The analyses allowed us to assess the actual repercussions of marginalisation on the built environment of the Granfonte neighbourhood. The phenomena detected at the urban scale and specified at the building scale reveal a complex and multifaceted nature: they can combine at multiple levels and have multi-scalar effects on the historic built environment.
5.1. The Problematic Nature of Abandonment
The close reading of the urban fabric has made it possible to assess the actual repercussions of the marginalisation phenomenon on the historic built-up area. The analysis of the building use reveals a diversified scenario corresponding to the evident fragmentation of residence: there are buildings in good condition and still inhabited—but often heavily transformed—alternated by empty, abandoned, and even ruined buildings. This investigation, shown in
Figure 5, thus clearly highlights the partially abandoned condition of the neighbourhood; in fact, abandoned dwellings account for about half of the total. The critical reading of the building fabric has also made it possible to identify the presence of some building units that should already be considered as in a state of ruin according to the classification adopted here, as they have roof collapses of more than 50%. Some of these appear to be affected by processes of a more advanced ruination as they are involved in serious structural instability. On the other hand, a few building units had a provisional roof following the collapse of the original one.
The observation of the distribution of abandoned and ruined buildings highlights certain peculiarities of the residential vacancy. If the abandonment—and sometimes the subsequent phase of advanced ruination—appears to be mainly scattered in the fringe blocks, it is more extensive in the heart of the neighbourhood.
The analytical filing carried out after the field work revealed important criticalities. For example, in the case of more complex aggregative conditions and progressive densification and superelevation, while the upper levels are still permanently inhabited, those downstream are in a state of abandonment for reasons stemming not only from ownership fragmentation but also from poor accessibility or lack of light and air. In such a scenario, the lower floors that are in a condition of prolonged abandonment may even present structural instabilities, which are sometimes resolved using hasty consolidation operations. This is aggravated by the presence of superelevation structures that continue to be inhabited. Abandonment thus generates multiple risk situations with repercussions on both the stability of the buildings involved and those adjacent and on the safety of public routes. The clearly precarious conditions have indeed forced the closure of some roads, but the instability of many buildings continues to represent a vulnus for the entire urban environment (
Figure 6).
5.2. The Loss of Built Fabric: Ruination and Demolition
The diachronic analysis on an urban scale has allowed us to reconstruct the transformation process of the historic urban fabric. This investigation highlighted the extent of the risks linked to the advancement of the deterioration process and to demolition practices. Based on the cartographic and photographic sources, we prepared a plan that summarised the evolution of the built fabric and presence of ruined buildings through a few significant phases (
Figure 7). Consultation of aerial photographs taken over the years made it possible to identify all buildings affected by the collapse of more than 50% of the roof. Maps indicate these conditions one by one.
The starting point for the analysis of the ruination process is the cadastral map of 1878. The critical restitution of this map on the basis of the current cadastre led to the recognition of the built areas that have retained an unaltered conformation until today. It is reasonable to assume that at this date, the neighbourhood was very dynamic and free of derelict areas and ruins. The 2001 census already showed that, similarly to what was found for abandonment, ruins also tend to be concentrated in certain areas, forming clusters. In particular, the upper part of the neighbourhood shows a conspicuous presence of ruined buildings that are concentrated particularly at the end of the blind street. Similarly, the aggregate at the bottom, also a dead-end street, approaches becoming a second cluster of ruins alternating with heavily structurally compromised buildings. The condition observed in 2013 clearly appears to have worsened. On the one hand, the number of ruined buildings increases in the vicinity of the two areas already affected by the phenomenon; on the other hand, the process of ruination advances, involving new areas. The snapshot to 2023 shows the extent to which the steady growth of ruined buildings is now a blight on the neighbourhood.
A comparison of the maps shows that, in addition to the rapid advancement of ruination, demolition is practised. Some areas are thus threatened not only by the rapid increase in ruins but also by the progressive (and definitive) loss of built fabric. The aerial photographs show that the collapse of even partial roofs most often constitutes a point of no return; in such situations, recovery or reconstruction are rare, and the building starts a process of incremental decay only mildly counteracted by the placement of a temporary covering. The next step is the total or partial demolition of unsafe parts justified on public safety grounds. The ordinances issued in recent years by the Municipality for reasons of public safety oblige owners to make buildings safe by demolishing unsafe parts and reinforcing the remaining structures. In actual practice, these interventions have been carried out without any particular care and even using mechanical means; some building units with serious structural instability have suffered such damage and collapse that they have (more or less justifiably) led to their total demolition.
In addition to demolition, there is an alternative way of securing the ruins. Following their partial demolition, the buildings have been filled with rubble and waste; they have then been used as supports for terraces and belvederes or as car parks or have been transformed into private vegetable gardens. In this sense, the intervention carried out between vico Vitale and via Risicato is emblematic (
Figure 8). The volumes of the houses on the slope, now devoid of roofs, were filled with soil and rubble, following the precarious securing of the facade walls downstream. This intervention is a symptom of an inappropriate way of dealing with the problem of ruins. In fact, in addition to definitively erasing the traces of pre-existing buildings, it is an inadequately designed consolidation operation that does not even ensure a safe response to the new function of being a retaining wall.
5.3. Drivers of Ruination
Following the diachronic reading, the survey of all ruined buildings inventoried allowed us to typify the phases of ruination (
Figure 9). We schematised it into seven damage configurations, expressing the progressive degeneration of the building in ruin until it reached the state of collapse and total ruination. The reconstruction of this process considered two main parameters that are easily detectable from the outside: the extent of the collapse of the roofs and walls. For the more advanced stages, we added another parameter concerning the collapse of internal structures and floors.
The first configuration identified is LR0 and concerns buildings affected by deterioration and minor roof or wall collapses, e.g., at openings made close to the guttering line. The on-site analysis made it possible to identify several forms that may be part of the initiation phase of the ruination process. The shapes identified may vary depending on the construction sequence as well as the inherent vulnerabilities [
81,
82].
The next phase, referred to as LR1, involves a first extension of the collapse of the roof up to 50%. Phase LR2 follows, with the total loss of the roof.
Phase LR3 involves the first collapses of the perimeter load-bearing walls, which is followed by phase LR4, in which more extensive collapses of the walls and the total collapse of the internal floors may also occur. It is worth pointing out, however, that the collapse of floors is a mode of damage that can also affect buildings that still have roofs. These are most frequently metal floors; only occasionally has the collapse of the barrel vault been verified. Phase LR5 is characterised by a very extensive collapse, which may affect the reading of the structural and formal configuration. Finally, phase LR6 consists of a definitive compromise of the configuration.
The extensive qualitative investigation of the built system and its state of preservation allowed for the identification of ruination exposure factors. They may not only depend on the building’s intrinsic vulnerabilities and physical condition, but also on the interaction between contiguous units and the degree of proximity to a ruin. The aggregative modalities typical of masonry architecture involve the construction of building units side by side without interruption and with the presence of at least one shared wall. Such structural contiguity inevitably conditions the mechanisms of degradation and instability triggered by the process of ruination of a building on its neighbouring units.
The study of the evolution of the phenomenon shows that these mechanisms may depend on the construction sequence or be amplified by the presence of intrinsic weaknesses (
Figure 10). The position of the openings made close to the masonry ridge represents a vulnerability. This results in a thin masonry portion that constitutes a weakening of the top of the façade. The position adjacent to a ruin can increase the risk of collapse of the top portion of the wall in case of loss of the roof. The extensive collapse of roofs may also weaken the shared wall and induce further collapses of the roofs or floors of adjoining cells. These mechanisms also occur in cases where the extensive collapse of the façade wall leads to the expulsion, even partial, of an incorporated cornerstone, triggering the collapse of a wall portion of the contiguous building. Moreover, as the collapse in the first unit progresses, the adjacent unit will be exposed to an increasing risk of collapse with respect to the shared wall.
The aerial photographs show the progression of the phenomenon just described. This is clear in the aggregate along via Cassarà. The extensive collapse of the roof of the corner cell, which acts as a link between the two parts of the aggregate, was followed by the collapse—at first minor and then extensive—of the roofs of the adjacent cells (
Figure 11).
The analysis of abandonment and ruination has highlighted some conditions related to the criticality of living in the neighbourhood that may have influenced the acceleration of these processes. In addition to the state of maintenance, intrinsic vulnerabilities, and type of interaction with neighbouring units, it is possible to detect an intrinsic propensity for abandonment for some building units. Given that the determining causes of the decline of these areas depend on a multiplicity of external factors, these processes would also be supported or amplified by reasons internal to the very nature of the buildings, which would be linked to some settlement and architectural features.
The first noticeable condition is the marginalisation of specific areas due to difficult accessibility.
Figure 12a shows that the more accessible building units tend to be more inhabited, while the less accessible are generally empty, abandoned, or in a state of ruin.
Figure 12b highlights how the relationship with the slope can influence the condition of use of the building unit.
Figure 12c,e show a prevalence of elevation and planimetric transformations in building units in use, while those that are transformed are in a state of abandonment or ruin.
Figure 12d highlights the relationship between the contiguity to ruins and the abandonment or ruination of building units.
On the basis of these analyses, it was possible to identify the following determinants of the phenomenon of ruination: characteristics linked to the building type and its degree of transformation, type of relationship with the morphology of the site, level of accessibility and use, type of interaction between adjoining units, and the condition of conservation of building components. The following paragraph will highlight the hierarchical scheme of factors favouring the ruination.
5.4. Database of the Characteristics of Building Units Exposed to the Ruination Process
The results of the observations and analyses have led to the construction of a database. It collects information regarding the general characteristics of the building unit (analysis of the configuration and typology with the relative transformations, analysis of the use, and accessibility) and the state of conservation of the structural elements and building components.
The database, contained in an Excel framework, highlights for each building unit the possible state of ruin, specifying its severity levels, and offers a specific description of the drivers previously identified (
Table 1).
Two damage thresholds indicate the level of ruination: the extent of collapse for roofs and walls. A score of 1–6 defines each ruination level of the by expressing a damage level of increasing severity, i.e., score 1—total roof collapse (100%) and wall collapse of 70%; score 2—total roof collapse (100%) and wall collapse of 51–70%; score 3—total roof collapse (100%) and wall collapse of 21–50%; score 4—total roof collapse (100%) and wall collapse of 20%; score 5—total roof collapse (100%) without wall collapse; score 6—roof collapse over 50% without wall collapse.
The factors related to ruination are declined through a hierarchical structure and are made explicit by a score indicating a level or type of the building characteristic (score 1-p) according to a range spanning from the most severe or problematic condition to the least severe or problematic condition.
Typological characteristics—period of construction: score 1—buildings before 1878; score 6—buildings after 2005;
Typological characteristics—B.U. layout related to the aggregate: score 1—isolated; score 2—corner; score 3—at the end; score 4—interlocked;
Typological characteristics—B.U. layout related to the proportions of the slope: score 1—substantial slope; score 2—steep slope; score 3—moderate slope;
Typological characteristics—B.U. position related to the slope: score 1—parallel position; score 2—perpendicular position;
Typological characteristics—plan changes: score 1—significant changes; score 2—medium changes; score 3—minor changes; score 4—no changes;
Typological characteristics—elevation changes: score 1—floor uplift; score 2—volume uplift; score 3—insertion of a terrace; score 4—insertion of a mezzanine/mezzanine; score 5—changes to the shape and position of the staircase;
Building use: score 1—abandoned; score 2—empty; score 3—use different from the original; score 4—in use;
Accessibility: score 1—difficult; score 2—medium-difficult; score 3—medium-easy; score 4—easy;
Elevation walls—No. of shared walls with contiguous B.U., No. of floors contiguous to ruins, level of contiguity to ruins, surface degradation, masonry slackening, out of plumb, expelled wall surface (cornerstones, wall covers), material disintegration: score 1—severe; score 2—medium; score 3—mild; score 4—absent. Walls collapse: score 1—more than 70%; score 2—70–51%; score 3—50–21%; score 4—20–1%; score 5—absent;
Floors: Floors collapse: score 1—100%; score 2—more than 70%; score 3—50–21%; score 4—20–1%; score 5—absent;
Roofs—Breaking of parts/water infiltrations: score 1—severe; score—medium; score 3—mild; score 4—absent. Roof collapse: score 1—100%; score 2—more than 70%; score 3—50–21%; score 4—20–1%; score 5—absent;
Openings: Reduced lintel: score 1—s < 70 cm; score 2—s = 70 cm. Efficacy of wooden lintel: score 1—not efficacious; score 2—efficacious;
Protruding volumes/elements—Gutter collapse: score 1—100%; score 2—more than 70%; score 3—50–21%; score 4—20–1%; score 5—absent.
5.5. Predicted Ruination Risk Scenario
The multivariate regression model defined in Equation (1) was used in order to investigate whether the data selected in the database identified a law for the ruination of the Granfonte neighbourhood in Leonforte. The analysed model is characterised by the dependent variable representing the level of ruination of the building unit ( and by the 24 independent variables representing the level of the building characteristics , as highlighted in the previous table.
The multivariate regression model was developed with the help of IMB SPSS statistical 27 software (Equation (2)).
The goodness-of-fit index of the model is expressed by the
value, which represents the proportion of the variance of
explained by the
explanatory variables. The variance explained by the model is 78.3% (
Table 2). The difference between
and
is modest and does not seem to suggest the presence of redundant independent variables.
The F-statistic used for hypothesis testing is significant for
p < 0.0001 (
Table 3).
The
t-statistics for the variables X
15, X
18, X
23, and X
24 all turn out to be associated with probability values of less than at least 0.05, so it can be concluded for them that all the regression coefficients are significantly different from zero; only the regression coefficient X23 has a probability of slightly more than 0.05 (
Table 4).
The most important independent variables are those highlighted with reference to the values of the standardised β coefficients. In particular, the independent variables in order of importance are X
18—Walls collapse, X
15—Out of plumb, X
21—Roof collapse, X
23—Wooden lintel efficacy, and X
24—Gutter collapse. The VIF statistic commonly used to detect the existence of multicollinearity in the regression model shows that all variables have a value ranging between one and five, which indicates a moderate correlation between a given predictor variable and other predictor variables in the model (
Table 5).
The analysis of the residuals, supported by the Durbin–Watson statistic, which was used to detect the presence of autocorrelation, shows a calculation value of 1.908, i.e., a normal value (
Table 6).
Equation (3) represents the ruination model of the building units for the Granfonte neighbourhood in Leonforte:
The estimated ruination model provides information on the current condition of building units and is able to provide a forecast scenario on the evolution of the phenomenon, highlighting the building units most exposed to future ruination. The model, therefore, can support the analysis of the risk of ruination of building units.
The most important independent variables identified on the basis of the values of the standardised β coefficients, i.e., with reference to the ruination drivers detected in the building units, X18—Walls collapse, X15—Out of plumb, X21—Roof collapse, X23—Wooden lintel efficacy and X24—Gutter collapse, can support the identification of the units whose condition may worsen in the future or which may not yet be affected by the ruination phenomenon and for which it may be activated.
Figure 13 takes up the study of the evolution of abandonment and ruination, proposing a comparison between the current state of ruination of building units and the forecast state as generated on the basis of the estimated model results.
The model shows a prediction scenario in which the state of ruination of some building units also extends to adjacent ones, i.e., the cases highlighted with the hatching in
Figure 13. This scenario shows how some building units that are not currently affected by the ruination phenomenon will be affected in the future, i.e., the cases highlighted with a white square in the same figure.
This scenario highlights the risks that depend on proximity to a building unit in a state of ruin, and overall, it shows a progressive extension of the ruination phenomenon into the historical fabric.