In recent years, the global photovoltaic (PV) market has experienced unprecedented growth. According to the International Energy Agency (IEA), the total installed capacity of PV systems exceeded 1300 GW in 2024, representing nearly a tenfold increase in just one decade. This dynamic pace of expansion, while crucial for the global energy transformation, has simultaneously revealed serious technical safety gaps, particularly regarding fire safety of PV installations.

Numerous cases of PV installation fires have been documented worldwide, both in rooftop systems of residential and commercial buildings, as well as in large-scale ground-mounted power plants. Empirical data unequivocally indicates an alarming trend: a 2021 report by the National Renewable Energy Laboratory (NREL) documents over 400 PV fires in the United States, with the vast majority (over 75%) originating in direct current (DC) system components, such as MC4 connectors, wires, junction boxes, and inverters. Similarly, in Poland, the Scientific and Research Centre for Fire Protection (CNBOP) documented over 260 cases of PV-related fires between 2015 and 2022, of which 85% involved building structures, and more than half resulted from improper installation and thermal phenomena.

The key problem, however, is not the presence of PV installations or the material properties of modules, but rather the inadequate methodology for assessing fire hazards. Current industry standards, such as UL 1703 (USA), IEC 61730 (international), or EN 13501-5 (Europe), focus almost exclusively on the behavior of PV modules when in contact with an external fire source. This approach, while theoretically consistent with the logic of material testing, is inadequate because it does not address the actual ignition mechanisms observed in the field, where the PV installation itself becomes the primary source of fire.

This paper proposes a fundamental paradigm shift in assessing the fire safety of PV systems. Instead of testing resistance to external fire, we propose a comprehensive approach focusing on resistance to internal ignition initiation. To this end, we have developed an engineering-justified diagnostic and design model that encompasses the analysis of the entire PV system as a potential ignition source. This model integrates advanced methods such as current-voltage (IV) curve analysis, thermal imaging (IR), heat flow modeling using Computational Fluid Dynamics (CFD) techniques, and detailed analysis of structural solutions related to mounting, materials, and passive cooling.

This section provides a critical analysis of existing fire safety standards for photovoltaic installations, compares them with empirical data on actual causes of fires, and presents key conclusions that form the basis for the proposed new testing model.

Current standards governing fire safety of photovoltaic systems primarily focus on material reaction to external fire, ignoring the potential of the PV installation itself as an ignition initiator. This approach, while rooted in traditional material testing, leads to fundamental gaps in assessing actual risk.

The following Table 1 summarizes the scope and key limitations of the analyzed standards in the context of actual fire hazards.

Analysis of empirical data from independent research institutions clearly indicates that the actual causes of fires in PV installations differ significantly from those that are the subject of assessment in current standards.

Based on the critical review of applicable standards and analysis of empirical data from actual PV fire cases, the following key conclusions can be formulated:

Empirical data from CNBOP, NREL, and Fraunhofer ISE unequivocally indicate that PV installation fires have different causes than those that current standards attempt to test. The approach based on propagation tests is methodologically flawed because it analyzes the effects of fire instead of its sources. This contradicts the basic principles of safety engineering, where the goal is to eliminate causal risk, not just increase resistance to secondary effects of an already existing hazard.

In response to the identified gaps in existing standards and discrepancies between tests and actual causes of fires in PV installations, we propose a realistic and holistic fire safety testing model. This model focuses on detecting and eliminating potential ignition sources and preventing fire propagation through passive protective measures. The methodology combines advanced diagnostic techniques with engineering analysis of the entire photovoltaic system.

Current-voltage characteristic (IV curve) analysis of photovoltaic modules is a powerful diagnostic tool that allows for the detection of subtle anomalies that may signal degradation or potential risk of electric arc formation long before they escalate to dangerous phenomena. In the context of fire safety, deviations from the nominal IV curve may indicate:

Methodology: In the proposed model, IV curve analysis should be conducted:

Interpretation of deviations from the reference IV curve requires specialized knowledge. Monitoring systems should be able to detect even small drops in current (Isc) or voltage (Voc) of the string, which may be early signals of problems. In particular, it is crucial to use inverters with built-in continuous monitoring of current curves. Such inverters automatically send all information to the central station or installer, providing a clear and transparent preview well in advance of any failures, which forms the foundation of proactive safety.

Thermography (infrared imaging) is a non-invasive and effective method for quickly detecting thermal anomalies - "hot spots" - in PV modules and other installation components. Hot spots, even those invisible to the naked eye, may be symptoms of serious problems, such as:

Methodology:

Thermography, combined with IV curve analysis, creates a powerful diagnostic duo, enabling early detection of hazards.

Computational Fluid Dynamics (CFD) modeling is an advanced numerical simulation technique that allows for detailed analysis of airflow and heat exchange around photovoltaic modules. In the context of fire safety, CFD simulations are key to:

Methodology:

CFD simulations enable the design of systems that passively manage heat, minimizing the risk of overheating and eliminating potential "thermal traps" at the design stage.

Direct current electric arc (DC arc fault) is the main and almost exclusive cause of fires initiated by photovoltaic installations. It arises from a loss of electrical circuit continuity under load, usually caused by a loose or damaged connection (connector, wire, junction box, cracked cell).

Arc Physics: A DC arc is a very high-energy phenomenon, characterized by extremely high temperature (thousands of degrees Celsius) and intense thermal and UV radiation. Its ability to initiate fire depends on:

Detectability: Standard protections (overcurrent breakers, residual current devices) are ineffective in detecting series DC arcs. AFCI (Arc Fault Circuit Interrupter) functions can detect some types of arcs, but their reliability and selectivity in real working conditions can be problematic. In the proposed model, a key role in arc detection is played by:

The proposed testing model requires system verification, which includes a comprehensive assessment of the entire PV installation - from modules, through the mounting system, wiring, to the inverter and protections. We do not test individual components in isolation, but their interactions in realistic conditions.

Elements of system verification:

This section presents the results of simulations, tests, and analyses that support the thesis that it is possible to practically eliminate the risk of fire in photovoltaic installations through the application of a holistic approach to design and monitoring. These data illustrate the effectiveness of the proposed methodology in identifying hazards and confirm the advantage of safe, passive structural solutions.

Detailed Computational Fluid Dynamics (CFD) simulations and heat transfer analyses were conducted to determine the critical distance required between a potential DC electric arc source and the combustible roof surface. The simulations modeled arc formation conditions in typical PV installations, considering various mounting heights of modules above the roof.

A series of tests were conducted on deliberately damaged modules (e.g., with micro-cracks, delamination, damaged connectors), simulating conditions leading to electric arcs. IV curves and point temperatures were monitored simultaneously.

The use of thermal imaging cameras during tests and simulations allowed for the visualization and quantification of temperature differences in modules and connections.

Material analysis and ignition tests were conducted for various types of modules and PV installation components, focusing on their susceptibility to fire initiation and flame propagation as a result of an internal DC arc.

Based on the simulations and tests conducted, and on the physical limitations of arc propagation and ignition, a re-estimation of the probability of fire in photovoltaic systems designed and installed according to the proposed model was made.

Key findings:

The analysis and data presented so far clearly indicate a fundamental divergence between current fire safety standards in photovoltaics and the actual mechanisms of fire occurrence. This disparity leads to a situation where the industry, instead of solving the problem at its source, focuses on strategies that, while seemingly enhancing "safety," in fact constitute masking the problem or reacting to an already existing threat, rather than preventing it. Such an approach, devoid of deep engineering justification and ignoring basic laws of physics, can be termed pseudoscientific in the context of fire risk management.

As demonstrated in Section 2, dominant standards (UL 1703, IEC 61730, EN 13501-5) focus on testing the resistance of PV modules to external fire – for example, how a panel will behave under the influence of a gas burner flame or sparking roof fragments. This is logical in the context of building material classification. However, empirical data from NREL, CNBOP, and other institutions (Section 2.2) clearly show that over 75% of PV fires originate inside the installation – in damaged connectors, cables, junction boxes, or cracked cells that generate a DC electric arc.

This dissonance is paradoxical: the industry and regulators test whether a module will survive a building fire, while the real problem is whether the module or its surroundings will cause a building fire. It's like testing whether a car will survive a collision, without analyzing the causes of engine failure that could lead to that collision. The lack of reflection on this fundamental contradiction indicates a deep methodological error that undermines the credibility of current certifications in the context of fire prevention.

True safety engineering is based on causal understanding and elimination of hazard sources. In the case of PV fires, the primary cause is a DC electric arc, whose ability to initiate fire is closely related to the physics of heat transfer and radiation.

Solutions based on passive protection, such as adequate clearance, Class A modules, and proper cable routing, are not only more effective but – as demonstrated (Section 4.4, 5.3) – do not involve higher costs, and even increase energy yields (20% more from bifaciality) and reduce insurance costs. The economic logic is on the side of safety.

Despite this, the industry shows considerable inertia and resistance to adopting these principles. This resistance stems from several factors:

This discussion leads to the conclusion that the current approach to PV fire safety is fundamentally flawed and requires radical change. It is not a matter of "improving" existing standards, but replacing them with a model based on the physics of fire initiation. The lack of understanding that the goal is for the photovoltaic installation not to cause a building fire, rather than being "resistant" to it, is the main obstacle.

The proposed methodology and proven results provide an engineering response to this gap. The pursuit of "nearly 100% safety" is achievable through:

This scientific publication has demonstrated fundamental gaps in the current approach to fire safety of photovoltaic installations, proposing a paradigm shift from "resistance to external fire" to "prevention of internal fire initiation." Based on the physics of DC electric arc formation, empirical analysis, and advanced simulations, specific, engineering-justified recommendations have been formulated, the implementation of which is key to achieving nearly one hundred percent safety in PV while simultaneously increasing economic efficiency.

Current fire safety standards (e.g., UL 1703, IEC 61730, EN 13501-5) are insufficient because they do not account for the actual mechanisms of ignition in PV installations. Recommendations:

The most effective way to prevent fires is to eliminate the physical ignition path. Recommendations:

Passive protection measures should be complemented by proactive monitoring. Recommendations:

The current classification is misleading and inadequate. Recommendations:

Even the best standards won't work without proper implementation. Recommendations:

Implementation of the above reforms will bring a number of long-term benefits that extend beyond fire safety itself. Recommendations:

This section contains a list of scientific literature, industry standards, and data sources that served as the basis for the analysis, conclusions, and recommendations presented in this publication.

To provide a comprehensive understanding of fire risks in photovoltaic installations and validate our proposed prevention methodology, we developed multiple risk analysis models. These models quantify the various failure modes, their interactions, and the effectiveness of our recommended safety measures.

Fault Tree Analysis is a deductive, top-down method used to identify all potential causes leading to a specific undesired event—in this case, a PV system fire. Our FTA model maps the logical chain of failures that must occur for a fire to start.

The updated FTA includes only PV system components (excluding energy storage systems) and expands the analysis to include:

This represents a complete logical model where each node can be physically eliminated through proper design.

FMEA systematically identifies potential failure modes within a system, their causes, and effects. Our FMEA for PV systems evaluates each component and potential failure based on:

The FMEA identifies the highest priorities as:

The risk map plots Severity against Occurrence, clearly showing:

The Risk Mitigation Matrix connects each failure mode with specific engineering solutions:

While the core ignition mechanisms in PV systems remain consistent across installations, the subsequent fire propagation patterns can vary significantly depending on roof type and construction. Understanding these patterns is essential for comprehensive risk assessment and designing effective passive safety measures.

From a physical perspective, the ignition mechanism and fire spread in PV systems follows the same fundamental pathway regardless of roof type:

The ignition mechanism is universal across all roof types. Differences between roofs primarily affect the speed and direction of fire spread after ignition has occurred. This fundamental understanding reinforces our central argument: "roof burn" tests are entirely useless as ignition predictors because they analyze only what happens AFTER a fault, not why it occurred.

To facilitate practical application of our research findings, we have developed a comprehensive design and implementation procedure for fire-safe PV systems. This procedural framework integrates all key safety elements into a structured workflow from initial design through commissioning and monitoring.

The design procedure flowchart illustrates each critical stage in creating a fire-safe PV system:

The procedure incorporates safety verification checkpoints (shown as dashed lines in the flowchart) including mounting height verification, elimination of combustible materials, thermal separation, connection torque control, and thermal detection. Each checkpoint represents a critical control point in the passive fire safety design.

This section provides a direct comparison between traditional PV installation approaches and our proposed fire-safe design methodology. The comparison highlights the critical differences in safety principles, materials, construction techniques, and risk levels.

The fire-safe approach is not only safer but often economically advantageous:

This section examines the current global regulatory landscape for PV fire safety and provides specific recommendations for regulatory bodies worldwide. Our analysis reveals significant gaps in existing frameworks and outlines a path toward comprehensive reform.

Fire safety regulations for PV systems vary significantly across regions, creating inconsistent safety standards and enforcement:

This analysis reveals the critical imbalance in current regulations, with excessive focus on external fire resistance while neglecting the more common internal ignition mechanisms.

We propose a new international standard framework that addresses the identified gaps and incorporates our research findings:

We recommend a phased approach to regulatory implementation:

While this publication presents a comprehensive framework for PV fire safety based on current technology and understanding, the field continues to evolve. This section outlines promising areas for future research and emerging technologies that may further enhance PV system safety.

Several emerging diagnostic technologies show potential for further improving early fault detection:

Material innovations may provide additional safety margins beyond current best practices:

Beyond component improvements, system-level approaches show significant promise:

To accelerate progress in these areas, we propose a structured collaboration framework between industry, academia, and regulatory bodies:

This collaborative approach would focus on:

Implementation of the fire safety principles outlined in this publication requires not just regulatory change but a comprehensive educational initiative to transform industry practices. This section outlines a structured educational framework targeting various stakeholder groups.

Different stakeholders require tailored educational approaches based on their roles and existing knowledge:

We propose development of the following educational resources to address identified needs:

To maximize effectiveness, we propose a phased implementation strategy:

To ensure the educational framework achieves its objectives, we propose the following metrics:

To address potential concerns about the economic feasibility of implementing our proposed safety measures, we conducted a comprehensive cost-benefit analysis across different installation scenarios. This analysis demonstrates that fire-safe PV design is not only safer but often economically advantageous when considering total lifecycle costs.

We analyzed the incremental costs of implementing fire-safe design principles in different installation scenarios:

These incremental costs primarily come from:

These initial costs are offset by several operational benefits:

When these factors are combined in a 25-year lifecycle cost analysis, the economic advantage becomes clear:

Net Present Value (NPV) calculations use industry-standard discount rates and include all operational benefits. The analysis shows that fire-safe design not only pays for itself but generates significant additional value over the system lifetime.

Beyond direct operational benefits, fire-safe design provides substantial risk mitigation value that is often underappreciated in traditional economic analyses:

For budget-constrained projects, we recommend the following prioritization of safety measures to maximize risk reduction per dollar invested:

This section presents detailed case studies of both PV fire incidents and successfully implemented fire-safe systems. These real-world examples provide concrete validation of our theoretical framework and practical guidance for implementation.

A particularly instructive case involves two adjacent warehouse buildings with similar rooftop PV systems installed by different contractors in 2020. Both systems were approximately 350kW in size and used the same inverter technology, but with significantly different safety approaches:

The thermal event in Building A originated from a corroded MC4 connector that developed a high-resistance connection. This led to local heating that ignited the PET backsheet, which then spread to adjacent modules. The fire-safe design in Building B cost approximately 2.2% more at installation but delivered superior performance and avoided costly damage.

Following a minor fire incident that caused limited damage, a 125kW commercial system was retrofitted with enhanced safety features in 2021. This provides valuable data on both retrofit feasibility and effectiveness:

Throughout the development of this research, we have encountered various objections and resistance from industry stakeholders. This section directly addresses the most common counterarguments against our fire safety framework, providing evidence-based responses that further strengthen our position.

The most common objection to implementing enhanced fire safety measures is economic in nature, suggesting that the additional costs make PV systems less competitive.

Some industry stakeholders have questioned the technical practicality of our recommendations, particularly for certain installation types.

Representatives from standards organizations and regulatory bodies have raised concerns about the practical implementation of our recommendations within existing frameworks.

Perhaps the most challenging resistance comes from stakeholders with vested interests in maintaining the status quo, including manufacturers of components that would be discouraged under our recommendations.

Some stakeholders have questioned specific technical aspects of our research methodology and conclusions.

Rather than viewing industry resistance as an obstacle, we see it as an opportunity for constructive dialogue and collaborative improvement. Our approach includes: