deliverable 2.2 conclusions about the needs of development...

THEME Environment (including Climate Change) Grant agreement for: Collaborative project, Small or medium scale focused research project

Sustainability and Performance assessment and Benchmarking of Buildings

SuPerBuildings

Dissemination level PU Public X RE Restricted to a group specified by the consortium CO Confidential, only for members of the consortium

Deliverable 2.2

Conclusions about the needs of development of performance levels and

benchmarking criteria and weighting methods

Actual release date : 2010-09-10 Lead contractor : Task leader: CSTB

D2.2 Conclusions about the needs of development of performance levels Page 1 of 20 and benchmarking criteria and weighting methods

D2.2 Final.doc (321536 bytes) 15/01/2013

Deliverable Administration & Summary WP1 SuPerBuildings

No & name D2.2 Conclusions about the needs of development of performance levels and benchmarking criteria and weighting methods

Status Final Due m08 Date 2010-09-10 Author(s) Sylviane Nibel, Jacques Chevalier, Alexandra Lebert

Description <Description of the Deliverable from the project plan> The objective of Task 2.2 “Benchmarking criteria and weighting methods” is to collect information and analyze the situation with regard to topics:

the degree of variations in indicators and performance levels for benchmarking that is needed for different kinds of buildings and in different areas of Europe;

the availability of aggregation and weighting methods whose results can be expressed with the aid of key figures or labels.

The conclusion identifies the needs of development according to these two topics.

Comments


D2.2 Final.doc (321536 bytes) 15/01/2013

TABLE OF CONTENTS

Table of contents ....................................................................................................... 2

Executive summary ................................................................................................... 3

1. Introduction ............................................................................................................ 4

2. Analysis of setting performance levels and benchmarking criteria in existing assessment tools ...................................................................................................... 5

2.1 Setting performance levels ...................................................................................... 5 2.2 Degree of variation for different building types and areas of Europe ........................ 6 2.3 The impossibility to share a performance scale for benchmarking ........................... 7 2.4 Conclusions and discussion .................................................................................... 8

3. Analysis of aggregation methods and weighting factors ................................ 10 3.1 General structure – aggregation levels .................................................................. 10 3.2 Aggregation methods ............................................................................................ 11 3.3 Determining weighting factors ............................................................................... 11 3.4 Degree of variation for different building types and areas of Europe ...................... 12 3.5 Examples of weighting and aggregation methodologies ........................................ 12

3.5.1 SBTool ................................................................................................................. 12 3.5.2 BREEAM – Environmental Category Weightings ................................................... 13 3.5.3 LENSE – 1000 points distribution .......................................................................... 14

3.6 Conclusions and discussion .................................................................................. 15

4. General conclusions: Needs of development of performance levels and benchmarking criteria and weighting methods .................................................... 16

5. References ........................................................................................................... 18


D2.2 Final.doc (321536 bytes) 15/01/2013

EXECUTIVE SUMMARY Task 2.2 of Work Package WP2 of the SuPerBuildings project aims at collecting information and analysing the situation with regard to availability and common understanding of performance levels for benchmarking, and aggregation and weighting methods, in order to come to conclusions about the needs for (further) development and/or harmonisation of (existing) sustainability indicators and assessment methods. The analysis is based on the study of eleven national building evaluation tools, i.e. BREEAM, BNB/DGNB, PromisE, HQE, Valideo, CASBEE, LEED, SBTool CZ, Klima:aktiv Gebaüdestandard, TQB and GPR Gebouw.

All these tools have performance rating scale at the building level and also at indicator level. But given the lack of some indicators, like the economic and social ones, a need of development of both performance levels and benchmarking criteria is also required. We need to define a methodology in order to create reference values for these indicators and calculation methods. This methodology must work from a benchmarking when both it is possible or inadequate.

Most of these tools have a single global score as the result of a total aggregation process with weighting factors. Most of them are based on a four-level aggregation scheme. Discussions may concern the consistency of the final set of criteria, establishing rules that clearly define weighting factors based on rational arguments and intermediate levels of aggregation.


D2.2 Final.doc (321536 bytes) 15/01/2013

1. INTRODUCTION Work Package 2 of the SuPerBuildings project, named “Establishment of the common starting point for the project”, has the following main objectives:

- to summarise the existing results on sustainable building indicators and assessment and rating systems;

- to establish a common starting point for the project; - to make conclusions about the main topics of sustainable building indicators and

related assessment methods that need further development. For this purpose, the following topics are viewed:

a) the availability of indicators that describe the potential environmental, social and economic impacts of buildings and the degree of common understanding about these indicators;

b) the availability of methods, which properly measure performance levels for each indicator, and the degree of common acceptance of these methods;

c) the degree of variations in indicators and performance levels for benchmarking that is needed for different kinds of buildings and in different areas of Europe;

d) the availability of aggregation and weighting methods whose results can be expressed with the aid of key figures or labels.

The objective of Task 2.2 “Benchmarking criteria and weighting methods” is to collect information and analyse the situation with regard to topics c) and d).


D2.2 Final.doc (321536 bytes) 15/01/2013

2. ANALYSIS OF SETTING PERFORMANCE LEVELS AND BENCHMARKING CRITERIA IN EXISTING ASSESSMENT TOOLS

This analysis is based on the study of eleven national building assessment tools, i.e. BREEAM, BNB/DGNB, PromisE, HQE, Valideo, CASBEE, LEED, SBTool CZ, Klima:aktiv Gebaüdestandard, TQB and GPR Gebouw. These have been reviewed and analysed from questionnaires filled by SuperBuilding’s partners.

2.1 Setting performance levels

All systems deliver a performance level, defined at the building level (at least to evaluate if the building could be certified or not). A building performance is always the result (in general, an aggregation, see 3.1) of several indicators of performance. This leads to reference values of performance at both building level and indicator level. Firstly, setting an evaluation framework is already a way to slant to a performance definition: Which thematic should be treated? What should be evaluated: the invested effort or the result? What should be considered as important or ignored? Frameworks must be built at the thematic level and at the building level. Same items may be different and tackled differently according to the cultures or the building type (see 2.2). Some tools voluntarily focus on few subjects (e.g. Klima:aktiv is essentially about energy), some try to cover the wide spectrum of sustainability. Thus, the prerequisite for setting a performance rating is to order indicators into categories and subcategories. The building performance is assessed from indicators and requirements, which constitute a set of criteria or a family of criteria that are the base of the aggregation process. These criteria could be qualitative variables (e.g. building aesthetics, purity of the buildings site, ecological value of land, etc.) or quantitative (e.g. heating demand, lifecycle cost analysis, etc.). Indicators or requirements could be a “solution-oriented” variable (e.g. use of construction material – amount of recyclable or renewable material; certified products; no coal, coke or electrical demand) or a performance assessment (e.g. CO2 emissions). At this time, there is no complete system based on life cycle approach (LCA, LCC…). But, some indicators seem to be LCA-based directly or indirectly (construction materials aspects, …). In general, tools are half way between a conception guide (sum of checklists) and performance scales. All tools mix these indicator categories, even if the general trend is to assess the building performance thanks to a core set of indicators (SBA; See Deliverable 2.1 Conclusion about the needs for development of sustainability indicators and assessments methods; 3.4.2. SBA Framework for Common Metris (2009)). As mathematics has prime numbers, the assessment of building performance brings out prime indicators.

The result value for an indicator or a requirement is compared to a range of reference values in order to judge the performance for this indicator. The result value for a qualitative indicator could be binary (e.g. Does the gross floor area inculde some space to park bikes? Yes or No), descriptive (i.e. nominal variable e.g. ecological value of the land) or could already include a primary judgment (i.e. ordinal variable, e.g. the accessibility measurement of different parts and equipments of the building for cleaning [...] can require an answer like ‘rather accessible, very accessible, etc.’). For a quantitative variable, this result value is discrete (e.g. number of bicycle racks) or continuous (e.g. energy consumption). The measurement of these indicators, usually marred by uncertainty could be based on gauging, calculation or on site


D2.2 Final.doc (321536 bytes) 15/01/2013

measurements. The uncertainty is scientific or/and can come from the appreciator judgment. For the most part of the tools, quantitative and qualitative indicators are translated into the same type of assessment (marks or grades).

A rating scale is defined by its perimeter (what is measured and how is it measured?), upper and lower bounds and a transfer function. The perimeter of an indicator could be physical, temporal, methodological, etc. (e.g. a CO2 emission value strongly depends on the chosen method and the LCI structure) and must be defined. The lower, intermediate and upper bounds of the rating scale can be respectively defined by the national regulation level (when the area is under law control), by national (or international) good current practice, and by national (or international) best available practice. The regulation level does not always correspond to the worst score for an indicator: sometimes (national) regulation is considered only as a prerequisite or conversely, to fulfill law requirements could be sanctioned by an intermediate mark (e.g. the GPR tool gives 6 points out of ten if the building is in accordance with the current regulation concerning one aspect). This correspondence between base level and regulation requirement allows a performance control, which is lacking for non-certified buildings. Best available practice seems to correspond to the better score for all systems. However, concerning some tools, a second condition defines the highest limit: a top-down approach is considered and only a percentage of the actually built buildings have to reach this level (e.g. the A/excellent level of each indicator of PROMISE must be achieved by 5 to 10 % of the new buildings; LEED wants to highlight only 25% of the top of the market). But in general, the way to define the performance rating scale for all indicators (or sub-indicators) seems empirical. Regulation level, current practice and best available practice seem to serve as benchmarks in most systems, but the definition of current and best available is difficult and generally a bit qualitative (based on scientific evidence). The transfer function, which consists of normalizing a performance value for each indicator into a score (on a rating scale) between the limits, seems purely practical as well. These transformation functions are not necessary linear, but could be exponential, following a Gaussian distribution, etc. For example, in the GPR Tool, the indicators of energy performance and shadow prices (materials impact) have a normalization function that does not follow a linear curve. A benchmarking has been done from existing buildings. These have been assessed from low sustainability to high sustainability and a curve was fit to the indicator scores for these buildings in order to obtain a transformation function. This function transforms indicator scores into a score between 0 and 10 (6 representing the standard practice of building code). More generally, the transformation functions will be different per building type and per theme. But, it is not always clear how the systems consider the effort to go from one class/score to another. Some systems have a number of levels built from feedbacks but others impose the same number of levels for each indicator: grades may be defined arbitrarily and do not have scientific and practical reality.

2.2 Degree of variation for different building types and areas of Europe

Three degrees of variation between requirements and indicators could be identified: difference between building types, between climatic zones and between systems.

All systems consider different types of buildings (residential, offices, schools, industrial, etc.) and different types of operation (new construction, large scale refurbishment). For some tools, each typology has its one version (GPR, LEED, CASBEE, HQE, etc.). These versions are more or less consistent between themselves (each sub-tool is representative of its temporal context and all tools are not yearly updated), but follow generally the same structure and the same general principles. Differences could be about the studied system: some preoccupations are considered specific to one building type or could be analysed deeper (e.g. indoor air


D2.2 Final.doc (321536 bytes) 15/01/2013

quality for healthcare). Differences could be about the setting of performance levels: energy requirement (on thermal regulation) could differ from a residential building to an office building, leading to different upper and lower limits. Even the transformation function could change from one type to another as well: e.g. statistical data might prove that it is harder to improve energy performance for a school than for a house. Finally, the level of detail could also depend on the age of the system (BREEAM has been created in 1990 and BNB in 2009). Climatic zone or geographical zone differences could be integrated into the performance scale, because results have to be analyzed in their context: benchmarking shall take into account variation in climate conditions (hygrometry, temperature, wind, sun…), seismic constraints, etc. Difference between systems (and so between countries) is various, because of the value scales multiplicity. Preoccupations and evaluated aspects are specific to each tool and reflect a culture. The document “D2.1 Conclusion about the needs for development of sustainability indicators and assessment methods” acknowledges that some preoccupations are widely shared (e.g. primary energy consumption), some are less common, others are very specific and some are completely absent (e.g. vibration pollution). Moreover, even if the same thematic is treated within different tools, it could be in a different way. For example, the issue of land use could be approached by the compactness of the building or the permeability of the soil, etc. Furthermore, some countries have their political action clearly “CO2 oriented” and some “energy saving oriented”. This has consequences on regulation requirements, so on performance scales. Some countries are ready for ambitious objectives and some are not.

2.3 The impossibility to share a performance scale for benchmarking

The question is the following one: “Despite this diversity of systems, is it possible to benchmark all these performance levels, for each indicator, in order to extract generic performance scales, valid for every country, without localization influence?” Obviously, these performance scales could be contextualized in a second part. The difficulty on this benchmark is not the data collection but the impossibility to compare accurately values between each other. To illustrate this difficulty, let’s take an example widely shared in the French construction sector: in people’s mind, PassiHaus and Minergie requirements are much higher than Effinergie French ones. People have in mind the following figures: 15 kWh/m²/year for Passivhaus, 42 kWh/m²/year for Minergie and 50 kWh/m²/year for BBC-Effinergie (French energy performance label). However, how to compare figures that do not cover the same perimeter, are not calculated with the same methodology and are not addressed to the same functional unit?


D2.2 Final.doc (321536 bytes) 15/01/2013

2,5822,85

15 kWhef/m²/an

Figure 1: Minergie, Effinergie and Passivhaus energy requirements. The previous figure illustrates the complexity of the comparison: In Switzerland, Minergie requirements integrate: heating, domestic hot water production, ventilation. The factor translating electricity final energy into primary energy is equal to 2. Consumption value is calculated with national standards. In France, Effinergie requirements integrate: heating, domestic hot water production, ventilation, lighting and auxilary devices (e.g. pumps). The factor translating electricity final energy into primary energy is equal to 2.58. And, in Germany, Passivhaus has two complementary requirements. One about heating only, that should not exceed 15 kWh of final energy/m²/year. The second requirement integrates heating, domestic hot water production, ventilation, lighting, auxilary devices and household appliances. The factor translating electricity final energy into primary energy is equal to 2.85. Moreover, reference units for areas are different from a country to another. In the tools studied in this project, the problem is exactly the same; figures cannot be compared because of:

- Disparity in system boundaries, - Disparity in calculation.

More generally, because of:

- the absence of shared and consensual indicators, - the imminent change in tools to LCA indicators (i.e. from energy consumption during

operation to energy consumption during the whole life cycle), it seems to be ineffective to compare currently used performance levels and benchmarks.

2.4 Conclusions and discussion

There are a few or no reference values for indicators given in standardization framework and environmental standards on building performance (PrEN15978). We need to define reference values for benchmarking for these indicators and calculation methods. For example, reference values will change when we will consider whole life cycle indicators.

We have to choose a philosophy:


D2.2 Final.doc (321536 bytes) 15/01/2013

- Do we want that two strictly identical buildings in the same place but answering two different briefs have the same performance ?

- How do we consider the management in the system (effects in assessment/calculation rules, a particular theme in the system, parallel assessment, involvement of stakeholders in benchmarking process, choice of weighting factors …)?

The following topics would need to be discussed: For new or system-specific indicators:

How to set performance levels when there are obviously lacks of experience and feedback data?

Standardized indicators (PrEN 15978): What are our needs in terms of reference values for benchmarking and how

lead this benchmarking? For non-quantifiable themes or indicators:

How to build a set of classes (mixing descriptive items and sub-indicators)? What are our needs in terms of common rules for that kind of approach?

Performance assessment: Is the performance assessment absolute or relative? How contextual factors are integrated in the assessment process? How political objectives (at national or EU scale) are integrated in the

assessment process?


D2.2 Final.doc (321536 bytes) 15/01/2013

3. ANALYSIS OF AGGREGATION METHODS AND WEIGHTING FACTORS

3.1 General structure – aggregation levels

All systems have several levels of aggregation. All of them have a reference performance level – at the building level - expressed as:

- A grade (A, B, C, D, E) or a number of stars, which is the most common: BREEAM and GPR propose five stars as levels; PromisE and Valideo display five letters as levels. These levels are generally accompanied by an appreciation (excellent, silver, etc.).

- A global mark (expressed in points) that is the sum of the points for each indicator (BNB/DGNB, TQB, etc.) or each preoccupation. These marks could be accompanied by categories or appreciation (e.g. Klima:aktiv: with 700 points a building obtains the Klima:activ Haus level and with more than 900 points the Klima:activ Passivhaus level).

- A ratio, the only case is the CASBEE tool: building environmental quality and performance on reduction of building environmental loadings associated with a class.

- An environmental profile (a building is HQE certified or not according to its 14-component environmental profile, which must be equivalent to or better than a minimum profile).

So, most of them, except HQE, have a single global score, expressed in letters, stars, wordings or points, as the result of a total aggregation process.

Tools can combine several performance scales with various levels of aggregation: for example, Valideo offers an environmental profile with point scores for each indicator (all results can be glanced over the spire chart); subtheme scores (in points), total score (expressed in %) and this last translated in a class (A, B, C, D, E) and an appreciation (outstanding, very good, good, etc.). Then, all of them have a reduced number of themes, issues, topics, items…. At this level, the number of themes is about 5 to 10. All systems also adopt a sub-level that we could call indicator level. As said by BRE, the number of indicators to treat each theme can vary a lot, so it is difficult to limit the number of indicators. Most systems have also a lower level beneath indicators that we could call partial indicators, sub-indicators, contributors…

As a synthesis, we can consider that all the systems are based on a four-level aggregation scheme:

Global score (points, grade, stars, word, letter…) Themes (or issues, or topics…) Indicators Sub-indicators

Tools – because of their conception and so their philosophy and their structure – accept more or less compensation between performances aspects. For example, a building could have a very good thermal insulation and require a very small energy quantity, but offer a bad natural light comfort: this situation could be more or less allowed from one tool to another. Each tool


D2.2 Final.doc (321536 bytes) 15/01/2013

has its minimum prerequisites, but some allow a global compensation (between environment, social aspects, for example) and some restrict these phenomena by having intermediate requirements (e.g. the HQE profile imposes minimal specification for each thematic in addition to minimal requirements at the indicator level). Aggregation allows grasping a global performance. But having a global performance in addition to intermediate indicators, assures transparency: for example SBTool has themes as intermediate notes before a global performance.

3.2 Aggregation methods

All systems use the weighted sum, a total agregation methode, to aggregate from one level to another. The agregation process as a part of aid for decision has in this context the purpose to certify a building and further to improve construction sustainability. Aggregation systems allow to make a balancing between numerous fragmented informations and a synthetic information (number of stars, etc.) comprehensible by stakeholders and in particular by decision makers. Agregation is done from a set of criteria. Criteria are here indicators and requirements mentioned before. In the multicriteria decision analysis domain, these must constitute a consistent criterion family. That means that each criterion must be reliable and pertinent and the set must be exhaustive, non-redundant and coherent1. These general multicriteria analysis requirements are clearly not reached by any of the tools. Different explanations can be put forward as system complexity or the infancy of them. Some authors present the decision aid as an “interface discipline” [1] that requires sharing competences.

Aggregation requires setting a rating scale or preference scale for each indicator. Thus, incommensurable criteria could be compared in the same “unit”. (See 2.1).

It is generally admitted that some criteria are more important than other. In most of the tools, criteria scores are weighted in the aggregation process. In some cases, the weights are all equal to one. CASBEE is the only system that considers at the higher level another aggregation method based on efficiency (ratio benefits / loadings). For weighting and scoring, every kind of marks is possible : percentages, letters or points are used. Sometimes, points operate both as scoring and weighting (the number of points between themes or indicators is different). Example: LEED.

3.3 Determining weighting factors

For most systems, weighting factors are given by a group of experts. Sometimes, weighting factors are given by public authorities or stakeholder groups. That depends on the governance of the system and the involvement of stakeholders in the definition of the benchmarking system. It should be interesting to establish rules that clearly state the rational arguments used to choose the weighting factors. Weighting factors could be affected with a top-down approach

1 For Ralph Keeney and Howard Raïffa, a criterion family must be exhaustive, operational, can be broken down

into smaller utility function, non-redondant and minimal. [ 2]


D2.2 Final.doc (321536 bytes) 15/01/2013

or conversely, with a bottom-up one. For example, the TQB weighting system has been built with the postulate that each of the five subjects represents the same number of points. For the BNB tool, indicators have been weighted from 1 to 3 according to the indicator importance and topics weighting is independent. Only one system considers partially an economic weighting (GPR).

3.4 Degree of variation for different building types and areas of Europe

All systems consider that it is or would be relevant to have different weighting factors for different types of buildings. In some systems, it seems to be also relevant to consider different weighting factors for different areas.

3.5 Examples of weighting and aggregation methodologies

3.5.1 SBTool The following figures (slides from Nils Larsson, june 2010) show the methodology used by SBTool for weighting and aggregation. The system uses a semi-objective weighting system. Weights for each parameter are based on degrees of probable duration of effect, combined with links to key impact indicators.

Weighting factors : duration and regional weight linked to the context


D2.2 Final.doc (321536 bytes) 15/01/2013

Excerpt from weighting sheet of 2010 SB Method – left hand section

Links of detailed criteria (in lines) with weighted impact categories (in columns)

3.5.2 BREEAM – Environmental Category Weightings There were a number of stages in the development of the revised BREEAM UK weightings as follows:


D2.2 Final.doc (321536 bytes) 15/01/2013

Stage 1 : Priorization of environmental impact categories by a panel of international environmental experts, + stakeholder survey

Stage 2 : Contribution of each impact of the construction sector to the total impact for the UK

Stage 3 : Combination of 1 and 2. Outside the UK, it’s necessary to adapt stage 2 to other countries or to Europe.

Stage 4 : Normalizing for the degree of influence, through 3 factors: • Influence factor (what the project team can influence)

• Local conditions factor (e.g. drought conditions) • Technological factor (e.g. water desalination process)

final normalisation to get a 100% scale.

. The current weightings for the BREEAM UK scheme

3.5.3 LENSE – 1000 points distribution The Lense methodology adopts a two stage approach to the use of weightings:

Category weightings (EU wide with country specific ‘fine tuning’)

Sub issue weighting (country specific) For initial testing of the structure it is proposed to have 1000 points available for distribution, 80% of which will be distributed according to the EU wide consensus. To enable an element of country specific fine tuning of the categories, according to regional and national priorities, the remaining 20% of these points will be distributed according to country specific consensus.


D2.2 Final.doc (321536 bytes) 15/01/2013

3.6 Conclusions and discussion

Benchmarking, weighting and aggregation methods shall be aligned with sustainable construction principles defined in ISO15392 (transparency, involvement of interested parties, holistic approach, etc).

Discussion may concern the following questions: Why not establishing rules that clearly define weighting factors based on rational

arguments? How to redistribute weights in case some indicators would not be applicable to the

building? Do we need to separate environmental, social and economic performance? How can we mix a determinist assessment with risk assessment? An interesting approach could be having a political weighting for the top levels and an

expert weighting for the low levels (BNB/DGNB)


D2.2 Final.doc (321536 bytes) 15/01/2013

4. GENERAL CONCLUSIONS: NEEDS OF DEVELOPMENT OF PERFORMANCE LEVELS AND BENCHMARKING CRITERIA AND WEIGHTING METHODS As there is a lack of some indicators, as economic and social ones, there is also a need of development of associated performance levels and benchmarking criteria. For indicators not frequently used, or specific indicators that would need a harmonization of the measurement/evaluation method, it is difficult for tool developers to determine reference values for the calibration of performance scales, because of the lack of experience and operational data. In terms of decision making, we shall build an exhaustive, non redundant and coherent family of indicators. Even if the trend is to develop more performance-based indicators than solution-based ones, some indicators cannot be easily measured by quantitative data. It is for instance the case of maintainability, flexibility, functionality, usability, protection from domestic accidents, social mixity, biodiversity, ecological value of the site, or climate change adaptation. We have to define the contents of those indicators, and to define the performance scale. A system of grades may be a solution, a class including several descriptive criteria, together with possible sub-indicators.

When setting performance levels, it is also important to avoid lobbying by industry or other actors.

Often, the contextual data is important to take into account (climate, socio-economic context, users’ expectations, technical opportunities and constraints, legal requirements…). According to the context, it will be possible or not to implement certain renewable energies, for instance. Do reference values defining the performance scales (benchmarks) depend on the context?

The majority of assessment systems allow a certain degree of freedom in the combination of solutions and performances that will lead to a given global score. How to take into account the relevance of the establishment of an issue hierarchy regarding the local context? (Here the word “local” relates to a small portion of territory, much smaller than a country).

Regarding weighting methods, we will have to pay attention to several points: - When an issue is composed of several indicators or parameters that could be

aggregated by calculation, it should be preferred to weighting factors (e.g. LCC). - When a comfort or health issue is defined by several sub-indicators or parameters, it is

often difficult to calculate a single indicator for this issue, even at a premise scale, and moreover at the building scale (so here we have a double problem to solve).

- In case some indicators or sub-indicators are not applicable within an issue, it is necessary to have rules in order to redistribute useless weighting factors to the other items.

- We have to avoid that the same ecological effects be multi-assessed; but it is important to distinguish between double-counting and the fact that an aspect (e.g. use of fossil energy resources) may have several and distinct environmental impacts (global warming, acidification, depletion of non renewable energy resources).

- The weighted sum method may be used, but together with rules that limit the “compensation” effect between good performances and bad ones.

- How can we mix a determinist assessment with risk assessment (on the same topic or on different topics)?


D2.2 Final.doc (321536 bytes) 15/01/2013

About overall/global aggregation, the answers given in the questionnaire lead to the following points:

- Ensure transparency of the aggregation process, with explicit weighting factors, - Aggregation at the top level (global score), plus visible results for major themes

(partially-aggregated results, multi-criteria profile, absolute values of core indicators…)

- Aggregation rules (weighting factors) defined by experts from various backgrounds and if possible a scientific-based method or at least an objective one (interesting examples of methods: iiSBE SBTool International, BREEAM).


D2.2 Final.doc (321536 bytes) 15/01/2013

5. REFERENCES 1. J.PICTET. Dépasser l’évaluation environnementale. Procédure d’étude et insertion dans

la décision globale. Presses Polytechniques et Universitaires Romandes. 187 pages. 1996.

2. A. SCHARLIG. Décider sur plusieurs critères. Panorama de l’aide à la décision multicritère. Presses Polytechniques et Universitaires Romandes. Collection Diriger l’entreprise. ISBN 2-88074-073-8. 1985. 304 pages.


D2.2 Final.doc (321536 bytes) 15/01/2013

6.

deliverable 2.2 conclusions about the needs of development...

Documents