**5. Flood risk assessment**

### **5.1 Vulnerability module**

The on-site and remote surveys allowed for the classification of the parameters affecting flood vulnerability, according to the methodology described in Section 2. **Figures 6a, b** and **7a, b** show the frequency distribution and the spatial distribution concerning, respectively, two of the most relevant parameters that compose the flood vulnerability index: the wall orientation, which determines the exposure component of the vulnerability, and the condition of the building, which falls under the sensitivity *Flood Risk Assessment in Urban Areas: The Historic City Centre of Aveiro as a Case Study DOI: http://dx.doi.org/10.5772/intechopen.109867*

**Figure 6.**

*Frequency distribution of (a) wall orientation and (b) building condition parameters.*

**Figure 7.** *Spatial distribution of (a) wall orientation and (b) building condition parameters.*

component. Regarding the former, all the investigated buildings fall into two categories, namely, "partial exposure with openings" or "full exposure with openings", with most of the buildings that present high level of exposure (class D) being located closer to the canal.

For what concerns the buildings' condition, their conservation state can be deemed as satisfactory. Almost 86% of the buildings assessed either are in a good condition (about 45%) or present minor conservation issues, mainly related with small cracks and moisture issues. The remaining 13% of structures face more serious degradation such as significant cracking, material loss, or settlement.

Regarding the heritage status, even though a great amount of the present building complexes are of architectural interest, only 3% of them are listed heritage with local/ municipal interest. To consider the intangible value of such assets, the methodology was modified, without altering its underpinning rationale, to include within category B the status of significant local heritage that was not included in the original formulation.

However, uncertainties arose for several buildings in the assessment of the remaining parameters, namely, construction material and age. This is due to the recent finishing that does not expose the original construction materials beneath and a lack of documentary evidence on the evolution of the building stock in the neighborhood. To address these uncertainties, several assumptions were made. Most of the

buildings with uncertain age were assumed to be built between 15th and 19th centuries, that is, category C. This option not only corresponds to the largest time range and, consequently, the highest probability for a building to be built during it but also is in accordance with the period of evolution of the Beira Mar neighborhood, completed during the 20th century (category B), which would be a less conservative assumption. On the other hand, buildings dating back to before the 15th century (category D) as well as to the 21st century (category A) were easily identifiable, due to their morphological and structural characteristics. Based on the aforementioned assumption, 81% of the buildings were classified as being built between the fifteenth and twentieth centuries.

Regarding the material, masonry and earthen constructions were often hardly distinguishable. Therefore, two cases were conceived. Case 1 has all uncertain construction materials classified as B (Masonry), while Case 2 has all uncertain construction materials graded as D (Earth). For the flood vulnerability assessment, Case 1 is selected, as it is more realistic based on the results obtained from the on-site survey.

Upon the classification, a clear predominance of masonry buildings emerged, accounting for 67% of the assets, assuming them as made of stone or bricks in the first scenario and of adobe in the second one. Nineteen percent of the buildings made of reinforced concrete are newer constructions.

Following this, the geographical distribution of the index is presented in **Figure 8**. The highest FVI value obtained for both the scenarios is 64.6. While this value is not negligible, in both cases, a very small percentage of buildings present a medium or high vulnerability.

Comparing the assessment of two different scenarios, it can be observed that Case 2, as expected, results in a higher vulnerability. Nonetheless, the overall vulnerability is slightly sensitive to the uncertainty in the classification of the material, as demonstrated by the very similar results both in terms of FVI values and distribution. Therefore, only Case 1 is considered for the risk assessment hereafter. The vulnerability is rather affected by the exposure component, with the most concerning buildings located along the canals.

Finally, the buildings are divided into three categories, namely, low, moderate, and high vulnerability (**Figure 9**). Index values of 20 and 40 are set as thresholds [14].

**Figure 8.** *Flood vulnerability index results: Distribution over the study area Case 1 (a) and Case 2 (b).*

*Flood Risk Assessment in Urban Areas: The Historic City Centre of Aveiro as a Case Study DOI: http://dx.doi.org/10.5772/intechopen.109867*

**Figure 9.**

*Distribution of flood vulnerability levels over the study area (a) and histogram with respect to the number of buildings (b).*

#### **5.2 Hazard levels**

The hazard component of the flood risk assessment was analyzed in collaboration with the University of the West of England, which provided flood extents, velocities, and water depth maps associated with return periods of 20 and 100 years—that is, frequent and rare flooding events, as per the Portuguese Decreelaw no. 115/2010 of 22 October. From these maps, the flood depth and velocity affecting each building were estimated, transforming the overlaid data (**Figure 10a**) into building-specific data (**Figure 10b** and **c**).

Within the investigated area, the estimated flood extent for 20- and 100-year return period is the same. Similarly, the flood velocity is equal in both cases, corresponding to 0.05 m/s. The flood depth, instead, significantly changes in the two scenarios. Indeed, for the 20-year return period, the buildings will be exposed to an average depth of 0.93 m and for the 100-year scenario, to an average depth of 0.98 m. Out of the 228 buildings affected by the flood in the two scenarios, a water height between 1.5 and 2.5 m is expected for 26 building under the 20-year return period and 28 under the 100-year return period. This is a significant value, but it represents a low percentage of buildings (5%) in comparison to the total amount.

Water height (y) and velocity (v) are used to calculate a single hazard indicator as follows:

$$\mathbf{H} = \mathbf{y}(\mathbf{v} + \mathbf{0}.5) \tag{2}$$

Based on the classes of hazard, defined in the Portuguese Floods Directive, five hazard levels are considered, namely, negligible, low, moderate, high, and extreme as shown in **Table 1**.

According to the criterion adopted in this matrix-based analysis, the level of flood hazard throughout the study area does not alter much passing from the 20-year to the 100-year peak flow scenario. By observing the distribution of percentages, the results can be deemed as satisfactory, portraying a low level of danger. In terms of spatial distribution (**Figure 11a** and **b**), it is possible to identify two blocks that can be particularly affected: one in the northern part of the city center and one in the southern, always in the flood-prone areas near the canal.

#### **Figure 10.**

*Hazard levels: (a) flood depth hazard map for 100 years return period; (b) flood depth per building for 20 years return period; (c) flood depth per building for 100 years return period.*

#### **5.3 Risk analysis**

By deploying the aforementioned levels of vulnerability and hazard, the flood risk matrix is obtained, as reported in **Table 2**, where the numbers represent the level of risk on a four-point scale.

*Flood Risk Assessment in Urban Areas: The Historic City Centre of Aveiro as a Case Study DOI: http://dx.doi.org/10.5772/intechopen.109867*


#### **Table 1.**

*Distribution of hazard levels for Aveiro.*

#### **Figure 11.**

*Spatial distribution of flood hazard levels (a) and values for the adopted 20-year peak flow scenario (b).*

**Figure 12** shows the risk maps for the two return periods. The flood risks for the 20-year return period (**Figure 12a**) and 100-year return period (**Figure 12b**) are remarkably similar. In a total of 495 surveyed buildings, 77.37% of them present low flood risk; 12.53 and 12.73% for 100 and 20 years, respectively, moderate; and 10.10 and 9.90% for 100 and 20 years, respectively, high. It is noted that for the 20-year return period, the moderate risk is slightly higher, raising more alarm for a dangerous situation in the near future.

It is observed that the buildings with a significant risk are located closer to the canal; they can be again distinguished into two blocks, and also, a significant number of them are located along one of the main streets, R. João Mendonça. Indeed, the


**Table 2.** *Flood risk matrix.*

**Figure 12.** *Spatial distribution of the flood risk results for 20 years (a) and 100 years (b).*

**Figure 13.** *Flood risk per building for a 20-year return period.*

buildings likely affected by the flood hazard, in the two return-period scenarios, are also the most vulnerable within the investigated area (**Figure 13**).

Overall, even though a percentage of about 10% of the buildings can be identified as being at high risk in the next decades, it is concluded that the flood risk for the Beira Mar neighborhood is relatively low, and in view of future mitigation strategies, the effort can be concentrated on the buildings located within the identified flood-prone area.
