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&amp;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: fritz.kleemann@tuwien.ac.at, 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: fritz.kleemann@tuwien.ac.at (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.