An application of multicriteria decision making to built heritage. The redevelopment of Venice Arsenale



The paper presents a multiple criteria model for the evaluation of the sustainability of projects for the economic re-use of historical buildings in Venice. The model utilizes the relevant parameters for the appraisal of sustainability, aggregated into three macro-indicators: intrinsic sustainability, context sustainability and economic-financial feasibility. The model has been calibrated by a panel of experts and tested on two reuse hypothesis of the Old Arsenal in Venice. Copyright © 2011 John Wiley & Sons, Ltd.


Mapping out the guidelines for the sustainable economic re-use of historic buildings cannot leave out of consideration the complexity of the objectives and methodologies for safeguarding cultural heritage. The historic, aesthetic and artistic characteristics of cultural assets make it difficult to apply a solely qualitative approach. The complexity of the investigation is also due to the public nature of these goods, not necessarily as far as the property right is concerned—many are privately owned—but rather to those related to historic, artistic and cultural values [Brosio, 1993]. The evolution of the concept of cultural heritage in Italian laws and regulations is very interesting. An important law for this matter, passed in 19391, deals with ‘moveable and immovable assets which are of artistic, historic, archaeological or ethnographic interest’, considered being objects which are aesthetically pleasing and, as such, should be safeguarded by appropriate legislation. Article 9 of the Constitutional Charter refers to these concepts and states: ‘The Republic […..] safeguards natural landscape and the historical and artistic heritage of the Nation’, affirming the central Government's sovereignty over the cultural heritage and the values of national identity [Giannini, 1976].

Italy's post-war cultural debate developed new views by proposing innovative laws and Commissions, including the Franceschini2 Commission, which first used the term ‘cultural heritage’ to describe ‘material evidence of civil value’. The cultural heritage assets are no longer simply aesthetically pleasing but also a palimpsest of a culture's history.

The cultural heritage and landscape is currently safeguarded by the ‘Codex of cultural heritage’,3 which, together with the prior law,4 defines cultural heritage. According to this definition, cultural heritage are assets that also encompass the qualities and attributes of objects that have ethnic, anthropological, archivistic or literary value for past, present or future generations.

In recent years, there has been an increasing interest concerning the economic value of the cultural heritage, defining the economic value not only in monetary terms but also in terms of a broader considerations, recognizing, for instance, the fact that the conservation of these assets also generates economic benefits to the society as a whole [Forte, 1977; Throsby, 2002].

Throsby defines cultural heritage as ‘an asset which embodies, stores or provides cultural value in addition to whatever economic value it may possess’ [Throsby, 2001, 2002]. The difference, however, between physical assets (from a strictly economic viewpoint) and cultural capital is indeed the concept of ‘culture’, which bestows the historic goods with an added qualitative dimension. It is this cultural quality that must be maintained and not simply the materials with which the asset is built.

In the scientific literature [Randall, 1991; Stellin and Rosato, 1998], the economic and cultural value of a historic asset are to be distinguished in two macro-categories which refer to two spatial and temporal dimensions. The difference lies in the use and non-use value:

  • The use value, linked to the benefits the consumer receives directly from the asset itself, is a contingent prerogative; it is the utility that the historic artefact offers the consumer from the very moment he comes into contact with it. For this reason synchrony must be created between the cultural asset and the user;

  • The non-use value, instead, does not have the same contingent obligation of the above and, as a result, does not require such close synchrony (but rather a diachrony) as it refers to the utility that the consumers perceive from the conservation of the cultural assets for themselves and for the future generations.


The valuation of the sustainability of the economic re-use of historical heritage is crucial on this discussion and helps to tailor safeguard and protection policies.

Starting with the well-known declination of the concept of economic, social and environmental sustainability, literature on the matter refers to a common premise according to which the ultimate objective of any type of intervention should be to develop local resources and, as a consequence, to contribute to enhancing the quality of life. This is a multi-dimensional concept in so far as ‘the quality of life’ touches several different economic and social aspects [Fusco Girard, 1987; Howarth, 1997].

The concept of sustainability was initially presented by the World Commission on Environment and Development [1987] with reference to the effects of development on environmental assets. Sustainable development was defined as ‘development that meets the needs of the present without compromising the ability of future generations to meet their own needs’.

As far as the cultural heritage and, in particular, architectural assets are concerned, the concept of sustainability is influenced by the environment and involves two main aspects: the sustainability of the material and formal transformation of the building and the sustainability of the new function that is to be installed therein. In other words, the objective of sustainability requires an equilibrium between the economic re-use of the asset and its conservation [Nijkamp and Voogd, 1989].

Current debate on the theories of restoration philosophies, which is particularly active in Italy today, follows two lines of thought.

The first is defined as critical restoration, and stems from the conviction that each intervention project represents a case of its own. Restoration must also transmit the asset to the future by guaranteeing and facilitating its interpretation without loosing sight of the fact that it is a ‘non-verbal criticism expressed in concrete non-verbal ways’ [Carbonara, 1987; Marconi, 1993].

On the other hand, we have the pure conservationists who support the conservation of each strata of material or matter that the building has accumulated over time. Under this approach the building becomes a sort of palimpsest where it is impossible to identify what exactly has to be conserved or removed: ‘The aim of restoration is to conserve both the matter and substance which represents an archive of what the building is actually made of’ [Dezzi, 1991].

An economic re-use project, attributing a new function to the building, often involves transforming, consolidating, adding and removing and may alter the various strata of existing materials and structures.

The decision not to use an asset, however, undermines the intrinsic value of the asset and use poses the threat of possible abandonment and subsequent loss of the asset on the whole.5

Often, however, historic architectural complexes are used for purposes that are completely different from those for which they had originally been built and the interventions required (especially in terms of standards and building regulations that need to be respected) might not always be compatible with the typology and structure of the architectural asset on which works are being carried out. Over-use or incompatible use can have similar consequences to those of abandonment and can gradually reduce the cultural value and historic evidence of the artefact.

Literature does not deal with the definition of what is, or is not sustainable as far as work carried out on historic buildings are concerned. One of the reasons for this silence might be sought in interdisciplinary character of the issue. In this paper a multiple criteria valuation model is proposed which is able to tackle interdisciplinary problems of valuation. The model is founded on a set of parameters measuring the performances of the reuse project.

From the informations codified in parameters, a set of indicators can been developed representing the different points of view with which the concept of ‘sustainability’ may be implemented in the case of restoration and reuse of historic buildings. These indicators should gear to the aim of identifying the limit of transformations, helping to identify the point at which the new use ceases to enhance the asset, and begins to consume and erode the original value. In the following paragraph, the quantitative framework utilized to implement such a model is presented.


Many methods have been proposed in the literature to approach multicriteria problems. Following Vincke et al. [1989], a commonly used classification distinguishes

  • Approaches derived from Multi Attribute Value Theory (MAVT);

  • Outranking approaches, like the ELECTRE family and its derivates;

  • Interactive approaches.

The MAVT methods compute a score for each alternative, using Aggregation Operators (AO) [Klement et al., 2000; Kolesàrovà and Mordelovà, 2001]. Many of the MAVT methods are based on common sense rules, tailored for not quantitative skills of the majority of the Decision Makers (DM). In this contribution, we propose a mathematically founded MAVT approach, which is at the same time easily to be understood by any DM given a suitably designed interface.

The most common aggregation operator is the (simple) Weighted Averaging approach (WA),6 which, for each alternative, computes the weighted average of the criterion score. It is a simple and intuitive compensative method, but no interaction among the criteria can be admitted, since it is based on the Independent Preference axiom. For this reason, many other methods were proposed. We limit to quote the Geometric Averaging (GA) which computes the geometrical averaging of the criterion scores. It can be usefully applied in strong conservative cases, since it gives a null global score if at least one criterion is null (thus impeding compensation). Another class of Aggregation Operators consists of the Ordered Weighted Averaging operators (OWA) introduced by Yager [1988, 1992]. It includes, as particular cases, the weighted averaging, and, as extreme situations, the Max and the Min operators. If the weights are obtained by a non-monotonic quantifier [Yager, 1993], the OWA operator implements linguistic statements as ‘at least’, ‘at most’, ‘at least the half’ and so on. The compensation operator introduced by Zimmermann [Von Altrock, 1995] uses a tuning parameter, representing thus more or less conservative situations. A different approach is obtained using a Fuzzy Expert System, but its design is not a simple task, since many effort needs to be devoted to the inference rules definition [Von Altrock, 1995].

More recently, the introduction of methods based on non-additive measures (NAM) helped to solve many theoretically cumbersome problems, and at the same time offers a wide range of possibilities of aggregation. Up to now, the multicriteria community considers these methods the most complete and mathematically well-founded MAVT approach. Roughly speaking, NAM consists of assigning a suitable weight to every possible coalition of the state of the criteria, and not only to a single criterion, as the WA approach. So the importance of a coalition of criteria can be greater, equal or less than the sum of the importance (weights) of each criterion included in the coalition. Both synergic and redundancy interactions among the criteria can be modelled in this way. If the importance of the coalition for each of them is equal to the sum of weights of the included criteria, the operator simplifies to the WA approach. In the other cases, a simple algorithm computes the score of the alternatives, considering the interactions among the criteria given by the non-additive measures. Moreover, some indices can be computed showing the tendency towards pessimism or optimism reflected in the valuation of the set of alternatives. It should be remembered that the NAM can be directly obtained by experimental data, or implicitly elicited from expert's judgements. In this contribution, we propose an implicit approach. The price to be paid with respect to WA or OWA consists of an increase in the number of parameters, which are equal to the number of all possible coalitions of criteria. For example, using only two possible states for each criteria, 4 criteria request 16 parameters, with 5 criteria 32 parameters, and with 6 criteria 64 parameters are needed. Verifying the absence of interaction between higher order coalitions, we can use a reduced order model where the number of parameters is strongly reduced [Grabish, 1997].

3.1. Non-additive measures

Let equation image. A non additive measure, [Marichal, 1998, 1999a, b] is a set function equation image, so that equation image the following conditions hold:

equation image

Such a measure is able to represent interactions among the criteria, giving a different weight to every possible coalition of them, and not only to a single one as in the case of the WA operator. The first and the third conditions limit the variability inside the domain [0,1], while the second condition is a monotonicity constraint, namely, if more criteria are satisfied, the global satisfaction cannot decrease.7

A non additive measure will be named as:

equation image

For an additive measure, no interaction is possible among the criteria and the linear superposition holds. For a sub-additive measure, a redundant effect is modelled, while the contrary holds for a super-additive effect (synergic effect).

3.2. The Choquet integral

Given a non-additive measure m, let equation image be the criteria values for a particular alternative, normalized in a common scale. We suppose that all the criteria are benefits (higher scores are more preferable than lower). The Choquet integral of the vector equation image with respect to the measure m is defined as follows:

equation image

being equation image an index permutation so that: equation image, x0 = 0, equation image and equation image.

It can also be written as:

equation image

This operator satisfies the following properties [Marichal, 1999a]:

  • (a)It coincides with the WA operator if the measure is additive with:
equation image

wi being the weight of the ith criterion.

  • (b)Every OWA operator is a Choquet integral if every subset of the same cardinality has the same measure:
equation image

For an intuitive explanation of the Choquet integral, see the example in Murofushi and Sugeno [1989].

3.3. The Möbius transform and the dual values

Given a non-additive measure m, its dual values can be obtained from the following biunivocal Möbius transform [Grabish et al., 2003; Marichal, 1998]:

equation image

The inverse transform is given by:

equation image

To be the dual of a non-additive measure, the 2n coefficients equation image need to satisfy:8

equation image

It can be verified that the Choquet integral can be written in the dual space as:

equation image

Moreover, if α(T)>0, the coalition T is synergic (in average), if α(T)<0, it is redundant, if α(T) = 0, there is no interaction and the Choquet integral collapses into the WA operator [Marichal, 1998, 1999a, b].

From a computational point of view, given n criteria, a non-additive measure requires the assignment of 2n coefficients, and this is very large as soon as n is greater than 5,6. In order to avoid this, the k-order models were introduced [Marichal, 1998], which assume interactions between subsets of cardinality less or equal to k, usually the second-order models are considered, that is, k = 2. Even though in many applications it can be reasonably assumed that there is no interactions between subsets with cardinality higher than 2, this hypothesis needs to be tested a priori.

3.4. Andness and orness measures

Given a non-additive measure, it is possible to compute an andness measure together with its complementary orness measure. If the andness measure is close to 1, it means that the measure set tends to the MIN operator, which is to the logical conjunction of the criteria value, showing a conservative tendency of the Decision Maker (pessimistic behaviour). Conversely, if orness = 1 we obtain the MAX operator, the logical disjunction, a totally compensative operator, corresponding to an optimistic behaviour. The computation of the orness index in the dual space is given by:

equation image


equation image

Both indices can be easily computed given the dual values of the measure.

3.5. Non-additive measures and the multi-linear operator

In the dual space, the Choquet integral computes, for each coalition, the minimum of the criteria values of the coalition. The MIN operator belongs to a wide class of operators, the triangular norm (T-norm), which satisfies a set of rationality properties and are widely used in the field of MCDA analysis, especially in the fuzzy logic applications [Klement et al., 2000]. Since the MIN is not compensative at all, some Authors proposed to substitute the MIN operator, in the dual space, with a smoother T-norm, [Kolesàrovà and Mordelovà, 2001; Klement et al., 2000; Despic and Simonovic, 2000; Fujimoto and Murofushi, 1997]. A natural choice can be the product of the values, that is a differentiable and partially compensative operator. We obtain the so-called multi-linear operator [Grabish et al., 2001]. In the dual space, substituting the MIN operator with the product, we obtain:

equation image

In the measure space the multi-linear operator has the following formulation [Marichal, 1999b]:

equation image

which represents a pseudo-Boolean function.

3.6. Identification of the measures

As said above, one of the most critical point in the evaluation is the assignment of the numerical values of the non-additive measure. Many methods were presented in literature, but most of them are based either on quite complex optimization algorithms, or on data mining techniques. In this case study, we preferred a user-friendly approach, and adopted a method based on a suitable questionnaire [Despic and Simonovic, 2000]. Let us suppose that the DM(s) judgements are in the scale [0,100], with the usual meaning for the numerical values, i.e. 0 = WORST, 50 = MEDIUM, 100 = OPTIMAL and so on. For each criterion two particular extreme cases are enhanced, the OPTIMAL and the WORST ones, conventionally indicated with 1 and 0 respectively from now on. An edge is a (fictive) scenario formed by a combination of (only) WORST and OPTIMAL evaluation. Each edge is nothing else that a question that is asked to the DM(s), which will assign his/(their) evaluation in the scale [0,100]. The edges are the vertices of an hyper-polyhedron in the criteria space. It is sufficient to define the values only in all those vertex to obtain the values of the measure, and this is the minimum amount of information. This simplification causes a poor statistical robustness, since it corresponds to the minimum number of interpolating points in an n-dimensional space, but given the unavoidable uncertainty, which is implicit in every human decision process, this does not seem to be a serious obstacle, considering the information gain that should be obtained explicitly considering all the possible interactions among the criteria. The advantages with respect to the WA approach are evident.

Table I reports an instance of the questions that need to be formulated in the case of 3 criteria. Referring to the case study, we are considering the node in the Sustainability Tree, which evaluates the Sustainability starting from Intrinsic Sustainability, Economic-Financial Sustainability and Context. The fourth column reports the DM evaluation (only one DM is simulated here). For a better comprehension, the fourth row implements the question:

Table I. The valuation table
Intrinsic sustainabilityContext sustainabilityEconomic and financial feasibilityEvaluation

‘How would you score an hypothetical case where the Economic-Financial Sustainability is OPTIMAL, and the two other criteria, Intrinsic Sustainability and Context, are WORST?’

After having fulfilled all the answers, a simple algorithm computes the dual values and passes such parameters to a procedure that implements the computation of the multi-linear aggregator for a real case. Moreover, the andness and the orness degrees can be computed and the behavioural nature of the DM can be obtained.

Assume, for the previous example with three criteria represented in Table I, the ‘weight’ of the first criterion to be equal to 30, the second to 20 (the second and the third empty cells in the last column of the Table), while the ‘weight’ of the coalition formed by the two criteria together to be equal to 70 (the last empty cell in the last column). Then, a synergic effect can be observed, since the ‘weight’ of the coalition is greater than the sum of the weights of the single criteria.

Evaluation in intermediate points would increase the statistical robustness, but the numerical complexity of the algorithm would increase significantly either. We feel that the edges evaluation and the multi-linear operator are a good compromise choice between theoretical complexity and operative usefulness. Other solutions, see for instance [Fujimoto and Murofushi, 1997], are difficult to be implemented and require a strong computational effort. Moreover, the same approach can be used in the case of multi-person decision scenario, where many Experts or Decision Makers cooperate in the assignment of the ‘weights’ of the criteria coalitions, and a measure of consensus could be easily defined and computed [Kacprzyk, 1987; Kacprzyk and Fedrizzi, 1988; Kacprzyk et al., 1992].


In the previous paragraphs, we illustrated that integrated conservation is defined as the best possible compromise in dealing with conflicting objectives. Therefore, the operative phase of the study concentrated on the definition of indicators for the evaluation of sustainability of alternative re-use projects for historic artefacts.

The design of a hierarchy model for the evaluation of the sustainability was based on the definition of criteria synthesizing the main characteristics, which could influence the evaluation of sustainability. This initial phase was completed by consulting experts in urban re-qualification and the re-use of historic buildings. The resulting proposed indicators take into consideration the effects of the intervention on the artefact by using three main points of view: the impact on the historic building (defining future re-use—and relative standards—to be hosted in the historic building); the social impact; the economic and financial feasibility (Table II).

Table II. Hierarchic structure (simplified) of the evaluation model (* nodes)
original image

Intrinsic sustainability: or the respect of the materials and typology of the building. This criterion is the synthesis of three sub-criteria:

  • Reversibility or the opportunity to restore the building to the state it was in before the modifications carried out with the re-use project;

  • Versatility or the possibility to eventually modify the function of the building proposed by the re-use project without major works;

  • Invasiveness or the degree to which the project interferes with the materials the historic building is made of.

Context sustainability, which refers to the extent to which the reuse project enhances the social, economic and environmental context of the building and its contribution to the local identity. The re-use project must, where possible, rebuild a relationship between the building and its social and environmental setting. It is also hoped that the project will produce positive externalities on circulation and bring economic advantages to the territory.

Economic and financial feasibility, which evaluates the project according to economic and financial principles. The model implies that the objective of sustainable re-use also depends on the project's financial efficiency of the economic activity. Moreover, the risk concerning the investment must also be taken into account.

After evaluation model structuring, to each criteria, sub-criteria and attribute was given a weight, which defines its contribution towards sustainability. In order to calculate the weight of each single characteristic, a questionnaire was prepared applying the edge's method described in the previous paragraph.

The questionnaire had a page for each of the nodes of the hierarchical tree, so that to each leave belonging to the node would be given a weight. The questionnaire was compiled by 11 experts.

Table III presents the average score of the evaluation given by the experts, their standard deviation and variation coefficient for the ‘sustainability’ node and for the extreme scenarios.9

Table III. The scores attributed to the ‘sustainability’ node
ScenarioSustainabilityAverage scoreStandard deviationVariation coefficient (%)
 Intrinsic sustainabilityContext sustainabilityEconomic and financial feasibility   

The data presented puts into evidence that the experts place at the first level the intrinsic sustainability of the re-use project (Scen. 1), the coherence with the social context is placed in the second step (Scen. 2) and, finally, they consider the economic aspects (Scen. 3). Another interesting result is that the importance given to the intrinsic sustainability is quite stable across the experts' valuations (V.C. 40%), but they gave quite different evaluation scores for the indicators regarding the ‘context sustainability’ criterion and the economic–financial feasibility (V.C. 70–80%).

Analysing the scores given to scenarios (5, 6 and 7), where ‘optimal’ judgements are given contemporarily to two criteria, it emerges that a ‘optimal’ judgement given to the ‘intrinsic sustainability’ criterion is sufficient to realize a good (approx. 60) and stable (V.C. 10%) score. Furthermore, in the other case the overall score is low (48) and variable across the experts (V.C. 25%).

For instance, the sustainability index (Is) can be computed from the criteria scores reported in Table III using the Choquet aggregation formula (see Section 3.2).10

equation image

where IS is sustainability; SI is intrinsic sustainability; C is context sustainability; FEF is economic and financial feasibility.

In the above equation the first three coefficients are simply the normalized values of rows 2, 3 and 4 of ‘Average score’ column in Table III. The other coefficients are the Möebius transform of the corresponding criteria coalition. For instance, the coefficient of the coalition (SI, C) is obtained applying the first formula of sub section 3.3 that is:

equation image

subsequently normalized dividing by 100.

Once the model has been calibrated with the value functions, the technician responsible for evaluating the sustainability of re-use projects expresses a judgement (0,100) for each parameters in which the various attributes of Table II have been disaggregated. This score is multiplied by the weight attributed to the parameter and by the weights assigned to the nodes higher up. In other words, giving a technical evaluation to each parameter associated with the project under examination, a comprehensive evaluation of the sustainability of the re-use project is realized.

The model is useful when there are several alternative projects to choose from, as it supplies a final sustainability score for each alternative and intermediate scores that refer to the criteria, sub-criteria and attributes. As described above, in order to assign weights to criteria, sub-criteria, attributes and parameters, the experts filled out a questionnaire and gave scores ranging from 0 to 100 to hypothetical scenarios.

The experts shared cultural knowledge in at least two fields: the conservationists were architects operating in the material restoration of historic buildings; the designers and planners were specialists in analysing and identifying the function that the historic building should be given and the economic evaluation of re-use.

It was thus useful to establish indices which would evaluate the attention toward conservation shown by each expert's judgement.

Andness and orness indices were used, where the index value may vary between 0 and 1 in both cases and takes on the following significance:

  • Total andness: the expert consulted considers that the sustainability of a project is guaranteed only if all the indicators are attributed the maximum score (andness index = 1; orness index = 0);

  • Total orness: the expert consulted considers that the sustainability of a project is guaranteed if one of the indicators is given the highest (andness index = 0; orness index = 1);

  • Mainly andness: the expert consulted considers that the sustainability of a project is guaranteed only if the majority of the indicators are attributed a high score (andness index>0.5; orness index <0.5);

  • Mainly orness: the expert consulted considers that the sustainability of a project is sufficiently guaranteed when one indicator rather than another receives a high score (andness index<0.5; orness index >0.5);

  • Additive measure: the expert consulted considers that the sustainability of a project depends on the sum of the scores assigned by the indicators, without there being any synergy between them (andness index = 0.5; orness index = 0.5).

Table IV shows average Andness/Orness indices for the 11 experts consulted. The majority of experts tended towards Andness behaviour in all the nodes examined which means that a project can be considered sustainable if at least two or more criteria are deemed ‘optimal’; thus it is not enough for the project to respect the historic building, but it must also be economically sustainable, and its reference context must be carefully considered. Furthermore, the Andness behaviour is higher for the ‘Intrinsic Sustainability’ criterion than in the other criteria.

Table IV. Indices of ‘andness’ and ‘orness’ for the most important criteria
 AverageStd. Dev.C.V.
Intrinsic sustainability
Economic and financial feasibility


The model presented in the previous paragraphs has been used for the valuation of the sustainability of alternative re-use hypothesis of the ancient Arsenale of Venice.

The Venice Arsenale is owned by the Italian government and is currently used primarily by the Italian Navy. About 45 hectares in size, the Arsenale accounts for about 15% of the area of the city of Venice, and is located in the Castello district. Founded in 1104, in its heyday the Arsenale employed roughly 20 000 workers and was said to produce one ship a day.

The Arsenale started to decline after World War I, and continued to decline at an even faster rate after World War II, when its buildings were progressively abandoned. In 1983 the Soprintendenza per i Beni Ambientali ed Architettonici of Venice (local of office of the state authority for cultural heritage conservation) started a series of conservation works. At present, the Italian Navy continues to own and occupy a large portion of the Arsenale. Research activities, shipbuilding, museums and exhibitions occupy other areas, but many buildings and areas remain unutilized.

Out of the analysis of the political debate on alternative options for the re-use of the Arsenal, two basic alternative directions could be extrapolated. The first one is pointing to installing ‘poor’ functions in the ancient buildings without considering the historic significance of the area, but well compatible with the historic building structures. The functions to be introduced are small artisans activities (carpenters, electricians, masons etc.) mostly already working within the historic centre but often under menace of expulsion because of pressings from the real estate market. The second option points to the introduction of ‘new’ uses somehow connected to the Arsenal's historic marine function, like a touristic marina. On the basis of these basic assumptions, two hypothetic projects or scenarios have been created in order to evaluate their sustainability.

5.1. 1st Scenario: area for artisans

In the first scenario it is assumed to use the buildings of the Arsenale for craftsmen's activities actually dispersed in the historic centre. The surfaces of water of the main dock and some of the buildings will be used for laying up small boats owned by Venetian residents.

It is presumed that the whole surface and all buildings, except those actually occupied by the Navy, will be used by artisan's activities. The re-conversion will take place after a restoration programme managed by the municipality, which will adapt the buildings to the requirements of craftsmanship and small manufacturing activities. The industries that are going to settle within the restored buildings will pay a rent ruled by medium- to long-term contract (around 20 years). The surface of the big dock (Darsena Grande) will be used for mooring of boats owned by Venetian residents. A limited number of buildings, including the covered docks, will be used for mooring and laying up of boats on high rise racks.

5.2. 2nd Scenario: marina

The second scenario refers to a proposal frequently presented in the past to use the historic Arsenale as a touristic marina for permanent and temporary mooring. The activities to be introduced regard, beyond the berths themselves (approx. 220 places), supplementary facilities comprising high quality shipyards, boats repair and laying up services, shops and services necessary for tourism, as retail stores for nautical equipment.

In this proposal the area's original shipyard vocation is taken up, expecting the nautical tourism to contribute to a revival of the traditions of this place in terms of boatbuilding. The berths of the main dock will be partly assigned on a permanent basis, 25% will be reserved for temporary mooring.

The historic buildings will house the facilities connected to the port such as marine shops, craftsmen activities and boatbuilding as well as a shipyard for the production of leisure time boats. A supermarket will be located in a position easy to access from the surrounding residential areas as well.

The open spaces, transformed in quays, are used as slipways for the marine activities and shipyards.

Some buildings on the southern front of the main dock will be transformed in to reception area with restaurants and bars, a yacht club, and rooms for small events, sailing schools etc. as well as services offering assistance for guests.

Introducing productive activities into the historic buildings does not represent a particular problem from the conservation point of view. Some more problems may be represented by the introduction of commercial facilities and supermarkets, which might ask for divisions of the inner spaces, with consequent modification of the typologies of the historic buildings.


The evaluation of the scenarios described above requires the assessment of the state of the indicators of the model in each re-use hypothesis. This assessment must regard the technical parameters that define the attributes.

6.1. CRITERION: intrinsic sustainability

6.1.1. Sub-criterion: reversibility

Reversibility of the interventions is not a major concern for re-use projects for the historic Arsenale, as the typological scheme of the buildings is easy to be adapted to the needs of productive activities. The open spaces inside the buildings allows, up to a certain extend, for the insertion of internal structures. These structures have to remain detached from the main structures in order to allow for the perception of the original shape of the building. The transformation of the shipyard buildings, which were initially open towards the waterfront, into closed buildings has already taken place during the 19th and 20th century, and will be reconfirmed by the project for the artisan's area. A problematic aspect of reversibility regards the lack of natural illumination of the original buildings, thus requiring transformation of parts of the coverage.

 Within the project for the Marina, the problems raised by the transformation are similar to those mentioned above, as the complex was created as a productive structure, and is relatively easy to be adapted to new uses of the same type. Within some limits the same can be said about the insertion of commercial services and restaurants, which might be practiced in a way similar to the productive activities, using detached structures inside the original buildings, emphasizing the technical and productive character of the context. The realization of support structures for the marina seems to be more complicate as the buildings have no lateral openings. It will thus be necessary to accurately distribute the functions inside each building. In no case an irreversible transformation of buildings is foreseen. As in the case of Scenario 1, the problems will be raised by the introduction of sanitary services in both scenarios.

With regards to finishing, no particular problems of conservation are to be expected given the industrial character of the buildings. In the case of the Marina, the lack of finishing can be transformed in added value, evocating the historic character of the area.

The introduction of new technical equipments will cause some problems, as transformation of roofs and/or openings will be required. In both cases, technical structures will be distributed and designed according to the requirements of the single enterprise, although the concentration of some facilities and some support services (reception, administration and canteen) in separate structures is planned (Table V).

Table V. The valuation of sub-criterion reversibility
CriterionIntrinsic sustainabilityScore
  Artisan's areaMarina
AttributeTypological structure90.873.3
 Conservation of characterizing elements90.070.0
ParametersPlasters and hangings100.095.0
 Thresholds, benches95.090.0
AttributeTechnical equipments77.565.5
ParametersRemovable housings75.070.0

6.1.2. Sub-criterion: versatility

The high grade of reversibility of both projects guarantees for a high grade of versatility, allowing eventually for the insertion of alternative productive uses. This is assured by inserting new structures and vertical connection as independent elements respect to the historic building, both from the static and the visual point of view. The Marina project, where the internal divisions to be introduced for restaurants, reception etc. might require more important transformations, results in a lower grade of versatility respect to alternative uses. In no case irreversible transformations of relevant parts of the existing structures are planned.

The adaptation of the buildings of the Arsenal to the necessities of small enterprises does not present particular problems for what regards the insertion of adequate technical structures. In analogous way the Marina project allows for the adaptation of the historic buildings by insertion of independent structures detached from the historic elements. Major difficulties might arise in this case of the restaurants and commercial facilities.

With respect to the type of use chosen in the first scenario, the Arsenal would regain its original productive destination, although from the symbolic point of view the significance of these new uses is quite different from the original one. The production of ships was a crucial activities for the maintenance of the geopolitical role of the Venice Republic as one of the major commercial and political forces in the Mediterranean, the craftsmen activities represent a mere support to the every-day maintenance of the city itself, without any strategic role for its economic base.

The symbolic value of the new use in the second scenario is quite high, and is consistent with the historic function of the complex. Similar to the period of the Venetian republic, the use of the Arsenal as Marina is coherent to the economic identity of the city, based today mainly on tourism.

For both scenarios, accessibility for pedestrians is determined by the original asset of the complex oriented to a maximum of control of the access to a strategic area for the military security of the Republic. Some new accesses have already been created during the transformations in the past two centuries, and only few further access points can be created if heavy transformations should be avoided. Furthermore, the area is situated in a peripheral location with respect to the city centre—and towards the principal accesses to the mainland. The accessibility within the complex is determined by the location of the single building with respect to the nearest access point, and can be in some cases very poor.

With respect to the accessibility by boat from outside for the first scenario, there are two accesses from public transport lines: one from the north and one from the south, which both connect to the pedestrian accesses to the area. The access for private boats can be considered very good.

Circulation inside the main dock may be made difficult by the presence of landing stages for the mooring of boats for both scenarios; access for boats to the port is optimal for the second scenario as the north-eastern opening of the main dock is easy to be reached from the lagoon. The entrance into the main dock is possible also for small ships (Table VI).

Table VI. The valuation of sub-criterion versatility
CriterionIntrinsic sustainabilityScores
  Artisan's areaMarina
AttributeType of re-use87.778.3
ParametersRigidity of installations8080
 Poss. surface removal9085
 Prevision of vertical connections9070
AttributeCongruity of technical installations with the standards required98.393.3
ParametersDedicated rooms100100
 Number of terminals10085
AttributeTypology of the historic complex65.075.0
ParametersHistoric character7080
 Congruity of technical installations with the standards required9893
 Dedicated rooms100100
 Outdoor spaces6070
ParametersPublic transport4040
 Parking spaces100100
 Access for disabled1010

6.1.3. Sub-criterion: invasiveness

The invasiveness of the structures under the first scenario is rather low, because the concept of detached structures to be introduced into the buildings guarantees for a good visibility of the original typological scheme. Albeit the convergence among traditional and new uses, not in all cases the coherence with traditional functions is assured, which may result in difficulties in re-establishing the original orientation of the buildings towards the water. The same can be said for the second scenario, although a stronger orientation towards the water surface is guaranteed by the specific functions foreseen. In the case of commercial services some important modifications of the distributional schemes will be necessary (Table VII).

Table VII. The valuation of sub-criterion invasiveness
CriterionIntrinsic sustainabilityScores
  Artisan's areaMarina
AttributeTypological scheme80.081.7
ParametersVisibility of the asset9085
 Functional coherence7090
 Changes in distribution8070
ParametersSubstitutions can be recognized9085
 Similarity of materials8560
 Removal of decay9070
AttributeFinishing and decorative elements100.0100.0
ParametersReconstructions can be recognized100100
 Removal of decay100100
AttributeTechnical equipments55.050.0
ParametersVisual impact5050

No substitution of structures is planned, but new structures may be necessary under both projects where the original buildings are lost. Technical equipments will be realised for both projects in a detached manner, which results in an elevated visual impact. For the artisan's project, a medium rate of compacting is expected, for the Marina project this rate will be medium–high.

6.2. Criterion: context sustainability

In the both scenarios, the scarce level of invasiveness will determine a substantial conservation of the urban landscape of the Arsenal. This is true for the buildings, but not for the outside areas and the water surface, which will be fragmented by the floating structures used for the mooring of small Venetian boats and for leisure boats in the second case. The impact of the re use on the surrounding area is limited, as no new uses will be introduced, and the area is substantially isolated towards the surrounding.

The decision to open the Arsenal to urban productive functions and to the moorings for the citizens will create a good level of consensus for the first scenario of the artisan's area.

Also under the marina project the Arsenal will be accessible to the citizens and to a somehow ‘noble’ function, reconnecting to the area's original function. These aspects will promote a positive perception of the project, whereas critical voices will note that the weight of the tourism in the urban economy will be further increased.

Judgements on the impacts on traffic foresee only scarce impacts for both scenarios.

The impact on the urban economy of the project described in the first scenario will be rather scarce. New uses in the Arsenal might be able to develop the urban economy and, as described in the second scenario, might be used to qualify the predominant sector of urban economy, tourism.

In the second scenario some positive effects may be expected in terms of re-qualification of the tourism sector on the surrounding areas (Table VIII).

Table VIII. The valuation of criterion context sustainability
CriterionContext sustainabilityScore
  Artisan's areaMarina
AttributeQuality of urban landscape80.096.7
ParametersMaintenance of landscape quality10090
 Maintenance of aesthetic quality90100
 Positive externalities on the built env.50100
ParametersSharing of functions with the community10030
 Public use10040
 Maintenance perception in the community7070
 Increase in perception of cultural value7070
AttributeImpacts on traffic97.592.5
 Private transport10090
 Public transport100100
 Natural and cultural paths9080
AttributeLocal economics60.990.9
ParametersBenefits for the community100100
 New economic activities induced by re-use10090
 Diversification of economic activities100100
 Natural and cultural paths9080

6.3. Criterion: economic and financial feasibility

The expected earning from the project described in the first scenario will be rather low, and public aid is needed for the restoration. These initial investments to be made by the municipality will only in part be covered renting the buildings. Also, the moorings for residents will have a low return. On the contrary, the attended earnings from the Marina project will be high as a high number of moorings for transit and of big boats is expected.

Under the Artisan's project mainly already existing functions will be transferred from other urban areas to the Arsenal. Consequently, the level of risk is low, but also the marina has low risk level as tourist activities in the Venetian context generally prove to be a quite sure form of investment.

The initiative for Artisans activities requires a high level of external financing for the restoration works and has low return rates to be expected albeit low management costs, whereas the marina initiative will guarantee for financial feasibility also without initial subventions, although management activities required will be higher (Table IX).

Table IX. The valuation of criterion economic and financial feasibility
  Artisan's areaMarina
AttributeExpected earning40.0100.0
 Financial feasibility50.090.0
 Onerousness of management of the new use90.060.0

This analysis shows that the evaluation of the sustainability of the hypothetical projects for the marina is similar to the sustainability of the project for the artisan's area. The project for the marina would ask for mayor transformations of the original buildings, resulting in a score on intrinsic sustainability which is slightly less favourable than for the artisan's area. The score for the context sustainability is slightly more favourable for the marina project, as positive impacts on the local economy outweigh the negative impacts expected in terms of social consensus and large boat traffic.

The score on the economic sustainability is favourable for the tourist marina project, as it can be expected to produce a sufficient return to cover expenses for restoring and maintenance of the structures (Table X).

Table X. The valuation of sustainability
 Artisan's areaMarina
Intrinsic sustainability0.6410.589
Context Sustainability0.8320.850
Economic & Financial Sustainability0.6580.804
Sustainability (overall)0.6490.675


The aim of the paper has been to present a procedure for the evaluation of the sustainability of projects for the economic re-use of historical buildings in Venice. A multiple criteria model for the analysis of alternative projects for re-use and to support the choice was set up. The model adopts a hierarchical approach that identifies the relevant indicators for the appraisal of sustainability, and groups them into three criteria: intrinsic sustainability, context sustainabilitya and economic-financial feasibility. The aggregation operator at each node of the hierarchical tree of the model computes a global evaluation based on non-additive measures and the multi-linear aggregation function. The measure values are implicitly obtained from a panel of experts who filled a questionnaire on hypothetical scenarios, allowing for the calibration of the value function with which to analyse the sustainability. The preference structure obtained permits the analysis of the conjunctive–disjunctive (andness–orness) behaviour of the experts.

Starting from the opinions expressed, indicators were then drawn up to estimate the level of conservativeness of the expert evaluations.

Operationally, the evaluation model was tested on two reuse hypothesis of the Arsenal in Venice. The evaluation model seems able to provide interesting results on the sustainability of the projects for re-use, correctly considering the environmental, social and economic components of the work and highlighting the strengths and weaknesses of the two type of re-use. Such analysis can be used in various ways.

Primarily, it can provide a useful support for the identification of the critical point, at the preliminary stage, of projects capable of combining conservation and economic improvement. Second, it can be a support for the selection of projects to be financed in that it allows the trade-off between economic use and conservation to be appraised and thus, implicitly, the cost of the conservation. Finally, it can provide a means of reading the projects for re-use, a kind of checklist of variables to be considered in the evaluation of the proposals.


The authors express their thanks to Valentina Zanatta and Mila Dallavalle, who collaborated in an early phase of the project. Their further thanks go to the following architects who contributed to compiling the questionnaire: Gianluca Frediani, Roberta Galli, Gabriele Ghiglioni, Alberto Guzzon, Raffaella Lioce, Danila Longo, Silvia Marini, Emilio Natarelli and Marco Storelli.

  • 1

    Law 1089 of 1 June 1939.

  • 2

    The Franceschini Commission operated from 1964 to 1966.

  • 3

    Legislative Decree bearing the ‘Codex of cultural heritage’, in accordance with article 10 of the Law no. 137 dated 6 July 2002.

  • 4

    In January 2000 the ‘Consolidated Law on natural and cultural heritage’ (TU 490, 1999) came into force; article 4 takes up the idea of cultural heritage as a testimony to civil value.

  • 5

    The European Charter of Architectural Heritage adopted by the Council of Europe [European Council, 1975] introduces the social and economic issues related to restoration and formalizes the concept of ‘integrated conservation’, or rather the integration of heritage into the ‘context of public life’, by means of restoration and appropriate use. In the same year, the Declaration of Amsterdam stated that the attribution of new functions should respect architectural characteristics and guarantee their survival; the conservation effort ‘must be based on the cultural and utilization value of the building’. Carlo Forte, on the other hand, claims that the conservation of the cultural heritage aimed at its integration into modern day life ‘constitutes a true and proper productive activity’ and an essential priority. The limited funding available do not allow conservation to be the sole finality of the intervention, but make it necessary that the building be put to compatible use. By doing so, the cultural capital will generate assets and services which will increase its social function and accessibility.

  • 6

    See Mousumi and Husain (2009) for an application to the decisions on heritage sites to be protected.

  • 7

    Violations of this constraint are accepted only in the case where a criterion is a benefit for one coalition, but a cost for another one. Nonmonotonic measures can capture this effect, but we will not this quite uncommon case.

  • 8

    P(S) is the power set of the set S.

  • 9

    It should be noted that the average of a set of non additive monotonic measures is a monotonic measure, too.

  • 10

    The same procedure has been applied for each node of the hierarchical tree shown in Table II.