UvA-DARE (Digital Academic Repository)
Improvement of the life cycle assessment methodology for dwellings Meijer, A.
Link to publication
Citation for published version (APA): Meijer, A. (2006). Improvement of the life cycle assessment methodology for dwellings
General rights It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)
Download date: 27 Nov 2018
Arjen Meijer / Improvement of the life cycle assessment methodology for dwellings
Tools that calculate the environmental performance of dwellings based on life cycle assessment (LCA) often do not address the indoor environment of houses. This thesis includes damage to the health of occupants caused by emissions from building materials and local traffic in the LCA methodology for dwellings, resulting in a more accurate description of the actual health impacts of dwellings. Further, the study determines the effects of several options for improving the environmental performance of a single-family row house, such as different types of solar panels, reduction of energy consumption, application of materials with low emissions to indoor air, reduction of radon entrance from the soil and measures to reduce the acoustic and air quality impact of local traffic. Comparison of the environmental performance of high-efficiency solar panels and of current silicon-based solar panels shows a similar performance, despite the higher conversion efficiency of the first. Comparison of the environmental performance of two sets of moderate and more radical measures to reduce the environmental impact of dwellings with the reference dwelling leads to a currently feasible reduction with a factor of 2.7.
2006
Improvement of the life cycle assessment methodology for dwellings Arjen Meijer
Improvement of the life cycle assessment methodology for dwellings
IOS Press BV Nieuwe Hemweg 6b 1013 BG Amsterdam The Netherlands Fax +31-20-6870019 E-mail:
[email protected]
Improvement of the life cycle assessment methodology for dwellings
ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Universiteit van Amsterdam op gezag van de Rector Magnificus prof. mr. P.F. van der Heijden ten overstaan van een door het college voor promoties ingestelde commissie, in het openbaar te verdedigen in de Aula der Universiteit op dinsdag 5 december 2006, te 14:00 uur
door Arjen Meijer geboren te Roosendaal en Nispen
Promotor: Prof. dr. L. Reijnders Copromotor: Dr. M.A.J. Huijbregts Overige leden promotiecommissie: Prof. dr. H.A. Udo de Haes Prof. dr. J.M. Verstraten Dr. J.C. van Weenen
Faculteit der Natuurwetenschappen, Wiskunde en Informatica Improvement of the life cycle assessment methodology for dwellings Arjen Meijer Thesis Universiteit van Amsterdam, Amsterdam, The Netherlands
Design: Cyril Strijdonk Ontwerpburo, Gaanderen DTP: Yvonne Alkemade, Delft Printed in the Netherlands by: Haveka, Alblasserdam ISBN 1-58603-690-4 NUGI 755 Subject headings: life cycle assessment methodology, LCA, sustainable building, indoor air quality Legal notice: the publisher is not responsible for the use which might be made of the following information. Copyright 2006 by Arjen Meijer. No part of this book may be reproduced in any form by print, photoprint, microfilm or any other means, without written permission from the copyrightholder.
Aan mijn ouders
Contents
1 1.1 1.2 1.3 1.4 1.5
I ntroduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 The factor X debate.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Environmental life cycle assessment.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Indoor air quality and building LCA.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Goal.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Outline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2 H uman health damages due to indoor sources of organic compounds and radioactivity.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Part 1 2.1 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2 Methodology.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.2.1 Calculation procedure.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.2.2 Fate factors.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.2.3 Effect factors.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.2.4 Damage factors.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.2.5 Other impact categories.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.2.6 Characterisation factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.3 Discussion.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.4 Conclusion.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Part 2 2.5 Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.6 Methodology.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.6.1 Calculation procedure.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.6.2 Characterisation factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.6.3 Concentrations in building materials.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.6.4 Material input for dwelling.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.7 Results.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.7.1 Material level.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.7.2 Dwelling level.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.8 Uncertainties and restrictions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 2.9 Conclusion.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Appendix 2.1 The calculation of the effective outgoing airflows.. . . . . . . . . . . . 43 Appendix 2.2 Background data.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Appendix 2.3 Human health damage scores of building materials.. . . . . . . . . 61
3 3.1 3.2 3.2.1
uman health damages due to road traffic.. . . . . . . . . . . . . . . . . . . . 63 H Introduction.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Methodology.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Traffic scenarios.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.2.2 3.2.3 3.2.4 3.2.5 3.3 3.3.1 3.3.2 3.4 3.5
Fate factors................ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Effect factors.............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Damage factors.. ......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sensitivity analysis...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results.. .................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scenarios.. ................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sensitivity analysis...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions............... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
67 70 71 72 72 72 72 73 76
Appendix 3.1 Supporting information.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4 4.1 4.2 4.2.1 4.2.2 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.4 4.5 4.5.1 4.5.2 4.6
ife cycle assessment of photovoltaic modules.. . . . . . . . . . . . . . . 81 L Introduction.. ............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Environmental life cycle assessment.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Methodology.............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Implications for early product development.. . . . . . . . . . . . . . . . . . . . 85 System description...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Product specification.. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Production of InGaP solar cells.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Production of mc-Si solar cells.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Module production...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Environmental profile.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Interpretation.. ........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 System optimisation.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Missing data.. ............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Conclusions............... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5 5.1 5.2 5.2.1 5.2.2 5.2.3 5.2.4 5.2.5 5.3 5.3.1 5.3.2 5.3.3 5.3.4 5.3.5 5.3.6
I mproved life cycle assessment of dwellings: a case study.. . . 97 Introduction.. ............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Dwelling scenarios...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Characteristics.. .......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Energy consumption.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Indoor air emissions.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Radon exhalation from soil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Local traffic.. .............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Damage scores........... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Total score................. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Materials................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Energy consumption.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Indoor air emissions.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Radon exhalation from soil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Local traffic.. .............. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.4 5.5 5.5.1 5.5.2 5.5.3
Results.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Discussion and conclusions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Limitations and uncertainties.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Further optimisations.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Conclusions.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Appendix 5.1 Material inventory.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6 C oncluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 6.1 Indoor environment.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 6.2 A factor X feasible for dwellings.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
References.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Summary.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Samenvatting.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Dankwoord.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Curriculum Vitae.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
[]
1 Introduction
This thesis deals with the environmental improvement of dwellings. The potential for such improvements has been subject to much discussion. A part thereof has been, and is, in the context of the factor X debate, in which X is a factor with which the economy, specific economic activities or products can be improved. As this thesis is linked to this debate, section 1.1 will briefly summarize the factor X debate in so far it is relevant to the subject of this thesis. An important matter in the factor X debate is: what is X actually referring to. In practice, X often refers to aggregate material inputs (in kg) or cumulative energy demand. However, this may not adequately reflect the actual environmental impact. As to dwellings one may argue that the evaluation of environmental performance on the basis of life cycle (impact) assessment (ISO, 1997) is in a first approximation a better choice than an environmental evaluation on the basis of aggregate material inputs or cumulative energy demand. Environmental life cycle assessment (LCA) is the subject of Section 1.2. However, as will be discussed in more detail in Section 1.3, when the work on this thesis started, an important aspect of dwellings, the indoor environment, was not or very inadequately represented in LCA-based tools for the assessment of buildings. This thesis addresses this deficiency. It presents tools that allow for the inclusion in life cycle assessment of 1. damages to human health caused by pollutants emitted from building materials to the indoor environment, and of 2. damages to the health of dwelling occupants caused by substances and noise emitted by local road traffic. With the improved life cycle assessment methodology, an LCA is performed of a Dutch single-family reference dwelling in order to determine what factor X is currently feasible for this dwelling. Furthermore, the possibilities are assessed to apply high-efficiency photovoltaic modules as an alternative for current silicon modules in dwellings. This is further elaborated in Sections 1.4 and 1.5 that deal with the goal and scope and the outline of the thesis.
1.1 The factor X debate
In the past years, a discussion started in industrialized countries about the technical possibilities to reduce the environmental impact of human activities. In this discussion, the reduction is represented as the so-called factor X (Reijnders, 1998). The factor X can refer to national economies, but also to specific economic activities (e.g. agriculture) or products (e.g. cars). In Europe, several national governments use a factor X in their policies. In the fourth national environmental policy plan of October 2001, the Dutch government opted for dematerialisation by a factor of 2 to 4 in 2030. Austria, Germany and
[]
[]
Denmark aim to increase resource productivity by a factor of 4 and Italy aims to reduce total material requirement by a factor of 10 by 2050 (European Environment Agency, 2005). When referring to national economies in the factor X debate, the IPAT model is used as general framework. With this model, the environmental impact of an economic activity can be determined as (Commoner, 1972): I = P·A·T
(1.1)
where I is the environmental impact of the economic activity; P is the world population; A is the per capita size of economic activity; and T is the environmental impact per unit of economic activity. The factor X is reflected in this model as a need for a certain reduction of environmental impact per unit of economic activity to obtain a certain environmental impact of the economic activity, given a change in world population and economic activity. For example: To obtain a halved environmental impact of an economic activity, given a doubled level of economic activity, the environmental impact per unit of economic activity needs to be a factor 4 lower:
/
/
I = P·2A·T 2 4
(1.2)
A factor X can be determined for a variety of objects ranging from national economies to individual products or services. Actual applications include economic sectors such as the ‘water chain’ (Jansen, 2003), categories of products such as high protein foods (Jansen, 2003) and specific ways of dealing with products, such as closed loop recycling (Morioka et al., 2005). There have been discussions about which factor X is desirable and/or feasible. Von Weizäcker et al. (1994) popularized the factor 4 as short-term goal for eco-efficiency for traditional products or activities. They gave 50 examples of products or activities with technical solutions to achieve a factor 4 reduction of environmental intensity. In 1994, the Factor 10 Club stated that the resource productivity must be increased by a factor 10 during the next 30 to 50 years (Factor 10 Club, 1995). This might be achieved with new products and services and new methods of manufacturing, as well as with changes in lifestyle. Other proposals focus on factors of 20 or larger (Jansen & Vergragt, 1992; Reijnders, 1996). The environmental impact in the IPAT model is commonly expressed as an indicator such as aggregate material consumption or cumulative energy consumption. These indicators reflect only a part of the total environmental impact. The environmental impact of a kilogram of mercury is different from that of a kilogram of wood. And cumulative fossil energy demand does not always adequately reflect the environmental performance of products
(Huijbregts et al., 2006). Especially when dealing with products, another possible indicator is based on life cycle assessment (LCA), which assesses the whole life cycle of the product or service and takes into account a variety of environmental impacts such as global warming, ozone layer depletion, toxicity and resource depletion.
1.2 Environmental life cycle assessment
According to ISO standardisation guidelines, an LCA study can be divided into four steps: goal and scope definition, inventory analysis, impact assessment and interpretation (ISO, 1997). In the goal and scope definition, the aim and the subject of an LCA study are determined and a ‘functional unit’ is defined. An example of a functional unit is ‘the production of 1000 litre coffee’ with, for instance, the aim to compare the environmental impacts of different types of coffee machines. In the inventory analysis, for each of the product systems considered, data are gathered for all the relevant processes involved in the life cycle. A product system can be considered as a combination of processes needed for the functioning of a product or service. In general, many processes (>1000) are modelled in the inventory. The outcome of the inventory analysis is a list of all extractions of resources and emissions of substances caused by the functional unit for every product system considered. In the impact assessment, it is first determined which impact categories will be taken into account and which extractions and emissions contribute to these impact categories. Impact categories correspond with environmental problems, such as acidification and global warming. There are two types of impact assessment methodologies (Jolliet et al., 2004). The first type determines the impact category indicators at an intermediate position of the source-impact pathway. An example of these midpoint indicators is the decrease of ozone in the stratosphere due to emissions of ozone depleting chemicals, which ultimately causes damage to human health and to ecosystems. The second type determines the impact category indicators at a level of actual damages to the environment. These endpoint indicators can easily be interpreted for further weighting, thus reducing the number of damage categories. An example of these endpoint indicators is damage to human health expressed as disability adjusted life years (DALYs). The CML2000 method is an example of an impact assessment method that uses midpoint indicators (Guinée et al., 2001), while the Eco-Indicator 99 method uses endpoint indicators (Goedkoop & Spriensma, 1999). Next, the magnitude of the potential impact of individual substances within each impact category is determined. This is done by multiplying the inventory list of emissions to the environmental compartments air, water and soil
[]
[]
and extractions with their corresponding characterisation factors. A characterisation factor represents the relative importance of the stressor to an impact category. In formula: Sj =
Q j,x,i · Mx,i i
(1.3)
x
where Sj is the impact score for impact category j per functional unit; Qj,x,i is the characterisation factor for impact category j of substance x emitted to compartment i; and Mx,i is the emission of substance x to compartment i per functional unit. In the next step of the impact assessment, the impact scores per impact category are normalised. This is the calculation of the magnitude of indicator results relative to a reference situation. The reference situation usually regards activities in a geographic area over a given period of time (Guinée et al., 2001). The normalisation is carried out by dividing the impact scores per impact category by the impact scores of the reference situation: Nj =
Sj nj
(1.4)
where Nj is the normalised impact score for impact category j per functional unit; and nj is the normalisation factor for impact category j. The last (optional) step of the impact assessment is the calculation of an environmental index by aggregation of the normalised impact scores. This can be done by attributing weighting factors to the different impact categories, which is commonly done for endpoint indicator methods: A =
Nj · wj
(1.5)
j
where A is the aggregated impact score per functional unit; and wj is the weighting factor for impact category j. The final phase in an LCA study is the interpretation of the results from the previous three steps, to draw conclusions and to formulate recommendations for decision-makers.
1.3 Indoor air quality and building LCA
For buildings, several environmental assessment tools based on LCA have been developed, such as Eco-Quantum, Greencalc or BEAT 2000 (Reijnders & Huijbregts, 2000; Erlandsson & Borg, 2003). These tools often do not or poorly address the indoor environment (Jönsson, 2000; Reijnders & Huijbregts, 2000; Erlandsson & Borg, 2003). This causes an incomplete assessment of the total environmental impact of the building, because of the long time of the use phase and the large time fraction that people spend inside buildings. It also
implies that for measures to improve the indoor air quality, such as the application of balanced ventilation, the environmental impact due to increased material and energy consumption is reflected in the calculations, but the beneficial effects of the measures on the health of the occupants are not taken into account. The absence of the indoor environment in life cycle assessment of buildings is also at variance with increasing concern about indoor air quality. During recent decades, measures have been taken to reduce the energy consumption in buildings. The thermal insulation of buildings has been improved and cracks have been filled. The air tightness of the building has been improved and the ventilation rate shows a decreasing trend (e.g. Stoop et al., 1998). A side effect of this ‘hunt for cracks’ is an increase of concentration of pollutants such as radon, formaldehyde and moulds in the indoor air. The study ‘Concern for tomorrow’ of the Dutch National Institute of Public Health and the Environment showed that for a number of substances that have harmful effects on human health, the concentration in indoor air exceeds the acceptable concentration (Langeweg, 1989). In the appendix ‘sustainable building’ of the Dutch National Environmental Policy Plans, requirements for environmental performance were set regarding the choice of building materials, stimulation of energy saving and indoor air quality. The Dutch ministry of Housing, Spatial Planning and the Environment declared 2004 to be the ‘year of the indoor environment’ to draw attention to the health problems associated with low indoor air quality. All this raises the importance of environmental assessment tools for buildings that include the quantitative assessment of indoor air quality. Indoor environmental issues are generally addressed quantitatively by methods similar to risk assessment. However, the integration of quantitative assessment of the indoor environment in environmental life cycle assessment has been a matter of discussion. Jönsson (2000) stated that there are methodological differences between these two types of assessment. LCA focuses on potential, predicted contributions to regional and global impact categories, while indoor climate assessment focuses on actual health effects on a local (indoor) level. LCA presents the results as impacts aggregated globally and over time, while indoor climate assessment is specific or restrictive as to time and space. Finally, indoor climate assessment is only applicable at a building level. Jönsson noted possibilities to address indoor climate as an impact category in LCA. However, characteristics of the indoor environment cannot be fully allocated to separate building materials alone; they are also determined by the construction and use of the building. Furthermore, timedependent effects, such as decreasing emission rates of VOCs from building materials, are difficult to implement in LCA. Therefore, according to Jönsson (2000), only limited parts of the indoor environmental issues can be addressed in building LCAs.
[]
[]
The SETAC working group ‘LCA in Building and Construction’ addressed the issue of indoor climate in building LCAs (SETAC, 1999). SETAC states that LCAs for building products might be performed for comparison of different products, or as an input for assessments of whole buildings. In the latter case, the LCA is split up in two parts: partial LCAs of the products and a partial LCA for the use phase of the building, in which a.o. indoor climate issues can be addressed (Paulsen & Borg, 2003). This overcomes the problem of the difference between global effects assessed in product LCAs and local effects assessed in building LCAs. The methodology proposed in this study avoids the difficulties raised by Jönsson (2000). Fate factors are calculated for a reference dwelling, so that emissions from building materials can be attributed to those building materials. The emission rate of harmful substances from building materials is averaged over the lifetime of the building material to convert actual non-constant emissions to continuous emissions. Combined with constant airflows in the building, this results in steady-state concentrations in the indoor environment. Furthermore, in line with the approach of Pennington et al. (2002), it is assumed that at low (indoor) exposure concentrations of a specific organic compound, the dose-response relationship for non-carcinogenic effects is linear with no threshold.
1.4 Goal
The goal of this thesis is twofold. The first goal is the improvement of the environmental life cycle assessment methodology for buildings. This will be done by incorporating damage to the health of occupants as a result of indoor emissions from building materials and emissions from local traffic. The second goal of this thesis is the determination of a factor X that is currently feasible for dwellings. Three dwelling scenarios are defined with different sets of measures to decrease the environmental impact of the dwelling. The environmental performance of these scenarios is calculated and compared using the improved LCA methodology. Furthermore, as energy consumption during the use phase of the dwelling appears to be an important factor in the comparison, one example of the possibilities of environmental improvement of energy consumption is assessed in detail: solar cells with a higher conversion efficiency than current silicon-based cells. A new type of indium-gallium-phosphide (InGaP) photovoltaic panels is compared with current multicrystalline silicon photovoltaic panels. InGaP/multicrystalline silicon tandem panels have an efficiency of up to 25% (The Helioz Project, 2002), which is higher than the 14.5% efficiency of multicrystalline silicon panels (Steeman et al., 1995).
1.5 Outline
This thesis consists of two parts. In the first part, the life cycle assessment methodology for dwellings is improved by adding local health effects. In Chapter 2, the health effects of emissions from building materials to the indoor air of the dwelling are determined. Characterisation factors are calculated for 36 organic compounds, radon and gamma radiation emitted by building materials applied in a Dutch reference dwelling. Human health damage scores per kilogram of building material for compartments of the Dutch reference dwelling are calculated using the method developed in this chapter. These scores are compared with the damage to human health due to the emissions occurring in the production and waste phase of these materials. In Chapter 3, a methodology is developed to calculate damages to human health of occupants caused by substances and noise emitted by local road traffic. Three traffic scenarios in residential areas are defined. Damage scores are calculated for these three scenarios based on emissions of noise, particulate matter (PM10), sulphur dioxide, benzene and benzo[a]pyrene. These damages are compared with the damage to human health due to the emissions occurring in the production, use and waste phase of the building materials used in the Dutch reference dwelling. In the second part of the thesis, two case studies are carried out. The first concerns solar cells, a component of dwellings that may reduce the environmental impact thereof. In Chapter 4, the environmental aspects and environmental potential as to further development of the InGaP (tandem) module to a more mature technology are assessed by comparing the environmental and energy-based profiles of a mechanically stacked InGaP/multicrystalline silicon (mc-Si) tandem module with those of its constituents i.e. a thin-film InGaP module and a mc-Si module. This is used as an assessment of the possibilities to apply InGaP/mc-Si tandem modules as an alternative for mc-Si modules in dwellings. In Chapter 5, an LCA methodology incorporating the effects of indoor emissions and local traffic on the health of the occupants is used to compare the environmental impacts of two packages of measures to improve the environmental performance of the Dutch single-family reference dwelling: one with moderate changes and one with more radical changes, if compared with current common practice. This results in a factor X, based on environmental impacts, that is currently feasible for dwellings. Finally, Chapter 6 deals with concluding remarks.
[]
[]
2 Human health damages due to indoor sources of organic compounds and radioactivity Part 1 Characterisation factors
Abstract Goal, Scope and Background. Methodologies based on life cycle assessment have been developed to calculate the environmental impact of dwellings. Human health damage due to exposure to substances emitted to indoor air are not included in these method‑ ologies. In order to compare this damage with human health damages associated with the rest of the life cycle of the dwelling, a methodology has been developed to calculate damages to human health caused by pollutants emitted from building materials. Methods. Fate, exposure and health effects are addressed in the calculation procedure. The methodology is suitable for organic substances, radon and elements emitting gamma radiation. The (Dutch reference) dwelling used in the calculation was divided in three compartments: crawl space, first floor and second floor. Fate factors have been calculated based on indoor and outdoor intake fractions, dose conversion factors or extrapolation from measurements. Effect factors have been calculated based on unit risk factors, (extrapolated) effect doses or linear relationship between dose and cancer cases. Damage factors are based on disability adjusted life years (DALYs). Results and Discussion. Characterisation factors have been calculated for 36 organic compounds, radon and gamma radiation emitted by building materials applied in a Dutch reference dwelling. For organic compounds and radon, the characterisation fac‑ tors of emissions to the second floor are 10‑20% higher than the characterisation fac‑ tors of emissions to the first floor. For the first and second floor, the characterisation factors are dominated by damage to human health as a result of indoor exposure. The relative contribution of carcinogenic and non-carcinogenic effects to the characterisa‑ tion factors is generally within one order of magnitude, and up to three orders of mag‑ nitude for formaldehyde. Conclusion. Health effects due to indoor exposure to pollutants emitted from build‑ ing materials appear to be dominant in the characterisation factors over outdoor ex‑ posure to such pollutants. The health effects of emissions of organic compounds and gamma radiation in the crawl space are very small, compared to the health effects of emissions into the other compartments. Using the characterisation factors calculated in this study, it is possible to calculate the human health damage due to emissions of substances and radiation emitted to indoor air and compare this damage with dam‑ ages to human health associated with the rest of the life cycle of the material. This is the subject of Part 2 of this research. Published as: Meijer, A., M.A.J. Huijbregts & L. Reijnders, 2005, Human health dam‑ ages due to indoor sources of organic compounds and radioactivity in life cycle impact assessment of dwellings – Part 1: Characterisation factors, in: International Journal of
[ 10 ]
[ 11 ] Table 2.1 Human health impact categories Human health impact category
Compound
Exposure
Carcinogenic effects
Organic compounds
Indoor and outdoor
Non-carcinogenic effects
Organic compounds
Indoor and outdoor
Effects of ionising radiation
Radioactive compounds
Indoor and outdoor
Life Cycle Assessment 10 (5): pp. 309-316.
Respiratory effects due to ozone creation
Organic compounds
Outdoor
Effects of climate change
All
Outdoor
Effects of ozone layer depletion
All
Outdoor
2.1 Introduction
In life cycle impact assessment (LCIA), several methods have been developed to calculate the impact of emissions of harmful components on human health (Guinée & Heijungs, 1993; Hertwich et al., 1998; Hofstetter, 1998; Krewitt et al., 1998; Goedkoop & Spriensma, 1999; Frischknecht et al., 2000; Huijbregts et al., 2000). These methods take into account outdoor sources of contamination. A method to evaluate the impact of indoor sources on human health due to indoor exposure is, however, still missing (Jönsson, 2000; Reijnders & Huijbregts, 2000). The reason for this absence is that LCIAs usually do not take into account local effects of products on users (Jönsson, 2000). However, environmental comparisons and improvements for building products may be biased by excluding the impact of indoor air pollution. For instance, human health damage scores of concrete compared to wood may be underestimated by excluding indoor air emissions of radon and gamma radiation. Another example is that the positive influence on human health of mechanical indoor air ventilation in buildings is not accounted for by disregarding impacts of indoor air pollution. The impact of indoor pollution on human health may be an important factor for the LCIA of dwellings, because people live in houses for a great part of their lives. The Dutch Health Council for instance estimated the number of casualties due to lung cancer as a result of exposure to radon in the Netherlands at 800 per year (Gezondheidsraad, 2002). In a review of radiation exposure in the Netherlands, it appears that nearly 50% of the total average annual dose per capita of the Dutch population originates from radon or gamma radiation from building materials (Eleveld, 2003). Apparently, the exclusion of indoor exposure to radioactive elements originating from building materials leads to an underestimation of the human health risks in the life cycle assessment of dwellings. Something similar may hold for organic pollutants. For instance, Sexton et al. (2004) showed an increased indoor concentration of fifteen organic compounds in three urban communities, if compared with outdoor concentrations. This article presents characterisation factors for 36 organic compounds, radon and gamma-radiating elements present in building materials. The characterisation factors are calculated for a Dutch reference dwelling (Novem, 1998; W/E Adviseurs, 1999). It is assumed that this dwelling is occupied by three persons. Fate, effects and damages are incorporated in the characterisation factor calculations (Goedkoop & Spriensma, 1999). Fate factors of organic compounds and radon are calculated using an indoor airflow and exposure model for dwellings. Exposure in both indoor and outdoor environment is
Source: Goedkoop & Spriensma (1999)
considered. Effect factors are calculated using unit risk factors for carcinogenic effects (Goedkoop & Spriensma, 1999), no observed effect levels (NOELs) and lowest observed effect levels (LOELs) for non-carcinogenic effects (Pennington et al., 2002), and epidemiological data for ionising radiation (Frischknecht et al., 2000). Damages to human health are expressed in disability adjusted life years (DALYs) (Hofstetter, 1998; Goedkoop & Spriensma, 1999; Frischknecht et al., 2000). The impact categories that are taken into account in this research are given in Table 2.1. Carcinogenic effects, non-carcinogenic effects and effect of ionising radiation are relevant for exposure in both indoor and outdoor environment. Respiratory effects caused by ozone creation, effects of climate change and effects of ozone layer depletion are only relevant for exposure in outdoor environment. Parameter values used in this methodology are given in Appendix 2.2.
2.2 Methodology
2.2.1 Calculation procedure
In the LCIA methodology, characterisation factors can be used to calculate the combined environmental damage occurring in the life cycle of a product (Heijungs & Hofstetter, 1996). In Figure 2.1, an overview of the steps from emission to damage is given. For compounds in building products, characterisation factors can be used to link (activity) concentrations to human health damage. For radon, the total amount of radon exhaled during the lifetime of the building material is used instead of activity concentration. The damage score for the use phase of building material p can then be calculated by Eq. (2.1): DSp,u =
Mx,p · Q x
(2.1)
x
where DSp,u is the damage score associated with the use phase of building material p (DALY·kgp-1); Mx,p is the (activity) concentration of compound x in building material p or the total amount of radon exhaled during the lifetime of building material p (kg·kgp-1 or Bq·kgp-1); and Qx is the characterisation factor
[ 12 ]
[ 13 ] kcWdjen_Y_jo
=Wbb_kc,D
kcWd^[Wbj^
;Yeioij[cgkWb_jo
IY[dWh_e'
IY[dWh_e(
DSh,t,s = DSh,t,tsa – DSh,t,tsm
H[iekhY[Z[fb[j_ed
JejWb
IY[dWh_e)
(5.6)
where DSh,t,s is the damage score for health damages due to local traffic in scenario s (y); DSh,t,tsa is the damage score for health damages due to local traffic in the actual traffic scenario (y); and DSh,t,tsm is the damage score for health damages due to local traffic in the traffic scenario with minimum damage to human health (y). Damage scores are calculated as differences between scenarios because of the non-linear relationship between traffic density and noise level, and because there are threshold values for the noise levels above or under which a change in noise level has no effect on the human health (Meijer et al., 2006). Damage scores have been calculated for health damages due to traffic emissions of particulate matter (PM10), sulphur dioxide (SO2), benzene, benzo[a]pyrene and noise (Meijer et al., 2006).
5.4 Results
In Figure 5.1, the damage scores per damage category are compared for the different scenarios. For all damage categories, scenario 1 has the highest damage score, followed by scenario 2 (a factor of 1.1 to 2.1 lower than scenario 1) and scenario 3 (a factor of 1.3 to 3.1 lower than scenario 1). The total damage score for scenario 2 is a factor of 1.7 lower than the total damage score for scenario 1, and the total damage score for scenario 3 is a factor of 2.7 lower than the total damage score for scenario 1. Damage to human health accounts for 54 to 68% to the total damage score, damage to ecosystem quality accounts for 1 to 3%, and resource depletion accounts for 31 to 44%.
CWj[h_Wbi
;d[h]o
?dZeehW_h[c_ii_edi
HWZed[n^WbWj_ed\hecie_b
BeYWbjhW\\_Y
In Figure 5.2, the contributions of the impact sources to the damage scores for the damage category human health are given for the three scenarios. Figure 5.2 shows that the relative contribution of the impact sources indoor air emissions, radon exhalation from soil and local traffic decreases from scenario 1 (66%) to scenario 2 (33%) to scenario 3 (15%). In scenario 1, local traffic has a contribution of 48%, mainly due to traffic noise levels and fine particulate matter emissions. Indoor air emissions cause 15% of the human health damage, with emissions of radon and gamma radiation from stony materials as the dominant source. Radon exhalation from soil has a relatively low contribution to human health damage in our calculations (3%). In scenario 2 and 3, the contributions of local traffic and radon exhalation from soil is reduced to < 1%, while the human health damage contribution of indoor air emissions is 32% and 14%, respectively. In all scenarios, the absolute value of the human health damage for materials is similar. The application of water-based acrylic paint and polypropylene instead of PVC lowers the health damage score, but the addition of installations such as photovoltaic panels, heat pumps and heat exchangers increases the health damage scores due to its impact on the wider environment. However, the increase of the damage score for materials due to these installations is much lower than the decrease in health damage associated with the lower fossil energy consumption in the use phase of the dwelling, resulting in a decrease of total damage to human health, particularly in scenario 3. The lower damage score for indoor air emissions in scenario 3 is mainly due to the reduction of radon emissions from concrete and sand-lime bricks because of the wooden skeleton of the dwelling. In scenarios 2 and 3, the contribution of local traffic to the total damage to human health is 0, because the damage to
[ 106 ]
[ 107 ]
human health due to local traffic is calculated as a difference rather than as actual values and the lower value is defined as zero value.
5.5 Discussion and conclusions 5.5.1 Limitations and uncertainties
The damage scores calculated in this study are representative for single-family dwellings in the Netherlands. They can be different for other types of buildings (e.g. apartments, offices) and for other countries. The methodology presented in this study can be applied to other situations, but the damage scores for the pollutants have to be recalculated and the inventory of materials and energy consumption has to be adapted. The calculations presented in this study are subject to uncertainties originating from the methodology used to calculate the damage scores. For the effect factors for non-carcinogenic effects of organic compounds, linear doseresponse relationships with no threshold have been assumed for low exposure concentrations, in line with the approach of Pennington et al. (2002). The effect levels, duration and severity of the health effects are uncertain (Meijer et al., 2005a). In the calculations of the damage scores associated with the construction and demolition of the building materials and of energy generation, further uncertainties originate in the averages of the material and energy consumption and emissions in the life cycle. Typical values used in the exposure and effect models applied may also diverge from actual values (Goedkoop & Spriensma, 1999). In general, only pollutants with a known effect on human health are considered. Pollutants for which no emission data or toxicological data are available are left out of the assessment. This can cause an underestimation of the total damage to human health. For emissions from road traffic, only particulate matter with a diameter smaller than 10 µm (PM10), sulphur dioxide (SO2), benzene and benzo[a]pyrene have been included in this study (Meijer et al., 2006). For these pollutants, emission data were available in the underlying CAR II model (Teeuwisse, 2003) and effect factors were available in the EcoIndicator 99 methodology in the hierarchist perspective (Goedkoop & Spriensma, 1999). The radon exhalation rates from soils range from 0.0002 to 0.07 Bq·m-2·s-1 in the Netherlands (Vaas et al., 1993). For the present study, a typical value of 0.016 Bq·m-2·s-1 was selected. If the maximum radon exhalation rate would have been used, a contribution of up to 13% of the total damage to human health is expected. In these cases, reducing the high radon exhalation rates by soil insulation can indeed be profitable to substantially decrease damage to human health.
Due to the uncertainties mentioned above, the damage scores of the different impact sources of the scenarios are uncertain as well. However, the conclusions that the damage scores of scenario 3 are lower than the damage scores of scenario 2, which are again lower than the damage scores of scenario 1, and that the differences between the scenarios would be smaller when the impact sources indoor air emissions, radon exhalation from soil and local traffic would be excluded from the comparison, do not change.
5.5.2 Further optimisations
The measures applied to dwellings considered in this study are relatively easy to implement in current Dutch building practice. There are several possibilities to further improve the environmental performance of the dwellings. First, the design of the dwelling can be adapted to profit optimally from solar irradiation and to minimize heat losses (e.g. Stahl et al., 1994; Florides et al., 2002). This can be done by additional capture of solar irradiation to supply additional heat and lighting at the south side and by applying small windows at the north side of the dwelling. The higher solar irradiation can lead to a lower need for additional energy supply, but overheating of the dwelling in the summer should be prevented. Secondly, measures can be applied to further reduce energy consumption, such as increase of the insulation thickness, application of low emittance double glazing and the use of heat exchangers to regain heat from waste water (Florides et al., 2002; Strootman, 2002; Huijbregts et al., 2003b). Improved insulation, however, may have a negative effect on the indoor air quality due to lower ventilation rates. Also solar panels with a higher electricity yield may be applied (e.g. Meijer et al., 2003; Mohr et al., 2006). The extent to which such measures may also improve the overall environmental performance of dwellings is a matter for further study.
5.5.3 Conclusions
In this study, damages have been calculated for three dwelling scenarios: a reference scenario (scenario 1), a scenario with moderate measures to increase the environmental performance (scenario 2) and a scenario with more radical measures to increase the environmental performance of the dwelling (scenario 3). From the calculations presented in this study, it can be concluded that when moderate measures such as photovoltaic panels, balanced ventilation with heat recovery and decrease of the impact of local traffic are applied to the Dutch reference dwelling, the environmental impact is a factor of 1.7 lower than the environmental impact of the reference scenario. When additional measures are applied, such as the use of solar collectors, heat pumps and a wooden skeleton of the dwelling, the environmental impact is a factor of 2.7 lower.
[ 108 ]
[ 109 ]
Appendix 5.1
The differences in damage scores are shown to be larger when the impact sources indoor air emissions, radon exhalation from soil and local traffic are included in the life cycle assessment of dwellings. Particularly, reduction of the impacts of local traffic, reduction of emissions of radon and gamma radiation from stony materials and reduction of energy consumption during the use phase of the dwelling reduce the damage to human health associated with dwellings. This stresses the significance of the inclusion of indoor exposure and local traffic in the LCA of dwellings.
Material inventory
[ 110 ]
[ 111 ] Table 5.4 Materials inventory in scenario 1 (including construction losses) (kg)
Table 5.4 Materials continuedinventory in scenario 1 (including construction losses) (kg)
Material
Material
Crawl space
First floor
Second floor
Outdoor
Crawl space
First floor
Second floor
Outdoor
ABS
–
1.1
7.6
–
Leaching, zinc
–
–
–
490
Acrylic paint
–
14
28
–
Lead
–
20
7.1
28
Alkyd paint
–
12
9.3
22
Meranti
–
130
190
220
1700
2100
Aluminium
–
13
120
17
Mortar
–
1300
Anodising layer
–
0.041
0.088
0.13
Multiply
–
190
490
700
Artificial stone
–
70
70
–
–
–
12
–
Bitumen
–
–
–
82
Other materials, unspecified
Brass
–
12
32
–
Paper
–
180
230
–
Pinewood
–
210
390
360
Pinewood, creosoted
–
–
–
2000
Plastic layer
–
80
–
–
Polyamide
–
0.33
1.3
–
Bricks
–
1400
1700
Cardboard
–
52
96
–
Cast iron
–
–
15
–
Ceramics
–
410
830
100
Chipboard
–
470
–
–
Chloroprene
–
0.0035
1.1
3100
–
Concrete
16000
16000
35000
Copper
4.3
20
100
–
Electronics
–
–
8.9
–
Enamel
–
2.2
3.0
–
EPDM
–
1.1
2.0
–
Expanded polystyrene 80
80
76
73
Glass
–
220
240
450
Glass wool
–
–
0.68
–
Glue
–
39
45
–
Glue, sand-lime bricks
–
270
340
Glue, water-based
–
30
61
–
Gypsum
–
2200
2500
–
Gypsum plaster
–
320
500
65
Hardboard
–
200
130
–
44
60
21
Leaching, creosotes
–
–
–
410
Leaching, lead
–
19
6.6
25
Leaching, copper
20000
43
–
Polybutylene
–
6.5
9.9
–
Polyester
–
2.5
7.6
–
Polyethylene (HDPE)
–
17
30
–
Polyethylene (LDPE)
–
–
52
46
Polypropylene
–
–
2.7
–
Polysulphide
–
13
17
20
18
31
38
35
PUR foam, air
–
–
10
–
PUR foam, pentane
–
7.3
33
9.7
Rock wool
–
50
64
Sand
Polyvinyl chloride
110
–
–
–
2400
4200
6100
–
Sand-lime bricks
–
17000
21000
2900
Stainless steel
–
0.13
54
4.9 320
Sand mortar
Steel
43000
480
340
660 630
–
Steel, enamelled
–
320
Steel, galvanized
–
49
160
–
Zinc
–
0.49
0.65
67
Sources: Novem (1998); W/E Adviseurs (1999)
[ 112 ]
[ 113 ] Table 5.5 Materials inventory in scenario 2 (including construction losses) (kg)
Table 5.5 Materials continuedinventory in scenario 2 (including construction losses) (kg)
Material
Material
Crawl space
First floor
Second floor
Outdoor
ABS
–
1.1
7.6
–
Acrylic paint
–
35
43
Alkyd paint
–
–
Crawl space
First floor
mc-Si solar cells
–
–
35
Meranti
–
0.58
0.58
Mortar
150
98 0.13
Second floor
Outdoor
–
8.1
130
91
220
–
1300
1700
2100
Multiply
–
180
490
700
Other materials, unspecified
–
–
12
–
Paper
–
180
230
–
Pinewood
–
220
460
360
Pinewood, creosoted
–
–
–
2000
–
–
Aluminium
–
13
Anodising layer
–
0.041
0.088
Artificial stone
–
70
70
–
Bitumen
–
–
–
82
Brass
–
12
32
–
Bricks
–
1400
1700
Cardboard
–
52
96
–
Plastic layer
–
80
Cast iron
–
–
15
–
Polyamide
–
0.33
1.3
–
Ceramics
–
410
830
100
Polybutylene
–
6.5
10
–
Chipboard
–
470
–
–
Polycarbonate
–
–
0.89
–
–
Chloroprene
–
3100
0.0035
1.1
Polyester
–
2.5
7.7
–
Concrete
16000
16000
35000
20000
Polyethylene (HDPE)
–
17
30
–
Copper
4.3
20
110
0.49
Polyethylene (LDPE)
75
–
52
65
Electronics
–
–
8.9
–
Polypropylene
11
8.6
6.1
18
Enamel
–
2.2
3.0
–
Polysulphide
–
13
17
20
EPDM
–
1.1
2.0
22
Polyvinyl chloride
–
16
32
5.8
Expanded polystyrene 80
110
110
130
–
PUR foam, air
–
–
10
Glass
–
220
240
640
Glass wool
–
–
–
PUR foam, pentane
–
7.3
38
9.7
0.68
Glue
–
39
45
–
Rock wool
–
–
3.7
3.7
Glue, sand-lime bricks
–
270
340
Sand
–
–
–
Glue, water-based
–
30
61
Gypsum
–
2200
2500
Gypsum plaster
–
320
500
Hardboard
– – 65
Sand mortar
43000
2400
4200
6100
–
Sand-lime bricks
–
17000
21000
2900
Stainless steel
–
0.13
56
4.9
480
340
660
320
Steel, enamelled
–
320
630
–
190
– 0.078
Steel
200
130
–
44
60
21
–
Leaching, creosotes
–
–
–
410
Steel, galvanized
–
66
Leaching, lead
–
19
6.6
25
Tin
–
–
–
Leaching, zinc
–
–
–
490
Zinc
–
0.49
0.65
Lead
–
20
7.1
28
Leaching, copper
–
43
67
Sources: Novem (1998); W/E Adviseurs (1999)
[ 114 ]
[ 115 ] Table 5.6 Materials inventory in scenario 3 (including construction losses) (kg)
Table 5.6 Materials continuedinventory in scenario 3 (including construction losses) (kg)
Material
Material
Crawl space
First floor
Second floor
Outdoor
Crawl space
First floor
Second floor
Outdoor
ABS
–
1.1
7.6
–
mc-Si solar cells
–
–
Acrylic paint
–
14
16
–
Meranti
–
130
Alkyd paint
–
–
0.58
0.58
Mortar
–
960
1200
2000
150
470
Multiply
380
940
1600
700
0.13
Other materials, unspecified
6200
–
12
–
Paint, based on linseed oil
–
21
27
35
Paper
9.8
180
230
–
Pinewood
18
1500
2700
920
Pinewood, creosoted
490
490
–
2000
Aluminium
–
13
Anodising layer
–
0.041
0.088
Artificial stone
–
70
70
–
Bitumen
–
–
–
82
Brass
5.0
Bricks Cardboard
12
36
2.8
–
1400
1700
3100
–
52
120
11
Cast iron
30
–
15
–
Cellulose
160
690
890
320
Ceramics
0.39
410
830
100
Chipboard
–
470
–
–
Chloroprene
–
0.0035
1.1
–
Concrete
11000
110
2500
13000
Copper
22
20
140
94
Electronics
–
–
8.9
–
Enamel
–
2.2
3.0
–
EPDM
–
1.1
2.1
89
Expanded polystyrene
–
0.023
2.7
–
Fibre mortar
–
75
40
110
–
220
240
1400
0.68
72
Glass
34
Glass wool
–
Glue, water-based
–
30
61
Gypsum
–
3200
5200
210
Gypsum plaster
–
46
170
–
Hardboard
200
130
–
44
60
21
–
Leaching, creosotes
–
–
–
410
Leaching, lead
–
19
6.6
65
Leaching, zinc
–
–
–
600
Lead
–
20
7.1
72
Leaching, copper
–
–
– 91
32 220
Plastic layer
–
80
–
–
Polyamide
–
0.33
1.3
–
Polybutylene
–
6.5
10
–
Polycarbonate
–
–
0.89
– –
Polyester
–
2.5
7.7
Polyethylene (HDPE)
160
17
30
–
Polyethylene (LDPE)
76
3.0
70
130
Polypropylene
11
8.6
6.1
18
Polysulphide
–
13
17
20
Polyvinyl chloride
–
13
28
5.8
PUR foam, air
–
–
–
10
7.4
14
37
10
Rock wool
–
29
50
6.3
Sand
–
–
–
Sand mortar
2100
4000
6100
Stainless steel
40
0.13
180
5.0
Steel
500
80
250
110
Steel, enamelled
14
320
630
–
Steel, galvanized
–
66
190
–
Tin
–
–
Zinc
–
PUR foam, pentane
0.49
43000 –
–
0.31
0.65
100
Sources: Novem (1998); W/E Adviseurs (1999)
[ 116 ]
[ 117 ]
6 Concluding remarks
In this study, the life cycle assessment (LCA) methodology for dwellings has been improved by incorporating damage to the health of occupants as a result of emission from building materials and local traffic to the indoor environment. The effects of measures to improve the environmental performance of the Dutch reference dwelling have been quantified using the improved LCA methodology. The possibilities have been assessed to apply high-efficiency photovoltaic modules as an alternative for current silicon modules in dwellings.
6.1 Indoor environment
The environmental life cycle assessment methodology for dwellings has been improved by incorporating the effects of emissions to indoor air from building materials and local traffic on the health of the occupants. Characterisation factors for 36 organic compounds, radon and elements emitting gamma radiation have been calculated for the Dutch reference dwelling. Damage factors have been calculated for 17 building materials using these characterisation factors. On a dwelling level, the damage to human health due to the emissions occurring in the use phase of the reference dwelling have the same order of magnitude as the damage to human health due to the emissions occurring in the production and waste phase of the building materials used in the same dwelling. Furthermore, damages to the health of occupants as a result of emissions of particulate matter (PM10), sulphur dioxide, benzene, benzo[a]pyrene and noise from local traffic have been calculated for differences between three traffic scenarios and for the Dutch reference dwelling. The change in human health damage due to road traffic noise and pollutants is, depending on the scenario, 1.5 to 2 times higher than the total damage to human health due to exposure to substances emitted by building materials to indoor air and due to the emissions occurring in the production and waste phase of the building materials used in the dwelling. This stresses the significance of the improvements in the LCA methodology for dwellings by taking into account the health effects towards dwelling occupants of emissions from building materials and local traffic. This research shows that the inclusion of human health damages related to the use phase of dwellings improves the extent to which the LCA of dwellings reflects the environmental benefits of sustainable building options. The effects of application of building materials with lower emission rates of harmful substances and the effects of lower traffic impacts are clearly visible in the improved dwelling LCA methodology. The same will hold for a change in ventilation characteristics, as these are directly related to the fate factors for organic compounds and radon (see Chapter 2). Indoor environmental issues are generally addressed by methods similar
[ 118 ]
[ 119 ]
to risk assessment. There are several methodological differences between risk assessment and LCA, such as the focus on actual, respectively potential damages, differences in geographical scale and differences in aggregation level (Jönsson, 2000). In this study, these two approaches have been brought together. The concentrations of harmful substances have been predicted based on average indoor airflows and constant averaged emission rates from building materials or local traffic. Furthermore, the fate factors have been calculated on a dwelling level, while the emissions rates have been estimated on a building material level. The indoor exposure model has been verified only to a limited extent. For radon, calculated indoor concentrations and actual concentrations found in a national survey in the Netherlands (Stoop et al., 1998) are found to be roughly the same. Other comparisons are hard to make, because most studies present data that are too incomplete to validate the methodology developed here. When datasets are available that includes detailed airflow characteristics, emission rates and concentrations, the model can be further verified. The damage scores have been calculated in this study for a new singlefamily row house. Generic characterisation factors applicable to all types of dwellings would make the integration of indoor environment in building LCA tools easy. Such generic factors make little sense, however, because the ventilation characteristics of the dwelling are different for each dwelling type. Nevertheless, characterisation factors and damage scores for building materials can be calculated separately for each dwelling type, for existing dwellings and for utility buildings. These damage scores can be used by users of the building LCA tools to calculate the environmental impact of their dwellings. When a type of dwelling that is not covered by these dwelling types is assessed, the fate factors in the indoor exposure model have to be recalculated. Furthermore, it has been assumed that the airflow characteristics and thus the fate factors do not differ significantly when other building materials are applied. However, when the building design changes considerably, for example by increasing the insulation thickness, the fate factors need to be recalculated. In the calculations of the damages to human health, linear dose-response relationships have been assumed for both carcinogenic and non-carcinogenic effects. Crettaz et al. (2002) and Pennington et al. (2002) proposed a dose-response method that assumes linearity below the effect dose affecting 10% of the individuals (ED10). An alternative is accounting for nonlinearity by calculating marginal changes in human response because of marginal changes in doseintake change in an ambient-background situation. Huijbregts et al. (2005) developed a method to calculate human health effect and damage factors for toxic pollutants, based on lognormal dose-response relationships. These effect and damage factors can be used as an alternative for the effect and damage factors as calculated by Crettaz et al. (2002) and Pennington et al. (2002).
There are several gaps in the methodology to calculate damages to human health caused by indoor exposure. Only pollutants with known effects on human health are considered. Pollutants for which no emission data or toxicological data are available have been left out of the present assessment. This causes an underestimation of the total damage to human health related to indoor exposure. Furthermore, only a limited number of ventilation characteristics are taken into account in the underlying indoor airflow model. Actual fate factors may thus differ from the fate factors calculated in Chapters 2 and 3. Only the dwelling is taken into account; the behaviour of the occupants, e.g. ventilation behaviour, smoking and emissions from consumer products, is excluded from the comparison. The environmental damages of emissions from building materials and local traffic to indoor air have been calculated as disability adjusted life years (DALYs). These damages can be added to damages to human health associated with the rest of the dwelling life cycle, if these damages are also calculated in terms of DALYs (e.g. Goedkoop & Spriensma, 1999). This eases the integration of the indoor air quality in dwelling LCAs. The methodology developed to characterize indoor emissions and local traffic can also be applied at a midpoint level. This requires recalculation of the effect factors and the introduction of reference substances. Also the aggregation level will be lower: several impact categories are required, such as ‘human toxicity’, ‘impacts of ionising radiation’ and ‘noise’ instead of one damage category ‘human health’.
6.2 A factor X feasible for dwellings
The improved LCA methodology has been applied to assess the effects of measures to reduce the environmental impact of dwellings. As one of the measures considered concerns the use of supposedly environmentally improved solar cells, an environmental comparison has been carried out between the production and use phase, except maintenance, of an indium-gallium-phosphide (InGaP) on multicrystalline silicon (mc-Si) tandem module, a thin-film InGaP module and a mc-Si module. The environmental impacts of InGaP/mc-Si tandem modules and of thin-film InGaP modules have the same order of magnitude as the environmental impact of mc-Si modules, despite the lower conversion efficiency of the latter. Therefore, mc-Si modules have been applied in the case study of the dwelling scenarios of Chapter 5. The environmental impacts of three dwelling scenarios have been compared using the improved LCA methodology. Besides material and energy consumption, indoor emissions from building materials, local traffic and radon exhalation from the soil have been included. The scenarios assessed are a reference scenario, a scenario with moderate measures and a scenario with more radical measures to improve the environmental performance of the
[ 120 ]
[ 121 ]
dwelling. When moderate measures are applied, the environmental impact, as estimated by the methodology developed in this thesis, is a factor of 1.7 lower than the environmental impact of the reference scenario. When more radical measures are applied, the environmental impact is a factor of 2.7 lower than the environmental impact of the reference scenario. To assess the influence of the new dwelling LCA methodology developed in this thesis, the three dwelling scenarios are compared with the standard and the improved LCA methodology. For the standard LCA methodology, the impacts of materials and energy production are taken into account only. In the improved LCA methodology as developed in this thesis, the impact sources indoor air emissions, radon exhalation from soil and local traffic are added. The results for human health damages are given in Figure 6.1. Figure 6.1 shows that the human health damages of the three dwelling scenarios differ within a factor of 1.3 when calculated with the standard LCA methodology. When the human health damages are calculated with the improved LCA methodology, the differences are larger (a factor of 3.1). This shows the relevance of the addition of damages to human health caused by pollutants and noise emitted to the indoor environment to standard dwelling LCAs. The methodology developed here also allows for evaluation of the impact of improvements in specific building components on the environmental performance of buildings. For instance: the environmental impact of the material input of photovoltaic modules for the application of 2 KWp of mc-Si modules in the reference dwelling is about 9% of the total environmental impact of the material input of the dwelling scenario with more radical measures. By using high-efficiency solar modules, a higher electricity production can be established with the same surface area of solar modules applied to the dwelling, potentially resulting in a further decrease of environmental impact due to energy consumption in the use phase of the dwelling. Application of concentrators may also be considered to increase the efficiency of solar modules with a relatively small investment of cheap materials with low environmental impact, compared to photovoltaic materials (Yamaguchi et al., 2005). It has been shown that for the Dutch reference dwelling, a factor of 2.7 is feasible as factor X when measures are taken such as replacement of concrete by wood, application of photovoltaic panels, solar collectors and heat pumps and reduction of the impact of local traffic. This factor is based on environmental impacts calculated with LCA, instead of total mass or cumulative energy consumption, thus giving a more appropriate view of the full environmental impacts. This factor X might be larger when additional measures are taken to improve the environmental performance of the dwelling, for example by increasing the insulation thickness and applying low emittance double glazing. These measures, however, may cause lower ventilation rates and thus may have a negative effect on the indoor air quality. Other measures are re-
:WcW][je^kcWd^[Wbj^ :7BO
