We
dle
r an
d
Hu
mm
el,
1947
We
dle
r an
d
Hu
mm
el,
1947
Sch
ulz
e,
1990
Hal
dim
ann
(199
1),
in
Gö
rg, 1
997
Gru
hle
r e
t
al.,
200
2
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l37
2.94
591.
7635
7.55
394.
5658
6.43
410.
0636
2.58
345.
0731
1.88
476.
2267
0.28
417.
2447
0.40
443.
6141
0.06
105.
89
Co
ncr
ete
9.13
22.4
821
2.11
294.
0327
7.04
155.
2637
.224
31.6
229
.60
174.
1227
0.96
51.7
110
2.10
128.
2610
2.10
107.
68
Gra
vel/
san
d11
.54
0.00
5.77
5.77
8.16
Bri
cks
253.
3436
9.68
61.8
238
.88
189.
1515
0.79
188.
3017
7.77
157.
0114
9.38
318.
7521
7.24
111.
5518
3.36
177.
7792
.60
Mo
rtar
/pla
ste
r88
.56
123.
9446
.59
26.7
685
.74
66.8
710
6.50
103.
1993
.12
95.2
639
.60
106.
2811
9.65
84.7
793
.12
30.8
3
Min
era
l fil
l2.
3274
.41
32.8
832
.88
32.8
832
.88
16.3
518
.96
19.6
842
.94
36.7
242
.01
32.0
732
.88
17.8
2
Slag
fil
l5.
586.
325.
575.
825.
580.
43
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s15
.84
15.8
415
.84
Pla
ste
r b
oar
ds/
gyp
sum
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.00
Gla
ss0.
360.
600.
440.
490.
700.
660.
480.
300.
300.
650.
750.
520.
490.
16
Ce
ram
ics
0.47
0.47
0.47
0.47
0.47
8.15
6.90
6.60
0.7
1.75
2.64
0.58
3.20
Nat
ura
l sto
ne
0.02
0.02
0.02
0.02
0.02
1.5
0.12
137.
1017
.35
0.02
48.3
9
Min
era
l wo
ol
0.02
0.02
0.02
0.02
0.02
0.13
1.40
0.23
0.02
0.52
Min
era
l wo
ol b
oar
ds
(Ce
me
nt)
asb
est
os
3.39
3.06
0.87
0.26
2.96
0.23
1.79
1.91
1.49
Org
anic
13.2
317
.23
8.20
4.57
14.1
918
.86
14.4
415
.40
15.1
111
.45
8.18
6.46
11.8
512
.24
13.2
34.
30
Wo
od
13.1
216
.82
7.80
4.15
13.7
916
.48
14.4
415
.40
15.1
19.
376.
885.
3911
.85
11.5
813
.12
4.37
He
rakl
it0.
001.
980.
990.
991.
40
Pap
er/
Car
db
oar
d0.
330.
110.
220.
220.
16
PV
C0.
100.
100.
10
Var
iou
s p
last
ics
0.34
0.34
0.35
0.34
0.34
0.14
0.24
1.08
0.40
0.34
0.28
Car
pe
t1.
200.
951.
081.
080.
18
Lam
inat
e
Lin
ole
um
0.04
0.04
0.04
0.04
0.04
0.04
0.04
0.00
Asp
hal
t
Bit
um
en
0.02
0.02
0.02
0.02
0.02
0.41
0.09
0.02
0.16
Po
lyst
yre
ne
Me
tal
0.85
2.82
3.53
6.84
7.91
2.23
4.65
4.77
4.33
4.18
11.4
97.
805.
025.
114.
652.
81
Iro
n/S
tee
l0.
852.
733.
446.
667.
822.
144.
634.
754.
314.
1511
.47
7.80
4.90
5.05
4.63
2.81
Alu
min
ium
0.02
0.02
0.11
0.02
0.02
0.04
0.02
0.04
Co
pp
er
0.00
0.06
0.06
0.06
0.06
0.06
0.03
0.02
0.12
0.05
0.06
0.03
Lead
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.00
Bra
ss0.
020.
020.
020.
020.
020.
00
Tota
l38
7.01
611.
8136
9.27
405.
9760
8.53
431.
1638
1.67
365.
2433
1.32
491.
8568
9.95
431.
5048
7.27
460.
9643
1.16
111.
63
1919
-194
5 -
Re
sid
en
tial
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-9
Tab
le S
6 M
ate
rial
inte
nsi
ties
for
com
mer
cial
an
d i
nd
ust
rial
bu
ild
ings
bu
ilt
bet
wee
n 1
919 a
nd
1945
Cat
ego
ry
Re
fere
nce
Bac
cin
i an
d
Pe
traz
a,
2006
Co
nst
ruct
io
n f
ile
s (n
r.
7)
Re
nz
et
al.,
2001
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Cas
e s
tud
y
WP
O
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l19
9.50
636.
3934
0.00
391.
9634
0.00
223.
0325
5.24
382.
7031
8.97
318.
9790
.13
Co
ncr
ete
79.8
021
6.38
116.
0013
7.39
116.
0070
.76
44.9
016
7.70
106.
3010
6.30
86.8
3
Gra
vel/
san
d
Bri
cks
58.8
029
5.74
224.
0019
2.85
224.
0012
1.50
169.
3919
7.80
183.
6018
3.60
20.0
9
Mo
rtar
/pla
ste
r56
.70
90.1
873
.44
73.4
423
.67
40.1
117
.20
28.6
528
.65
16.2
0
Min
era
l fil
l32
.81
32.8
132
.81
Slag
fil
l
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s0.
380.
380.
38
Pla
ste
r b
oar
ds/
gyp
sum
0.17
0.17
0.17
Gla
ss0.
490.
490.
490.
270.
270.
27
Ce
ram
ics
0.47
0.47
0.47
Nat
ura
l sto
ne
4.20
0.02
2.11
2.11
2.96
Min
era
l wo
ol
0.14
0.14
0.14
Min
era
l wo
ol b
oar
ds
0.05
0.05
0.05
(Ce
me
nt)
asb
est
os
0.14
0.14
0.14
Org
anic
4.20
7.11
9.00
6.77
7.11
2.42
3.77
51.6
027
.69
27.6
933
.82
Wo
od
4.20
6.63
9.00
6.61
6.63
2.40
3.62
51.6
027
.61
27.6
133
.93
He
rakl
it0.
070.
070.
07
Pap
er/
Car
db
oar
d
PV
C0.
010.
010.
01
Var
iou
s p
last
ics
0.35
0.35
0.35
Car
pe
t
Lam
inat
e
Lin
ole
um
0.04
0.04
0.04
Asp
hal
t
Bit
um
en
0.02
0.02
0.02
0.14
0.14
0.14
Po
lyst
yre
ne
0.00
0.00
0.00
Me
tal
6.42
5.66
6.00
6.03
6.00
0.38
5.82
5.82
5.82
Iro
n/S
tee
l6.
305.
476.
005.
926.
000.
425.
795.
795.
79
Alu
min
ium
0.12
0.12
0.12
0.03
0.03
0.03
Co
pp
er
0.12
0.06
0.09
0.09
0.04
0.00
0.06
0.03
0.03
0.04
Lead
0.01
0.01
0.01
Bra
ss
Tota
l21
0.12
649.
1635
5.00
404.
7635
5.00
223.
7126
4.84
434.
3034
9.57
349.
5711
9.83
1919
-194
5 -
Co
mm
erc
ial
1919
-194
5 -
Ind
ust
rial
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-1
0
Tab
le S
7 M
ate
rial
inte
nsi
ties
for
resi
den
tial
bu
ild
ings
bu
ilt
bet
wee
n 1
946 a
nd
1976
Cat
ego
ry
Re
fere
nce
Cas
e s
tud
y
PW
G
Cas
e s
tud
y
DN
G
Co
nst
ruct
io
n f
ile
s (n
r.
4)
Co
nst
ruct
io
n f
ile
s (n
r.
8)
Co
nst
ruct
io
n f
ile
s (n
r.
12)
Co
nst
ruct
io
n f
ile
s (n
r.
21)
Co
nst
ruct
io
n f
ile
s (n
r.
22)
Co
nst
ruct
io
n f
ile
s (n
r.
25)
Co
nst
ruct
io
n f
ile
s (n
r.
28)
Co
nst
ruct
io
n f
ile
s (n
r.
33)
Co
nst
ruct
io
n f
ile
s (n
r.
36)
Co
nst
ruct
io
n f
ile
s (n
r.
37)
Co
nst
ruct
io
n f
ile
s (n
r.
39)
Co
nst
ruct
io
n f
ile
s (n
r.
40)
Co
nst
ruct
io
n f
ile
s (n
r.
57)
Co
nst
ruct
io
n f
ile
s (n
r.
63)
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Sch
ulz
e,
1990
Sch
ulz
e,
1990
Sch
ulz
e,
1990
Alb
rech
t,
1984
LV B
üro
Dip
l.-I
ng.
F.
Wö
rne
r in
Gö
rg (
1997
)
Ble
ngi
ni,
2009
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l42
8.67
395.
9666
5.90
395.
0859
2.59
540.
5836
7.30
383.
8860
2.78
490.
0116
2.22
509.
9863
5.96
183.
2348
8.44
270.
3640
1.42
457.
5447
2.22
423.
8742
4.39
434.
4438
9.06
442.
2754
3.46
432.
5353
1.39
431.
8280
0.98
369.
0338
4.00
453.
2743
2.53
129.
31
Co
ncr
ete
396.
3361
.67
374.
1321
0.30
188.
6923
4.27
336.
0833
2.36
363.
0231
2.47
115.
6927
5.99
230.
7196
.31
215.
4814
4.49
100.
0036
6.13
297.
3912
7.57
276.
4215
3.24
237.
0814
5.65
280.
1018
8.45
207.
2938
5.92
264.
2633
3.36
237.
8524
1.57
237.
0894
.55
Gra
vel/
san
d16
.32
10.5
631
.67
8.57
6.23
9.95
2.97
45.3
616
.45
10.2
514
.60
Bri
cks
0.00
241.
0820
1.57
117.
2127
4.20
214.
284.
409.
7416
1.93
108.
7314
9.00
296.
0056
.00
188.
7893
.55
176.
565
.414
9.5
50.4
150.
162
.315
1.0
150.
013
5.0
224.
019.
3527
5.64
17.5
080
.10
131.
4914
9.00
86.7
9
Mo
rtar
/pla
ste
r8.
2874
.62
78.6
427
.53
117.
8680
.22
14.6
525
.80
66.1
856
.85
2.90
69.9
197
.78
13.2
681
.14
20.5
811
0.32
69.4
381
.70
115.
2371
.54
101.
1166
.87
119.
0110
0.24
99.4
986
.36
22.5
171.
118.
5566
.05
68.5
771
.54
40.7
5
Min
era
l fil
l4.
9439
.23
4.94
4.94
5.79
4.24
4.94
4.94
4.94
4.94
4.94
4.94
4.94
9.61
12.4
917
.97
21.9
517
.89
20.5
416
.72
19.5
40.
000.
000.
000
5.28
9.26
4.94
9.15
Slag
fil
l11
.16
3.72
3.72
3.72
3.72
3.72
3.72
3.72
3.72
3.72
3.72
4.40
3.72
2.24
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s3.
481.
812.
652.
651.
18
Pla
ste
r b
oar
ds/
gyp
sum
0.00
0.05
0.15
0.15
0.15
0.15
3.09
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
1.78
0.40
0.15
0.80
Gla
ss0.
460.
540.
260.
550.
540.
500.
290.
680.
350.
660.
670.
680.
182.
870.
890.
430.
500.
330.
781
0.63
0.41
0.65
0.54
0.54
Ce
ram
ics
1.90
3.43
1.98
0.09
1.98
1.98
2.20
0.32
1.98
1.98
1.98
1.98
1.98
1.98
1.98
1.98
0.93
0.78
0.95
0.75
1.25
8.48
1.95
1.98
1.64
Nat
ura
l sto
ne
0.28
1.80
0.00
0.00
936
.57
7.94
1.04
14.4
5
Min
era
l wo
ol
0.04
1.47
0.51
0.02
0.51
0.51
0.80
0.02
0.51
0.51
0.51
0.51
0.51
0.51
0.02
0.51
2.85
9.49
3.27
5.49
3.25
3.16
3.04
5.33
1.36
2.13
1.95
0.33
0.86
0.73
1.69
0.62
2.10
Min
era
l wo
ol b
oar
ds
0.08
0.08
0.08
(Ce
me
nt)
asb
est
os
1.49
0.12
3.11
0.00
3.49
2.14
0.00
6.54
4.12
4.88
6.32
3.04
0.00
0.09
0.10
0.02
32.
221.
812.
35
Org
anic
4.14
4.85
8.12
3.99
3.48
9.49
13.9
711
.19
5.35
16.9
447
.71
4.91
6.87
15.0
410
.30
26.2
33.
915.
496.
546.
866.
506.
324.
568.
884.
064.
404.
864.
0121
.05
1.39
12.0
09.
476.
509.
01
Wo
od
2.27
4.16
7.13
3.65
2.49
8.49
12.9
04.
294.
3515
.49
42.7
23.
925.
8812
.82
9.42
25.2
33.
912.
006.
546.
866.
506.
324.
567.
102.
071.
753.
222.
4017
.46
0.44
12.0
08.
015.
888.
40
He
rakl
it0.
234.
170.
411.
231.
510.
821.
83
Pap
er/
Car
db
oar
d0.
000.
330.
280.
330.
110.
060.
190.
200.
14
PV
C0.
520.
080.
960.
520.
520.
44
Var
iou
s p
last
ics
0.07
0.85
0.35
0.85
0.85
1.04
0.73
0.85
0.91
0.85
0.85
0.85
0.85
0.85
0.85
3.50
1.78
0.08
0.38
0.12
0.36
0.35
0.83
0.85
0.71
Car
pe
t0.
021.
201.
890.
650.
901.
971.
101.
050.
75
Lam
inat
e0.
330.
330.
33
Lin
ole
um
0.02
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.00
Asp
hal
t
Bit
um
en
1.04
0.01
0.11
0.11
0.11
2.00
0.11
0.11
4.11
0.11
0.11
0.11
0.11
0.38
0.10
0.54
0.24
1.21
0.59
0.11
1.02
Po
lyst
yre
ne
0.23
0.23
0.23
Me
tal
7.91
3.50
4.28
8.37
3.77
3.58
6.17
5.00
7.01
7.75
3.58
5.81
8.34
4.83
3.35
4.67
13.8
821
.48
13.0
78.
2313
.01
12.6
413
.68
10.6
62.
918.
475.
5023
.26
6.99
14.6
54.
188.
407.
015.
16
Iro
n/S
tee
l7.
583.
304.
208.
203.
713.
526.
104.
936.
957.
683.
515.
748.
284.
633.
294.
6113
.88
21.4
813
.07
8.23
13.0
112
.64
13.6
810
.66
2.88
8.43
5.46
23.2
26.
9314
.61
4.00
8.34
6.95
5.20
Alu
min
ium
0.22
0.15
0.04
0.14
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.17
0.03
0.03
0.01
0.06
0.03
0.06
Co
pp
er
0.11
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.04
0.04
0.04
0.06
0.02
0.18
0.04
0.03
0.04
Lead
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.00
Bra
ss
Tota
l44
0.72
404.
3167
8.30
407.
4559
9.84
553.
6538
7.44
400.
0761
5.14
514.
7021
3.51
520.
7065
1.17
203.
1050
2.10
301.
2641
9.22
484.
5249
1.83
438.
9644
3.90
453.
4040
7.29
461.
8155
0.43
445.
4054
1.74
459.
0982
9.02
385.
0740
0.18
471.
1445
3.40
125.
90
1946
-197
6 -
Re
sid
en
tial
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-1
1
Tab
le S
8 M
ate
rial
inte
nsi
ties
for
com
mer
cial
an
d i
nd
ust
rial
bu
ild
ings
bu
ilt
bet
wee
n 1
946 a
nd
1976
Cat
ego
ry
Re
fere
nce
Cas
e s
tud
y
KES
Pat
Cas
e s
tud
y
ZA
Co
nst
ruct
io
n f
ile
s (
nr.
58)
Bac
cin
i an
d
Pe
traz
a,
2006
Re
nz
et
al.,
2001
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l40
5.53
293.
2664
4.99
352.
8034
3.00
407.
9235
2.80
138.
3933
6.60
336.
6033
6.60
Co
ncr
ete
228.
4627
3.85
301.
1828
4.40
137.
0024
4.98
273.
8566
.10
254.
9025
4.90
254.
90
Gra
vel/
san
d
Bri
cks
130.
1324
6.24
21.6
020
6.00
150.
9916
8.06
98.7
930
.50
30.5
030
.50
Mo
rtar
/pla
ste
r43
.60
93.0
443
.20
59.9
543
.60
28.6
651
.20
51.2
051
.20
Min
era
l fil
l2.
322.
322.
32
Slag
fil
l
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s
Pla
ste
r b
oar
ds/
gyp
sum
14.6
40.
167.
407.
4010
.24
Gla
ss0.
530.
581.
500.
870.
580.
55
Ce
ram
ics
1.70
0.11
0.90
0.90
0.90
0.79
Nat
ura
l sto
ne
1.55
0.63
3.60
1.93
1.55
1.52
Min
era
l wo
ol
0.55
0.03
1.33
0.64
0.55
0.66
Min
era
l wo
ol b
oar
ds
0.57
0.09
0.33
0.33
0.34
(Ce
me
nt)
asb
est
os
0.00
0.09
0.05
0.05
0.06
Org
anic
2.81
7.57
23.7
13.
608.
009.
147.
578.
47
Wo
od
2.20
0.69
20.9
83.
608.
007.
093.
608.
23
He
rakl
it0.
070.
070.
07
Pap
er/
Car
db
oar
d
PV
C0.
210.
330.
270.
270.
09
Var
iou
s p
last
ics
0.07
0.13
2.64
0.95
0.13
1.47
Car
pe
t
Lam
inat
e
Lin
ole
um
0.07
0.02
0.05
0.05
0.03
Asp
hal
t3.
653.
653.
65
Bit
um
en
2.67
2.67
2.67
Po
lyst
yre
ne
0.27
0.10
0.18
0.18
0.12
Me
tal
4.77
8.00
5.74
7.38
2.95
5.77
5.74
2.03
12.5
612
.56
12.5
6
Iro
n/S
tee
l4.
527.
595.
657.
202.
805.
555.
651.
9712
.40
12.4
012
.40
Alu
min
ium
0.09
0.32
0.06
0.18
0.12
0.15
0.12
0.10
Co
pp
er
0.16
0.10
0.02
0.04
0.08
0.07
0.06
0.16
0.16
0.16
Lead
0.00
0.00
0.00
0.00
0.00
Bra
ss
Tota
l41
3.11
308.
8467
4.44
363.
7835
3.95
422.
8236
3.78
145.
4534
9.16
349.
1634
9.16
1946
-197
6 -
Ind
ust
rial
1946
-197
6 -
Co
mm
erc
ial
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-1
2
Tab
le S
9 M
ate
rial
inte
nsi
ties
for
resi
den
tial,
com
mer
cial,
an
d i
nd
ust
rial
bu
ild
ings
bu
ilt
bet
wee
n 1
977 a
nd
1996
Cat
ego
ry
Re
fere
nce
Co
nst
ruct
ion
file
s (n
r. 5
)
Co
nst
ruct
ion
file
s (n
r. 9
)
Co
nst
ruct
ion
file
s (n
r. 1
7)
Co
nst
ruct
ion
file
s (n
r. 3
5)
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
Gru
hle
r e
t
al.,
200
2
LV
Mas
sivb
au
Gm
bH
Do
rnd
orf
in
Gö
rg (
1997
)
LV
Mas
sivb
au
Gm
bH
Do
rnd
orf
in
Gö
rg (
1997
)
Alb
rech
t,
1984
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Cas
e s
tud
y
RZ
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Bac
cin
i an
d
Pe
traz
a,
2006
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l41
0.57
730.
6559
0.15
301.
4041
1.28
367.
9868
5.15
398.
0859
3.09
431.
4943
4.40
486.
7543
1.49
87.4
539
4.36
374.
4038
4.38
384.
3814
.12
165.
6016
5.60
165.
60
Co
ncr
ete
204.
3130
2.03
307.
9124
9.44
203.
3816
8.54
342.
5816
4.88
345.
7938
5.92
343.
0027
4.34
302.
0398
.65
377.
2231
9.80
348.
5134
8.51
40.6
014
5.80
145.
8014
5.80
Gra
vel/
san
d15
3.34
47.7
27.
312.
9752
.84
27.5
224
.68
2.44
2.44
2.44
#DIV
/0!
Bri
cks
143.
2231
6.65
219.
8815
.45
143.
2778
.65
105.
0899
.31
9.35
42.3
511
7.32
102.
2046
.15
5.48
19.5
012
.49
12.4
99.
919.
009.
009.
00
Mo
rtar
/pla
ste
r57
.17
104.
1749
.80
19.0
647
.34
84.2
779
.93
40.7
512
3.93
22.5
049
.05
61.6
349
.80
44.6
51.
2931
.20
16.2
416
.24
21.1
53.
603.
603.
60
Min
era
l fil
l14
.41
35.1
184
.83
00.
0026
.87
14.4
1
Slag
fil
l24
.47
24.4
724
.47
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s
Pla
ste
r b
oar
ds/
gyp
sum
0.00
0.00
6.93
13.0
71.
650.
003.
610.
824.
194.
194.
19
Gla
ss0.
550.
550.
550.
370.
000.
61.
031.
000.
580.
550.
240.
860.
860.
86
Ce
ram
ics
1.85
1.85
1.85
0.09
2.3
8.18
0.75
2.41
1.85
3.92
0.81
0.81
0.81
Nat
ura
l sto
ne
0.78
0.78
0.78
26.5
50.
669.
006.
420.
7813
.22
1.74
3.90
2.82
2.82
1.52
7.20
7.20
7.20
Min
era
l wo
ol
0.85
1.10
1.10
3.92
1.24
1.40
10.2
6.88
3.34
1.32
0.32
0.32
0.32
Min
era
l wo
ol b
oar
ds
0.00
0.00
0.00
(Ce
me
nt)
asb
est
os
1.85
3.53
1.35
0.00
1.68
1.60
0.01
0.01
0.01
Org
anic
5.98
5.00
6.73
53.7
86.
598.
434.
898.
7212
.41
4.01
11.7
511
.66
6.73
3.83
2.09
0.00
1.04
1.04
1.48
Wo
od
4.85
3.80
5.35
45.8
84.
947.
023.
266.
059.
782.
4011
.75
9.55
5.35
4.14
1.23
1.23
1.23
He
rakl
it0.
240.
424.
721.
790.
420.
000.
000.
00
Pap
er/
Car
db
oar
d0.
760.
670.
110.
510.
670.
350.
000.
000.
00
PV
C0.
120.
120.
12
Var
iou
s p
last
ics
0.94
0.94
0.94
0.34
1.65
1.40
1.63
0.60
0.40
0.36
0.92
0.94
0.13
0.11
0.11
0.11
Car
pe
t1.
151.
490.
901.
181.
150.
30
Lam
inat
e
Lin
ole
um
0.01
0.01
0.01
0.01
0.01
Asp
hal
t
Bit
um
en
2.84
0.16
0.07
0.24
0.83
0.20
0.09
0.63
0.63
0.63
Po
lyst
yre
ne
0.18
0.18
0.18
Me
tal
5.45
5.89
6.83
5.12
25.9
412
.64
13.0
56.
425.
3823
.26
11.7
311
.06
6.83
8.19
10.1
015
.84
12.9
712
.97
4.06
14.5
814
.58
14.5
8
Iro
n/S
tee
l5.
285.
716.
665.
0525
.94
12.6
413
.05
6.38
5.35
23.2
211
.50
10.9
86.
668.
199.
4115
.60
12.5
012
.50
4.38
14.4
014
.40
14.4
0
Alu
min
ium
0.14
0.14
0.14
0.03
0.12
0.14
#DIV
/0!
0.54
0.54
0.54
Co
pp
er
0.03
0.03
0.03
0.03
0.04
0.03
0.04
0.23
0.06
0.03
0.10
0.15
0.24
0.20
0.20
0.06
0.18
0.18
0.18
Lead
0.01
0.01
0.01
0.01
0.01
0.01
Bra
ss
Tota
l42
2.00
741.
5360
3.71
360.
3044
3.80
389.
0470
3.10
413.
2261
0.88
458.
7645
7.88
509.
4845
7.88
86.4
540
6.55
390.
2439
8.39
398.
3911
.53
180.
1818
0.18
180.
18
1977
-199
6 -
Ind
ust
rial
1977
-199
6 -
Re
sid
en
tial
1977
-199
6 -
Co
mm
erc
ial
20
16 J
ou
rna
l o
f In
du
str
ial E
co
log
y –
ww
w.w
ile
yo
nlin
elib
rary
.co
m/jo
urn
al/jie
S-1
3
Tab
le S
10 M
ate
rial
inte
nsi
ties
for
resi
den
tial
an
d c
om
mer
cial
bu
ild
ings
bu
ilt
aft
er 1
997
Cat
ego
ry
Re
fere
nce
Fin
al B
ill
WH
A P
GH
.
LCA
Dat
a JA
SLC
A D
ata
IBO
1
LCA
Dat
a
IBO
2
LCA
Dat
a
IBO
3
LCA
Dat
a
IBO
4
LCA
Dat
a
IBO
5
LCA
Dat
a
IBO
6
LCA
Dat
a
IBO
7
LCA
Dat
a
IBO
8
LCA
Dat
a
IBO
9
LCA
Dat
a
IBO
10
Gru
hle
r e
t
al.,
200
2
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Cas
e s
tud
y
KES
OP
Co
nst
ruct
io
n f
ile
s (n
r.
6)
Co
nst
ruct
io
n f
ile
s (n
r.
64)
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Me
an v
alu
e M
ed
ian
St
and
ard
de
viat
ion
Mat
eri
al [
kg/m
³]
Min
era
l49
5.71
479.
9137
3.95
361.
9631
5.51
213.
8850
2.60
446.
3027
4.58
418.
3942
6.86
379.
3536
8.45
389.
0437
9.35
86.0
131
9.58
158.
9343
9.27
305.
9331
9.58
140.
67
Co
ncr
ete
484.
5645
8.53
355.
8227
4.72
292.
6216
6.20
475.
8242
7.31
230.
9939
8.32
408.
1833
9.27
205.
6634
7.54
355.
8210
6.15
310.
5713
5.51
421.
3028
9.12
310.
5714
4.10
Gra
vel/
san
d0.
603.
252.
845.
2722
.04
11.1
43.
285.
541.
8910
.13
6.60
4.28
6.40
#DIV
/0!
#ZA
HL!
Bri
cks
57.7
215
.98
88.2
553
.98
57.7
236
.28
10.4
01.
836.
116.
116.
06
Mo
rtar
/pla
ste
r4.
011.
506.
422.
522.
632.
253.
232.
112.
312.
792.
9574
.54
8.94
2.71
20.7
02.
4412
.59
7.51
7.51
7.18
Min
era
l fil
l3.
663.
116.
226.
1624
.64
2.26
7.68
4.91
8.47
Slag
fil
l
Ho
llo
w b
rick
s
Foam
ed
cla
y b
rick
s
Pla
ste
r b
oar
ds/
gyp
sum
8.49
9.53
10.1
411
.12
12.5
814
.35
5.39
5.75
9.08
7.08
12.3
06.
949.
409.
312.
837.
142.
950.
513.
542.
953.
35
Gla
ss0.
692.
612.
291.
621.
950.
890.
520.
791.
260.
900.
780.
771.
260.
890.
690.
510.
620.
880.
670.
620.
19
Ce
ram
ics
0.95
0.52
0.47
0.09
2.84
0.09
0.09
0.09
0.09
0.09
0.09
2.19
0.63
0.09
0.93
0.82
0.34
0.34
0.50
0.34
0.28
Nat
ura
l sto
ne
0.47
0.47
0.47
0.63
0.63
0.63
0.63
0.00
Min
era
l wo
ol
0.42
0.99
0.90
1.34
3.01
4.57
1.16
2.96
3.13
1.88
0.83
1.11
1.86
1.25
1.27
0.54
4.75
1.20
2.16
1.20
2.26
Min
era
l wo
ol b
oar
ds
(Ce
me
nt)
asb
est
os
1.30
1.30
1.30
Org
anic
2.77
8.74
18.4
311
.09
9.14
43.0
115
.26
10.3
541
.64
13.8
58.
699.
546.
6115
.32
10.3
512
.60
5.67
16.9
32.
988.
535.
677.
40
Wo
od
1.55
4.95
6.18
4.25
4.59
36.6
51.
913.
3333
.24
1.98
1.47
2.67
5.44
8.32
4.25
11.9
30.
6216
.37
0.84
5.94
0.84
9.03
He
rakl
it0.
070.
000.
070.
000.
303.
370.
630.
071.
34
Pap
er/
Car
db
oar
d0.
090.
060.
020.
060.
060.
03
PV
C0.
180.
180.
18
Var
iou
s p
last
ics
0.34
1.12
5.89
6.41
4.12
6.02
6.69
4.01
3.03
9.99
5.70
5.07
1.17
4.58
5.07
2.69
0.84
0.56
0.93
0.77
0.84
0.19
Car
pe
t0.
060.
060.
06
Lam
inat
e
Lin
ole
um
0.05
0.04
0.05
0.05
0.00
Asp
hal
t5.
225.
771.
334.
105.
222.
42
Bit
um
en
0.97
1.08
0.43
0.32
0.35
0.50
1.69
1.94
1.88
1.50
1.80
1.13
1.08
0.66
3.66
1.21
2.43
2.43
1.73
Po
lyst
yre
ne
0.78
1.69
1.23
1.23
0.65
0.38
0.38
0.38
Me
tal
19.7
59.
2618
.99
14.5
014
.29
8.41
22.2
820
.31
4.17
20.4
318
.42
15.5
014
.63
15.4
615
.50
5.42
9.97
4.78
11.2
68.
679.
973.
43
Iro
n/S
tee
l19
.45
9.19
18.5
314
.26
14.2
18.
2422
.12
20.1
64.
1020
.17
18.2
515
.32
14.0
015
.23
15.3
25.
389.
514.
4611
.09
8.35
9.51
3.47
Alu
min
ium
0.27
0.04
0.38
0.22
0.05
0.07
0.13
0.07
0.04
0.17
0.15
0.15
0.42
0.17
0.15
0.13
0.22
0.29
0.14
0.22
0.22
0.08
Co
pp
er
0.03
0.03
0.08
0.03
0.03
0.10
0.03
0.08
0.03
0.09
0.03
0.03
0.21
0.06
0.03
0.05
0.24
0.03
0.03
0.10
0.03
0.12
Lead
0.00
0.00
0.00
0.00
0.00
Bra
ss
Tota
l51
8.23
497.
9041
1.37
387.
5533
8.94
265.
3154
0.14
476.
9732
0.39
452.
6745
3.97
404.
3938
9.69
419.
8141
1.37
80.7
933
5.22
180.
6445
3.51
323.
1233
5.22
136.
84
Aft
er
1997
- R
esi
de
nti
alA
fte
r 19
97 -
Ind
ust
rial
Aft
er
1997
- C
om
me
rcia
l
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-14
Expert’s estimation of the average height of basement floors and roofs
Data about the average height of roofs and basement floors for different building categories
was compiled based on expert’s estimation. As there is no other source of information this
approach represents the only possibility to predict the overall gross volume. Details about the
estimation are shown in Table S11.
Table S11 Expert estimation about average height of basements floors and roofs of
different building categories (sample size: 9)
Category
Average height
basement floor(s)
[m]
Average height
roof [m]
standard
deviation
basement
standard
deviation
roof
Before 1918
Residential 3.75 2.42 0.40 0.34
Commercial 2.21 2.42 1.96 0.34
Industrial 0.57 1.83 1.40 0.90
1919-1945
Residential 3.50 2.42 0.38 0.34
Commercial 1.79 2.42 1.56 0.34
Industrial 0.71 1.83 1.75 0.90
1946-1976
Residential 3.43 2.33 0.42 0.37
Commercial 1.79 1.75 1.56 0.75
Industrial 0.57 1.08 1.40 0.45
1977-1996
Residential 4.43 2.25 1.37 0.48
Commercial 2.00 1.58 1.75 0.61
Industrial 0.50 0.50 1.22 0.50
After 1997
Residential 5.21 2.25 1.96 0.75
Commercial 2.43 0.67 2.43 0.75
Industrial 0.50 0.42 1.22 0.45
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-15
Number of data sources for different materials of a building
Table S12 shows the share of the different building categories of the overall building stock.
Furthermore, the number of data sources used for the generation of specific material
intensities for different selected materials is apportioned.
Table S12 Number of data sources for different materials within a building
Bricks/Mortar Concrete Steel (constructional ) Wood (constructional) Wood (other) Metals (other)
Residential 24.7 12 + (8) 12 + (8) 12 + (8) 12 + (8) 6 + (8) 6 + (5)
Commercial 6.9 8 + (1) 8 + (1) 8 + (1) 8 + (1) 6 + (1) 6 + (0)
Industrial 1.2 2 + (1) 2 + (1) 2 + (1) 2 + (1) 2 + (1) 2 + (1)
Residential 5.4 6 + (7) 6 + (7) 6 + (7) 6 + (7) 4 + (7) 7 + (7)
Commercial 0.9 1 + (2) 1 + (2) 1 + (2) 1 + (2) 6 +(2) 7 + (2)
Industrial 0.7 1 + (1) 1 + (1) 1 + (1) 1 + (1) 0 + (1) 2 + (1)
Residential 16.9 16 + (15) 16 + (15) 16 + (15) 16 + (15) 3 +(15) 3 +(15)
Commercial 3.8 2 + (3) 2 + (3) 2 + (3) 2 + (3) 3 +(3) 3 +(3)
Industrial 2.3 0 + (1) 0 + (1) 0 + (1) 0 + (1) 3 +(1) 3 +(1)
Residential 9.6 4 + (7) 4 + (7) 4 + (7) 4 + (7) 1 + (7) 1 + (7)
Commercial 6.5 1 + (1) 1 + (1) 1 + (1) 1 + (1) 1 + (1) 1 + (1)
Industrial 2.7 0 + (1) 0 + (1) 0 + (1) 0 + (1) 1 + (1) 1 + (1)
Residential 5.1 12 + (1) 12 + (1) 12 + (1) 12 + (1) 12 + (1) 1 + (1)
Commercial 3.3 3 + (0) 3 + (0) 3 + (0) 3 + (0) 1+ (0) 1+ (0)
Industrial 0.5 0 + (0) 0 + (0) 0 + (0) 0 + (0) 0 + (0) 0 + (0)
9.5 - - - - - -no information
Number of data collected + (data from literature)Volume in
building
stock %
Category
1977-1996
After 1997
Before 1918
1919-1945
1946-1976
2016 Journal of Industrial Ecology – www.wileyonlinelibrary.com/journal/jie
S-16
References
Albrecht, R., L. Paker, S. Rehberg, and Y. Reiner. 1984. Umweltentlastung durch@: okologische Bau-
und Siedlungsweisen [Environmental relief through ecological building and settlement
patterns]: Sl.
Baccini, P. and A. Pedraza. 2006. Bestimmung von Materialgehalten in Gebäuden. In Bauwerke als
Ressourcennutzer und Ressourcenspender in der Langfristigen Entwicklung urbaner Systeme.
[Identification of material contents in buildings. In: Buildings as resource consumer and
resource donor] Zürich: vdf Hochschulverlag AG.
Blengini, G. A. 2009. Life cycle of buildings, demolition and recycling potential: A case study in Turin,
Italy. Building and Environment 44(2): 319-330.
Görg, H. 1997. Entwicklung eines Prognosemodells für Bauabfälle als Baustein von
Stoffstrombetrachtungen zur Kreislaufwirtschaft im Bauwesen, Schriftenreihe WAR -
Wasserversorgung · Abwassertechnik · Abfalltechnik · Umwelt- und Raumplanung der TH
Darmstadt. Darmstadt, Germany: Verein zur Förderung des Instituts WAR -
Wasserversorgung · Abwassertechnik · Abfalltechnik · Umwelt- und Raumplanung der TH
Darmstadt.
Gruhler, K., R. Böhm, C. Deilmann, and G. Schiller. 2002. Sofflich-energetische Gebäudesteckbriefe -
Gebäudevergleiche und Hochrechnungen für Bebauungsstrukturen.[Material and energetic
building characterization] Band 38 vols, IÖR-Schriften. Dresden, Germany: Institut für
ökologische Raumentwicklung
Rentz, O., A. Seemann, and F. Schultmann. 2001. Abbruch von Wohn- und Verwaltungsgebäuden -
Handlungshilfe, edited by L. f. U. Baden-Württemberg. Karlsruhe, Germany.
Schulze, H.-J., P. Walther, and M. Schlüter. 1990. Gebäudeatlas-Mehrfamilienwohngebäude der
Baujahre 1880-1980 [Building atkas - Multi family houses built 1880-1980]: Bauinformation.
Wedler, B. and A. Hummel. 1947. Trümmerverwertung: technische Möglichkeiten und wirtschaftliche
Bedeutung: Wilhelm Ernst & Sohn.
3rd
Paper
Using change detection data to assess amount and composition of demolition waste from
buildings in Vienna
Fritz Kleemann, Hubert Lehner, Anna Szczypińska, Jakob Lederer, and Johann Fellner
Resource Conservation and Recycling
DOI: 10.1016/j.resconrec.2016.06.010
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
Resources, Conservation and Recycling xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Resources, Conservation and Recycling
journa l homepage: www.e lsev ier .com/ locate / resconrec
Full length article
Using change detection data to assess amount and composition ofdemolition waste from buildings in Vienna
Fritz Kleemann a,∗, Hubert Lehner b, Anna Szczypinska b, Jakob Lederer a, Johann Fellner a
a Christian Doppler Laboratory for Anthropogenic Resources, Institute for Water Quality, Resource and Waste Management, Technische Universität Wien,Karlsplatz 13/226, A-1040 Vienna, Austriab Vienna City Administration, Municipal Department 41 – Urban Survey, Muthgasse 62, A-1190 Vienna, Austria
a r t i c l e i n f o
Article history:Received 18 March 2016Received in revised form 13 June 2016Accepted 13 June 2016Available online xxx
Keywords:Construction and demolition wasteDemolition waste generationUrban miningImage matchingChange detectionGIS
a b s t r a c t
Major waste streams in urban areas result from the demolition of buildings. In the case of lack of data ondemolition waste generation at the municipal level, the quantity and composition of demolition wastesfrom buildings can be estimated by multiplying the volume of demolished buildings, which is takenfrom statistical data sets, by their material composition. However, statistical data sets about the numberand thus total volume of buildings demolished are often incomplete. This paper presents an alternativeapproach to validating demolition statistics (number and volume of buildings demolished) and subse-quently demolition waste generation by applying change detection based on image matching to the casestudy of the city of Vienna, Austria. Based on this technique, building demolition activities not reportedto statistical municipal departments can be identified. Results show that in the city of Vienna, demolitionstatistics yield a total volume of 1.7 M m3/a demolished building volume, while change detection basedon image matching yields a total volume of 2.8 M m3/a. Consequently, demolition waste generation fig-ures solely based on statistical data probably underestimate the total waste generation, which can havesignificant consequences for the estimation of landfill space and recycling plant capacity required. Forthis reason, the approach presented is not only a useful tool for validating existing data on demolitionwaste generation and demolition statistics, but can also be used when these data sets are not existent atall.
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
Waste resulting from construction and demolition activitiesgreatly contributes to overall waste generation in most industri-alized societies. The European Commission (2014) estimates that25–30% of all waste generated in the EU can be attributed to con-struction and demolition waste (CDW). Due to the high potentialto reduce the use of primary resources and landfill space, recyclingtargets for CDW for all member states were defined by the EuropeanParliament and Council of the European Union (2008). Accordingly,a minimum of 70% (by weight) of non-hazardous construction anddemolition waste (soil and stone excluded) shall be prepared forreuse or be recycled or undergo other material recovery by the year2020. In order to assess whether this target has been reached, dataon waste generation and waste flows recycled are required. In the
∗ Corresponding author.E-mail address: [email protected] (F. Kleemann).
case that it turns out that recycling targets are not met, recommen-dations can be given to increase the recycling rates in the CDWsector based on the CDW generation data retrieved. Contrary toother waste streams (e.g. packaging waste, WEEE, etc.), CDW man-agement is mostly realized at a regional rather than a national level,and only a very small share of the overall CDW stream (e.g. metals)is subject to the national and international market. The reason isthat the mineral fraction, which represents the bulk of the CDW,is characterized by a rather small trade and transport radius dueto its large quantity and the low value per mass unit if comparedto waste metal scrap, for instance. Hence, regional and local dataon CDW generation needs to be established, verified and broughttogether to obtain coherent information for CDW management andrecycling, which in turn is necessary to analyse current practices inmanaging CDW and to detect optimization potentials.
Demolition waste (DW) from buildings plays a crucial role whenit comes to meeting recycling targets for two reasons. On the onehand, buildings are much more diverse in their material composi-tion compared to most civil infrastructure (e.g. roads, railways, pipe
http://dx.doi.org/10.1016/j.resconrec.2016.06.0100921-3449/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
2 F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx
networks), making potential recycling rates of materials hardlyassessable and predictable due to lack of data. On the other hand, incities, which are the main drivers of material consumption and DWgeneration in urbanized societies, the bulk of the material stock inthe built environment is located in buildings rather than civil infras-tructure (Tanikawa and Hashimoto, 2009). Despite this fact, in somecountries like Austria, no differentiation between DW from build-ings and civil infrastructure is provided in the DW statistics, andthus no information is available to what extent building demolitioncontributes to overall CDW generation (BMLFUW, 2011).
Research addressing CDW management has resulted in numer-ous publications in recent years (Yuan and Shen, 2011). The focus ofstudies ranges from single buildings to global material flow stud-ies of different materials, which in many cases also consider thebuilt environment. The following paragraph distinguishes betweenstudies modelling CDW on a national or regional level and thosefocusing on methods meant to be applied on a building scale. Themethods applied in the studies vary considerably and range fromanalysing historic statistical data to utilising building informationmodelling for the prediction of CDW.
Studies focusing on CDW generation on a national or regionallevel include the work of Hashimoto et al. (2009), where materialstocks in buildings and infrastructure and thereof resulting wastestreams are estimated for Japan. For the calculation of waste frombuildings, the demolished floor space and material per m2 of floorspace are used. Bergsdal et al. (2007) present a procedure to projectwaste flows entering the Norwegian waste management system sothat it meets future requirements in terms of capacity and tech-nical facilities. The model considers different activity levels in theconstruction industry. Waste generation factors based on stocksand flows of materials are used to estimate the waste amount.Through Monte Carlo simulation uncertainties are considered tomake results more robust. Fatta et al. (2003) estimated quantitiesof CDW for Greece based on the number of demolition licenses andcertain assumptions about the amount of waste generated per m2
demolished building area. The study further focuses on scarcity ofsuitable landfill space and hazardous substances in CDW. For a largescale region in the United States, Cochran and Townsend (2010)used a material flow analysis approach to estimate CDW based onstatistics about construction materials consumed in the past andassumptions about the lifetimes of these materials. Through thisapproach the CDW amount and composition for the year 2002 waspredicted.
Tanikawa and Hashimoto (2009) developed a 4d GIS (geo-graphical information system) to investigate spatial and temporalcharacteristics and changes in the accumulation of material inbuildings and infrastructure. This study focuses more on a regionallevel investigating two case areas in Japan and the UK. Hu et al.(2010) model input and output flows as well as stocks of the urbanresidential building system in Beijing city based on a survey oftypical residential buildings. For the Lisbon Metropolitan Area inPortugal, De Melo et al. (2011) describe how CDW generation canbe estimated based on construction activity and waste load move-ments of different stakeholders involved in construction activitiesin the region. The estimation aims at improving CDW managementinfrastructure and avoiding illegal dumping in the region.
Studies focusing on a building level include Solís-Guzmán et al.(2009), who developed a model to estimate the amount of wastegenerated during both construction and demolition. The model isbased on the investigation of the bill of quantities and coefficientsto estimate demolished volume, wreckage volume, and packagingvolume. The model was applied to two case studies (one construc-tion and one demolition project). Cheng and Ma (2013) facilitatebuilding information modelling (BIM) to estimate building materi-als. Their system allows extracting information about materials insingle buildings and estimating waste amounts to efficiently plan
recycling and reuse procedures and to calculate disposal fees. Theyapply their method to a building in Hong Kong. A prerequisite forthe application of the method is that all relevant materials of therespective building are recorded in BIM.
While in many of the studies mentioned a lot of effort is putinto developing models to simulate material stocks and flows in thebuilt environment, the basis of the calculation, e.g. material com-position of buildings, often seems not to be considered of equalimportance. Moreover, the actual lifetime of buildings dependson numerous factors, making reliable predictions about averagelifetimes almost impossible and thus calls into question the over-all applicability of building models based on lifetime assumptions(Kohler and Yang, 2007). In addition to that, the decoupling of CDWmanagement from its regional context appears inappropriate aslong current management practice remains at a local level. In orderto generate data on the amount and quality of DW occurring ina city, detailed knowledge on the material composition of build-ings, on the one hand, and about the demolition activity, on theother, is required. As shown in Kleemann et al. (2014), however,the quality (incl. completeness) of statistical data on the mate-rial composition of DW from buildings is often poor. With regardto data on demolition activities, only rudimentary documentationis frequent. Consequently, alternative approaches to generatingdata on DW generation and composition are required in orderto assess the quality of existing data. Thus, the objective of thisarticle is to present such an alternative, and, to our knowledge, anovel approach, exemplarily applied to the case study of the cityof Vienna, the capital city and cultural and economic centre ofAustria, with about 1.8 million residents (representing around 20%of the population in Austria). The study at hand aims at estimatingthe amount and composition of DW generated through demolitionactivities in the building sector. Information about the demolitionactivity is based on statistics (Subsection 2.1) and on data derivedfrom remote sensing techniques, which is a data source that has sofar not been used to verify reported DW quantities (Subsection 2.2).The generation of data on the specific material intensities of differ-ent building categories through different approaches is explainedin Subsection 2.3.
2. Materials and methods
Different data sources are used in this study to estimate theamount and composition of waste resulting from building demo-lition. The demolition activity (demolished gross volume [m3]) ofdifferent building categories in Vienna is evaluated based on (i)statistics and (ii) image matching based change detection data.Subsequently, specific material intensities (kg/m3 gross volume)for 15 different building categories (which discriminate betweenconstruction period and utilization) are assigned to the demol-ished building volume of the respective categories to assess theamount and composition of the demolition waste. All data sourcesare processed in a GIS model. Fig. 1 summarises the data genera-tion approach schematically, which is described in more detail inthe following subsections.
2.1. Statistics-based analysis of the demolition activity
The demolition statistics provided by the municipal buildingauthority (MA 37) is based on the collection and documentation ofdemolition notifications and contains only information about theaddress of the buildings being demolished, but not about any fea-tures of the respective building itself. Hence, in order to obtain dataabout the size and type of each building, different municipal GISdata sets are combined. The data set, provided by the municipaldepartment for urban survey (MA 41), contains terrestrially and
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx 3
Fig. 1. Schematic description of the procedure involved in assessing the amount and composition of demolition waste arising from buildings based on statistical data (lefthand side) and image matching based change detection (right hand side) combined with a GIS building model.
photogrammetrically measured data on the size of each building(area [m2] and height [m]), which subsequently allows the calcu-lation of the gross volume (GV [m3]). Data on construction periodand utilization (residential, commercial, or industrial) of all build-ings in Vienna are collected through data sets of different municipaldepartments (city development and planning (MA 18), architectureand city design (MA 19), district planning and land use (MA 21)) andspatially joined to the data set provided by the MA 41. The result-ing data set is referred to as building model in this paper. Addresspoints of the MA 21 are used to locate the potential demolition sites(addresses in the demolition statistics). Orthophotos of differentyears, which are processed from the annual aerial image campaignof the municipal department for urban survey (MA 41), are thenused to verify whether a demolition took place and which partsof a building were subject to the demolition. The building modelallows the determination of size, construction period and utiliza-tion of the buildings that have been demolished. Fig. 2 shows a largedemolition site in Vienna before (left) and after (middle) demoli-tion as well as the polygons of the buildings of the building model(right).
The resulting data, providing information about size (expressedas m3 GV), construction period, and utilization of the demolishedbuildings, are subsequently combined with specific material inten-
sities (given in kg/m3 GV) of the different building categories (seeSubsection 2.3) to assess the overall amount and composition ofDW generated.
2.2. Demolition activity analysis based on remote sensing imagematching
The recording of observed demolition activities randomly car-ried out by members of TU Wien indicated that not all buildingsdemolished in Vienna are covered by the demolition statisticsdescribed in Subsection 2.1. To verify the statistics based dataon demolition activities in the city of Vienna, a second datasource is utilized. So-called image matching is currently used bythe municipal department for urban survey (MA 41) in order todetect changes between the building model and the current build-ing stock. Image matching received a boost with the publicationof semi-global matching by Hirschmüller (2005, 2008). Scientifi-cally, image matching is still being improved by applying differentalgorithms depending on the purposes. Some studies use imagematching as an economic alternative to laser scanning (Brenner,2005). Deriving building models based on image matching datais an established technique and is used e.g. to detect the degreeof damage after seismic events (Turker and Cetinkaya, 2005), to
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
4 F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx
Fig. 2. Data collection based on demolition statistics, orthophotos and building model (black framed polygons).
retrace the development of urban areas (Paparoditis et al., 1998),or to reconstruct historic settlement changes (Nebiker et al., 2014).Dense image techniques are meanwhile available in many pho-togrammetric standard software products and enable to derivedigital surface models from oriented aerial imagery (Haala, 2013).A digital surface model (DSM) is a digital height model, which con-tains the earth’s surface and all natural and man-made objects onit, such as buildings, vegetation, etc. Digital terrain models (DTM),on the other hand, just represent the bare ground surface. The basicfunction of image matching is to search for corresponding pointsin overlapping images. Due to the known intrinsic of the cameraand the extrinsic orientation of the aerial images, it is possibleto calculate the three-dimensional position of the correspondingpoints in space. This yields a 3D point cloud, which is subsequentlyused to derive a regular height raster of the surface. Difficulties ofthe method mainly regard matching errors in urban canyons dueto the fact that those areas are hardly visible in more than oneaerial image. Shaded areas are sometimes lacking enough textureinformation and thus result in bumpy surfaces. Further problemsoccur on water surfaces due to texture changes on wave patternsand specular reflections which are different from different viewangles. By using many highly overlapping images these effects canbe reduced in the surface generation process.
The municipal department for urban survey (MA 41) employsstandard image matching software with a non-standard workflowin order to reduce detection errors in the DSM, as mentioned before,and in order to derive a quality layer, which highlights areas withinsecure height information. Both the DSM and the quality layerhave a spatial resolution of 25 cm. So far these products have beengenerated for the aerial image campaigns of 2013 and 2014. Theheight accuracy of flat and well textured surfaces in the DSMs isabout ±15 cm.
For the specific analysis for this study a change detection pro-cedure for demolished buildings is set up as follows: a heightdifference model of two DSMs of 2013 and 2014 is calculated bysubtracting the older surface model from the more recent one.Areas with higher height values in 2013 than 2014, such as demol-ished buildings, thus, result in areas with negative values. Thequality layers of both years are used to exclude differences, whichmight occur due to detection errors or insecure height informationin the one or the other DSM. In a next step all areas with values equalor lower than −3 m are converted into polygons in order to createsingle objects for the further processing. The threshold of 3 m wasused to cover changes in the buildings stock of about one floor ormore. The authors are aware that some minor demolition projects(e.g. carports, small garden sheds, etc.) might be neglected; how-ever, all relevant demolition activity is considered by this approach
and the number of polygons to be processed is still manageable. Thenumber of polygons is further reduced by considering only poly-gons with an area of minimum 30 m2. Thereby, smaller objects,especially cars, are excluded but small garden cottages typical forsome areas of the city of Vienna are still considered. To excluderemaining vegetation in the polygon data the so-called Normal-ized Difference Vegetation Index (NDVI) of the orthophoto of 2013is used. On the basis of surface classes for water bodies and glasshouses remaining artefacts from these two classes are removedfrom the polygon data.
Subsequently, the final polygons (Fig. 3a) (city surfaces largerthan or equal 30 m2 with height differences of more than 3 m) arevisually checked using the orthophotos of 2013 and 2014 in orderto verify whether a demolition took place or not (Fig. 3b and c).In cases in which a demolition of a building could be verified, thebuilding model and the joined metadata of the buildings are usedto identify size (m3 GV), construction period and utilization of thebuilding demolished (Fig. 3d).
Finally, the resulting data set (GV, construction period and uti-lization of demolished buildings) is combined with specific materialintensities of the respective building categories (Subsection 2.3)to draw conclusions on the amount and composition of DW frombuildings.
The aerial imagery used for the analysis was taken between 16and 24 July 2013, and between 06 and 07 June 2014, respectively.A subsequent period of about 10.5 months can be covered by thisdata. To allow comparison of the data from the analysis with thestatistical data, a projection to a period of one year is carried out.
2.3. Material composition of building categories
As described in the previous two sections, the assessment ofthe amount and composition of demolition waste from buildingsdepends, on the one hand, on the overall demolition activity (num-ber and size of buildings demolished) and, on the other hand, onthe specific material composition of the buildings torn down. Forthe latter the authors have generated a comprehensive databasecontaining specific material intensities (given in kg/m3 GV) for15 different building categories, which cover the vast majority ofbuildings present in the city of Vienna (Kleemann et al., 2016). Thisdatabase has been compiled through the following steps:
• Detailed on-site investigation of buildings prior to demolition:14 buildings of different utilization and construction period havebeen investigated regarding their material composition priorto their demolition (Kleemann et al., 2014). Built-in materialshave been quantified through analysing available documents
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx 5
Fig. 3. Determination of demolition activities in Vienna using polygons of potentially demolished buildings, based on aerial imagery taken in two different years, in conjunctionwith a GIS building model.
(construction plans, experts’ reports) and detailed on-siteinvestigation with selective sampling and analyses.
• Evaluation of construction plans of demolished buildings: Avail-able building files (mainly construction plans) of around 40buildings reported to being demolished in Vienna have beenanalysed. Depending on the quality of the documents, buildingmaterials used for walls, ceilings and sometimes also for roofor floor constructions could be determined. Data gaps, mostlyregarding materials of low concentration (e.g. plastics, copper,aluminium) have been filled by using specific material intensitiestaken from the case studies analysed in detail.
• Data on new buildings: In order to complete the data set, thematerial composition of newer buildings, which have usually notyet been demolished but might be in the future (e.g. buildingsconstructed after the year 2000), has been compiled. This datahas been derived from documents such as life cycle assessments,tendering documents, construction plans, and accounting docu-ments of newly constructed buildings in Vienna.
• Literature: To put all collected data in context to other stud-ies carried out, data about the material composition of relevantbuildings reported in the literature have been included in thedatabase set up.
Table 1 shows the specific material intensities for differentbuildings used in this study. The data is based on a study conductedby Kleemann et al. (2016).
The specific material intensities used in this research werespecifically collected for buildings of the city of Vienna and arebelieved to be relatively robust, as for most building categories sev-eral building have been characterized for their composition. It hasto be stated, however, that uncertainties arise at different stagesof the collection and analysis of this data and through the catego-rization of buildings. The analysis of available documents underliesuncertainties as the documents might differ from the actual build-ings and errors may occur during the analysis itself. The sameapplies for the on-site investigations carried out; beside all themeasurements in some cases assumptions have to be made. By cat-egorising the buildings the singularity of each building is neglected,which again implies uncertainty and should be kept in mind whenusing this data.
3. Results
3.1. Volume of demolished buildings
Based on the statistics from the building authority about thedemolition activity in 2013 and 2014, the orthophotos of the yearsbetween 2012 and 2014, the polygons of potentially demolishedbuildings based on the height difference model of 2013 and 2014,and the building model from 2012 it was possible to calculate thedemolished gross volume of different building categories.
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
6 F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx
Tab
le
1Sp
ecifi
c
mat
eria
l in
ten
siti
es
(giv
en
in
kg/m
3gr
oss
volu
me)
of
dif
fere
nt
buil
din
g
cate
gori
es
in
Vie
nn
a
(rou
nd
ed
to
two
sign
ifica
nt
dig
its)
.
Mat
eria
l [kg
/m3
BR
I]–1
918
1919
–194
5
1946
–197
6
1977
–199
6
1997
–
resi
den
tial
com
mer
cial
ind
ust
rial
resi
den
tial
com
mer
cial
ind
ust
rial
resi
den
tial
com
mer
cial
ind
ust
rial
resi
den
tial
com
mer
cial
ind
ust
rial
resi
den
tial
com
mer
cial
ind
ust
rial
Min
eral
390
430
280
410
340
320
430
350
340
430
380
170
380
320
290
Con
cret
e
22
23
48
100
120
110
240
270
250
300
350
150
360
310
270
Gra
vel/
san
d
5.8
10
28
2.4
4.3
4.3
Bri
cks
220
270
170
180
220
180
150
170
31
100
12
9 58
34M
orta
r/p
last
er
92
81
52
93
73
29
72
44
51
50
16
3.6
2.7
3.2
Min
eral
fill
33
33
33
4.9
2.3
14
4.9
4.9
Slag
fill
14
5.6
3.7
24Fo
amed
clay
bric
ks0.
6
0.38
16
0.38
2.6
Plas
ter
boar
ds/
gyp
sum
0.15
1.4
0.15
0.15
7.4
0.82
4.2
9.3
3
6.2
Gla
ss
0.26
0.32
0.68
0.49
0.27
0.54
0.58
0.55
0.86
0.89
0.62
0.76
Cer
amic
s
0.47
0.54
0.09
2
0.58
2
0.9
1.8
0.81
0.09
2
0.34
0.22
Nat
ura
l sto
ne
0.01
7
0.01
7
46
0.01
7
2.1
1
1.6
0.78
2.8
7.2
0.47
3.8
Min
eral
woo
l
0.01
7
0.21
0.01
6
0.01
7
0.62
0.55
1.3
0.32
1.3
1.2
1.3
Min
eral
woo
l boa
rds
0.05
3
0.04
7
0.04
7
0.08
4
0.33
(Cem
ent)
asbe
stos
0.28
0.13
0.14
1.9
0.14
1.8
0.04
6
1.6
0.00
57O
rgan
ic
19
3.7
5.8
13
7.1
28
6.5
7.6
7.6
6.7
1
1
10
5.7
5.6
Woo
d
18
3.3
5.8
13
6.6
28
5.9
3.6
3.6
5.4
1.2
1.2
4.3
0.84
2.1
Her
akli
t
0.27
0.06
5
0.99
0.82
0.06
5
0.06
5
0.42
0.06
7
0.03
4Pa
per
/Car
dbo
ard
0.22
0.2
0.67
0.05
6
0.02
8PV
C
0.2
0.17
0.00
72
0.1
0.00
72
0.52
0.27
0.27
0.12
0.12
0.18
0.15
Var
iou
s
pla
stic
s0.
34
0.13
0.06
9
0.34
0.85
0.13
0.13
0.94
0.11
0.11
5.1
0.84
2C
arp
et
1.1
1.1
1.2
0.05
7
0.02
9La
min
ate
0.33
Lin
oleu
m
0.04
0.04
0.04
0.03
0.04
5
0.04
5
0.01
0.04
6
0.02
3A
sph
alt
3.7
3.7
5.2
2.6
Bit
um
en
0.02
3
0.05
8
0.14
0.02
3
0.14
0.11
2.7
2.7
0.2
0.63
0.63
1.1
2.4
1.4
Poly
styr
ene
0.00
28
0.00
28
0.23
0.18
0.18
0.18
1.2
0.38
0.53
Met
al
2.8
4.2
8.7
4.7
6.2
5.9
7 5.
9
12
6.9
14
15
15
9.7
13Ir
on/S
teel
2.7
4.1
8.6
4.6
6
5.8
6.9
5.7
12
6.7
13
14
15
9.5
13A
lum
iniu
m
0.01
9
0.05
1
0.03
1
0.01
9
0.12
0.02
8 0.
033
0.12
0.12
0.14
0.54
0.54
0.15
0.22
0.3
Cop
per
0.06
3
0.07
1
0.02
0.06
3
0.09
2
0.03
1 0.
026
0.06
7
0.16
0.02
8
0.2
0.18
0.02
6
0.02
9
0.07
8To
tal
410
440
290
430
360
350
450
360
350
460
400
180
410
340
310
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx 7
Fig. 4. Demolished building volume (m3 gross volume) of the different categories in 2013 and 2014 based on statistics of the building authority.
In 2013 around 300 demolition notifications were documentedby the building authorities, with a total demolished gross volume of1.9 M m3. In 2014 demolition notifications amounted to 240, whichwas equivalent to a total demolished building volume of 1.5 M m3.Fig. 4 shows the distribution of demolished building volume amongthe building categories for the years 2013 and 2014. About 2.3% and8.6%, respectively, of the overall demolished building volume couldbe assigned to neither a utilization nor a construction period cate-gory (indicated in Fig. 4 by no category assignable – N/A). For thesebuildings it has been assumed that their age and utilization are con-sistent with the total demolished building volume. For buildings forwhich only information about the construction period was avail-able, the utilization pattern has been assumed to be in accordancewith the data available for the respective construction period. Ananalogical procedure has been applied to buildings with informa-tion on utilization but with no data on the construction period. Fig. 4shows that the distribution of demolished building volume amongthe building categories varies from year to year as very few demo-lition projects already strongly influence the overall distribution.
As an alternative method to evaluate the demolition activityin Vienna, change detection based on image matching was used.To facilitate the data, each polygon of a potentially demolishedbuilding had to be checked manually. Although changes smallerthan 30 m2 and less than 3 m in height were excluded computa-tionally, many polygons represented not buildings but e.g. train
wagons, containers, vegetation, trucks or soil movement. Thesepolygons had to be excluded, reducing the number of polygonsfrom around 2600 to around 500. As the manual checking is verylabour intensive, the computational procedure should be improvedwhen applying the method on a regular basis. Different optionsshould be tested in this regard. In case the threshold of 30 m2 isincreased, it is necessary to check whether buildings are neglectedby doing so and what influence this might have on the final results.Another possibility would be to exclude polygons of certain shapes(width-length-ratio) in order to detect e.g. trucks, train wagons andcontainers.
In Fig. 5 results based on the change detection results are shownand compared to average results of the statistics from 2013 and2014. As mentioned, the aerial imagery used for this study wastaken between July 2013 and June 2014. To make the results com-parable to the average results of the yearly statistics, the 10.5 monthperiod between the aerial image campaigns used in the changedetection analysis was projected to one year. For the projected 1.5month it is assumed that the demolition activity and the shares ofbuilding categories being demolished are equal to the actual changedetection results (10.5 month period). This seems appropriate as notrend in demolition activity could be observed analysing the noti-fication statistics. In contrast to construction activities, which arestrongly influenced by the season, demolition works are also carriedout during winter.
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
8 F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx
Fig. 5. Gross volume of demolished buildings (given in m3/a) based on statistics of the building authority (one year average of 2013 and 2014), and image matching basedchange detection (one year equivalent of data from July 2013–June 2014).
The results based on the change detection analysis differ fromthe statistics results, with a total gross volume of demolished build-ings of 2.8 M m3/a and 1.7 M m3/a, respectively. This significantdifference clearly indicates that demolition activities are only partlycovered by the statistics of the building authority. This is also sup-ported by the fact that some demolition projects were observedby the project team and could not be found among the demolitionnotifications. As the statistic is based on the demolition notifica-tions, it can be assumed that a considerable amount of demolitionprojects are not reported. For some building categories (e.g. resi-dential 1977–1996, commercial 1919–1945, industrial 1977–1996)however, the statistics give larger demolished building volumes incomparison to the change detection analysis (Fig. 5). This observa-tion is explained by the fact that the timeframe of both approachesis not equal, meaning that demolition projects might have beenreported but have been demolished prior or after the aerial imageryhas been taken (July 2013 and June 2014).
Fig. 6 underlines the fact that a few big demolition projects havea significant influence on the total demolished building volume byshowing that 6–7% of the biggest demolition projects represent 50%
of the total demolished building volume. Furthermore, the resultsindicate that the demolition of smaller buildings is underrepre-sented in statistics.
3.2. Amount and composition of demolition waste from buildings
By combining the GV of demolished buildings with specificmaterial intensities for the respective building categories, theamount and composition of demolition waste can be estimated.Table 2 summarises the results of different approaches applied inthis study for assessing the demolition activities in Vienna (statis-tics versus image matching based change detection). The results ofthe image matching based change detection approach imply wastegeneration from building demolition in the city of Vienna of about1.1 M t/a (or 610 kg/cap/a). This represents about 0.3% of the overallmaterial stock embedded in buildings (210 tons/cap according toKleemann et al., 2016). In comparison, the estimated annual con-sumption of materials for constructing buildings amounts to about1600 kg/cap/a (based on data on new buildings of the municipalbuilding authority (MA 37)), indicating an annual growth in the
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx 9
0
5
10
15
20
25
30
0 100 200 30 0 400 500
x 10
0,00
0 m
³ gro
ss v
olum
e
Number of dem oli �on projects
Sta�s�cs 2013
Sta�s�cs 2014
Chan ge detec�on analysis
Data source
Fig. 6. Cumulative curve of demolished building volume (m3 gross volume) of demolition projects ranked by size.
Table 2Quantity and composition of waste from demolished buildings in Vienna (based on different data sources).
Material [tons] Demolition activity (volume)based on statistics for 2013
Demolition activity (volume)based on statistics for 2014
One year average demolitionactivity (volume) based onstatistics for 2013 and 2014
Demolition activity (volume)based on image matchingbased change detection(projected for a period of oneyear)
Mineral 750000 540000 640000 1000000Concrete 310000 240000 270000 500000Gravel/sand 5000 3000 4000 4700Bricks 280000 190000 230000 330000Mortar/plaster 110000 75000 91000 140000Mineral fill 22000 15000 19000 27000Slag fill 7300 5100 6200 9000Expanded clay bricks 2400 1500 2000 2200Plaster boards/gypsum 3800 2000 2900 5100Glass 800 540 670 1200Ceramics 1200 780 1000 1500Natural stone 8200 7300 7700 11000Mineral wool 550 290 420 600Mineral wool boards 150 56 100 140(Cement) asbestos 820 500 660 780Organic 21000 13000 17000 27000Wood 15000 10000 12000 20000Reed 420 270 340 440Paper/Cardboard 110 60 86 81PVC 370 240 310 490Various plastics 590 370 480 630Carpet 410 240 330 340Laminate 63 39 51 57Linoleum 54 29 42 66Asphalt 2100 650 1400 2500Bitumen 1700 840 1300 2400Polystyrene 180 75 130 190Metal 13000 13000 13000 22000Iron/Steel 13000 12000 12000 21000Aluminium 210 310 260 470Copper 140 150 150 270Total 780000 560000 670000 1100000
Viennese buildings stock of 2.9 M t/a, which represents about 0.8%of the overall material stock embedded in buildings.
The bulk of the material (96%) is mineral, mainly representedby concrete (44%), bricks (28%) and mortar (13%). Organic mate-rials and metals constitute a very small share (4%) of the overallmaterial composition of DW from buildings, of which wood andsteel are the major contributors. With regard to the recyclability ofmaterials it is important to mention that the bond in which they areused is crucial. Materials in old buildings tend to be more easily sep-arated (e.g. bricks, mortar, wood) than materials such as reinforcedconcrete or various composite materials. Currently, most of the DWfrom Vienna is transported and recycled outside the city. Concreteaggregate is usually used in road construction, and crushed bricks
find application as a plant substrate or are utilized as raw mate-rial in the cement industry. Recycling in the sense that e.g. recycledconcrete is used in new concrete as aggregate is rare. Most organicmaterial is thermally treated, and in some cases wood is used inwood board industry. Due to their value, metals are recycled atvery high rates.
The comparison in Table 2 makes it clear that the statisticalbased approach neglects a relevant part of demolition activitytaking place in Vienna. As all polygons of potentially demolishedbuildings were checked and compared with orthophotos, an over-estimation of the demolition activity via image matching basedchange detection can be excluded.
Please cite this article in press as: Kleemann, F., et al., Using change detection data to assess amount and composition of demolitionwaste from buildings in Vienna. Resour Conserv Recy (2016), http://dx.doi.org/10.1016/j.resconrec.2016.06.010
ARTICLE IN PRESSG ModelRECYCL-3287; No. of Pages 10
10 F. Kleemann et al. / Resources, Conservation and Recycling xxx (2016) xxx–xxx
4. Discussion
The approaches described in this paper show to what extentremote sensing can contribute to monitor demolition activitiesin the city of Vienna. In combination with material data of dif-ferent building categories it furthermore allows the estimation ofamount and quality of waste occurring from demolished buildings.The results are believed to be well-founded; however, estimationsabout the demolished building volume as well as the amount andcomposition of resulting demolition waste underlie uncertainties.Uncertainties regarding the demolished building volume mainlyresult from the quality and topicality of the GIS data comprised inthe used building model. Amount and quality of related demolitionwaste additionally underlie uncertainties which result from thespecific material intensities used for different building categories.The material intensities are based on different sources of informa-tion ranging from document analysis and on-site investigation tothe use of data from relevant literature. Through the categorizationof buildings, generalization was necessary, which by nature impliesuncertainties. Nevertheless for the overall amount of demolitionwaste an uncertainty of below 15% is estimated. For single materi-als especially of lower intensity (e.g. Copper), typically much higheruncertainties are to be accepted, when predicting their overall gen-eration. What advances the study is the fact that quite topical data isused. Whereas the different GIS data is continuously updated by themunicipality, the database containing specific material intensitieswill be improved with ongoing research.
5. Conclusions
The system described can practically be implemented by thebuilding authority as a monitoring tool and estimations aboutwaste streams can be compared with national waste statistics.Due to a new electronic data management system implemented inAustria, all waste related data is collected centrally. Federal statesreceive the relevant data from this central source and can analyseit according to their needs. For the relevant time span, no data wasavailable for comparison at the time the study was carried out. Thisshows that the proposed method can be carried out fast enoughto be used as comparative values for these statistics. Knowledgeabout demolition activities and resulting waste streams in Viennaallow the amount of materials available for recycling and secondaryuse to be predicted. By combining this information with data onplanned buildings and construction (city planning) in the city, thisprojection may help to coordinate the potential use of the recyclingmaterial. The city of Vienna has constantly been growing over thelast few years and a further growth of about 10% in the next decadeis predicted by Statistik Austria (2014). Therefore, it is believedthat construction and demolition activities will remain at a highlevel. For the management of CDW these approaches can providecomparative values to validate statistical data and to help detectinconsistencies.
Acknowledgements
The work presented is part of a large-scale research initia-tive on anthropogenic resources (Christian Doppler Laboratoryfor Anthropogenic Resources). The support of the “WienerStadtbaudirektion – Gruppe Umwelttechnik und behördliche Ver-fahren” is highly appreciated. The financial support of thisresearch initiative by the Federal Ministry of Science, Research,
and Economy and the National Foundation for Research, Tech-nology and Development is gratefully acknowledged. Industrypartners co-financing the research centre on anthropogenicresources are Altstoff Recycling Austria AG (ARA), Borealisgroup, voestalpine AG, Wien Energie GmbH, Wiener Kommunal-Umweltschutzprojektgesellschaft GmbH, and Wiener Linien GmbH& Co KG.
References
BMLFUW, 2011. Bundes-Abfallwirtschaftsplan 2011. Bundesministerium für Land-und Forstwirtschaft, Umwelt und Wasserwirtschaft (BMFLUW), Vienna.
Bergsdal, H., Bohne, R.A., Brattebø, H., 2007. Projection of construction anddemolition waste in Norway. J. Ind. Ecol. 11, 27–39.
Brenner, C., 2005. Building reconstruction from images and laser scanning. Int. J.Appl. Earth Obs. Geoinf. 6, 187–198.
Cheng, J.C., Ma, L.Y., 2013. A BIM-based system for demolition and renovationwaste estimation and planning. Waste Manage. 33, 1539–1551.
Cochran, K.M., Townsend, T.G., 2010. Estimating construction and demolitiondebris generation using a materials flow analysis approach. Waste Manage. 30,2247–2254.
De Melo, A.B., Goncalves, A.F., Martins, I.M., 2011. Construction and demolitionwaste generation and management in Lisbon (Portugal). Resour. Conserv.Recycl. 55, 1252–1264.
European Commission, 2014. Construction and Demolition Waste (CDW).European Commission.
European Parliament, Council of the European Union, 2008. Directive 2008/98/ECof the European Parliament and of the Council of 19 November 2008 on wasteand repealing certain Directives.
Fatta, D., Papadopoulos, A., Avramikos, E., Sgourou, E., Moustakas, K., Kourmoussis,F., Mentzis, A., Loizidou, M., 2003. Generation and management of constructionand demolition waste in Greece—an existing challenge. Resour. Conserv.Recycl. 40, 81–91.
Haala, N., 2013. The landscape of dense image matching algorithms. In: Fritsch, D.(Ed.), Photogrammetric Week ’13. Wichmann, Stuttgart, Germany, pp.271–284.
Hashimoto, S., Tanikawa, H., Moriguchi, Y., 2009. Framework for estimatingpotential wastes and secondary resources accumulated within an economy—acase study of construction minerals in Japan. Waste Manage. 29, 2859–2866.
Hirschmüller, H., 2005. Accurate and efficient stereo processing by semi-globalmatching and mutual information. June 2005, San Diego, CA, USA In: IEEEConference on Computer Vision and Pattern Recognition, vol. 2, pp. 807–814.
Hirschmüller, H., 2008. Stereo processing by semi-Global matching and mutualinformation. IEEE Trans. Pattern Anal. Mach. Intell. 30 (February (2)), 328–341.
Hu, D., You, F., Zhao, Y., Yuan, Y., Liu, T., Cao, A., Wang, Z., Zhang, J., 2010. Input,stocks and output flows of urban residential building system in Beijing city,China from 1949 to 2008. Resour. Conserv. Recycl. 54 (12), 1177–1188.
Kleemann, F., Lederer, J., Aschenbrenner, P., Rechberger, H., Fellner, J., 2014. Amethod for determining buildings’ material composition prior to demolition.Build. Res. Inf., 1–12.
Kleemann, F., Lederer, J., Rechberger, H., Fellner, J., 2016. GIS-based analysis ofVienna’s material stock in buildings. J. Ind. Ecol. (in press).
Kohler, N., Yang, W., 2007. Long-term management of building stocks. Build. Res.Inf. 35, 351–362.
Nebiker, S., Lack, N., Deuber, M., 2014. Building change detection from historicalaerial photographs using dense image matching and object-based imageanalysis. Remote Sens. 6, 8310–8336.
Paparoditis, N., Cord, M., Jordan, M., Cocquerez, J.-P., 1998. Building detection andreconstruction from mid-and high-resolution aerial imagery. Comput. Vis.Image Underst. 72, 122–142.
Solís-Guzmán, J., Marrero, M., Montes-Delgado, M.V., Ramírez-de-Arellano, A.,2009. A Spanish model for quantification and management of constructionwaste. Waste Manage. 29, 2542–2548.
Statistik Austria, 2014. Vorausberechnete Bevölkerungsstruktur für Wien2013–2075 laut Hauptszenario.
Tanikawa, H., Hashimoto, S., 2009. Urban stock over time: spatial material stockanalysis using 4d-GIS. Build. Res. Inf. 37, 483–502.
Turker, M., Cetinkaya, B., 2005. Automatic detection of earthquake-damagedbuildings using DEMs created from pre-and post-earthquake stereo aerialphotographs. Int. J. Remote Sens. 26, 823–832.
Yuan, H., Shen, L., 2011. Trend of the research on construction and demolitionwaste management. Waste Manage. 31, 670–679.