SEARCH

SEARCH BY CITATION

Keywords:

  • buildings;
  • construction;
  • demolition;
  • industrial ecology;
  • recycling;
  • renovation

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

In this article we have elaborated a consistent framework for the quantification and evaluation of eco-efficiency for scenarios for waste treatment of construction and demolition (C&D) waste. Such waste systems will play an increasingly important role in the future, as there has been for many years, and still is, a significant net increase in stock in the built environment. Consequently, there is a need to discuss future waste management strategies, both in terms of growing waste volumes, stricter regulations, and sectorial recycling ambitions, as well as a trend for higher competition and a need for professional and optimized operations within the C&D waste industry. It is within this framework that we develop and analyze models that we believe will be meaningful to the actors in the C&D industry. Here we have outlined a way to quantify future C&D waste generation and have developed realistic scenarios for waste handling based on today's actual practices. We then demonstrate how each scenario is examined with respect to specific and aggregated cost and environmental impact from different end-of-life treatment alternatives for major C&D waste fractions. From these results, we have been able to suggest which fractions to prioritize, in order to minimize cost and total environmental impact, as the most eco-efficient way to achieve an objective of overall system performance.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

The architectural, engineering, and construction (AEC) industry is a major contributor to the overall waste generation in Norway. Much of this waste (95%) is technically recyclable (GRIP/Økobygg 2001), but is today not recycled for various reasons. In this article we investigate the environmental and economic performance of different waste handling options in a future construction and demolition (C&D) waste recycling system of Trondheim, Norway. The baseline for the system examined is 2003.

There have been several attempts to describe, quantitatively (Bossink and Brouwers 1996; Dantata et al. 2005; Davidson and Wilson 1982; Müller 2006; Touran et al. 2004; Wilson 1975; Yost and Halstead 1996) or qualitatively (Chung and Lo 2003; Eriksson et al. 2005; Reijnders 2000), waste handling and recycling systems on a regional level, but to our knowledge, few have attempted to combine the quantitative aspect of both waste projection and the corresponding environmental impact for C&D waste recycling systems (Bohne 2005; Symonds et al. 1999).

Eco-efficiency analysis (BCSD 1993; Keffer et al. 1999; Schmidheiny 2000; Sturm and Schaltegger 1989; Verfaille and Bidwell 2000) is a tool, primarily developed for production processes and firms, where value-added and environmental impact are reported mostly at the corporate scale. When we are dealing with recycling systems, the picture gets more complicated (Brattebø 2005; Huppes and Ishikawa 2005a, 2005b), both for the estimation of value-added and for environmental influence, because these systems involve numerous companies, products, and material fractions, as well as open loop recycling options where the variables are not easily determined, and allocation problems will often arise. This article examines the calculation of eco-efficiency for recycling systems, and how eco-efficiency analysis can be utilized as a tool in decision making processes for investigations of recycling systems dealing with products with a long service life and thus slow turnover. Eco-efficiency, as used in this article, differs from traditional cost-benefit analyses, in the sense that in eco-efficiency analysis, one does not use a welfare function in which environmental aspects can be traded with and expressed in the same monetary terms as non-environmental ones (Huppes and Ishikawa 2005a, 2005b).

To make well considered predictions about the future, it is necessary to know something about the past (Bergsdal et al. 2007; Bohne 2005; Horvath 2004; Torring 2001). Bergsdal and colleagues (2007) describe a method for the projection of a future generation of C&D waste in Trondheim, Norway, from 1995 to 2018. We have made use of these calculations (see figure 1) as input to our calculations in this article. In order to have a reference to the scale of figure 1, the population in Trondheim equals 150,000 persons in 2002, and the total building area is estimated to 46 million square meters (m2).1

image

Figure 1. Projections of construction and demolition (C&D) waste in Trondheim during 1995–2018, projected on the basis of history and stock dynamics of existing buildings (from Bergsdal et al. 2007). EPS = Environmental Priority System.

Download figure to PowerPoint

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

System Borders and Allocation Between Co-Products

In systems modeling and modeling of environmental impacts (life cycle assessment or LCA) in particular, co-production (the joint production of two or more products from the same process or system) has been seen as presenting a problem to the modeling, and the traditional solution has been co-product allocation (the partitioning and distribution of the environmental exchanges of the co-producing processes over its multiple products according to a chosen allocation key) in parallel to cost allocation (Udo de Haes 2002; Weidema 2000; Weidema and Norris 2002).

Weidema and others (Udo de Haes 2002; Weidema 2000; Weidema and Norris 2002) have demonstrated how co-product allocation can be avoided by expanding the system to also include product system B: “the co-producing process (and its exchanges) shall be ascribed fully (100%) to the determining co-product for this process (product A)” (Weidema 2000) (figure 2).

image

Figure 2. System expansion (from A to A+B) for the allocation of influence among co-products, as input to the calculation of eco-efficiency in construction and demolition (C&D) waste recycling systems. Boxes and arrows with dark shading denote system boundaries for this study. Arrows denote transport between processes. ri represents the ratio of a given waste fraction that enters an alternative end-of-life treatment, and γi is the factor for how much virgin material that is replaced by ri. The underlying layers represent further system expansions.

Download figure to PowerPoint

In the recycling system for C&D waste, it is the enterprise owner (of the construction or demolition project, for example, of the building under construction) or the entrepreneur in system A who in general determines where and what to do with the waste. However, local and central governments often seek to influence these decisions through regulation and economic instruments. This reopens the discussion about allocation among co-products. Here we have chosen to follow Weidema's (2000) recommendation of system expansion, as our interest is in the overall system, and our target is to maximize the overall system performance.

Waste Handling and Environmental Impact

The environmental impact of waste management options is the result of processing and disposal methods and transportation types and distances, for all disposal, recycling, and reuse options in the system. Moreover, recycling and reuse should generate positive downstream environmental benefits when recycling and reuse is applied, due to avoided emissions, and some of these benefits should also be allocated to the environmental performance of the initial C&D waste system. Different end-of-life treatments can be ranked in a general hierarchy according to their environmental impact (Bohne and Brattebø 2003; Reijnders 2000), in which direct reuse is ranked higher than recycling, which in turn is better than energy recovery.

In practice, some of these alternatives are either not preferred, due to lack of market demand, or because they are not possible or too expensive to follow (Reijnders 2000; Symonds et al. 1999; Wilson 1975). Some of these processes demand facilities that are expensive to build and maintain. It is therefore of public interest to know as much as possible about the future waste generation and its possible corresponding environmental impact (Bohne and Brattebø 2003; Chung and Lo 2003; Horvath 2004).

Environmental impacts for the different alternatives for end-of-life treatment are calculated on the basis of data from many different sources using LCA methodology (Kotaji et al. 2003; PRé Consultants 2002). A problem with these kinds of calculations for recycling systems is that we deal with a wide range of products, of differing ages, and from many different producers. Available data do cover products within systems with long service life, but only to a limited extent. Thus, there is a clear need for the development of knowledge and methods, which is our motivation for this research. Hence, it is a challenge to make use of appropriate system borders, cut-off rules, and allocation rules when doing the analysis. It is the aggregated environmental system impact over time that is of importance and should be used as a design criteria, rather than aggregated volume or weight parameters, which often are used in industry's evaluations of system performance in the AEC sector.

Figure 1 and 3a shows that the brick and concrete by far will be the largest fraction and also the fastest growing fraction in the forthcoming 15 years (Bergsdal et al. 2007). This does not necessarily mean, however, that this is the most important fraction to deal with to reduce environmental impact (figure 3b). In order to examine this question, we need to determine the aggregated total environmental impact (Ψ*j) for each waste fraction (j) during the whole period we are studying and then for the different waste fractions, including all transportation and end-of-life treatment activities, over the time period in question.

image

Figure 3. Aggregated: (A) Waste generation, (B) environmental impacts, and (C) costs of construction and demolition (C&D) waste handling for selected waste fractions in the city of Trondheim from 2003 to 2018.

Download figure to PowerPoint

The basis for the calculation of environmental impact in this article is life cycle inventory data from the EcoInvent database, using the SimaPro software package and the EcoIndicator 99 method for impact assessment (Huijbregts et al. 2003). Thus, the life cycle inventory for each of the process or subprocess steps is found in the EcoInvent database (with the exception of reuse of wood), and environmental impact per tonne,2ψ, for the different processes and/or subprocesses is then calculated using the SimaPro software package. The environmental impact for each process is shown in table 1. The values for environmental impact are given in eco-points (Pt.). We have chosen EcoIndicator 99 because this is a single value indicator, which makes it easy to communicate the results to decision makers. Other weighting methods may be used, but the principle remains the same.

Table 1.  Environmental impact and cost data used in the calculations, 2003. All costs with negative values are expenses while positive cost values are used when the process generates income.
Waste fractionProcessUnitDatabasevalues (Pt.)Database-key EI 99Costs (€)
  1. Units: tkm = tonne-kilometers ≈ 0.7 ton-mile; MJ = megajoule = 106 joules (J, SI) ≈ 239 kilocalories (kcal) ≈  948 British Thermal Units (BTU); kg = kilogram ≈ 2.204 pounds (lb). 1 Euro (€) = 7.785NOK; N.A. = not applicable.

  2. aLocal transport.

  3. bLong distance transport.

  4. cAlthough Norwegian building stock uses a mixture of 30.5% fuel oil and 69.5% electricity for heating purposes, we have used oil in our calculation, because we are interested in the residential facilities capable of utilizing waterborn district heating for substitution. Thus the results are only valid where district heating is available.

  5. dDifficult in practice.

  6. eConverted from cubic meters (m3) to kilograms (kg) using a factor of 500 kg/m3.

  7. fWood: 2.74 MJ/kg wood incinerated.

  8. gCardboard: 3.23 MJ/kg cardboard incinerated.

  9. iPlastic: 7.03 MJ/kg plastic incinerated.

  10. hPCB containing windows are entering a special waste stream at a gate fee of €186.- (NOK 1450.-).

TransportLorry 28tatkmEcoInventEIN SYSX065738017713.24E−02−0.385
Lorry 40tbtkmEcoInventEINSYSX065738017732.26E−02−0.257
HeatFuel oilcMJEcoInventEIN SYSX065738014310.00326−0.027
BrickLandfillkgEcoInventEIN SYSX065738018320.00389−0.143
RecyclingkgEcoInventEIN SYSX065738019660.00234−0.010
ReusekgEcoInventEIN SYSX065738019670.00419 0.257
Virgin gravelkgEcoInventEIN SYSX065738004600.0014−0.016
Virgin brickkgEcoInventEIN SYSX065738004870.0143−0.128
ConcreteLandfillkgEcoInventEIN SYSX065738018370.00402−0.143
RecyclingkgEcoInventEIN SYSX065738019700.00247−0.010
GravelkgEcoInventEIN SYSX065738004600.0014−0.016
WoodLandfillkgEcoInventEIN SYSX065738018790.222−0.143
IncinerationkgEcoInventEIN SYSX065738019550.0219−0.061
ReusedkgOEN.A.0.0002 0.128
Virgin woodkgeEcoInventEIN SYSX065738023120.115−0.283
Heat2.74 MJfOCN.A.0.0053855−0.059
GypsumLandfillkgEcoInventEIN SYSX065738018590.00389−0.143
RecyclingkgEcoInventEIN SYSX065738019750.00234−0.103
Virgin gypsumkgEcoInventEIN SYSX065792030370.037−0.385
CardboardLandfillkgEcoInventEIN SYSX065738020360.0241−0.143
IncinerationkgEcoInventEIN SYSX065738019300.0257−0.039
RecyclingkgEcoInventEco SysX10984500028−0.0271 0.019
Virgin cardboardkgEcoInventEIN SYSX065738015360.105−0.385
Heat3.23 MJgOCN.A.0.0063487−0.069
GlassLandfillkgEcoInventEIN SYS065738018960.00107−0.143
RecyclinghkgEcoInventEco SysX10984500030−0.082−0.039
Virgin (brown glass)kgEcoInventEIN SYSX065738007810.0541−0.128
PlasticLandfillkgEcoInventEIN SYSX065738020410.0348−0.143
IncinerationkgEcoInventEIN SYSX065738019370.0386−0.039
RecyclingkgEcoInventEco sysX10984500033−0.128−0.039
Virgin plastic filmkgEcoInventEIN SYSX065738016920.138−1.285
Heat7.03 MJiOCN.A.0.013818−0.151
Metals (steel)LandfillkgEcoInventEIN SYSX065738019070.00107−0.143
RecyclingkgEcoInventEco SysX10984500038−0.258 0.050
Virgin steelkgEcoInventEIN SYSX065738010720.464−1.541

The total environmental impact (Ψj) of a given waste fraction (j) is the product of the environmental impact per tonne (transport included) (ψj) and the corresponding weight of the waste fraction in tonnes (wj):

  • image(1)

where j is the waste fraction in question, wj is the weight of the waste fraction in tonnes, and ψj is the environmental impact per tonne from the given mixture of end-of-life treatment options that are applied to the given waste fraction. (Notations are summarized in a glossary at the end of the article.)

For most waste fractions, there are several end-of-life alternatives to consider. The total environmental impact of a given waste fraction is the sum of all environmental impacts from all end-of-life treatment alternatives for this waste fraction, see equation (2):

  • image(2)

where j is the waste fraction in question and ri,j is the share of this waste fraction that is sent to an end-of-life treatment alternative, i (see figure 2). The aggregated total environmental impact of a given waste fraction owner of given end-of-life treatment options, Ψ*j, is then calculated by summarizing the total environmental impact over the years of interest (t), equation (3):

  • image(3)

where wj,t is the annual waste generation of a given waste fraction, t is the year of interest, and * denotes that it is an aggregated number.

Calculating Ψ*j for all waste fractions will thus identify which fractions should be of greatest concern to minimize environmental impact from the C&D waste system (figure 3b).

Economic Data

Decisions cannot be made from environmental data alone. A system owner would prefer to optimize his system for the best performance possible in order to maximize the return from his investments. However, systems that include reuse or recycling, which are in fact reproduction systems, are complex systems composed of numerous products and stakeholders. Due to this complexity, economic efficiency is often hard to measure.

For C&D waste, a typical recycling chain involves several stakeholders with diverging interests, each of whom seeks to maximize their own profit and thus is less driven by a wish to reduce environmental impact. The system owner is (often) a municipality, who also is (in part) responsible for the policies affecting the system. Given these constraints, we can categorize the economic data in three categories: (1) the source, (2) the type, and (3) the availability of the data (table 2).

Table 2.  Availability and source of economic data
Category123
SourcePublic taxesPrice listsPersonal communication and/or best estimates
AvailabilityVery goodVery good/goodGood to unavailable
Type of dataStaticDynamicRapidly changing
Data qualityVery poorVery poor/goodGood to fair

It is most often the values for processing expenses that are unavailable due to market competition (category 3), but actual prices on transportation and waste delivery for large deliveries have also been found to be lower than the official prices (and therefore shift from category 1 to category 2). Hence, for the same reason, accurate data can be hard to get. It is also a problem that some economic values have a considerable dynamic variation, such as the price on transportation, which shifts with fuel prices. We have therefore used historically observed data and corrected those data according to the corresponding statistical index (Statistics Norway 2004). By using actual costs as seen by the stakeholders (transfer payments and taxes) as the economic indicator (table 1), we have managed to limit our data sources to category 1 and 2.

The economic data is calculated the same way, by summarizing the actual transfer costs and taxes for each process step or subprocess using the same system borders and allocation rules as the environmental data (figure 3c).

Transport Distances

Most of the C&D waste (by weight) is handled by different companies within Trondheim, but some of the fractions have to be shipped long distances outside the city if they are to be recycled. Table 3 takes into account the nearest recycling facilities for these fractions and the transport methods and distances for each fraction. We have used these transport distances in our calculations even though waste fractions may from time to time be sent to other more far distant places.

Table 3.  Distances to the nearest recycling facilities for waste fractions that are not being recycled in Trondheim
Recycling ofWhereDistanceRecipientTransport by
GypsumDrammen539 kmGyprochTruck
CardboardRanheim15 kmPeterson ASTruck
GlassStjørdal33 kmGlava ASTruck
PlasticsFolldal197 kmFolldal GjenvinningTruck
MetalsMo i Rana482 kmFundia Armeringsstål ASTruck

Sensitivity Analysis

In the calculations, we use weight estimates combined with either environmental impact data or economic values. For these calculations to be useful in the actual decision making processes, knowledge of the uncertainty and sensitivity of the data is important. There are three sources of uncertainty in the calculations: (1) weight, (2) life cycle inventory (LCI) data, and (3) cost data, which thus need investigation.

The calculations of waste generation have previously been published by Bergsdal and colleagues (2007). Here Monte Carlo simulations were run to find a probable standard deviation for our expression, which returned an uncertainty of ±5%–10%.

For the environmental impact data, we base our data on life cycle inventory data from EcoInvent (Ecoinvent 2007; PRé Consultants 2002), and it should be stated that the level of inaccuracy is not widely published with LCI data.

In the scientific literature (Kotaij et al. 2003; PRé Consultants 2002), it is acknowledged that CO2 emissions are well studied and therefore have lower uncertainties, while emissions, such as dust and noise, are less studied and therefore possess greater uncertainties. We have therefore chosen a single standard error range of ±5% for the LCI data used in our calculations, which is an accepted approach to uncertainty of LCI data (Huijbregts et al. 2003).

The economic data (i.e., costs) are well known in our system and only need to be corrected for dynamic variations over time. In our case, we found that the transportation and fuel prices were the most relevant factor with respect to variations. We have shown graphically how a ±50% variation in transport distances will affect the dynamic eco-efficiency calculations (figure 3).

Eco-Efficiency in Recycling Systems

Eco-efficiency (BCSD 1993; Keffer et al. 1999; Schmidheiny 2000; Sturm and Schaltegger 1989; Verfaille and Bidwell 2000) was first mentioned by Sturm and Schaltegger in 1989, “The aim of environmentally sound management is increased eco-efficiency by reducing the environmental impact while increasing the value of an enterprise.” Later the Business Council (now the World Business Council) for Sustainable Development described how to achieve eco-efficiency in a report released just prior to the 1992 Earth Summit in Rio de Janeiro.

The term can be expressed mathematically as (Keffer et al. 1999):

  • image(4)

The World Business Council for Sustainable Development (WBCSD) (BCSD 1993; Keffer et al. 1999; Schmidheiny 2000; Verfaille and Bidwell 2000) and the United Nations Conference on Trade and Development (UNCTAD) (Sturm and Upasena 2003) advocates for using internationally standardized economic indicators when calculating eco-efficiency. Value-added is proposed as the indicator of choice for product or service value. Because eco-efficiency was designed primarily for measuring efficiency improvements in production systems within a company, both value-added and environmental influence should be known, at least for internal purposes. When we are looking at recycling systems, however, we cannot use the term value-added in the same way as at the firm level. With a system of many stakeholders who seek to make profit along the way, this picture gets more complicated.

Even so, some of this profit does not necessarily increase the value of the material in question, but arises from the stakeholders' performance of services such as collection, transportation, sorting, and processing. Processing activities in recycling systems, in fact, despite an increase in value for the material, normally lead to a considerable downcycling3 of the material at the same time as the stakeholder makes profit. However, the alternative of no processing would of course be worse, because this leads to even less value in the market. We have therefore rewritten the formulae (equation 5) to include all economic transactions (for the extended system):

  • image(5)

We use the term costs to denote all economic transactions when the material is transferred from one process to another. Equation 5 can be expressed mathematically as:

  • image(6)

where ɛj is the eco-efficiency of a given waste fraction within the system, κi,j is the process costs of a given waste fraction and end-of-life process alternative on a per tonne basis, ψi,j is the environmental impact of a given waste fraction and end-of-life process alternative on a per tonne basis, i is the different end-of-life processing alternatives, and j is the different waste fractions. Figure 2 shows that we include all processes in the overall system when calculating κj and ψj in equation 6. Hence we will avoid the difficulties of allocation.

However, for eco-efficiency to have any meaning as a tool for decision making, we need to measure the change in eco-efficiency between different end-of-life treatment options, or waste handling scenarios, for each of the different waste fractions (Bohne and Brattebø 2003; Huisman 2003; Saling et al. 2002). Thus, what we want to measure is the relative change in eco-efficiency (ɛ′) of a proposed alternative end-of-life treatment option or set of options (b) compared to the current practice (a):

  • image(7)

where a is a given mixture of end-of-life process alternatives that are made use of in the current (reference) system and b is a given mixture of end-of-life process alternatives that are made use of in the proposed alternative system.

However, eco-efficiency is a one-dimensional number (Euro/Pt.) that conceals valuable information from decision makers, especially when more than one alternative process is to be considered. Another issue is that the value of eco-efficiency will increase if the cost increases. Hence we will have to rearrange this parameter in order to better communicate information the way we prefer.

We will, therefore, follow Huisman (2003) in his attempt to visualize the change in eco-efficiency similar to the BASF method (Saling et al. 2002). As with the BASF method, we visualized eco-efficiency by plotting the numerator (κj′) and denominator (φj′) for a given waste fraction (j) in an xy-plot as shown in figure 4, where a positive value for the numerator expresses increased economic value (the y-axis), and a negative value for the denominator expresses less environmental impact (the x-axis). But unlike the BASF plot, we do not rescale the values to a normalized value, thus the values in our plot represent the net gain or losses in cost or environmental impacts.

image

Figure 4. Relative change in eco-efficiency for the different end-of-life treatments for selected fractions of construction and demolition (C&D) waste in Trondheim, Norway. Be aware of the large variations of scale between the different waste fractions in the figure. The bars indicate the sensitivity to transportation work in the analysis (±50%), both on the environmental impact and the economic benefit.

Download figure to PowerPoint

By comparing several end-of-life process alternatives this way, that is, by plotting the results in the same graph (keeping the reference process, a, constant), decision makers can make better decisions as to what solution to follow. Another interesting feature of this two-dimensional figure is that policy makers here can test how different policies will affect the eco-efficiency of the different end-of-life treatments within the system.4

Scenarios

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

The next step is to set up alternative scenarios for the distribution of C&D waste fractions between various end-of-life treatment options. Scenario 0: TP5 (TP = today's practice) assumes the continuation of the current end-of-life practice during the whole period. Scenario 1: NAP is the recommendations for waste handling as proposed in the Norwegian National Action Plan (NAP) 2005 (GRIP/Økobygg 2001). Scenario 2: Maximum Recycling is the result if as much as possible of the C&D waste is directed towards recycling, and Scenario 3: Maximum Energy Recovery is the result if as much as possible of the C&D waste is directed towards energy recovery. Table 4 shows the distribution factors among the different end-of-life alternatives used in these scenarios.

Table 4.  Distribution of construction and demolition (C&D) waste fractions (on a fraction mass basis) between different end-of-life treatment options in Trondheim
C&D wasteScenario 0Scenario 1: NAPScenario 2: Maximum RecyclingaScenario 3: Maximum Energyb
LandfillRecyclingEnergyReuseLandfillRecyclingEnergyReuseLandfillRecyclingEnergyReuseLandfillRecyclingEnergyReuse
  1. aIn Scenario 2: Maximum Recycling of all fractions (except wood) = 90%. Wood is incinerated for energy recovery.

  2. bIn Scenario 3: Maximum Energy, all fractions with an energy potential are sent to energy recovery. All other fractions are following the suggestions of the National-Action-Plan 2005.

  3. cOnly gypsum from construction waste is recycled. Gypsum from renovation or demolition waste is not possible to recycle with current technology.

Concrete & Brick0.700.300.000.000.200.800.000.000.100.900.000.000.200.800.000.00
Wood0.600.000.390.010.200.000.700.100.100.900.000.000.100.000.900.10
Gypsumc0.950.050.000.001.000.000.000.000.750.250.000.001.000.000.000.00
Cardboard0.500.300.200.000.200.700.100.000.100.900.200.000.100.000.900.00
Glass0.800.200.000.000.200.800.000.000.100.900.000.000.200.800.000.00
Plastics0.400.200.400.000.100.800.100.000.100.900.400.000.100.000.900.00
Metals0.100.900.000.000.100.900.000.000.100.900.000.000.100.900.000.00

The distribution of waste fractions demonstrates that an ambitious shift away from landfill towards recycling for concrete/brick, gypsum, cardboard, and plastics must be realized, if the current end-of-life practice (Scenario 0) is to be replaced by Scenario NAP. Likewise, wood waste will have to be redirected from landfilling towards energy recovery and direct reuse.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

In order to examine the eco-efficiency of C&D waste systems at the local level in a city, we have estimated future waste projections for the city of Trondheim. Trondheim is the third largest city in Norway, with a population of 150,000 inhabitants and a building structure that is characterized by many detached family houses made of wood covering a large area, together with larger residential and office buildings made of concrete in the center of the city and in some clusters around the center of the city.

The projections of local C&D waste fractions for Trondheim, as given in figure 1, are accumulated for the whole period 2003–2018, and the aggregated amounts are shown in figure 3A. One can clearly see the dominant role of the concrete and brick fraction, in addition to wood wastes, even for a city with a majority of the buildings made of wood.

On the basis of the data in figure 3A and table 3, it is now possible to calculate the Net Present Value (Euro) and environmental impacts (Pt.) for each waste fraction during the 2003–2018 period. Cost data are obtained directly from the actors in the system, and environmental impact data are obtained from LCA software (PRé Consultants 2002) by using the Eco-Indicator 99 valuation method (see table 1).

For the calculations of net present value, we have used the interest rate from the Norwegian State Obligations (4% for obligations with 10 years running time) to estimate the present value of future costs. This is in order to avoid problems due to depreciated values of costs over time in the calculations.

The results given in figure 3B and  3C show the estimated environmental impact (Pt.) and net present value (Euro) for each waste fraction6 in Trondheim for the next 157 years, when Scenario 0: TP, Scenario 1: NAP, Scenario 2: Maximum Recycling, and Scenario 3: Maximum Energy Recovery are applied for the whole period.8

It can be seen that the concrete and brick fraction by far dominates the material composition of the C&D waste (figure 3A). This trend is to some extent also reflected in the system costs (figure 3C), while its corresponding environmental impact is less obvious (figure 3B). If we look at figure 3B alone, we would suggest that the wood is the fraction worth focusing on from an environmental point of view. As can be seen from figure 3B, the picture is somewhat different from the picture in figure 3A. In figure 3B, wood and to some extent concrete and brick are the two most important fractions from an environmental point of view. This also corresponds well with how developed, or mature, the recycling systems for these fractions are today.

To make sound decisions on what to do with the different waste fractions, we need to compare the eco-efficiency for the different end-of-life options for each of the fractions against each other. This will then be a basis for further sound decisions in the overall C&D waste management system.

Figure 4 shows two-dimensional relative plots of the eco-efficiency for the different end-of-life alternatives studied. We have provided plots for each of the waste fractions, and the data show how different end-of-life treatment options position themselves relative to the current treatment practice (as in Scenario 0), which is always represented by the origin. The difference in environmental impact, between a given treatment option or scenario and the current treatment scheme, is given along the x-axis, where reduced impact gives a position to the right of origin. The difference in cost is expressed on the y-axis, as the net economic benefit. Thus a reduced cost—these are relative numbers—is the sum of end-of-life options that are part of both the current situation and the suggested alternative. In order to compare on a straightforward basis, the results are presented on a per tonne basis (i.e., Euro/tonne and Pt./tonne, by using the Eco-indicator 99 method). The better treatment options will always position themselves in the upper right corner of the plot.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

The results shown in figures 3 and 4 reveal some interesting issues. Let us first take a look at the results from figure 3.

Although the concrete and brick fraction is by far dominating the waste generation and the cost of waste handling, it does not have a correspondingly high impact on the environment. A second interesting finding is that the figures show that wood has the largest environmental impact of all fractions, and it is also the fraction that has the largest possibilities for further environmental savings (actually, the possible reduction in environmental impact from wood is larger than for the other fractions together).

One of the reasons for this is that the concrete fraction is unique in the sense that there is no real recycling, but a downcycling to crushed aggregates, which then is used as a substitute for gravel. This in turn has implications for our calculations, as concrete production is left outside our system borders, while virgin gravel (which has much lower environmental impact) is included. The concrete system as modeled in this study, therefore, systematically underestimates the environmental impact and costs, because only waste handling (and not production) is included. Reuse of concrete is not a realistic alternative within the next few decades and is therefore omitted, as a consequence of the construction techniques used 30–60 years ago.

Wood, on the other hand, is incinerated with heat recovery, and the heat is substituting for oil through district heating. Wood, in contrast to concrete, is also an organic fraction, which causes considerable environmental impact if landfilled. And lastly, there is the density; in volume—one tonne of wood is almost five times more voluminous than concrete. Thus from an environmental point of view, one needs to focus on better collection and handling practices of wood debris, so that a larger portion is delivered to energy recovery. From an economic approach, concrete, wood, brick, and gypsum, in that order, are where the potential for increased return is possible.

It is worth mentioning here that, in theory, it is possible to increase environmental savings even more for the wood fractions if the wood is reused instead of being sent to recycling or energy recovering. Such a shift can also be more profitable for stakeholders. However, this is difficult due to the need for altered construction practices and the required handling of materials for reuse. If such a system should be implemented at a large scale, and not as is done today, be a work training facility, the costs would also increase beyond what we have used in our calculations, and as a result, the corresponding eco-efficiency would decrease.

Figure 4 shows how the different end-of-life alternatives and scenarios will perform on a per tonne basis in a two-dimensional plot. For all but the metal fraction, there is still an opening for both environmental gains and profits (on an overall systems level).

For the brick fraction, the figure reflects the current situation of 30% recycling, in which most of the brick is delivered for recycling as a gravel substitute (together with concrete) and not for reuse. There is still an unexploited niche in the reuse of bricks, but it is questionable to what extent it is possible to industrialize this in a wood-based city such as Trondheim. Today this part of the system is handled by work at a training facility and by scrap dealers in Norway. A ±50% change in transport has negligible effects on system performance.

For the concrete fraction, there is a clear economic and environmental benefit from recycling, and the figure reflects the current situation of 30% recycling. Because much of the recycling potential is still not realized, other issues are presumed to be of higher economic importance to the decision maker. Due to the environmental gains from recycling, a ±50% change in transport has some impacts on environmental performance.

Wood gives some of the more interesting results, due to the fact that this is one of the fractions composed of renewable material and that it can involve all end-of-life solutions. As previously mentioned, there is a great potential for better environmental performance, if more wood is reused, recycled, or incinerated with energy recovery. As with concrete and bricks, other factors are presumed to dominate the decision processes for stakeholders, because more wood is not actually being delivered, for instance, for energy recovery. A ±50% change in transport has negligible effects on system performance.

Figure 4 shows that the recycling of gypsum is the most environmentally friendly and economic solution for stakeholders located in Trondheim. Recycling however is only an option for fresh wastes such as stubs and cutoffs from construction and renovation activities with current technologies. Thus, most waste gypsum, which arises from demolition, will still end up in the landfill in Trondheim.

The gypsum fraction is also relatively sensitive to a ±50% change in transport. This is due to the fact that there are only two recycling facilities for gypsum in Norway, located in Drammen and Fredrikstad, and that the long distant transport is both costly and contributes significantly to the overall environmental performance.

Cardboard faces a classic situation in which energy recovery competes with recycling. Here contamination and convenience determines what end-of-life options to follow. Figure 4 reflects a functioning recycling system not yet optimized. Recycling is the favorable environmental and economic choice, however, other factors presumably dominate the decision processes for stakeholders, because more cardboard is not actually being delivered for recycling. A ±50% change in transport has no effect on system performance.

Glass is an example of a recycling system in its early stage, with a great deal of unrealized potential (figure 4). A ±50% change in transport has negligible effects on system performance. As with cardboard, plastics are in a situation in which energy recovery competes with recycling, but with one important difference. Here we are dealing with a nonrenewable resource, and although the economic potential is on the same order, there is a significant difference in environmental potential. The environmental impact for the system indicates clearly that recycling should be favored, the opposite conclusion of the economic indicator. Policymakers should, therefore, restructure their use of policy incentives, such that recycling becomes more favored than energy recovery with regard to overall system costs. A ±50% change in transport has no effect on system performance.

Metals from C&D debris have an image of a mature recycling system driven by the economic value of a material. Almost all the environmental potential is therefore realized. Metals are also the only waste fraction in which one gets paid when delivering waste. A ±50% change in transport has no effect on system performance.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

There is a need for municipalities and governments to make more considered decisions on environmental issues, with regard to the long-term management of waste handling systems and natural resources. We have shown that long-term waste generation models combined with environmental and economic information can be a powerful tool in such regard.

Total environmental impact and eco-efficiency calculations can be used by system owners and stakeholders to evaluate their options and their system performance, as well as to identify which waste fractions to focus on and which end-of-life alternative to give priority.

Of special interest to waste handling systems is the possibility of generalizing potentially aggregated environmental and economic effects of different policies regarding end-of-life treatment alternatives. Important to decision makers will be how different system alternatives meet given policy targets. Our model is able to simulate such issues.

However, even though we have demonstrated that recycling and/or reuse often are the most eco-efficient choices in C&D waste systems, we know that they are many times not followed in practice for a variety of reasons. We assume that this is often due to the fact that other economic processes outweigh the benefits of sound waste handling decisions. Time penalties for delays in construction or demolition projects are an obvious example, and these issues need more investigation.

Our reflection on this research method is that the approach looks very promising. Our way of presenting specific and aggregated results on eco-efficiency in C&D waste systems seems intuitive and attractive with respect to communication with stakeholders as a basis for decision making. However, there are two aspects that need improvement. First, one needs to refine the dynamic model estimating future waste generation. This model should be based on more detailed examination and data of the building stock, including its material composition, lifetime distribution, and dynamic aging phenomena. This would give more robust projections for waste generation.

Second, we need more precise data on important processes in the C&D waste system, including environmental and economic indicators. Only then will it be possible to offer models that are really meaningful to the industry in decision making. As for now, one can, to some extent, change the outcome of the analysis by using a different LCI database.9 However, there is reason to believe that improved data (on a standardized form) on a national level will solve some of these issues.

The C&D sector is so important in terms of its waste amount, that such research work should be given high priority.

Notations

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors
Euro=

Euro (currency).

Pt.:=

Eco-point, environmental performance indicator based on the EcoIndicator 99 method.

NAP(2005) =

The National Action Plan for recycling of C&D waste. Overall target of 70% recycling within 2005.

ɛ″=

eco-efficiency (Euro/Pt.).

ɛ′=

Relative eco-efficiency (Euro/Pt.).

Ψ*=

Aggregated-total environmental impact (Pt.).

Ψ=

Total environmental impact (Pt.).

ψ=

Environmental impact (Pt./tonne).

ψ′=

Relative environmental impact (Pt./tonne).

κ=

Costs (Euro/tonne).

κ′=

Relative Costs (Euro/tonne).

w=

Waste generation (tonnes).

r=

Recycling ratio.

t=

Time (years).

j=

Waste fraction.

i=

End-of-life treatment alternative.

a=

Processes of product system A.

b=

Processes of product system B.

Notes
  • 1

    One square meter (m2, SI) ≈ 10.76 square feet (ft2).

  • 2

    All tons are metric and thus spelled tonne; one tonne (t) = 1000 kilograms (kg, SI) ≈ 1.103 short tons.

  • 3

    Downcycling is the recycling of a material into a material of a lesser quality (Wikepedia contributors 2008).

  • 4

    Be aware that the origin of the plot also changes (relatively) by introducing new policies.

  • 5

    Scenario 0 is the reference point for our examination, because it is equal to the current practice. Hence this scenario will be represented by κa and ψa in equation 7, and the origin location in the eco-efficiency plots (figure 4). Likewise, Scenario 1: NAP, Scenario 2: Maximum Recycling, and Scenario 3: Maximum Energy Recovery will be represented by parallel sets of κb and ψb values in equation 7 and located away from origin in the plots.

  • 6

    In figure 3, the concrete and brick fractions are combined to one fraction because of the lack of data on waste generation. It is, however, believed that the brick and concrete fraction is dominated by concrete, thus numbers on environmental impact and net present values for concrete are used.

  • 7

    This work was done in 2005.

  • 8

    We want to remind the reader that these are calculations for the extended system and are not representative for the individual stakeholders, but for the system as a whole, see figure 2.

  • 9

    The EcoInvent database is developed for continental Europe and, as such, does not necessarily represent activities performed in Norway.

Acknowledgments

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

The authors want to thank Øyvind Spjøtvold and Aage Heie, Norsas, for valuable discussions and help in data collection during this work. We also want to thank Anders Strømman, Ingve Simonsen, Bård Skaflestad, and Glen Peters, all from Norwegian University of Science and Technology (NTNU), for guidance and discussions about calculations and MATLAB programming.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors
  • BCSD (Business Council for Sustainable Development). 1993. Getting eco-efficient. Report of the Business Council for Sustainable Development, First Antwerp Eco-Efficiency Workshop, November, Geneva.
  • Bergsdal, H., R. A. Bohne, and H. Brattebø. 2007. Projection of construction and demolition waste in Norway. Journal of Industrial Ecology 11(3): 2739.
  • Bohne, R. A. and H. Brattebø. 2003. Eco-efficiency in Norwegian C&D waste recycling systems. In ILCEDES, Kuopio, Finland, 1st–3rd December.
  • Bohne, R. A. 2005. Eco-efficiency and performance strategies in construction and demolition waste recycling systems. Doctoral thesis 2005: 22, Norwegian University of Science and Technology .
  • Bossink, B. A. G. and H. J. H. Brouwers. 1996. Construction waste: Quantification and source evaluation. Journal of Construction Engineering and Management 122(1): 5560, March.
  • Brattebø, H. 2005. Toward a methods framework for eco-efficiency analysis? Journal of Industrial Ecology 9(4): 911.
  • Chung, S. and C. W. H. Lo. 2003. Evaluating sustainability in waste management: the case of construction and demolition, chemical and clinical wastes in Hong Kong. Resources, Conservation, and Recycling 37(2): 119145.
  • Dantata, N., A. Touran, and J. Wang. 2005. An analysis of cost and duration of deconstruction and demolition residential buildings in Massachusetts. Resources, Conservation, and Recycling 44: 115.
  • Davidson, T. A. and D. G. Wilson. 1982. U.S. building-demolition wastes: Quantities and potential for resource recovery. Conservation & Recycling 5(2–3): 113132.
  • Ecoinvent. 2007. The ecoinvent Database. http://www.ecoinvent.ch. Accessed December 2007.
  • Eriksson, O., M. C. Reich, B. Frostell, A. Björklund, G. Assefa, J.-O. Sundquist, A. Baky, and L. Thyselius. 2005. Municipal solid waste management from a systems perspective. Journal of Cleaner Production 13(3):241252.
  • GRIP/Økobygg. 2001. Nasjonal handlingsplan for bygg-og anleggsavfall.[National Action Plan for Construction and Demolition Waste]. Technical Report, Oslo : GRIP/Økobygg .
  • Horvath, A. 2004. Construction materials and the environment. Annual Review of Environment and Resources, 29(1): 181204.
  • Huijbregts, M. A. J., G. A. Norris, R. Bretz, A. Ciroth, B. Maurice, N. Mahasenan, B. von Bahr, and B.P Weidema. 2003. A framework for evaluating data uncertainty in life cycle inventories. In Code of Life-Cycle Inventory Practice, pp. 113121, edited by A. S. H.de Baufort-Langeveld, R.Bretz, R.Hishier, M. A. J.Huibregts, P.Jean, T.Tanner, and G.van Hoof. Brussels , Belgium : Society of Environmental Toxicology and Chemistry (SETAC).
  • Huisman, J. 2003. The QWERTY/EE concept. Ph.D. thesis, Delft , Netherlands : Delft University of Technology .
  • Huppes, G. and M. Ishikawa. 2005a. A framework for quantified eco-efficiency analysis eco-effciency and its terminology. Journal of Industrial Ecology 9(4): 2539.
  • Huppes, G. and M. Ishikawa. 2005b. Why eco-efficiency? Journal of Industrial Ecology 9(4): 25.
  • Keffer, C., R. Shimp, and M. Lehni. 1999. Eco-efficiency Indicators & Reporting. Geneva: WBCSD, April.
  • Kotaji, S., A. Schuurmans, and S. Edwards. 2003. Life-Cycle Assessment in Building and Construction: A State Of-The-Art Report. Brussels , Belgium : Society of Environmental Toxicology and Chemistry (SETAC).
  • Müller, D. B. 2006. Stock dynamics for forecasting material flows—case study for housing in The Netherlands. Ecological Economics 59(1): 142156.
  • PRé Consultants. 2002. SimaPro 5 User Manual. Amersfoort, the Netherlands: PRé Consultants.
  • Reijnders, L. 2000. A normative strategy for sustainable resource choice and recycling. Resources, Conservation, and Recycling 28(1-2): 121133.
  • Saling, P., A. Kicherer, B. Dittrich-Krämmer, R. Wittlinger, W. Zombik, I. Schmidt, W. Schrott, and S. Schimdt. 2002. Eco-efficiency analysis by BASF: The method. Journal of Life Cycle Assessment 7(4): 203218.
  • Schmidheiny, S. 2000. Eco-efficiency. Creating more value with less impact. Geneva: WBCSD, August.
  • Statistics Norway. 2004. Cost index for road goods transport, March 2004. http://www.ssb.no. Accessed March 2004.
  • Sturm, A. and S. Schaltegger. 1989. Ökologieinduzierte Entscheidungsinstrumente des Managements[Ecologically induced decision instruments of the management]. WWZ: Discussion Paper Nr. 8914, Basel : WWZ.
  • Sturm, A. and S. Upasena. 2003. A Manual for the Preparers and Users of Eco-efficiency Indicators. UNCTAD Technical Report, http://www.unctad.org/isar. Accessed August 2004.
  • Symonds, ARGUS, COWI, and PRC Bouwcentrum. 1999. Construction and demolition waste management practices, and the economic impact. Report to DGI, European Commission.
  • Torring, M. 2001. Management of Concrete Demolition Waste. Dr. ing thesis 2001:65, Norway : Norwegian University of Science and Technology .
  • Torring, M. 2001. Management of Concrete Demolition Waste. Dr. Torring thesis 2001:65, Norwegian University of Science and Technology .
  • Touran, A., C. Christoforou, N. Dantata, and J. Wang. 2004. An estimating system for construction and demolition waste management. Journal of Solid Waste Technology and Management 30(2):8189.
  • Udo de Haes, H.A. 2002. Industrial ecology and life cycle assessment. In Handbook of Industrial Ecology, chapter 12, pp. 138148, edited by R. U.Ayres and L. W.Ayres. Northampton , MA : Edward Elgar Publishing.
  • Verfaille, H. A. and R. Bidwell. 2000. Measuring eco-efficiency. A guide to reporting company performance. Geneva: WBCSD.
  • Weidema, B. 2000. Avoiding co-product allocation in life-cycle assessment. Journal of Industrial Ecology 4(3): 1134.
  • Weidema, B. and G. A. Norris. 2002. Avoiding co-product allocation in the metals sector. In Presentation for the ICMM International Workshop on Life Cycle Assessment and Metals, Montreal, Canada, April 15th–17th.
  • Wikipedia contributors, “Downcycling,” Wikipedia, The Free Encyclopedia, http://en.wikipedia.org/w/index.php?title=Downcycling&oldid=138225285. Accessed January 7, 2008.
  • Wilson, D. G. 1975. The resource potential of demolition debris in the United States. Resource Recovery and Consumption (1): 129140.
  • Yost, P. A. and J. M. Halstead. 1996. A methodology for quantifying the volume of construction waste. Waste Management & Research 14(5): 453461.

About the Authors

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Scenarios
  6. Results
  7. Discussion
  8. Conclusion
  9. Notations
  10. Acknowledgments
  11. References
  12. About the Authors

Rolf André Bohne is a Postdoctoral fellow at NTNU in the Department of Civil and Transport Engineering/Industrial Ecology Programme. Håvard Bergsdal is a Ph.D. student at NTNU in the Department of Hydraulic and Environmental Engineering/Industrial Ecology Programme. Helge Brattebø is a professor at NTNU and currently head of the Department of Hydraulic and Environmental Engineering and Master of Science program director in the Industrial Ecology Programme.