Chile extends from latitudes 17°30’S to 56°30’S, is approximately 4,200 km long and 117 km wide with a surface of 756,096 km², and has an estimated population of just over 17,8 million by 2014. From north to south, Chile can be roughly divided into five macro-regions: the arid Great North (Norte Grande), the semiarid Little North (Norte Chico), Mediterranean climate Central Zone (Zona Central), the rainy Southern Zone (Zona Sur), and the chilly and wet Austral Zone (Zona Austral). From east to west, four zones are identified: the Andes (Cordillera de los Andes), central valley, the Chilean Coastal Range (Cordillera de la Costa), and the coastal plain. The main economic activities of the country are copper mining, agriculture, fishing, and forestry.

Two different approaches are described to construct an exposure model for Chile: using statistical data to construct a national exposure model for the whole country, and using remote digital surveys to construct a detailed exposure model for three specific cities. The latter was done to validate the national exposure model as this model was built with information obtained from census data and other public databases.

From the available information in the census and public databases, and from the information we were capable of collecting when conducting the remote digital surveys, 18 typologies are considered in both National and Detailed Exposure Model. The 18 typologies corresponds to:

• Four for reinforced concrete structures
• Ten for masonry structures
• Two for timber houses
• One for adobe houses
• One for informal or self-construction houses

The description of the 18 typologies for the exposure models in Chile and the related GEM taxonomies are:

To access to the complete report of the Exposure Model for Chile called Exposure Model: Buildings typology distribution at the census block level and replacement costs click in the following link: chile_exposure_model_building_distribution_and_replacement_costs.pdf

Three different sources at different resolution levels were used to build the national exposure model: the 2002 and 2012 Population and Housing Census, and the 2002-2014 Unique Edification Statistics Form (UESF).

2002 Census

Data from the 2002 census with a census block resolution is available at request from the National Institute of Statistics (INE, http://www.ine.cl/). Census-blocks are the smallest subdivisions of a municipality for census purposes. The census is a nationwide dwelling-per-dwelling survey, and the information is obtained from the response of a dwelling owner or an adult responsible of answering the survey. Therefore, non-experts provide technical information of the dwelling, such as exterior walls material. Since census data is collected per dwelling, it is easy to relate it to a structure when the dwelling is a house (usually, one house corresponds to one dwelling), but poses a major difficulty to relate dwellings in apartment buildings to a structure. Neither the number of stories per building nor the number of apartments per story are registered in the census. Hence, assumptions are required to relate dwellings to structures. The following data from the 2002 census was used to build the exposure model:

• location of the dwelling (census-block location with urban or rural classification)
• type of dwelling (house, apartment in building, emergency house, informal house, or other)
• exterior wall material (reinforced concrete or stone, brick masonry, structural panels or blocks, timber, adobe, recycled materials, or other)

The main drawback of the 2002 census is that it is outdated, and therefore, it does not represent the current building stock in the country.

2002-2014 UESF

The Unique Edification Statistic Form (UESF) is a database that contains all construction permits issued by Chilean communes between January 2002 and September 2014 for both new houses and apartment buildings, and modifications of existing buildings. However, only the information of new houses and apartment buildings was used. This information is aggregated at the communal level, and it is publicly available upon request at INE. The technical quality of this database, to classify the structure into a structural typology, is better than that of the census, since the construction permit is signed by the real state owner and the architect or engineer in charge of the construction project. This permit is reviewed and approved by the commune, or by an inspector from MINVU. In Chile, the correlation between issued construction permits and actually built structures is relatively good, so the construction permits are considered as a good approximation of structures actually built between 2002 and 2014. The information from the UESF used to build the exposure model was the following:

• location of the dwelling (municipality)
• number of structures (houses or apartment buildings) and dwellings per construction permit
• type of structure, (house or apartment building)
• number of stories of each structure
• exterior wall material (reinforced concrete, hollow clay brick, handmade clay brick, concrete block, timber, adobe, or others)
• percentage of main exterior walls material in all exterior walls.

The two drawbacks of the UESF data to construct the national exposure model are: (i) the database does not have information of construction permits issued before 2002, which could have been used to relate dwellings from 2002 census to apartment buildings; and (ii) it does not include information on emergency and informal constructions and settlements, which are usually built without construction permits, and are related to vulnerable inhabitants.

2012 Census

In 2012, a census was performed by INE. After its preliminary results –aggregated at communal level– were published, politicians and experts heavily criticized the methodology used for the census. An external expert review showed that the methodology led to mistakes in data collection, with a high national omission rate of responses of at least 9.3%. As a result, the 2012 census was declared as non-valid by the government, and the detailed data at a census-block resolution was not published. Given the lack of accuracy of this database, for buildings stock purposes it was only used to obtain the number of emergency and informal houses at the communal level.

The number of structures per typology of the National Exposure Model was estimated using a combination of the 2002 Census, the UESF, and the 2012 Census.

The proposed methodology uses the 2002 Census information to obtain the total number of dwellings per census block up to the year 2002. The key aspect of the methodology is to relate dwelling information from 2002 Census with the 18 proposed structures typologies. To establish this relationship, the story distribution (percentages) from UESF data for masonry and reinforced concrete buildings, building technology and masonry unit type distribution for masonry structures are extrapolated to Census 2002 data.

For structures built from 2002 to 2014, the proposed methodology uses the UESF data to estimate the number of structures for each building typology to complete the inventory. Since the UESF is available at communal level, the number of structures per typology at a commune was uniformly distributed in the 2002 census blocks. If there was no information available in the UESF data for a certain commune and a certain type of structure, the distribution at the regional level is considered to distribute the existing dwellings prior to 2002 in that particular commune. If there is no information available at the regional level in the UESF data, then the national distribution is considered for that particular commune.

The 2012 Census is used to estimate the number of emergency and informal houses, because these housing types are not thought as permanent. As this Census has the information at communal level, the total number of these types of dwellings per commune was uniformly distributed throughout the 2002 census blocks in that commune.

The methodology to estimate the number of masonry and reinforcend concrete buildings, and masonry houses is explained in sections 2.2.1, 2.2.2, and 2.2.3 of the Exposure Model: Buildings typology distribution at the census block level and replacement costs report, respectively. Appendix A, B and C of the report present examples to obtain estimated number of masonry buildings, reinforced concrete buildings and masonry houses in certain communes, respectively. For reinforced concrete, timber and adobe houses the methodology to estimate the number of houses is similar to the methodology used for masonry houses. The aforementioned sections and appendixes are included below.

section_2.2.1_number_of_masonry_buildings.pdf

section_2.2.2_number_of_reinforced_concrete_buildings.pdf

section_2.2.3_number_of_masonry_houses.pdf

appendix_a_example_for_masonry_buildings.pdf

appendix_b_example_for_rc_buildings.pdf

appendix_c_example_for_masonry_houses.pdf

The average occupants per type of dwellings (i.e. RC apartments, RC houses, masonry apartments, masonry houses, timber houses, adobe houses, emergency houses and informal housing) in each commune was obtained using 2012 Census data. Then, the total occupants per typology can be estimated by multiplying this average with the number of dwellings of the exposure model.

The average floor area for structures of each typology was estimated using UESF data for each commune, because census data does not provide information on floor areas of dwellings. For multi-story buildings, the average floor area considers the built area of all stories. For emergency timber houses, and for informal houses an average floor area is not estimated because the total cost of these structures are estimated directly with the number of structures.

Ministry of Housing and Urbanism (MINVU) defines the costs per unit of area for five quality categories of residential structures: from superior to inferior.

For the National Exposure Model, it is not possible to calculate the category for each structure based on any of the statistical data used. Typical structures were considered for the main building materials (reinforced concrete, masonry, timber and adobe), and it was possible to establish the best and worst qualities for each type of structure with its reposition costs related.

As the communes in the whole country does not have the same reposition costs, it is necessary to differentiate costs between them. The Internal Revenue Service (SII) has information of base appraisal values for residential structures with the same five categories than MINVU. This base appraisal value is modified by four factors dependent on the structure’s location, special conditions of the structure, depreciation, and a commercial coefficient applicable to structures built in commercial zones. The factor that depends on the structure’s location is assigned for each commune, from 1.0 to 0.4. To estimate the replacement cost of structures in each commune, the cost associated to best quality of construction was assumed for structures in communes with factor 1.0, and the cost associated to worst quality of construction was assumed in communes with factor 0.4. For other communes, the replacement cost is extrapolated according to its factor.

Replacement cost for emergency and informal houses are not defined by MINVU. Then, it was assumed to be the same as the construction cost of a new emergency house: an average of USD 1,280. Replacement costs for informal houses are considered the same as for emergency houses because after an earthquake, damaged informal houses would be replaced with emergency houses by the authorities or helping institutions.

A total of 4.26 million residential structures, with 5.3 millions of dwellings, were identified in the National Exposure Model. The participation of each construction material in the total number of structures is:

To see more detailed information about the number of structures and dwellings per typology click here

To access to the National Exposure Model database at census block level click here.

Approximately a total of 17.1 million of inhabitants were estimated for residential structures according to the Natiomal Exposure Model. The participation of inhabitants per typology is:

To see more detailed information about the inhabitants per typology click here

Approximately a total of 396 million m² are built for residential structures according to the National Exposure Model. The participation of each typology in the floor area built is:

To see more detailed information about the surface built per typology click here

More than 108 billion USD corresponds to the total reposition cost for all the residential structures in Chile. The participation of the reposition cost per typology is:

To see more detailed information about the reposition costs per typology click here

The participation in the Global and Detailed models of each typology in the total number of structures per city is the following:

... see more detailed information about the number of structures per typology

…page under construction …