PSI - Issue 84

Kelly Camarena et al. / Procedia Structural Integrity 84 (2026) 999–1006

1002

class S3 and exposure class XS2/XD2; Ordinary Portland cement with standard mix composition and a service life target: 100 years. Table 1 presents the values for the random variables considered for chloride and carbonation induced corrosion and their respective statistical distribution, while Table 2 list the constant parameters employed.

Table 1. Random Variables Mechanism

Parameters

Units

Value (1)

Statistical distribution

Chloride-induced Concrete cover

mm

u = 35; s = 8

Normal Normal

Temperature

°C

(2)

Diffusion Coefficient

mm 2 /year

u = 63.093; s = 12.62

Log-normal

Carbonation induced

Concrete cover

mm

u = 35; s = 8

Normal

CO 2 concentration Regression parameter ( ) - Error term ( )

kg/m

u =0.0082 + Ci (3) ; COV = 0.12 Normal

3

u = 1.25; s = 0.35 u = 315.5; s = 48

Normal Normal

(mm²/years)/(kg/m³)

Notes: 1.

u: mean, s: standard deviation.

2. The mean and standard deviation for the temperature varies across Veneto, data obtained from ARPAV and INEMAR Veneto (2021). 3. The value of 0.0082 kg/m³ represents the background CO₂ concentration, as defined by the fib Model Code, while Ci denotes the CO₂ concentration corresponding to each municipality

Table 2. Constant parameters Chloride-induced

Carbonation-induced

Parameters

Units

Zone 1 Zone 2 Zone 3 Parameters

Units

Value

Critical chloride content ( Ccr )

wt.-%/c

0.9 0.3

0.9 0.9 Reference relative humidity (RHref) 0.3 0.3 Environmental function exponent (fe) 5 7.76 Environmental function exponent (ge)

%

65

Water-binder ratio ( w/b )

- - - -

- -

5

Relation between Cs and w/b ( A c )

2.57

2.5

Curing factor ( k c ) Ageing exponent ( a ) Reference time ( t 0 )

0.79 0.79 0.79 Reference time (t0) 0.65 0.50 0.37 Time of wetness (ToW)

years

0.0767

- -

0.26

years 0.0767 0.0767 0.0767 Probability of driving rain (p_SR) K 293 293 293 Inverse effective carbonation resistance (R ACC (-1) )

0.1

Reference temperature ( T ref )

(mm²/years)/ (kg/m³)

977.67

Environmental variable ( b e )

K 4800 4800 4800 Period of curing (tc)

days

28

3.1. Probabilistic assessment and spatial analysis

Given the often-skewed nature of service life data, the results were fitted to a lognormal distribution, a model frequently employed in probabilistic risk assessments for its theoretical and practical advantages. The results are expressed in terms of key percentiles being the 5th percentile a representation of a high probability of early failure (pessimistic scenario), 50th percentile that represents median service life and is considered the most probable outcome and 95th percentile representing an outcome of high durability (optimistic scenario). The percentiles were estimated using the following expression: =exp( ( ) + −1 ( ). ( ) ) . For the results analysis, intensity maps representing the time to corrosion initiation will be produced using GIS tools, enabling a clear visualization of spatial variability and facilitating the identification of the most vulnerable zones within the Veneto region.