| Front page | | Contents | | Previous | | Next |

State of LCA in Denmark 2003

3 What is new in EDIP2003

• 3.1 EDIP2003 versus EDIP1997
• 3.2 Definition of the functional unit
• 3.3 Market-based system delimitation
• 3.4 Normalisation, weighting and spatial characterisation
• 3.5 Working environment

3.1 EDIP2003 versus EDIP1997

The Danish LCA methodology and consensus project has introduced a number of important changes and improvements to the methodology known as EDIP1997. As the project was finalised in 2003, it was decided to name the EDIP methodology as it is now EDIP2003.

In table 3.1 is summarised what is new in EDIP2003 compared to EDIP1997, and it is indicated in which guidelines the topics are dealt with in more detail.  In the following a brief introduction is given to these topics.

Table 3.1  What is new in EDIP2003

1)  These new normalisation and weighting factors are regarded as an update of EDIP1997 and not as a part of EDIP2003, since spatial differentiation has now been introduced.

Besides these guidelines a new report dealing with calculation of emissions and environmental impacts from landfilling of waste is also being finalised. This report has been developed outside the methodology and consensus project and will not be described further here. Reference is made to [Hansen et al. 2004].

3.2 Definition of the functional unit

The functional unit is a key concept in LCA. The functional unit is the reference unit, which is being assessed and compared in an LCA. This unit can briefly be characterised as a quantified description of the output or service delivered by the product system/systems assessed in an LCA. The service delivered by a coffee machine may e.g. be described as 5 cups of coffee 2 times a day 300 days a year during 5 years.

Experience shows, that the definition of the functional unit may often determine the result of an LCA.

With the guideline on Product, functional unit and reference flows in LCA  [Weidema et al. 2004] a procedure has been developed that can assist in assuring that the functional unit is correctly defined.

Some of the considerations presented in this guideline have previously been presented in a simplified version in the publication: Håndbog i miljøvurdering af produkter - en enkel metode (Handbook in environmental assessment of products - a simple methodology)  [Pommer et al. 2001].

3.3 Market-based system delimitation

Market-based system delimitation is an important new element that can be compared to introducing a economic way of thinking in LCA.

Traditionally in LCA one has used data, which describe the present production processes taking place. Occasionally is used an average of several different processes. E.g. electricity is often calculated as a weighted average of hydropower, nuclear power and coal-based power.

It has now been recognized that the most correct approach is to use data for production processes that reflect the real technological consequences and thereby also the real environmental consequences of the decisions made based on an LCA.

The fundamental issue is that when an LCA identifies e.g. a certain material as being advantageous, a demand for this material is created. In a free market the response to this demand will typically be an increased production by the manufacturer, who is the most competitive and is not constrained with respect to the size of the production. Therefore the data to be used in the LCA should be data from this manufacturer and not necessarily data from the present suppliers of the material.

The consequences of this line of thinking is best illustrated by an example:

A company situated in Europe is mainly using electricity based on hydropower. In an LCA the company must, however, take into account that hydropower is a technology that is constrained in Europe. By and large there are no opportunities left for increasing the production capacity for hydropower in Europe. If the company in an LCA is assuming that it needs more electricity, the company also has to accept that this extra electricity will be produced by other technologies than hydropower. As e.g. nuclear power is also constrained in Europe - in this case the constrain is political - the company will in most cases have to assume that the extra electricity the company needs will be based on fossil fuels like coal or natural gas. In the LCA the company therefore has to rely on coal or natural gas technology and not hydropower.

In order to carry out a reliable LCA, it is thus necessary to know the market conditions for the central products and  production processes.

With the guideline on Geografical, technological and temporal delimitation in LCA  [Weidema 2004] a tool has been developed to evaluate these market conditions and identify the production processes, relevant in the LCA in question. These production processes is here known as the affected  production processes.  It is emphasized that the time perspective can be very important in these assessments (reference is made to section 4.2 - subsection on trend analysis/projection).

It must also be recognised that this development of methodology is new, and that the experience available so far is limited. In the following is briefly discussed some of the most important issues, which may cause hesitation.

The affected production processes cannot be determined with certainty
In several cases one may likely be in doubt of which production processes are the right ones to consider. In these cases  it will be correct to identify several scenarios which include the possible processes. If e.g. it is assumed realistic that the affected production processes for electricity will be either coal technology or solar battery technology, it will be correct to establish 2 scenarios - one scenario based on coal technology and another based on solar battery technology.

Is it still relevant to use data from the present production processes?
Yes - as stated in the guideline on Geografical, technological and temporal delimitation in LCA it should be considered relevant to use data from the present production processes in the following situations:

• As educational introduction to LCA, because the approach in some cases may seem more simple. Only data from companies belonging to their own chain of products are needed.
• In the initial analysis of environmental impacts in the life cycle of a product aimed at determining the most significant impacts and thereby to decide on which issues it is relevant to focus in order to improve the environmental performance of the product. However, actual changes in design and selection of materials should be determined based on market-based system boundaries.

Environmental product declarations
Internationally as well as in Denmark efforts are in these years invested in defining common guidelines for environmental product declarations. Among other issues efforts are invested in developing a new ISO standard on environmental product declarations. The current trend of this work seems to be that the guidelines will not require market-based system delimitation.  A discussion of advantages and drawbacks of using market-based system delimitation compared to the approach of basing the LCA on the present production processes is integrated in the report Market information in life cycle assessment [Weidema 2003].

Does the market-based system delimitation allow that the LCA will endorse initiatives not necessarily recognised by the market?
The market-based system delimitation normally requires that actions without a real positive environmental impact cannot be endorsed in an LCA. If a company is choosing "green" electricity,  this can only be endorsed, if the action of the company implies that the production capacity of "green" electricity is also expanded in reality. If such an expansion does not take place, the use of "green" electricity by the company will only have as a result that other users of "green" electricity will convert from "green" electricity to e.g. coal-based technology.

Correspondingly a company using secondary raw materials for its production will only be able to endorse this, if a real extra collection of second-hand materials is taking place as a consequence of the company's choice. Otherwise the result of the company's choice will be that the use of secondary raw materials by others will be displaced by the use of virgin raw materials.

According to the guideline on Geografical, technological and temporal delimitation in LCA it is, however, allowed to count on the signal value of the action under the precondition that it is clearly stated that the signal value is taken into account.

This means that it is acceptable in the LCA to count on the fact that the company by the action of buying "green" electricity is sending a signal of that a market for such products exists. The signal may actually contribute to ensure that an expansion of the production capacity of "green" electricity or the collection of second-hand materials will be realized. Either because some market actors see the opportunity of a good business, or because the issue is becoming subject to political initiatives as e.g. legislation.

In case signal values are taken into account, it will be most correct to include 2 scenarios - with and without signal value - as it cannot be known for sure, whether the development called for by the signal will actually be realized.

3.4 Normalisation, weighting and spatial characterisation

Also within these fields some news are introduced:

• Based on the same calculation principles as in EDIP1997 new normalisation and weighting factors for different geographical areas have been developed.
•  The option of using spatial characterisation has been made available.

Normalisation and weighting factors
In EDIP1997 only one set of normalisation and weighting factors was given, which was based on Danish data apart from the global effect categories, i.e. global warming and stratospheric ozone depletion.

With the guideline on Impact categories, normalisation and weighting in  LCA  [Stranddorf et al. 2006] it is now possible to choose between 3 sets of factors covering the following geographical areas:

1.      Denmark

2.      EU-15

3.      The world

The new normalisation factors correspond to the reference year of 1994, whereas the new weighting factors correspond to the reference year of 2004.

Companies hereby have the opportunity of making calculations considering conditions on export markets as well as in Denmark for the non-global impact categories (i.e. all categories apart from global warming and stratospheric ozone depletion). With respect to global warming and stratospheric ozone depletion the normalisation is continuously based on the global emissions.

The updated normalisation and weighting factors cover all impact categories in EDIP1997 apart from the categories addressing waste for landfilling. These last categories are not updated, as it is foreseen that they at least partly will be superfluous when a coming report (reference is made to [Hansen et al. 2004]) on assessment of emissions and environmental impacts from landfilling is ready.

That an environmental impact is normalised means that the size of the impact - i.e. typical the size of the emission - has been related to the total load per year within a specific geographical area and thereby also a specific number of humans. The environmental impact thereby for all impact categories can be expressed in person-equivalents. This step in principle makes it possible to compare the size of the impact between impact categories and see whether e.g. global warming or acidification in a relative perspective is the most important impact in the actual case.

The weighting is following the normalisation and means that the normalised environmental impacts are multiplied by a factor that reflects the decision-maker's position on the importance of the different environmental impacts.

The weighting factors established in EDIP have all been based on politically accepted targets of reduction to the extent such targets have been established at all.

It is emphasized that the fact that the weighting reflects the decision-makers position on the importance of the different environmental impacts means that the decision-maker is allowed to define his own weighting factors.

If a company e.g. in its own environmental policy is stressing that hazardous substances are not to be used in the product or in the manufacturing process it would be logical for the company to assign extra high weighting factors for impact categories dealing with human or environmental toxicity.

Spatial characterisation
The possibility of using spatial characterisation is also an important new element in EDIP and one of the central differences between EDIP1997 and EDIP2003.  Spatial characterisation means that the LCA will take the geographical location of the emissions into account. Furthermore, changes have been introduced to the models behind some of the impact categories. This has reduced the uncertainty related to estimates of the environmental impact for a number of regional and local impact categories, like acidification, photochemical ozone formation and eutrophication. It is noted that eutrophication in EDIP2003 is divided in the water environment and the terrestrial environment.

The characterisation factors given in EDIP1996 and EDIP1997 are mainly  estimated from the basic properties of the chemical substances. Thus no attention has been paid as to where the emission is taken place in geographic perspective, and one is therefore talking about a so-called site-independent characterisation of environmental impacts.

In reality the geographical location of the emission may significantly influence the seriousness of the impact to the environment. This is related not only to where the pollution is transported, but also to the sensitivity of the local environment receiving the pollution to an extra load.

The acidification effect of an SO₂ emission in countries like Norway and Sweden is thus about 1000 times worse than for a similar emission in Greece. It may therefore be of high importance in LCA whether the emission takes place in Scandinavia or in Greece.

Similar differences may be observed for the impact categories photochemical ozone formation and eutrophication in the terrestrial environment.

In the guideline on Spatial differentiation in life cycle impact assessment   [Hauschild & Potting 2005] is described when it is relevant to undertake spatial characterisation, and how this is actually done.

As stated in the guideline, spatial characterisation is so far not included in an LCA computer programme and it will until then be complicated to undertake the calculations. It can be assumed that spatial characterisation will be integrated in the LCA computer programmes that is being approved by the Danish LCA Center in the coming years (reference is made to section 2.4). Until then it is recommended that spatial characterisation is used as a manual procedure in those cases, in which it is assumed to have a considerable influence on the outcome of an LCA, i.e. in which the impact categories acidification, photochemical ozone formation and eutrophication are assumed to be of considerable importance and  the most important processes during the life cycle dominantly are located in Europe.

Even if it has not been possible to develop spatial characterisation for the impact categories human toxicity and eco-toxicity, the guideline includes calculations on test-substances which illustrate the potential significance of spatial characterisation in these cases.

It is, furthermore, possible to undertake a so-called site-independent characterisation of environmental impacts based on the new models. The guideline on spatial differentiation in LCA does also include calculation factors for site-independent characterisation. These factors correspond to the site-dependent characterisation factors and may be used for processes in LCA , when the geographical location is not known.

The advantages of undertaking site-independent characterisation according to EDIP2003 are that a more precise assessment of the real environmental impacts is obtained, and that the uncertainties related to the spatial differentiation are now known. These uncertainties are stated in the guideline.

Since this methodology development is new there is so far only little experience to build on. In the following some important issues are briefly addressed.

When is it recommended to use spatial characterisation?
It is recommended to use spatial characterisation when:

§  The impact categories acidification, photochemical ozone formation and eutrophication are assumed to be significant regarding the outcome of an LCA

§  The most important processes during the life cycle are dominantly located in Europe, and it is known in which country in Europe the processes are taking place.

For what production processes can spatial characterisation be carried out?
Spatial characterisation can be carried out for all production processes taking place in Europe, assuming that it is known in which country the processes are located. For production processes located outside Europe spatial characterisation is not possible, and one has to use the calculation factors that are designed for site-independent characterisation.

If the country of production is unknown, one also has to use site-independent characterisation.

When is it recommended to use site-independent characterisation according to EDIP2003?
Site-independent characterisation according to EDIP2003 can be used as an alternative to the traditional calculation according to EDIP1997. Generally it is recommended to use site-independent characterisation according to EDIP2003, when the following preconditions are fulfilled:

• The most important processes during the life cycle are dominantly located in Europe - this condition is due to the fact that the calculation factors for EDIP2003 correspond to an average for the whole of Europe.
• The impact categories acidification, photochemical ozone formation and eutrophication (in the soil and water environment) are assumed or assessed to be significant regarding the outcome of the LCA in question.
• The ordinary weighting factors in EDIP are not used - as these weighting factors correspond to the data foundation of EDIP1997 and therefore not relevant to EDIP2003.

When is it recommended not to use spatial characterisation?
It is recommended that spatial characterisation should not be used when the LCA is forming the foundation for:

• Environmental product declarations
• Criteria for environmental labelling
• Marketing in general

The argument for not recommending spatial characterisation in these cases  is that use of spatial characterisation in principle could have as a result that companies are choosing to situate production plants at places, where the environment can withstand a high level of pollution rather than choosing a more environmentally-friendly technology. This attitude should be characterised as environmental dumping and as a misuse of spatial characterisation.

In these situations either site-independent characterisation according to EDIP2003 or the traditional calculation based on EDIP 1997 must therefore be used.

New normalisation factors under EDIP2003
As the impact categories acidification, photochemical ozone formation and eutrophication are characterised and calculated in a new way in EDIP2003, it has been necessary to develop new normalisation factors. The new factors are presented in the guideline.

3.5 Working environment

The wish of being able to integrate the working environment in LCA is based on the notion that it cannot be considered acceptable if improvements of the outer environment are obtained at the expense of the working environment.

The original version of EDIP (EDIP1996) therefore included a methodology for assessment of the working environment. This methodology, however, has never been used to any extent noteworthy - likely due to the requirements of the methodology regarding collection and preparation of data.

It is therefore a pleasure that the project has succeeded in developing a new methodology for assessment of the working environment in relation to LCA. This methodology replaces the original methodology in EDIP1996. The methodology is described in the guideline on The Working Environment in LCA - a new approach [Schmidt et al. 2004].

The new methodology is a trade assessment methodology, which combines statistical data on production in industrial trades with statistical data on the number of work related damage and accidents reported for each trade.  Thereby a yardstick is obtained on the working environment load per unit produced in these trades. This yardstick allows that the working environment load is compared between the trades and thereby an assessment of whether the load will increase or decrease, when a material is replaced by another.

The methodology should be regarded as a screening methodology having the advantage that it is simple and objective.

It is emphasized that the methodology is developed specifically for LCA and that the methodology cannot replace workplace assessments and other types of working environment assessments carried out according to the law on working environment.

It can be assumed that the methodology will be integrated in the LCA computer programmes that will be approved by the Danish LCA Center in the coming years (reference is made to section 2.4). Until then is recommended that the methodology is used as a manual procedure in the cases, in which the decision-maker finds it important to consider the working environment in the assessments.

| Front page | | Contents | | Previous | | Next | | Top |

Version 1.0 October 2007, © Danish Environmental Protection Agency