Performance-Based Damage Quantification and Hazard Intensity Measures for Vertical Forest Systems on RC Buildings

Abstract

The European building stock is aging and needs renovation. Holistic renovation approaches, including Vertical Forest (VF) systems, are emerging as sustainable alternatives to demolition and reconstruction. This paper reviews and defines missing reliable damage and hazard intensity measures for the holistic renovation of existing reinforced concrete (RC) buildings with VF systems. Based on an extensive literature review and preliminary studies, including empirical multiparametric system evaluation assessments, Monte Carlo simulations, and System-Theoretic Process Analysis (STPA), combined structural, non-structural, vegetation, and human comfort components are examined. Key damage indicators are identified, including interstory drift ratio, residual deformation, concrete and reinforcement strains/stresses, and energy dissipation, and their applicability to VF-integrated structures are evaluated. Green modifications are found to have higher risk profiles than traditional RC buildings (mean scores from Monte Carlo method: 9.72/15–11.41/15 vs. 9.47/15), with moisture management and structural integrity as critical concerns. The paper advances the understanding of hazard intensity measures for seismic, wind, and rainfall impacts. The importance of AI-driven vegetation monitoring systems with 80–99% detection accuracy is highlighted. It is concluded that successful VF renovation requires specialized design codes, integrated monitoring systems, standardized maintenance protocols, and enhanced control systems to ensure structural stability, environmental efficiency, and occupant safety.

1. Introduction

As the majority of building stock in Europe has aged beyond its original design lifespan [1,2], there is a need for proper interventions. The assessment of building seismic vulnerability has gained importance in recent years, with researchers using various methods for preliminary evaluation by analyzing structural integrity and earthquake resistance [3]. The holistic demolition and reconstruction (DaR) of massive RC buildings may not be a viable solution in some cases, due to high economic cost, environmental impact, and social disruption [4,5,6]. Studies indicate that the DaR process relative to renovation could potentially have two to six times higher relative cost [4]. The large-scale demolition of RC stock may have a negative impact on the environment, despite the recycled nature of some materials, as the carbon footprint from demolition equipment, waste transportation, and production of new concrete is significant [5]. Furthermore, finding temporary housing and commercial spaces while maintaining community cohesion can be challenging [6].

Consequently, the seismic and energy upgrade of RC buildings is a more feasible solution and has been proposed by many studies [7,8,9,10,11,12]. Additionally, in the concept of reducing the carbon footprint, studies and building construction techniques have proposed the incorporation of Vertical Forest (VF) systems into existing buildings [13,14,15]. VF systems may include (see Figure 1) green facades (vegetation rooted in the ground and either using the wall itself to climb or independent support systems), green walls (type of living wall system where the plants, substrate, and structural support are directly attached to the building wall), green terraces and green roofs (plants growing on the horizontal terraces, which are built at different heights and levels), and vertical forests (groups of trees which grow on cantilevered balconies around the envelope of a building) [16]. Despite the beneficial impact of VFs, the characteristics of the systems have not been extensively examined as a holistic solution but more like separate subsystems [17,18,19].

Recent advances in artificial intelligence (AI) have shown promising applications in structural health monitoring and damage assessment. Convolutional Neural Networks (CNNs) have demonstrated significant capability in addressing missing measurement data recovery in structural monitoring [20], while hybrid approaches combining CNNs with Bidirectional Gated Recurrent Unit (BiGRU) and squeeze-and-excitation mechanisms have proven effective in reconstructing structural acceleration responses under varying environmental conditions [21]. Furthermore, innovative combinations of data augmentation techniques with adaptive optimization neural networks (DA-ASAPSO-CNN) have enhanced damage identification capabilities in steel bridges, particularly when dealing with limited data samples [22] Deep learning approaches, such as the combination of Deep Convolutional Neural Network (DCNN) with Long Short-Term Memory (LSTM) networks, have shown remarkable accuracy in modeling nonlinear temperature-induced structural responses, with a prediction accuracy of 99.8% [23]. These AI-based methodologies provide promising tools for enhancing the monitoring and assessment capabilities of vertical forest systems integrated with RC buildings.

The current review aims to properly define reliable damage and hazard intensity measures for holistic renovation of existing RC buildings, mainly in European region buildings from approximately 1920 onwards [1,2], incorporating vertical forest systems. More than one hundred high-quality technical publications were identified and screened through an exhaustive search of hazard identification and damage quantification to thoroughly evaluate the integration of VF systems on reinforced concrete buildings. A systematic search of scientific databases like Scopus, Google Scholar, and Web of Science were conducted within 1945–2025 range. Structural performance, monitoring systems, and risk assessment were part of secondary keywords, apart from fundamental keywords that were connected to vertical forests and reinforced concrete buildings. Peer-reviewed journal articles, conference proceedings, technical standards, and documented case studies that provided empirical data or rigorous analysis were of higher priority. Multidisciplinary literature that integrated structural engineering, environmental performance, monitoring systems, and well-being of individuals were brought within narrow focus. Publications were evaluated based on usability on VF systems, practical applicability, data quality, and methodological rigor, keeping real-world practice-oriented outcomes along with validated performance metrics of interest. Firstly, the most reliable global damage indicators for existing RC buildings are defined. Subsequently, the study separates the structural and non-structural components and then inspects the vegetation and forestry components, simultaneously. This review greatly considers the human factor, as ultimately the occupants’ comfort is a crucial factor that is sometimes neglected. In the Appendices, preliminary analyses of Monte Carlo for global behavior of the VF systems and System-Theoretic Process Analysis (STPA) for specific systems (VF balcony and vertical wall) based on the literature review and the empirically assessed multiparametric interrelations are presented.

2. Damage Measures and Methods

The research focused mainly on RC frames with infills. Based on the literature review, reliable global damage measures for performance-based seismic analysis commonly used by engineers include: Interstory Drift Ratio (IDR), Residual Deformation (RD), Concrete and Reinforcement Strains/Stresses (CaRSS), and Energy Dissipation (ED). These parameters are all measurable experimentally as well. While these measures can be applied to other categories with appropriate modifications, they are discussed in detail in the Structural Components section. The summarized results can be found in

Table 1. Frequently used damage measures and methods.

Damage Measure	Description	Applications	Measurement Methods
Interstory Drift Ratio (IDR)	A unitless quantity calculated as the relative horizontal displacement between adjacent floors (Δ) normalized by the story height (h). It relates elastic and inelastic structural behavior and defines performance levels.	- Performance-based design criteria - Threshold definitions for operational, life safety, and collapse prevention levels	- Linear Variable Differential Transformers (LVDTs) - Draw-wire sensors - Laser-Based Displacement Sensors - GPS technology (for high-rise buildings and long-term monitoring)
Residual Deformation (RD)	Permanent deformations in structural systems after seismic events, influenced by material nonlinearity, geometric changes, and cyclic deterioration.	- Probabilistic evaluations of seismic demands - Multi-objective optimization for robust structural systems - Post-event functionality assessment	- Advanced computational models (path-dependent constitutive behavior, geometric nonlinearities) - Probabilistic assessments
Concrete and Reinforcement Strains/Stresses (CaRSS)	Measures of deformation and internal forces in concrete and reinforcement, indicating potential failure zones (e.g., cracking, yielding).	- Ductility and damage evaluation - Identification of damage zones - Targeted retrofits	- Strain gauges - Fiber optic sensors - Wireless and non-invasive sensors (future innovations)
Energy Dissipation (ED)	The ability of a structure to absorb and dissipate seismic energy through mechanisms like plastic hinge formation, reinforcement yielding, and frictional sliding.	- Hysteresis loop analysis - Equivalent viscous damping ratios - Design frameworks (e.g., Eurocode 8) - Optimization of VF system	- Hysteresis loop analysis - Equivalent viscous damping ratios - Advanced materials and innovative designs for ED systems

.

These damage measures, among others, were subsequently utilized in advanced elaborations including empirical, Monte Carlo, and STPA methods to address, for the first time -to the authors’ knowledge-, combined effects related to VF cases. The methodology and results of these elaborations are presented in Section 4 and Appendix A, Appendix B and Appendix C.

2.1. Structural Components

The concept of IDR has been integrated into structural engineering for decades and is a fundamental aspect of performance-based seismic design. It is a unitless quantity (denoted often as d, sometimes defined as percentage, and calculated as the relative horizontal displacement between adjacent floors Δ, normalized by the corresponding story height h, as expressed in Equation (1):

d = Δ/h

(1)

This normalized quantity assists in the development of performance-based design criteria, since it directly relates elastic and inelastic structural behavior and enables threshold definitions corresponding to different performance levels. Standards like Eurocode 8 [24] and ASCE 41 [25], with some different notation, include levels ranging from Operational Level to Collapse Prevention that guide engineers during design and assessment analysis. Experimentally, the IDR of a structure can be measured with Linear Variable Differential Transformers (LVDTs), Draw-wire sensors, Laser-Based Displacement Sensors [26], or, for high-rise buildings and long-term monitoring, through the use of GPS technology.

The residual deformations in structural systems may not be properly assessed without consideration of material nonlinearity, geometric changes, and cyclic deterioration during seismic events [27]. Holistic prediction requires, among others, advanced computational models, path-dependent constitutive behavior, and geometric nonlinearities.

Subsequently, probabilistic evaluations are proposed for proper assessment of uncertainties in seismic demands, as the link between residual deformations and serviceability criteria for performance-based design is significant [28,29]. Additionally, in the concept of robust structural systems, multi-objective optimization algorithms have been extensively examined, e.g., Generic Algorithms (GA) [30], Sunflower Optimization Algorithm (SFO) [31], Particle Swarm Optimization (PSO) [32], and Simulated Annealing (SA) [33,34]. Furthermore, the post-event functionality and seismic performance methodologies are enhanced, as ultimately, the structural performance is a crucial factor that is sometimes neglected [35,36].

As the Concrete and Reinforcement Strains/Stresses (CaRSS) are significant indicators of RC structures under seismic loading [37], there is a need for proper measurements of deformations and internal forces. The holistic assessment of concrete and reinforcement may identify potential failure zones, among others, with signs of cracking or yielding, which are crucial factors for ductility and damage evaluation. Additionally, strain gauges, fiber optic, piezoelectric sensors, and acoustic emission techniques have been proposed as the real-time data monitoring of the behavior of materials, including retrofitted materials, is significant [38,39,40,41,42,43]. Furthermore, higher-order probabilistic methods like Gaussian Bayesian Networks have proven efficient in the structural system’s damage detection capability, thus enhancing identifications and location of probable zones of failure with statistical modeling approaches [44].

Subsequently, the thresholds for CaRSS are defined by design codes, despite the challenging nature of sensor installation in structures with vertical forest (VF) systems. The large-scale research indicates that CaRSS applications have a significant impact on identifying damage zones and supporting targeted retrofits [45,46,47]. Furthermore, future innovations in wireless and non-invasive sensors may enhance RC structure maintenance while satisfying the particular requirements of VF-integrated buildings [48].

As the Energy Dissipation (ED) in RC buildings plays a crucial role in structural behavior [49], there is a need for proper assessment of its ability to absorb and dissipate seismic energy. The holistic ED process may not be the ultimate solution without understanding the mechanisms, among others, of plastic hinge formation, reinforcement yielding, and frictional sliding, as they have a significant impact on reducing stress during earthquakes. The studies indicate that high ED capacity relative to conventional solutions could potentially have more relative seismic resilience [49,50]. The large-scale quantification through hysteresis loop analysis and equivalent viscous damping ratios has been proposed by many different studies [51,52].

Consequently, the implementation of ED in design frameworks like Eurocode 8 is a more feasible solution and has been proposed through correlation with reduced deformation and damage. Additionally, in the concept of VF systems optimization, proper ED assessment is required, despite the complex nature of added mass and vegetation dynamics. Furthermore, the incorporation of advanced materials and innovative designs into existing ED systems [53] enhances the optimization capacity while maintaining the VF benefits, as ultimately, the holistic consideration of both seismic performance and vegetation requirements is a crucial factor that is sometimes neglected.

2.2. Non-Structural Building Envelope and VF Façade Systems

The advanced analytical frameworks for cladding displacement and connector damage are proposed, as there is a need for proper modeling of interactions between structural deformations and façade responses. The holistic multiscale modeling techniques have been extensively examined, combining finite element analyses with global response models [54,55], as the precise predictions of, among others, damage mechanisms, clamping loss, and shear deformation, are significant. The study [56] indicates that probabilistic evaluations enhance connection designs, ensuring resilience, durability, and cost-efficiency.

Consequently, the seismic performance assessment of glazing and airtightness is a more feasible solution through finite element and fluid-structure interaction models. The large-scale integration of simulations may have a significant impact on thermal, acoustic, and structural properties [57], despite the complex nature of material brittleness and airflow patterns.

Additionally, in the concept of planter stability under seismic loads, nonlinear analyses and probabilistic methods have been proposed [58,59]. The optimization designs enhance resistance to seismic actions while maintaining aesthetic and maintenance requirements, thus satisfying both safety and functionality.

2.3. Vegatation and Forestry Components

Computer vision and Artificial Intelligence (AI) models have a significant impact in automatic and real-time plant health monitoring [60]. There is an extensive review on how to use different AI models for detecting and monitoring plants stress [61]. Throughout its lifespan, a plant on a daily basis undergoes different stresses from factors, among others, drought, temperature, and nutrients. Deep learning models like Convolutional Neural Networks (CNNs) [62] or Deep Convolutional Neural Networks (DCNNs) [63,64] can be used for image analysis and feature extraction. They are particularly effective in processing complex plant images and can detect variations in plant physiology. Traditional Machine Learning models like Support Vector Machines (SVMs) [65] can be used to classify types of plant stresses with accuracy as high as 99% in some cases. For water stress discrimination Random Forest (RF) [66] models can be used. For real-time and effective monitoring changes over time, Recurrent Neural Networks (RNNs) [67] can and has successfully been applied for water stress detection. It is important to mention that these damage monitoring models depend on the application and type of stress being monitored; the models can achieve accuracy rates between 80–99%. A summary of the content and additional references can be found in

Table 2. Summary of computer vision and AI methods for plant health assessment.

Model/Method	Main Purpose	Category	Reference	Performance Range % (Based on the Metric Used)
Computer Vision and AI (General)	Automatic and real-time plant health monitoring	General Approach	[60,68,69]	80–90%
Convolutional Neural Networks (CNNs)	Image analysis and feature extraction for plant physiology	Deep Learning	[62,70,71]	80–99.5%
Deep Convolutional Neural Networks (DCNNs)	Advanced image analysis and feature extraction	Deep Learning	[63,64]	79.8–99%
Recurrent Neural Networks (RNNs)	Real-time monitoring of changes over time, water stress detection	Deep Learning	[67]	38.8–92.3%
Support Vector Machines (SVMs) and Random Forest (RF)	Classification of plant stress types, Water stress discrimination	Traditional ML	[65,66,72,73]	65.3–97%

.

2.4. Occupants Satisfaction and Comfort Measures

The occupants’ satisfaction is an ambiguous measurement, as it directly relates to the human factor and primarily collects qualitative data (personal opinions of occupants) that often requires proper analysis to be interpreted into quantitative data for mathematical analysis, though this is not always necessary (e.g., Decision Trees). Initially, proper assessment of occupants through detailed questionnaires is proposed. Several validated questionnaires can be used, among others, Building Use Studies (BUS) Questionnaire [74]: measures occupant satisfaction systematically; Questionnaire on User Satisfaction (QUIS) [75]: originally developed for human-computer interactions satisfaction studies, but can be adapted to include built habitats. The latter must undergo validation and reliability testing to ensure that the survey accurately measures its intended outcomes. Different techniques are applied, among others, (a) Confirmatory Factor Analysis (CFA) [76]: Validates the factor structure of a questionnaire; (b) Cronbach’s Alpha [77]: assesses the internal consistency reliability of survey scales and defines how closely related a set of items are as a group; (c) Test-Retest [78]: evaluates the stability of responses over time. Afterwards, the data can be converted into quantitative data (if not already) using methods like (i) thematic analysis with numerical tagging that is systematic coding and categorization approach to identify patterns, (ii) rating scales based on qualitative descriptions that define numerical scales e.g., 1–5 to qualitative descriptions like “Very Dissatisfied” = 1 − “Very Satisfied” = 5, (iii) dichotomization simplifying qualitative data into binary categories (e.g., yes/no, present/absent). Then the data can be fed into statistical methods like Structural Equation Modeling (SEM) [79], Hierarchical Linear Modeling (HLM) [80], or even with Neural Networks and Deep Learning techniques [81] for further analysis. Summary of the content and additional references can be found in

Table 3. Overview of Methods and Models for Occupant Satisfaction Assessment and Analysis.

Model/Method	Main Purpose	Category	References
Building Use Studies (BUS)	Systematically measures occupant satisfaction	Assessment Tool	[74]
Questionnaire on User Satisfaction (QUIS)	Originally for human–computer interactions, adaptable for built environments	Assessment Tool	[75]
Confirmatory Factor Analysis (CFA)	Validates the factor structure of questionnaires	Validation Method	[76,82]
Cronbach’s Alpha	Assesses internal consistency reliability of survey scales	Validation Method	[77,83]
Test-Retest	Evaluates stability of responses over time	Validation Method	[78,84]
Thematic Analysis	Systematic coding and categorization to identify patterns	Data Conversion	[85,86]
Rating Scales	Converts qualitative descriptions to numerical scales (e.g., 1–5)	Data Conversion	[87,88]
Dichotomization	Simplifies qualitative data into binary categories	Data Conversion	[89,90]
Structural Equation Modeling (SEM)	Advanced statistical analysis of relationships	Analysis Method	[79,91,92]
Hierarchical Linear Modeling (HLM)	Multi-level statistical analysis	Analysis Method	[80,93]
Neural Networks/Deep Learning	Complex pattern recognition and analysis	Analysis Method	[81]
Decision Trees	Analysis of qualitative data without conversion needed	Analysis Method	[94,95,96]

.

3. Hazard-Related Intensity Measures and Methods

3.1. Seismic Intensity Measures

Traditional seismic demand assessment using intensity measures relies on strong mathematical models representing probabilistic correlations between characteristics of ground motion and parameters of structural response [97,98,99,100]. The Peak Ground Acceleration (PGA), despite its computational efficiency and widespread use, may not be sufficient for complex dynamic behaviors, multi-degree-of-freedom systems and higher mode contributions. The study from [101] indicates that Spectral Acceleration (Sa) at the fundamental period enhances response predictions in probabilistic frameworks. Additionally, in the concept of comprehensive assessment, Peak Ground Velocity (PGV) has been proposed for nonstructural components and façade systems through advanced signal-processing methodologies [102]. Subsequently, the proper understanding and quantification of these intensity measures requires systematic examination. Different approaches are presented, among others.

3.1.1. Seismic Accelerometer Recording and Microzoning

Basic research into seismic microzoning and site amplification is important for understanding earthquake hazards on both micro and macro scales. The Swiss Seismological Service has extensively examined detailed models of site amplification based on geophysical measurements and earthquake records [103]. The holistic models may include, among others, local site response knowledge at approximately 245 seismic stations and geospatial prediction methods for amplification factors across diverse areas.

Subsequently, the quantification of site-specific amplification is a more feasible solution through the integration of geophysically measured data and recorded ground motions. Studies indicate that techniques incorporating empirical spectral modeling methodologies have been successful in forecasting the observed site amplification factors at instrumented sites [104]. The large-scale interpolation of amplification may have a significant impact on larger areas despite the complex nature of site condition proxies such as topographic slope and geological classification.

Additionally, in the concept of site response analysis at the building level, a proper understanding of local soil conditions affecting ground shaking intensity has been proposed [105]. The multi-scale approach is a crucial factor that is sometimes neglected, as ultimately, the combination of instrumentation data with site characterization provides valuable insight into both regional hazard mapping and building-specific analyses.

3.1.2. DYFI and Q-Felt Maps

As the collection of earthquake impact data requires proper citizen science approaches, the USGS has developed the ‘Did You Feel It?’ (DYFI) system. Studies indicate that since its establishment in the 1990s, the DYFI system has extensively examined data collection methods, among others, through millions of individual intensity reports and digital questionnaires [106]. The large-scale production of Community Internet Intensity Maps may have a significant impact on macroseismic data resolution, despite the conventional nature of postal surveys.

Subsequently, the search engine query analysis for mapping felt areas is a more feasible solution. The study from [107] indicates that real-time monitoring of earthquake-related keywords could potentially have a significant impact on quick estimation of earthquake effects distribution. The Q-Felt Map system has been extensively examined through machine learning to assess the reliability of more than 1 million queries within 3 min following an earthquake, as the correlation with official intensity maps in populated areas is significant.

Additionally, in the concept of earthquake alert systems, proper incorporation of various data sources has been proposed [106]. The thorough integration of social and crowd-source information with conventional seismic observations is an important feature that requires further attention. In addition, an apt illustration can be taken from the case of the Swiss Seismological Service, describing an organized scheme of information regarding earthquakes through a blending of instrumental recordings, macroseismic observations, and quick impact evaluation tools [105].

3.1.3. Remote Sensing and Building Collapse Detection

Remote sensing has become a commonly used means for fast post-earthquake damage assessment, particularly in identifying collapsed buildings. The study by [108] investigated a total of 25 different remote sensing features while identifying collapsed buildings. The results showed that Homogeneity, Energy, Local Entropy, Local Standard Deviation, and Gradient were helpful in finding the rapid damage effect. They could apply such features to large-scale imagery with high recognition accuracy featuring F1-scores in the range of 0.71–0.94.

Deep learning approaches have achieved state-of-the-art performance regarding automated building damage assessment. For detecting collapsed buildings in post-earthquake remote sensing images, an enhanced YOLOv3 model was developed by [109]. By replacing a conventional Darknet53 neural network with a lightweight ShuffleNet v2 backbone and some other modifications in terms of the loss function, they attained detection speeds as high as 29.23 frames per second at 90.89% accuracy. Those were significant gains over the original YOLOv3 in both speed and accuracy. Similarly, there has been an evolution of computer vision-based frameworks with advanced models proposed for crack detection in concrete and dimension identification by a three-tier approach of deep learning with image processing to precisely estimate damage [110].

With reference to disaster response capabilities, the integration of detection technologies with established assessment methods presents significant advantages. The study from [108] indicates that framework effectiveness depends on various parameters that include, among others, image resolution and population density around the epicenter. Building on these results, their system detects structural failures within minutes after post-disaster image acquisition. In terms of performance validation, the work of [109] demonstrated that the improved deep learning approach maintains high accuracy under difficult conditions like fluctuating lighting and complicated backgrounds.

3.2. Hazard Intensity Measures for Environmental and Climate Aspects

3.2.1. Wind Intensity Considerations

The determination of wind impact for integral façade vegetation structures involves a thorough analysis with complex mathematical frameworks describing atmospheric boundary layer behavior and structural behavior, respectively. Successful estimation of wind loads is dependently determined with simulations of turbulence and application of statistical postprocessing techniques [111,112]. In addition, in relation to proper correlation factors, a thorough analysis of wind intensity factors and system performance has been developed, thus creating a sound determination between deterministic and stochastic atmospheric actions. Computational fluid dynamics (CFD) simulations can have a significant impact with regard to façade stability and vegetative behavior [113]. In addition, multi-criterion decision-support frameworks, utilized through air resistance analysis techniques, promote aerodynamic, structural, and vegetative viability [114,115].

3.2.2. Rainfall Intensity Analysis

The analysis of rain loads in integral buildings can fall short if not accompanied by a thorough consideration of complex mathematics capable of effectively describing the coupled behavior between hydraulic and structural dynamics. The proper integration of hydrological modeling with structural analysis has been proposed [116,117], as ultimately, the system responses under combined loading conditions are significant factors. Studies indicate that analytical frameworks incorporate, among others, time-distribution patterns and spatial variability through statistical methods [118,119]. Subsequently, probabilistic theories have been extensively examined for proper correlation between rainfall parameters and system performance. The multimodal assessment methodologies enhance drainage system quantification while maintaining the balance between infiltration phenomena and numerical simulations. Furthermore, the post-earthquake response evaluation requires systematic examination of environmental loading and system damage states.

4. Preliminary Studies Based on the Literature Review

Besides the remarkable reviews concerning combined structural and energy upgrades [120] or integrated comfort analyses [121], the subject of Vertical Forest analysis is highly interdisciplinary and multidisciplinary, and needs further investigation. To the authors’ knowledge, this is the first time the Vertical Forest Renovation is the central focus of the review, gathering and directly utilizing VF specific damage and hazard assessment methods. The scope and depth of the presented review manages to identify and evaluate VF-related critical parameters of the renovation process. From the literature review, it was concluded that proper questionnaires may provide relevant information for later application to complex mathematical and probabilistic models. Thus, three preliminary studies were conducted to support this review: (a) empirical multiparametric assessment through expert opinion table rating system (see Appendix A), (b) Monte Carlo analysis (see Appendix B), and (c) two specific domain and length STPA analyses (see Appendix C).

Subsequently, five different empirical aspects (damage assessment, analysis assessment, modeling assessment, retrofit assessment, and structural health monitoring assessment) were provided, identifying and rating different multiparametric limit states and the indirect contribution of different parameters to VF renovation. The rated risks range from 1 to 3 (Level 1 for allowable deteriorations without immediate action, Level 2 for further investigation requirements, and Level 3 for dangerous conditions requiring urgent response), producing three tables that allow for some preliminary, but very significant, quantification elaborations that further utilize the review outcomes. The study provides detailed tables in Appendix A (

Table A1. Expert assessment of damage causes and severity in Common Reinforced Concrete Buildings.

Common Reinforced Concrete Buildings			Levels					Total
Structural Performance
Causes of Damages		Damages	EO1	EO2	EO3	EO4	EO5
Material	Poor concrete quality	Cracking	2	2	2	3	3	12
		Reduced load-bearing capacity	2	2	2	3	3	12
		Surface spalling	2	2	2	3	3	12
		Permeability issues	2	2	2	3	3	12
	Corrosion	Reinforcement deterioration	2	2	2	3	3	12
		Loss of structural integrity	2	2	2	3	3	12
		Water leakage	2	2	2	3	3	12
Design	Non-compliance with the earthquake regulations	Structural cracking	2	2	2	3	2	11
	Short column	HRF of structural elements (columns, beams and joints), Loss of structural integrity during design EQ	2	2	2	3	2	11
	Strong beam- weak column	HRF during design EQ	2	2	2	2	2	10
	Insufficient shear capacity	HRF load-bearing capacity during design EQ	2	2	2	2	2	10
	Soft and weak story	HRF of Soft story failure, Joint failures during design EQ	2	2	2	2	2	10
	Weak attachment to the frames	HRF of elements like infill walls, ceilings and facades during design EQ	2	2	2	2	2	10
Reinforcement details	Insufficient reinforcement ratio	Reinforcement slippage	3	3	3	3	3	15
	Insufficient transverse reinforcement	Shear cracking	3	3	3	3	3	15
	Smooth stirrups	Loss of structural integrity	3	3	3	2	3	14
	Lack of crossties	Buckling of longitudinal reinforcement	3	3	3	3	3	15
	Lack of transverse reinforcement in connection region	Reduced load-bearing capacity	3	3	3	3	3	15
	90° hooks instead of 135°	Joint and connection failure	3	3	3	3	3	15
Others	Adjacent buildings effect	Collision damage, Structural instability, Cracking	2	3	2	2	2	11
	Insufficient determination of soil properties	Structural tilting, Cracks in walls and floors, Foundation failure	2	3	3	2	2	12
	Infill wall damages	Loss of lateral stability, Loss of lateral stability	3	3	2	3	2	13

HRF = High Risk of Failure/Partial or Full collapse/Falling, EQ = Earthquake, EO = Expert Opinion.

,

Table A2. Expert assessment of damage causes and severity in Greenery-Covered Buildings (Structural Performance).

Greenery-Covered Buildings		Levels					Total
Structural Performance
Causes of Damages	Damages	EO1	EO2	EO3	EO4	EO5
Root penetration (growth/action)	Wall and foundation integrity	2	2	2	2	3	11
	Potential instability issues	2	2	2	2	3	11
	Cracks	2	2	2	2	3	11
	Potential damage to technical features (e.g., pipelines)	2	2	2	2	3	11
Plants direct attachment/green layer removal	Wall and foundation integrity	2	2	2	2	2	10
	Lifted blocks	2	2	2	2	2	10
	Cracks	2	2	2	2	2	10
	Scarring	2	2	2	2	2	10
	Plaster removal	2	3	2	2	2	11
	Local wetness	2	3	2	2	2	11
High level of moisture (due to evapotranspiration)	Mold development	3	3	3	3	3	15
	Indoor and outdoor dampness	2	3	2	3	3	13
	Material covers damages	3	3	3	3	3	15
	Slow drying	3	3	3	3	3	15
High level of moisture (due to watering/irrigation systems)	Water leaks	3	3	3	3	3	15
	Water pooling	3	3	3	3	3	15
	Dripping water	3	3	3	3	3	15
	Indoor and outdoor dampness	3	3	3	3	3	15
	Mold development	3	3	3	3	3	15
	Corrosion	3	3	3	3	3	15
	Streaking	3	3	3	3	3	15
	Material covers damages	3	2	3	3	3	14
Soil moisture changes	Wall and foundation integrity	2	2	2	1	2	9
	Potential instability issues	2	2	2	1	2	9
	Subsidence	2	2	2	1	2	9
Soil microbial flora and organic acids	Concrete erosion	2	3	2	2	2	11
Overgrown vegetation	Overtaking space	2	2	2	2	2	10
	Blocking windows	2	2	2	2	2	10
	Cascading branches	2	2	2	2	2	10
	Technical features obstruction (e.g., drip grooves, pipes)	2	2	3	3	2	12
Additional weight and stresses (plants/planting media)	Potential instability issues (e.g., reinforced concrete planting balcony)	1	2	3	2	2	10
Fire	Structural integrity	3	3	3	3	3	15
	Burned materials	3	3	3	3	3	15
	Collapse	3	3	3	3	3	15
Wind	Structural integrity	2	3	2	3	3	13
	Cracks	2	1	2	3	3	11
	Collapse	3	3	2	3	3	14
	Blown out	2	2	2	3	3	12
	Debris	3	2	2	3	3	13
Earthquake	Structural integrity	3	3	3	3	3	15
	Cracks	3	3	3	3	3	15
	Collapse	3	3	3	3	3	15
	Spalling	3	3	3	3	3	15
	Instability	3	3	3	3	3	15

EO = Expert Opinion.

,

Table A3. Expert assessment of damage causes and severity in Greenery-Covered Buildings (Greenery performance).

Greenery-Covered Buildings		Levels					Total
Greenery Performance
Causes of Damages	Damages	EO1	EO2	EO3	EO4	EO5
High level of moisture (due to evapotranspiration)	Pests	3	3	3	3	3	15
	Degradation of planting media	3	3	3	3	3	15
Low level of moisture (due to watering/irrigation systems)	Plant stress	3	3	3	3	3	15
	Dry vegetation	3	3	3	3	3	15
	Dead plants	3	3	3	3	3	15
	Wilting	3	3	3	3	3	15
	Gaps	3	3	3	3	3	15
	Degradation of planting media	3	3	3	3	3	15
Improper maintenance	Overgrown vegetation	3	3	3	3	3	15
	Dead biomass	3	3	3	3	3	15
	Bio debris	2	3	3	3	3	14
	Gaps	2	3	3	3	3	14
	Degradation of planting media	3	3	3	3	3	15
	Loss	3	3	3	3	3	15
	Compaction	3	3	3	3	3	15
Inadequate space	Distorted vegetation growth	2	2	3	2	2	11
Environmental stressors (harsh sunlight, strong wind)	Plant stress	3	3	3	3	3	15
	Dehydration	2	3	3	3	3	14
Fire	Burnt vegetation	3	3	3	3	3	15
Wind	Uprooting	3	3	3	3	3	15
	Mechanical stress	3	3	3	3	3	15
	Dehydration	3	3	3	3	3	15
	Reduced plant growth rate	3	2	3	3	3	14
	Wilting	3	3	3	3	3	15
	Bended trees	3	3	3	3	3	15
	Falling bio debris	3	2	3	2	2	12
Earthquake	Planter anchorage and instability	3	3	3	3	3	15
	Tree pot anchorage and instability	3	3	3	3	3	15

EO = Expert Opinion.

and

Table A4. Expert assessment of damage causes and severity in Greenery-Covered Buildings (Human Welfare).

Greenery-Covered Buildings		Levels					Total
Human Welfare
Causes of Damages	Damages	EO1	EO2	EO3	EO4	EO5
Dead plants	Aesthetics	3	3	3	3	3	15
	Utility	3	3	2	3	3	14
	Energy needs (increased during summer)	3	3	2	3	3	14
Overgrown vegetation	Insects	3	3	3	3	3	15
	Aesthetics	2	2	3	3	3	13
	Energy needs (increased during winter)	2	2	3	3	3	13
	Utility	2	2	3	3	3	13
	Reduced sunlight/view	2	2	3	3	3	13
	Living space occupation	2	2	3	3	3	13
High moisture level	Health issues	3	3	3	3	3	15
	Pests	3	3	3	3	3	15
	Insects	3	3	3	3	3	15
	Mold development	3	3	3	3	3	15
Maintenance	Cost	2	2	3	2	3	12
	Time	2	2	3	2	3	12
	Potential falls	3	3	3	3	3	15
	Work at height	2	3	3	2	3	13
Non-native species	Ecological imbalances	2	2	3	2	3	12
Fire	Safety	3	3	3	3	3	15
	Evacuation obstruction	3	3	3	3	3	15
	Fire vertical propagation	3	3	3	3	3	15
	Chimney effect	3	3	3	3	3	15
Wind	Safety	3	3	3	3	3	15
	Fallen objects/bio-debris	3	3	3	3	3	15
	Turbulence	3	3	3	3	3	15
Earthquake	High repairing cost	3	3	3	3	3	15

EO = Expert Opinion.

) that document the potential causes and severity of damage in four key areas: (i) Common Reinforced Concrete Buildings, Greenery-Covered Buildings (ii) with focus on structural performance, (iii) with focus on greenery performance (iv) with focus to Human-Related Factors. These tables attempt to provide a thorough and systematic documentation of each category. The common buildings have been extensively examined for, among others, material-related issues, design defects, and reinforcing details. Additionally, in the concept of green buildings, the tripartite model has been proposed while maintaining the balance between structural performance, greenery efficacy, and human welfare, as root invasion, moisture management, and occupant safety are significant factors. The scoring mechanism through risk categorization enhances proper assessment, among others, low (5–8), medium (9–12), and high (13+) priorities (maximum risk value is considered 15), as ultimately, the maintenance planning and future designs are crucial factors that require systematic examination. Furthermore, the framework integration with traditional structural engineering and sustainability has been examined through multi-expert evaluations while maintaining reliability and consensus between Common Reinforced Concrete Buildings and Greenery Covered Buildings parameters.

The Monte Carlo analysis application in the current study aimed to assess expert evaluations, as the probability distribution of risk levels in different system subcomponents is significant. Such analytical methodology enables the identification of probabilities for each subcomponent falling into three predefined risk levels and, consequently, results in computing the average risk value for the whole system. The methodology converts discrete damage ratings into continuous intervals with partial overlapping: rating 1 corresponds to [0.0, 1.3], rating 2 corresponds to [0.7, 2.3], and rating 3 corresponds to [1.7, 3.0]. Additionally, this enables more detailed modeling of expert estimates and takes into consideration uncertainty in boundary regions.

The simulation run overall takes 10,000 runs to produce a reliable output in terms of statistics. During each run, the process samples randomly from the pool of ratings by every expert in each category for the various types of damages. These discrete ratings, having thus been sampled, undergo conversion into continuous values based on overlapping ranges and are subsequently summed up to develop a total damage score. The approach establishes a sound distribution of possible outcomes while maintaining the variability due to expert opinions. Furthermore, the test proceeds with deterministic calculations in which the risk level falls into three categories: low risk, between 5 and 8 points; medium risk, between 8 and 12 points; high risk, between 12 and 15 points.

For each type of damage, the simulation calculates statistical indicators, among others, average scores, standard deviations, and 95th percentile values. These measures provide thorough information about central tendency and variance in terms of assessments of risk, both for normative and extreme case identifications. The last step represents the consolidation of results for overall risk profile indication. In this, individual risk assessments are combined to compute mean overall risk probabilities while presenting results through visualization methods, including risk score distribution, risk zone demarcation, and crucial statistical measures.

The visualization of Common RC Buildings distribution is depicted in Figure A1 and statistical results are denoted in

Table A5. Statistical results of Monte Carlo analysis distribution results for Common RC Buildings based on Table A1.

Group	Damage Type	Mean Score	Standard Deviation	95th Percentile	Low Risk (%)	Medium Risk (%)	High Risk (%)
Material	poor concrete cracking	9.18	1.34	11.36	19.73	78.69	1.58
	poor concrete load bearing	9.19	1.32	11.37	19.12	79.47	1.41
	poor concrete spalling	9.19	1.34	11.38	19.29	79.00	1.71
	poor concrete permeability	9.20	1.33	11.38	18.89	79.74	1.37
	corrosion reinforcement	9.19	1.34	11.38	19.21	79.35	1.44
	corrosion structural	9.19	1.31	11.37	18.56	80.05	1.39
	corrosion water leakage	9.21	1.33	11.41	18.16	80.24	1.60
Design	earthquake structural cracking	8.36	1.25	10.44	39.04	60.77	0.19
	structural element failure	8.36	1.25	10.48	39.78	60.01	0.21
	partial collapse	7.51	1.03	9.21	67.56	32.44	0.00
	load bearing capacity	7.49	1.03	9.20	68.32	31.68	0.00
	story failure	7.51	1.04	9.23	67.70	32.30	0.00
	element falling	7.49	1.02	9.17	68.19	31.81	0.00
Reinforcement	slippage	11.76	0.83	13.16	0.00	61.58	38.42
	shear cracking	11.76	0.83	13.12	0.00	61.04	38.96
	structural integrity loss	10.91	1.15	12.70	0.88	81.84	17.28
	buckled reinforcement	11.77	0.85	13.15	0.00	59.93	40.07
	reduced load capacity	11.75	0.84	13.14	0.00	61.48	38.52
	connection failure	11.75	0.85	13.15	0.00	61.60	38.40
Others	collision damage	8.35	1.27	10.47	40.20	59.63	0.17
	structural tilting	9.17	1.32	11.34	19.22	79.27	1.51
	lateral stability loss	10.05	1.30	12.11	6.26	87.45	6.29
Overall Score		9.47	0.25	9.88	0.00	100.00	0.00

.

The visualization of Greenery-Covered Buildings (Structural Performance) distribution is depicted in Figure A2 and Figure A3 and statistical results are denoted in

Table A6. Statistical results of Monte Carlo analysis distribution results for Greenery-Covered Buildings (Structural Performance) based on Table A2.

Group	Damage Type	Mean Score	Standard Deviation	95th Percentile	Low Risk (%)	Medium Risk (%)	High Risk (%)
Root	penetration wall foundation integrity	8.35	1.26	10.47	39.81	59.96	0.23
	penetration potential instability	8.33	1.26	10.41	40.61	59.18	0.21
	penetration cracks	8.35	1.26	10.47	39.71	60.05	0.24
	penetration technical features damage	8.33	1.26	10.42	40.12	59.63	0.25
Plants	attachment wall foundation integrity	7.49	1.03	9.22	68.45	31.55	0.00
	attachment lifted blocks	7.52	1.05	9.24	67.35	32.65	0.00
	attachment cracks	7.50	1.03	9.21	68.17	31.83	0.00
	attachment scarring	7.49	1.04	9.23	67.90	32.10	0.00
	attachment plaster removal	8.33	1.24	10.40	39.85	60.03	0.12
	attachment local wetness	8.35	1.24	10.43	39.63	60.14	0.23
Moisture Evapotranspiration	mold development	11.75	0.84	13.11	0.00	61.15	38.85
	dampness	10.04	1.31	12.11	6.80	87.21	5.99
	material covers damages	11.74	0.84	13.12	0.00	62.15	37.85
	slow drying	11.76	0.85	13.14	0.00	60.77	39.23
Moisture Irrigation	water leaks	11.76	0.84	13.15	0.00	61.32	38.68
	water pooling	11.74	0.84	13.12	0.00	61.21	38.79
	dripping water	11.76	0.83	13.14	0.00	61.29	38.71
	dampness	11.75	0.84	13.11	0.00	61.59	38.41
	mold development	11.74	0.84	13.11	0.00	61.60	38.40
	corrosion	11.74	0.84	13.12	0.00	61.83	38.17
	streaking	11.76	0.84	13.13	0.00	60.83	39.17
	material covers damages	10.91	1.16	12.72	1.00	81.02	17.98
Soil Moisture	wall foundation integrity	6.65	1.25	8.69	85.35	14.65	0.00
	potential instability	6.64	1.25	8.70	85.92	14.08	0.00
	subsidence	6.65	1.25	8.70	85.89	14.11	0.00
	soil microbial concrete erosion	8.34	1.26	10.40	39.72	60.14	0.14
Overgrown Vegetation	overtaking space	7.50	1.04	9.22	68.30	31.70	0.00
	blocking windows	7.51	1.03	9.21	67.52	32.48	0.00
	cascading branches	7.51	1.03	9.22	68.18	31.82	0.00
	technical features obstruction	9.21	1.33	11.38	18.21	80.24	1.55
Additional Weight	potential instability	7.48	1.55	10.05	62.94	36.92	0.14
	structural integrity	10.91	1.18	12.73	1.04	80.80	18.16
Fire	burned materials	11.76	0.84	13.14	0.00	60.69	39.31
	collapse	11.77	0.84	13.14	0.00	60.47	39.53
Wind	structural integrity	10.01	1.31	12.10	6.82	87.12	6.06
	cracks	8.34	1.69	11.06	41.97	57.06	0.97
	collapse	10.91	1.16	12.73	0.84	81.60	17.56
	blown out	9.21	1.35	11.38	18.91	79.57	1.52
	debris	10.05	1.30	12.13	5.95	87.77	6.28
Earthquake	structural integrity	11.74	0.84	13.14	0.00	61.92	38.08
	cracks	11.75	0.83	13.12	0.00	61.37	38.63
	collapse	11.76	0.84	13.15	0.00	61.32	38.68
	spalling	11.74	0.84	13.11	0.00	61.90	38.10
	instability	11.74	0.85	13.16	0.00	61.79	38.21
Overall Score		9.72	0.17	9.99	0.00	100.00	0.00

.

The visualization of Greenery-Covered Buildings (Greenery Performance) distribution is depicted in Figure A4 and statistical results are denoted in

Table A7. Statistical results of Monte Carlo analysis distribution results for Greenery-Covered Buildings (Greenery Performance) based on Table A3.

Group	Damage Type	Mean Score	Standard Deviation	95th Percentile	Low Risk (%)	Medium Risk (%)	High Risk (%)
High Moisture Evapotranspiration	pests	11.76	0.84	13.12	0.00	60.84	39.16
	planting media degradation	11.76	0.84	13.14	0.00	60.65	39.35
Low Moisture	irrigation plant stress	11.74	0.83	13.12	0.00	61.94	38.06
	irrigation dry vegetation	11.75	0.84	13.14	0.00	61.64	38.36
	irrigation dead plants	11.76	0.84	13.16	0.00	61.30	38.70
	irrigation wilting	11.75	0.84	13.15	0.00	61.41	38.59
	irrigation gaps	11.74	0.84	13.14	0.00	61.49	38.51
	irrigation planting media degradation	11.75	0.83	13.11	0.00	61.55	38.45
Improper Maintenance	overgrown vegetation	11.75	0.83	13.10	0.00	61.46	38.54
	dead biomass	11.74	0.83	13.10	0.00	61.87	38.13
	bio debris	10.90	1.16	12.73	1.08	81.23	17.69
	gaps	10.90	1.17	12.73	0.99	81.54	17.47
	planting media degradation	11.75	0.82	13.11	0.00	62.03	37.97
	loss	11.74	0.85	13.14	0.00	61.57	38.43
	compaction	11.76	0.84	13.15	0.00	60.90	39.10
Inadequate	space distorted vegetation growth	8.35	1.25	10.46	39.64	60.17	0.19
Environmental Stressors	plant stress	11.75	0.83	13.14	0.00	61.29	38.71
	dehydration	10.90	1.16	12.71	1.16	81.23	17.61
Fire	burnt vegetation	11.75	0.83	13.13	0.00	61.57	38.43
Wind	uprooting	11.75	0.84	13.11	0.00	61.25	38.75
	mechanical stress	11.73	0.83	13.10	0.00	62.22	37.78
	dehydration	11.75	0.84	13.13	0.00	60.89	39.11
	reduced plant growth rate	10.90	1.14	12.68	0.91	81.86	17.23
	wilting	11.76	0.84	13.14	0.00	60.93	39.07
	bended trees	11.75	0.84	13.15	0.00	60.65	39.35
	falling bio debris	9.20	1.34	11.41	18.93	79.24	1.83
Earthquake	planter anchorage instability	11.76	0.84	13.13	0.00	60.90	39.10
	tree pot anchorage instability	11.75	0.84	13.15	0.00	61.68	38.32
Overall Score		11.42	0.18	11.71	0.00	99.97	0.03

.

The visualization of Greenery-Covered Buildings (Human Welfare) distribution is depicted in Figure A5 and statistical results are denoted in

Table A8. Statistical results of Monte Carlo analysis distribution results for Greenery-Covered Buildings (Human Welfare) based on Table A4.

Group	Damage Type	Mean Score	Standard Deviation	95th Percentile	Low Risk (%)	Medium Risk (%)	High Risk (%)
Dead Plants	aesthetics	11.76	0.84	13.15	0.00	60.55	39.45
	utility	10.90	1.16	12.74	0.90	81.66	17.44
	energy needs summer	10.90	1.17	12.70	1.16	80.90	17.94
Overgrown Vegetation	insects	11.75	0.83	13.13	0.00	61.16	38.84
	aesthetics	10.05	1.31	12.16	6.32	87.00	6.68
	energy needs winter	10.06	1.30	12.15	6.21	87.20	6.59
	utility	10.08	1.29	12.12	5.94	87.94	6.12
	reduced sunlight view	10.06	1.32	12.16	6.56	86.77	6.67
	living space occupation	10.05	1.30	12.12	6.25	87.43	6.32
High Moisture Level	health issues	11.75	0.82	13.09	0.00	61.40	38.60
	pests	11.76	0.84	13.13	0.00	60.74	39.26
	insects	11.74	0.84	13.14	0.00	61.18	38.82
	mold development	11.74	0.85	13.14	0.00	61.62	38.38
Maintenance	cost	9.19	1.33	11.34	19.03	79.65	1.32
	time	9.23	1.32	11.39	18.18	80.33	1.49
	potential falls	11.75	0.84	13.15	0.00	61.09	38.91
	work at height	10.05	1.29	12.11	6.39	87.59	6.02
Fire	ecological imbalances	9.19	1.34	11.37	18.92	79.56	1.52
	safety	11.76	0.85	13.15	0.00	61.14	38.86
	evacuation obstruction	11.75	0.84	13.12	0.00	61.37	38.63
	vertical propagation	11.72	0.84	13.09	0.00	62.61	37.39
	chimney effect	11.75	0.84	13.14	0.00	61.27	38.73
Wind	safety	11.75	0.84	13.13	0.00	61.44	38.56
	fallen objects bio debris	11.77	0.84	13.13	0.00	60.13	39.87
	turbulence	11.75	0.84	13.14	0.00	61.14	38.86
Earthquake	high repairing cost	11.74	0.85	13.15	0.00	61.62	38.38
Overall Score		11.00	0.21	11.34	0.00	100.00	0.00

.

Additionally, in the concept of hazard analysis, the STPA has been extensively examined for vertical forest applications, as there is a need for proper identification of specific hazards in complex systems. The study from MIT by Leveson and Thomas indicates that safety requires a control-based approach rather than component failure analysis [122]. Subsequently, two main systems have been proposed for detailed examination, among others, the VF balcony system and the living wall system as depicted in Figure 2.

The VF balcony system integration requires proper assessment of structural elements, supportive plant structure, and plant care systems while maintaining the balance between load monitoring, drainage management, and safety protocols. The relevant implementation tables can be found in Appendix C from

Table A9. Accident (Loses) definition for VF RC Buildings Balcony system.

Accident (Loses)	Description
A-1: Loss of human life or injury	A-1.1: People injured/killed by falling plants/containers
	A-1.2: People injured/killed by structural collapse
	A-1.3: People injured/killed by slipping on wet surfaces
	A-1.4: Injuries/fatalities during maintenance activities
A-2: Property damage	A-2.1: Damage to balcony structure
	A-2.2: Damage to apartment interior due to water/nutrient leakage
	A-2.3: Damage to lower floors/apartments
	A-2.4: Damage to personal property on balcony
A-3: Environmental damage	A-3.1: Plant death/ecosystem disruption
	A-3.2: Chemical/nutrient runoff affecting environment
	A-3.3: Excessive water waste
	A-3.4: Soil contamination from system failures
A-4: Mission losses	A-4.1: Failure to maintain healthy vertical forest
	A-4.2: System abandonment due to maintenance issues
	A-4.3: Loss of property value
	A-4.4: Loss of system certification/compliance status

,

Table A10. Hazards definition for VF RC Buildings Balcony system.

Hazards	Description
H-1: Structural Integrity Hazards	H-1.1: Balcony load capacity exceeded
	H-1.2: Uneven load distribution creates structural stress
	H-1.3: Structural deterioration from water/chemical exposure
	H-1.4: Plant anchoring system failure
H-2: Plant System Hazards	H-2.1: Over-saturated growing medium
	H-2.2: Nutrient imbalance in delivery system
	H-2.3: Root system compromises structural elements
	H-2.4: Plants grow beyond designed size/weight limits
H-3: Control System Hazards	H-3.1: Incorrect sensor readings lead to system malfunction
	H-3.2: Failure of automated irrigation/nutrient delivery
	H-3.3: Monitoring system failure masks dangerous conditions
	H-3.4: Control system provides incorrect commands
H-4: Environmental Interaction Hazards	H-4.1: High wind conditions affect plant/container stability
	H-4.2: Extreme weather compromises system integrity
	H-4.3: Excessive heat/cold damages system components
	H-4.4: Water accumulation creates unsafe conditions

,

Table A11. Identification of UCSs for VF RC Buildings Balcony system.

Controller	Control Action	Not Providing Causes Hazard	Providing Causes Hazard	Wrong Timing/Order	Stopped Too Soon/Applied Too Long	Related Hazards
Load Monitoring System	Emergency System Shutdown	System continues operating under dangerous load conditions	Unnecessary shutdown when load is within safe limits	Delayed shutdown after critical load detected	Partial shutdown leaving some systems running	H-1.1, H-1.2, H-4.2
	Load Warning Alerts	No warning of dangerous loading conditions	False alarms causing unnecessary evacuations	Delayed warnings after critical conditions	Warning stops before condition resolved	H-1.1, H-1.2, H-3.3
	Weight Distribution Control	Uneven load distribution not corrected	Incorrect load redistribution causing stress	Adjustment during peak usage times	Redistribution process incomplete	H-1.2, H-2.4, H-4.1
	Maximum Load Limit Enforcement	Overloading condition not prevented	System restricts necessary maintenance access	Limit enforcement during emergency operations	Temporary override remains active too long	H-1.1, H-1.2, H-2.4
	Dynamic Load Response	Wind/weather load effects not managed	Excessive movement restrictions applied	Response during stabilization period	Dynamic response ends prematurely	H-1.2, H-4.1, H-4.2
Plant Care System	Irrigation Control	Plants die from lack of water	Oversaturation causing excess weight	Watering during high wind conditions	Continuous watering leading to saturation	H-2.1, H-4.4, H-3.2
	Nutrient Delivery	Plant health deterioration	Excess nutrients damaging structure	Delivery during rain events	Extended delivery causing accumulation	H-2.2, H-3.2, H-3.4
	Growth Rate Monitoring	Excessive growth not detected	False growth alerts triggering pruning	Delayed growth assessment	Monitoring interruption during growth season	H-2.4, H-2.3, H-3.1
	Root System Control	Root invasion not prevented	Excessive root restriction	Root check during critical growth	Partial root management	H-2.3, H-1.3, H-3.3
	Environmental Adjustment	Climate stress not mitigated	Excessive environmental modification	Adjustment during plant dormancy	Incomplete climate adaptation	H-4.2, H-4.3, H-3.4
Maintenance Staff	Plant Pruning	Excessive growth exceeding loads	Excessive pruning damaging plants	Pruning during flowering/growth	Incomplete pruning cycle	H-2.4, H-4.1, H-1.1
	System Inspection	Component deterioration undetected	Invasive inspection damage	Inspection during peak loads	Partial system check	H-3.3, H-1.3, H-2.3
	Equipment Maintenance	System failure from wear	Improper maintenance procedures	Maintenance during critical operations	Incomplete maintenance routine	H-3.2, H-3.4, H-3.1
	Safety Check Protocol	Safety hazards unidentified	Disruptive safety procedures	Checks during emergency conditions	Abbreviated safety assessment	H-3.3, H-4.4, H-1.2
	Drainage Cleaning	Blockage leading to water damage	Aggressive cleaning damaging pipes	Cleaning during heavy water use	Partial drainage clearance	H-4.4, H-2.1, H-1.3
Structural Monitoring System	Crack Monitoring	Structural damage undetected	False crack detection alerts	Delayed crack assessment	Monitoring gaps during stress	H-1.3, H-3.3, H-1.2
	Deflection Assessment	Critical deformation missed	Incorrect deflection readings	Assessment during loading	Incomplete deflection analysis	H-1.1, H-1.2, H-3.3
	Strain Measurement	Stress concentration undetected	False strain readings	Measurement during movement	Partial strain assessment	H-1.2, H-3.1, H-3.3
	Vibration Monitoring	Dangerous resonance missed	False vibration alerts	Monitoring during stabilization	Interrupted vibration analysis	H-4.1, H-4.2, H-3.1
	Material Degradation Check	Corrosion/damage undetected	Invasive material testing	Check during peak stress	Incomplete degradation assessment	H-1.3, H-3.3, H-2.3
Building Management	Maintenance Schedule	System deterioration unchecked	Excessive maintenance disruption	Schedule during peak usage	Incomplete maintenance cycle	H-3.3, H-2.4, H-1.3
	Emergency Response	No response to critical conditions	Unnecessary system shutdown	Delayed emergency procedures	Premature procedure termination	H-3.3, H-4.2, H-1.1
	Resource Allocation	Critical supplies not provided	Excess resource commitment	Resource delivery timing wrong	Partial resource provision	H-3.2, H-2.2, H-2.1
	Inspection Protocol	System problems unidentified	Overly invasive inspections	Inspection during critical ops	Incomplete inspection coverage	H-3.3, H-1.3, H-1.2
	Access Control	Unauthorized maintenance allowed	Excessive access restrictions	Access during critical operations	Temporary access left active	H-3.3, H-1.1, H-2.

,

Table A12. Identification of main UCSs Causal Factors and proposal of Safety Requirements for VF RC Buildings Balcony system.

System Component	Main Unsafe Control Actions	Primary Causal Factor	Safety Requirement
Load Monitoring System	Not providing shutdown when load exceeds critical threshold	Sensor calibration drift	System must initiate automatic shutdown at 85% of maximum design load
	No warning issued when dangerous loading condition exists	Communication system failure	Redundant warning systems with independent power supplies must be maintained
	Uneven load distribution not corrected	Load distribution sensor malfunction	Real-time monitoring system must verify load distribution every 15 min
	Overloading condition not prevented	Incorrect load calculation algorithm	Automated load monitoring system must prevent addition of new loads when at 90% capacity
	Wind/weather load effects not managed	Weather sensor system failure	Automatic response system must adjust for wind loads exceeding 30 km/h
Plant Care System	Excessive water delivery causing oversaturation	Moisture sensor failure	Moisture content must be maintained between 40–70% with automated cutoff
	Excess nutrients damaging structure	Nutrient concentration sensor malfunction	Nutrient delivery must stop if concentration exceeds prescribed limits
	Excessive growth not detected	Monitoring camera system failure	Weekly physical verification of plant size against growth limits
	Root invasion not prevented	Root barrier system degradation	Monthly inspection of root barrier integrity with documentation
	Climate stress not mitigated	Temperature sensor failure	Environmental parameters must be maintained within specified ranges
Maintenance Staff	Excessive growth exceeding design loads	Lack of maintenance schedule adherence	Mandatory monthly pruning schedule with documentation
	Critical component deterioration undetected	Inadequate inspection training	Quarterly certified inspection by qualified personnel
	System failure from lack of maintenance	Poor maintenance documentation	Computerized maintenance management system with alerts
	Safety hazards unidentified	Incomplete safety checklist	Daily safety checks with electronic verification
	Blockage leading to water damage	Missed cleaning schedule	Weekly drainage system cleaning with flow verification
Structural Monitoring System	Structural damage undetected	Monitoring sensor malfunction	Automated crack detection system with weekly calibration
	Critical deformation missed	Incorrect baseline measurements	Monthly professional structural assessment with documentation
	Stress concentration undetected	Strain gauge failure	Redundant strain monitoring system with automatic alerts
	Dangerous resonance missed	Vibration sensor calibration drift	Daily vibration monitoring with automated threshold alerts
	Corrosion/damage undetected	Inspection tool malfunction	Quarterly material integrity testing with certification
Building Management	System deterioration unchecked	Poor schedule management	Computerized maintenance tracking with mandatory deadlines
	No response to critical conditions	Emergency protocol confusion	Clear emergency response procedures with annual drills
	Critical supplies not provided	Inadequate inventory tracking	Automated inventory management system with minimum stock alerts
	System problems unidentified	Lack of systematic approach	Standardized inspection protocols with electronic checklists
	Unauthorized maintenance allowed	Access control system failure	Multi-factor authentication for system access with logging

,

Table A13. Risk matrix (see Figure 3) numbering explanation and risk level calculation examples for VF RC Buildings Balcony system.

Level	(Consequences)—Scale 1–5			(Likelihood)—Scale 1–5
1	Negligible	- Sensor calibration drift within acceptable range - Minor plant growth issues - Temporary irrigation interruption - Small moisture sensor misreading - Documentation delays		Very Unlikely	- Rare maintenance oversight - Highly unlikely system failure - Almost no environmental stress - Exceptional preventive measures in place - No historical precedent
2	Minor	- Partial nutrient system malfunction - Minor load distribution variance - Temporary monitoring system interruption - Plant pruning delays - Non-critical sensor failure		Unlikely	- Annual maintenance issues - Uncommon sensor drift - Limited environmental exposure - Strong preventive measures - Rare occurrence in similar systems
3	Moderate	- Complete irrigation system failure - Multiple sensor system failures - Significant plant die-off - Moderate structural stress detected - Control system malfunction		Possible	- Monthly calibration needs - Occasional sensor issues - Moderate environmental exposure - Standard preventive measures - Known to occur in similar systems
4	Significant	- Major structural deflection - Critical monitoring system failure - Complete plant system failure - Severe water/nutrient leakage - Multiple control system failures		Likely	- Weekly maintenance required - Regular sensor adjustments - Frequent environmental stress - Basic preventive measures - Common in similar systems
5	Severe	- Complete structural failure - Total system collapse - Life-threatening structural damage - Complete control system breakdown - Catastrophic water damage to building		Very Likely	- Daily system stress - Constant environmental exposure - Multiple risk factors present - Minimal preventive measures - Frequently observed issues measures
Risk Level Calculations (Examples)
Low Risk (L) Scenarios			Medium Risk (M) Scenarios		High Risk (H) Scenarios
L1 (1,1): Minor sensor calibration drift with rare occurrence			M4 (2,2): Plant care system requiring recalibration		H10 (5,2): Critical structural monitoring system failure
L2 (1,2): Temporary irrigation interruption during routine maintenance			M5 (3,1): Single control loop failure with backup systems active		H12 (4,3): Major load distribution anomaly

and

Table A14. Accident (Loses) and Hazard definition for RC Building Living Wall system.

Accident Category and ID	Accident (Loss) Description	Hazard ID	Associated Hazards
A.1 Structural Integrity
A.1-1	Building structural damage due to excessive load	H.1.1	Plant mass exceeds structural design limits
A.1-2	Facade material deterioration	H.1.2	Root system penetrates waterproofing membrane
A.1-3	Support system failure	H.1.3	Cable/mesh attachment points fail or deteriorate
A.2 Water-Related
A.2-1	Internal water damage to building	H.2.1	Irrigation system leakage into building envelope
A.2-2	Foundation damage from water accumulation	H.2.2	Drainage system blockage or failure
A.2-3	Mold/moisture damage to building materials	H.2.3	Inadequate ventilation between plants and wall
A.3 Plant System
A.3-1	Complete plant die-off requiring system replacement	H.3.1	Critical irrigation system failure
A.3-2	Uncontrolled plant growth damaging building systems	H.3.2	Plant growth monitoring system failure
A.3-3	Invasive plant spread to unintended areas	H.3.3	Containment system breach
A.4 Human Safety
A.4-1	Injury from falling plant material or debris	H.4.1	Dead plant material accumulation
A.4-2	Injury during maintenance operations	H.4.2	Unsafe access for maintenance
A.4-3	Health issues from mold/biological agents	H.4.3	Harmful biological growth in plant system
A.5 Environmental Control
A.5-1	Building energy efficiency loss	H.5.1	Gap formation between wall and plant coverage
A.5-2	Building envelope thermal damage	H.5.2	Excessive heat buildup between plants and wall
A.5-3	Local ecosystem disruption	H.5.3	Introduction of invasive species or pests
A.6 Operational
A.6-1	System maintenance cost overrun	H.6.1	Control system malfunction
A.6-2	Building operation disruption	H.6.2	Integration failure with building systems
A.6-3	Aesthetic value loss	H.6.3	Uneven or patchy plant growth

. The living wall system incorporating mechanical and biological components are examined in

Table A15. Identification of UCSs for RC Building Living Wall system.

Controller	Control Action	Not Providing Causes Hazard	Providing Causes Hazard	Wrong Timing/Order	Stopped Too Soon/Applied Too Long	Related Hazards
Facility Management	Set Maintenance Schedule	No maintenance schedule established	Schedule too infrequent or intensive	Maintenance scheduled during inappropriate weather conditions	Schedule terminated before critical checks completed	H.1.2, H.2.2, H.3.1, H.4.2
	Define Operating Parameters	No system parameters defined	Parameters set outside safe limits	Parameters changed during critical operations	Parameters modified before verification complete	H.2.1, H.3.1, H.5.1, H.6.1
Maintenance Staff	Perform Plant Maintenance	Required pruning not performed	Excessive/improper pruning performed	Maintenance during extreme weather	Incomplete maintenance operation	H.3.2, H.3.3, H.4.1, H.6.3
	Check Support Structure	Structure inspection skipped	Incorrect load assessment	Delayed inspection after storm events	Inspection stopped before completion	H.1.1, H.1.3, H.4.1
	Maintain Irrigation System	Required cleaning skipped	Improper system modification	Maintenance during peak usage	Partial system maintenance	H.2.1, H.2.2, H.3.1
Horticultural Specialist	Plant Care Actions	Required treatment not provided	Wrong treatment applied	Treatment during stress conditions	Treatment stopped mid-process	H.3.1, H.3.2, H.3.3
	Growth Management	Growth limits not enforced	Incorrect growth training	Late intervention in growth pattern	Incomplete growth management	H.1.1, H.3.2, H.6.3
Central Controller	Irrigation Control	No water provided	Excessive water provided	Watering at wrong time	Watering duration incorrect	H.2.1, H.2.2, H.3.1
	Environmental Monitoring	No monitoring active	Wrong parameters monitored	Delayed response to conditions	Monitoring interrupted	H.3.1, H.5.1, H.5.2
Irrigation Controller	Water Flow Control	No water flow when needed	Excessive water flow	Wrong timing of water delivery	Water flow duration incorrect	H.2.1, H.2.2, H.3.1
	Nutrient Dosing	No nutrients provided	Excessive nutrients provided	Wrong timing of nutrient delivery	Nutrient cycle incomplete	H.3.1, H.3.2, H.6.3

,

Table A16. Safety Constraint for RC Building Living Wall system.

Hazard	Control Action	Unsafe Control Action	Safety Constraint	Safety Requirements	Verification Method
H.1.1: Plant mass exceeds structural limits	Growth Management	UCA-1: No load monitoring	SC-1: Must prevent excessive loading	SR-1.1: Load monitoring system	- Load cell data
		UCA-2: Wrong growth control		SR-1.2: Growth limits protocol	- Growth measurements
		UCA-3: Missed inspections		SR-1.3: Regular inspections	- Inspection logs
H.1.2: Root penetration	Root Management	UCA-4: No root barriers	SC-2: Must contain root system	SR-2.1: Root barrier system	- Barrier integrity tests
		UCA-5: Wrong root pruning		SR-2.2: Root monitoring	- Root imaging
		UCA-6: Missed checks		SR-2.3: Inspection schedule	- Check records
H.2.1: Water leakage	Irrigation Control	UCA-7: Excess water flow	SC-3: Must prevent water infiltration	SR-3.1: Leak detection	- Leak sensors
		UCA-8: Wrong timing		SR-3.2: Flow control	- Flow meters
		UCA-9: No monitoring		SR-3.3: Moisture monitoring	- Moisture data
H.2.2: Drainage failure	Drainage Management	UCA-10: No cleaning	SC-4: Must maintain drainage	SR-4.1: Flow monitoring	- Flow rates
		UCA-11: Wrong maintenance		SR-4.2: Cleaning protocol	- Maintenance logs
		UCA-12: Blocked flow		SR-4.3: Blockage detection	- Sensor data
H.3.1: Irrigation system failure	System Control	UCA-13: No water supply	SC-5: Must maintain irrigation	SR-5.1: Backup systems	- System status
		UCA-14: Wrong settings		SR-5.2: Alert protocols	- Alert logs
		UCA-15: System shutdown		SR-5.3: Emergency response	- Response times
H.3.2: Uncontrolled growth	Growth Control	UCA-16: No pruning	SC-6: Must control growth	SR-6.1: Growth monitoring	- Growth data
		UCA-17: Wrong technique		SR-6.2: Pruning schedule	- Maintenance records
		UCA-18: Missed areas		SR-6.3: Coverage checks	- Coverage maps
H.4.1: Falling debris	Maintenance	UCA-19: No removal	SC-7: Must prevent debris fall	SR-7.1: Debris monitoring	- Inspection reports
		UCA-20: Wrong timing		SR-7.2: Removal protocol	- Removal logs
		UCA-21: Incomplete check		SR-7.3: Safety checks	- Safety audits
H.4.2: Unsafe access	Access Control	UCA-22: No safety gear	SC-8: Must ensure safe access	SR-8.1: Safety protocols	- Safety records
		UCA-23: Wrong procedure		SR-8.2: Access systems	- Training logs
		UCA-24: Missed checks		SR-8.3: Safety training	- Audit reports
H.5.1: Energy efficiency loss	Coverage Control	UCA-25: Coverage gaps	SC-9: Must maintain coverage	SR-9.1: Coverage monitoring	- Coverage data
		UCA-26: Wrong density		SR-9.2: Density control	- Thermal imaging
		UCA-27: No monitoring		SR-9.3: Gap prevention	- Efficiency tests
H.6.1: Control system failure	System Management	UCA-28: No monitoring	SC-10: Must maintain control	SR-10.1: System redundancy	- System logs
		UCA-29: Wrong settings		SR-10.2: Fault detection	- Error reports
		UCA-30: System conflicts		SR-10.3: Backup protocols	- Backup tests

and

Table A17. Risk matrix (see Figure 3) numbering explanation and risk level calculation examples for RC Building Living Wall system.

Level	(Consequences)—Scale 1–5			(Likelihood)—Scale 1–5
1	Negligible	- Temporary sensor drift within range - Minor plant health issues - Short irrigation interruptions - Small drainage blockages - Documentation delays		Very Unlikely	- Rare maintenance issues - Highly unlikely failures - Minimal environmental stress - Strong preventive measures - No historical precedent
2	Minor	- Localized root barrier stress - Partial drainage system blockage - Minor load distribution issues - Limited plant growth problems - Non-critical control issues		Unlikely	- Annual maintenance needs - Uncommon sensor issues - Limited environmental exposure - Regular inspections - Rare in similar systems
3	Moderate	- Complete irrigation failure - Multiple sensor failures - Significant plant die-off - Moderate structural stress - Water management issues		Possible	- Monthly maintenance needs - Regular sensor calibration - Moderate environmental exposure - Standard maintenance - Known in similar systems
4	Significant	- Major structural loading issues - Critical monitoring failure - Extensive water damage - Severe root penetration - Multiple system failures		Likely	- Weekly maintenance needs - Frequent adjustments - High environmental stress - Basic preventive measures - Common in similar systems
5	Severe	- Complete structural failure - Total system collapse - Severe building damage - Complete control failure - Critical water infiltration		Very Likely	- Daily system stress - Constant exposure - Multiple risk factors - Minimal prevention - Frequently observed
Risk Level Calculations (Examples)
Low Risk (L) Scenarios			Medium Risk (M) Scenarios		High Risk (H) Scenarios
L2 (2,1): Temporary irrigation interruption during maintenance			M6 (2,3): Nutrient delivery system temporary malfunction		H15 (5,3): Complete plant support system failure
L3 (3,1): Single non-critical sensor failure with backup			M8 (4,1): Non-critical monitoring system interruption		H16 (4,4): Severe structural stress detection
					H12: Major water infiltration (4,3)
					H10 (5,2): Critical structural monitoring system failure

. The Control Structure scheme for both Balcony and Living Wall can be found in Figure A6 and Figure A7, respectively. Study [15] indicates that self-climbing vegetation support and automatic irrigation mechanisms have been extensively examined, as the proper integration of watertight membranes and control systems is significant. Furthermore, environmental monitoring enhances building envelope integrity while maintaining the balance between plant vitality, seasonal changes, and meteorological conditions. The results from STPA were used to create a Risk Matrix to evaluate the risk level of each system, as depicted in Figure 3.

As depicted in Figure 3, a simple tool for rating potential risks in both VF RC Building Balcony and RC Building Living Wall structures is the risk evaluation matrix. In this arrangement, risks are organized according to a dual-axis scheme that considers both consequence severity and likelihood of occurrence, creating a 5 × 5 grid pattern in its totality. Each cell in the grid holds a level of danger, represented with a letter (L, M, or H) and a number; L for low danger (in green), M for medium (in yellow), and H for high (in red). The grid not only portrays an intensification of danger through severity of impact but through an events’ likelihood, as well. For instance, any events with a 5 (Very Likely) rating will have at least a rating of medium (M5 regardless of consequence level). The matrix accurately reflects how a combination of consequence and likelihood can produce a range of risks, starting with a minimum rating (L1) for events with negligible consequences and a high chance of not happening, and culminating in a rating of H25 for events with high consequences and a high chance of happening.

5. Discussion

As the integration of Vertical Forest systems into existing Reinforced Concrete buildings has been extensively examined, there is a need for proper assessment of structural performance, environmental sustainability, and occupant well-being. The holistic damage quantification and hazard intensity measurement may not be the ultimate solution through individual subsystem analysis. The reliable global damage indicators have been proposed, among others, (a) Interstory Drift Ratio (IDR), (b) Residual Deformation (RD), (c) Concrete and Reinforcement Strains/Stresses (CaRSS), and (d) Energy Dissipation (ED). Studies indicate that structural metrics require integration with vegetation components, as the added mass and vegetation-induced damping are significant factors for dynamic behavior in some cases. Furthermore, the non-structural elements assessment, through façade systems and planter stability, has a significant impact on proper evaluation.

Consequently, occupant satisfaction and comfort are more feasible solutions for performance assessment. The proper methodologies for qualitative data analysis have been proposed while maintaining the balance between renovation outcomes and living conditions. Additionally, in the concept of reducing monitoring complexity, AI-driven plant health assessments and real-time structural sensors enhance seismic performance evaluation. The environmental hazard intensity measures, among others, seismic, wind, and rainfall impacts, have been extensively examined through comprehensive risk assessments. The large-scale integration of these approaches requires systematic examination, given that the aesthetic advantages, structural safety, and occupant well-being are critical factors urging for further attention.

As the four main categories of the Monte Carlo analysis results have been extensively examined, there is a need for proper risk distribution patterns evaluation. The study indicates that common RC buildings analysis may not be the ultimate solution without proper categorization, as 100% of cases cluster in the medium risk category with a mean score of 9.47/15. Subsequently, the reinforcement-related issues show high risk classification while maintaining 35.3% of cases and mean scores of 11.15/15. The slippage, shear cracking, and connection failures are significant factors that require systematic examination.

The STPA analysis results require proper evaluation of two main configurations, among others: (a) VF balcony system and (b) living wall system. The study from the balcony system indicates that four key hazard categories have been proposed, as the structural integrity, plant systems, and control mechanisms are significant factors. Additionally, in the concept of risk matrix analysis, the high risk scenarios through structural monitoring system failure (H10, 5,2) have been extensively identified. Subsequently, the living wall system analysis shows six major accident categories while maintaining the balance between structural integrity and operational problems. The water-related hazards have been proposed for proper assessment, as ultimately, the water infiltration (H12, 4,3) and critical system failure (H16, 4,4) are crucial factors that require systematic examination.

The STPA uncovered a variety of unanticipated failure modes, specifically in the behavior between different subsystems that might not have been identified through traditional analysis methods. The Control System Interactions investigation uncovered unforeseen cascading interactions between the automation of plant maintenance and structural monitoring systems. In particular, automated irrigation responses during windy conditions may mask indicators of structural deterioration by temporarily alleviating symptoms visible to the naked eye while potentially exacerbating loading issues (H-3.3, Table A11). For Bio-Structural Feedback, one key observation was the complex interrelationship between biological and structural monitoring systems, suggesting that degradation of the root barrier system may cause failure in moisture sensors and, in turn, have a cumulative effect towards integrity monitoring (H-2.3, H-3.1). Additionally for Time-Sensitive Activities, the analysis showed unanticipated risks with regards to maintenance activity timing; such routine maintenance operations, namely pruning or inspections performed during high demand times, may cause overall system destabilization (Table A11 and Table A12). These unexpected failure modes have led to enhanced safety requirements and constraints as detailed in Table A12 and Table A16, particularly in the areas of integrated system monitoring and control.

As both analyses require proper integration, there is a need for correlation patterns assessment. The study indicates three main aspects, among others:

• Structural Integrity Assessment: The Monte Carlo analysis shows medium to high risk categories with a mean score of 9.72/15 while maintaining correlation with STPA structural hazards findings (H-1.1 through H-1.4). Furthermore, seismic considerations have been extensively examined through high risk scenarios in both methods (Monte Carlo: 11.75/15, STPA: H10).
• Moisture Management Correlation: The proper evaluation of moisture-related risks has been proposed, as the Monte Carlo analysis indicates persistent high risk scores (11.74/15–11.75/15). Additionally, in the concept of living wall systems, STPA water hazard findings (H.2.1 through H.2.3) enhance risk understanding.
• Control Systems Requirements: The large-scale implementation of control systems may have a significant impact on risk mitigation, as both analyses indicate. The proper monitoring and control actions have been extensively examined while maintaining the balance between probabilistic assessment and specific safety protocols.

6. Conclusions

As the holistic renovation of RC buildings through VF systems requires proper evaluation, there is a need for systematic assessment of damage quantification and hazard intensity measures. The study indicates that structural components, among others, IDR, RD, CaRSS, and ED, have been extensively examined while maintaining the balance between conventional solutions and green modifications.

Subsequently, the proper integration of non-structural elements and VF façade systems has been proposed, as ultimately, the cladding displacement, glazing performance, and planter stability are crucial factors that require systematic examination. Additionally, in the context of vegetation components, AI-driven plant health monitoring using different models—such as (a) CNN, (b) DCNN, and (c) RNN—may, in some cases, enhance detection accuracy to between 80% and 99%.

The proper assessment of occupants’ satisfaction has been extensively examined through validated questionnaires while maintaining the balance between qualitative and quantitative data analysis. Furthermore, the hazard intensity measures for seismic, wind, and rainfall impacts may have a significant impact on proper evaluation methods.

Subsequently, the analytical results from Monte Carlo and STPA analyses indicate several crucial findings, among others:

• The green modifications relative to traditional RC buildings could potentially have higher risk profiles (mean scores 9.72/15–11.41/15 versus 9.47/15).
• The moisture management and structural integrity have been proposed for systematic examination.
• The proper control systems implementation enhances risk mitigation strategies.

Additionally, in the concept of future developments, the study indicates several directions that require proper assessment, among others:

• The specialized design codes development, as the current standards may not be sufficient.
• The holistic monitoring systems integration through advanced technologies.
• The standardized maintenance protocols enhance both structural and biological components.
• The large-scale implementation of control systems.

While this review presents extensive information on VF systems in RC buildings, it is necessary to highlight some gaps in the knowledge that need to be addressed in the future. The risk assessments aim, for the first time, to take into account combined effects for VF installations; therefore future long-term performance data will be very helpful for rigorous recalibrations of estimated risks. Future reviews may need to pay special attention to complex interactions between structural and biological components. The STPA analysis focused on specific subsystems, i.e., the balcony and living wall, and might require extension to cover other VF configurations. Also, the environmental impact assessments mainly consider Mediterranean climate conditions, and the results might need to be adjusted to suit different climatic zones.

Furthermore, vertical forest renovations require proper consideration of seismic resilience, as ultimately, the earthquake-related parameters are crucial factors that sometimes largely interact with VF components and holistic system resilience assessment. Sustainable urban transformation through VF systems has been extensively reviewed, with special attention to structural stability, environmental efficiency, and occupant safety issues.