The European project Climate for Culture, which investigated the potential impact of climate change on Europe’s cultural heritage assets, particularly on historic buildings, provided high-resolution risk maps that identify the most urgent outdoor risks for European regions until 2100 but also risks for indoor collections [
13]. These maps are the output of climate change scenarios coupled with building simulations at the European scale and serve as a powerful tool for preventive conservation and decision-makers that deal with cultural heritage [
19]. The maps, through the colour codes, show the level of risk, both for outdoor and indoor environments, for 16 building types and 19 environmental variables. The results of the project can be used to understand how the climatic changes affect the buildings in natural conditions (without the use of indoor heating, ventilation, and air conditioning (HVAC) systems) in relation to their geographical location, building size, window size, and constituting material.
2.1.1. Description of Buildings and Climatic Data
In the CfC project, the general assessment and map creation process has been carried out using the following specifications:
• Emission scenario
The impact of climate change on historic buildings was evaluated using the high-resolution regional climate model REMO [
20] which provides climate change projections for entire Europe at 12.5 km spatial resolution. Two emission scenarios were applied in the project. The first is the mid-line A1B scenario [
21], which considers a CO
2 emission increase until 2050 and a decrease afterwards. The second is the more recent Representative Concentration Pathway 4.5 Emission Scenario (RCP4.5) of the Intergovernmental Panel on Climate Change (IPCC) assessment report 5 (AR5) [
22]. This scenario is based on long-term, global emissions of greenhouse gases, short-lived species, and land-use-land-cover, which stabilizes radiative forcing at 4.5 watts per metre squared (approximately 650 ppm CO
2 equivalent) in the year 2100 without ever exceeding that value.
• Locations
Climate data assessment and simulations were calculated for a regular grid that covers entire Europe including the Mediterranean region.
• Time windows
The climate data were produced for all the climate-induced variables from hourly data elaboration over two 30-year time windows: 2021–2050 (Near Future) and 2071–2100 (Far Future), maintaining the period 1961–1990 (Recent Past) as a reference period (
Table 1).
• Future indoor climates and risk assessment
The outdoor climate influences the cultural heritage structures, both in terms of outdoor and the indoor environment [
13]. The future climate predictions explained above were used to create the risk maps for the outdoor environmental variables, which provide important information for decision makers to plan outdoor adaptation measures. These climate change predictions linked with building simulations allow the estimation of indoor climate variables (temperature
T, relative humidity
RH) and indoor damage variables for mechanical, chemical and biological decay using an automated method [
23]. The risk induced indoor by climate change is assessed by the combination of indoor climate data with the damage functions of the variables [
24].
• Buildings
Indoor climates of historic buildings were modelled and simulated following two different approaches. The first consisted of the development of a full-scale multizone dynamic hygrothermal whole building simulation while the second used a simplified hygrothermal building model. The first model gives more detailed results about the temperature and relative humidity inside the building, but it has a high development cost and takes long simulation time. The simplified model provides reliable results within a short simulation time and for this reason, it was applied in the CfC project to predict indoor temperature and relative humidity. It has the restriction to be effective to buildings without active HVAC systems and to request all the necessary measured values for the parametrisation of the model [
13]. Through this model, it was possible to perform simulations on 16 generic sacred buildings, virtually located in all the grid cells, for producing indoor climate data and risk maps. The general layout of buildings is composed of a rectangular floor plan, a gable roof, and long walls in the North-South direction with windows only on the long walls. Each of the buildings is unconditioned and their matrix is a combination of their volume (small/large), window area (small/large), structure (heavyweight/lightweight), and moisture buffering capacity (MBP) (low/high) as given in
Table 2.
2.1.2. Indoor Deterioration Variables
The variables, according to the CfC results, for each mechanism of indoor deterioration (mechanical, chemical, and biological) and assigned to different building materials (wood, masonry, and concrete), are reported in
Table 3 and
Table 4 [
25].
A short explanation for the indoor decay-linked mechanisms, according to the CfC deliverable D4.2 [
24], is given as follows:
• Mechanical risk for wooden elements: panels, jointed elements, cylindrical elements
The
RH of the air affects the moisture content (MC) in a wooden element. As the moisture content changes, so do the dimensions of the element, which set up internal stresses that lead to deformations. At low stresses, the wood behaves elastically, with reversible deformations while above a certain threshold of strain (the yield point), the deformation becomes plastic, the change is not reversible anymore and the material fails. The damage functions used in CfC for this type of elements are based on Marco Martens’ interpretation [
26] of studies by Mecklenburg, Bratasz, and Jakiela [
27,
28,
29]. Different response times are used in the algorithm to smooth out the
RH fluctuations in order to represent better the moisture changes experienced in the substrate of different building elements in wood. The strains induced by the expected changes are calculated and a final assessment is made to evaluate if the resultant strain falls in the area of elastic (green code), plastic (orange code), or failure (red code) response.
• Mechanical risk for masonry and concrete
Salt-crystallization cycles. Damage from salt crystallization occurs at the interface between air and the object, or beneath the surface of the object. The surface gets covered by a mass of small crystals that destroy the visual integrity or disfigure the natural appearance of masonry or concrete. When this occurs below the surface, the visible result is surface disruption and loss of material. The damage function for stone weathering is studied from Grossi et al. [
30] and predictions in the context of climate change are discussed in the atlas of Noah’s Ark project [
31] and reported by Lankester and Brimblecombe [
32]. The damage function used in CfC for calculating the number of cycles counts the transition that occurs in a range around 75%
RH (independently from the temperature) as this is the threshold of deliquescence of the sodium chloride.
Thenardite-Mirabilite cycles. Similarly, the porous stone might be destroyed due to the pressure exerted during the transition from the thenardite (Na2SO4) to the mirabilite (Na2SO4 × 10H2O) that occurs with the inclusion of 10 molecules of water in the hydrated crystal. Mirabilite exerts a very high crystallization pressure on the porous wall causing the damage of stone. A pressure of about 10 MPa occurs when the RH increases across value described by a critical RH = 0.88 × T + 59.1. Repeated cycles may accumulate stress and in the long-term, they may cause severe decay. The damage function used in CfC counts the transition that occurs in the thenardite-mirabilite system and estimates a green code for up to 60 cycles; orange code for 60–120 cycles and red code over 120 cycles.
Freeze-thaw cycles. When water goes from liquid to solid phase within a porous masonry element or in a structural crack, it increases in volume, which can cause damaging stress. If this stress is repeated in a cyclic way, the brick or stone becomes weaker, and eventually delaminates and spalls. The theoretical background of freeze-thaw cycles is discussed by Camuffo [
33] and in the atlas of Noah’s Ark project [
31]. The damage function counts the number of cycles between
T < −3 °C (freeze) and
T > +1 °C (thaw) that occur in one year. The results of CfC maps indicate a green code for up to 30 freeze-thaw cycles, orange code for cycles between 30 and 60 and red code for more than 60 cycles during the year.
Frosting time is considered the total amount of time (in hours) during the year when the air temperature (outdoor or indoor) is below zero degrees Celsius. The effect of frosting time over cultural heritage materials has been studied by Camuffo [
33]. Separately, this variable is not helpful to predict material damage but it may serve as an indicator for further investigations. Frosting time can be a useful parameter in sub-zero temperature zones (many zones in the Scandinavian countries) where it determines the penetration risk of the ice front through the building wall. The level of risk according to CfC maps is estimated green for up to 2400 h/year, orange for frosting time between 2400 and 4800 h/year, and red for more than this value.
• Chemical risk
Lifetime Multiplier (LM) is the ratio between the predicted lifetime of the material subjected to the environmental conditions and the predicted lifetime at standard conditions of 20 °C and 50%RH. When LM > 1, the material will last longer than the standard conditions (green code) while for LM < 1, the rate of deterioration is greater and the lifetime shorter. The level of LM < 0.5 (half of lifetime), is defined as the threshold of high risk and is illustrated in red.
The calculation of the
LM for different types of materials is done using the Equation (1) derived by Michalski [
34]:
where
RH is the relative humidity [%],
T is the absolute temperature [K],
Ea is the activation energy [J mol
−1], and
R is the constant of gas (8.314 [J mol
−1 K
−1]).
In the calculations, the value of activation energy (the least possible amount of energy which is required to start a reaction) is considered 59.24 kJ mol−1 for wood and 42.5 kJ mol−1 for masonry and concrete. The values are taken as average because the activation energy can vary for a different range of materials. The equation does not consider the effects of very low or very high RH but it can be a good indicator of the decay rate if the LM will increase or decrease in the future.
• Biological risk
Mould growth is an extensive problem that implicates the human health and the integrity of the material. The effects on heritage items can vary from light powdery dust to severe stains, which weaken and disintegrate the substrate material. It is assumed that at temperatures above zero degrees Celsius and relative humidity above 70% the mould spores can germinate. The rate of growth depends on the climatic conditions, type of material but also the accumulation of dirt and dust in case of inorganic materials. The CfC maps have been developed using the Sedlbauer isopleths system [
35] and they consider a growth rate of less than 50 mm/year as safe (green), a growth rate between 50 and 200 mm/year as possible damage (orange), and an annual growth rate greater than 200 mm as damage (red).
Insects can be another cause of damage to heritage items. The damage can be caused by certain moths and beetles and some forms of insects such as silverfish and booklice. The risk of damage from insects depends on relative humidity for some species and on temperature for most insect types. The key factors in assessing risk are climatic conditions, type of insect, and the vulnerability of the organic material such as wood. The insects’ activity is present in temperatures of 5–30 °C but below 15 °C, their damage is limited [
36]. The results for the CfC project have been achieved by calculating the annual degree-days over 15 °C ((days × (
T − 15)) with
RH > 75% and
T < 30 °C for humidity dependent insects and
T < 30 °C for temperature dependent ones.