A Comparative Study of Major Risk Assessment (RA) Frameworks in Geologic Carbon Storage (GCS)

Abstract

Carbon Capture and Storage (CCS) technology presents a practical solution for reducing industrial carbon dioxide (CO₂) emissions through underground anthropogenic CO₂ storage in depleted hydrocarbon reservoirs. The long-term storage efficiency faces several CO₂ leakage challenges that need to be addressed in the planning phase of the CCS project. Thus, effective risk assessment (RA) methodologies are crucial for ensuring safety, regulatory compliance, and public acceptance of CCS projects. This review examines RA parts and their corresponding technical and non-technical challenges. The analysis critically compares over 20 qualitative, semi-quantitative, quantitative, and hybrid RA techniques employed throughout GCS operations. Available quantitative RA tools do not deliver dependable results because they require technical data that become available late in the CCS project development process. Qualitative approaches work well for the initial screening of storage sites with limited data available, yet quantitative methods enable quantification of CO₂ leakage. For the first time, a comparative analysis of two integrated assessment tools is presented in this paper. The techniques achieve success based on high-quality data and analysis of existing technical and non-technical challenges which this paper examines. The comparative analysis outlines the limitations and advantages of every methodology studied and emphasizes the need for integrated hybrid frameworks to boost decision-making in the RA process. Future research should focus on creating or improving existing hybrid frameworks for late-stage RA while utilizing qualitative frameworks in the initial site screening stage to advance GSC’s safe and effective implementation.

1. Introduction

As countries seek sustainable ways to reduce greenhouse gas emissions, CCS has become one of the demanding strategies in the changing global energy sector [1]. The increasing concentrations of anthropogenic CO₂ in the atmosphere pose a threat to global efforts to prevent climate change. Recent studies investigate the full spectrum of CCS, reviewing capture technologies alignment with net-zero goals [2], deployment challenges [3], global innovation trends and advances in CCS techniques [4,5], concise policy and economic roadmaps for future implementation [6], as well as operational insights and lessons from reservoir management and monitoring [7], which collectively underscore the multidimensional progress and importance of CCS development.

The Earth’s atmosphere now has the highest recorded CO₂ levels, which reached 424.61 parts per million (ppm) as the annual average and 427.09 ppm during February 2025 [8]. The increase in global CO₂ emissions since 1990 resulted in more than a 60% surge, reaching 37.41 billion metric tons (BMt) CO₂ in 2024 (Figure 1). Operational CCS facilities worldwide have increased their capture capacity from 28 MtCO₂ per year in 2014 to approximately 50 MtCO₂ per year in 2024. While net-zero emission targets demand approximately 1000 MMtCO₂ capture capacity, the predicted increase in CO₂ capture capacity for 2030 stands at 435 MMtCO₂. Global CCS investment needs to grow its current value by 45 times through yearly expenditures averaging USD 500 billion during the 2024–2030 period to meet the emissions targets [9,10].

CCS has proven itself a vital industrial technology by securely storing CO₂ within geological formations for emission reduction. Under U.S. leadership, 231 CCS projects exist at different development levels, and 17 current operational facilities benefit from policy support through the Inflation Reduction Act (IRA) (

Table 1. The most extensive CCS facilities in the U.S. by the Global CCS Institute [11].

The Most Extensive CCS Facilities in the US	CO2 Capture Capacity (MMT/yr)
ExxonMobil Chute Creek Gas Processing Plant	7
Longfellow WTO Century Plant	5
Great Plains Synfuels Plant and Weyburn-Midale	3
Petra Nova Carbon Capture	1.4
ADM Illinois Industrial Carbon Capture and Storage	1
Air Products and Chemicals Valero Port Arthur Refinery	0.9
Coffeyville Gasification Plant	0.9
Contango Lost Cabin Gas Plant	0.9

) [11,12]. Expanding CCS operations still faces multiple obstacles based on long-term storage security, economic viability, and public perception. The success of CCS heavily relies on site selection, regulatory compliance, financial and political stability, technological capability of monitoring, and storage integrity of CO₂ [11].

Figure 1. Annual CO₂ emissions worldwide from 1940 to 2024 in billion metric tons (BMT/y) [13].

Figure 1. Annual CO₂ emissions worldwide from 1940 to 2024 in billion metric tons (BMT/y) [13].

Another crucial component of the CCS project’s success is risk assessment (RA), which guarantees the safe storage of CO₂ for millennia. RA frameworks help us to evaluate potential technical hazards such as CO₂ leakage through legacy wells, geologic faults, cement integrity issues, casing corrosion, leakage from plugs, brine migration, induced seismicity, etc., and non-technical challenges such as public perception, financial security, regulatory compliance, etc. [14,15,16,17,18,19,20,21,22,23,24,25]. However, uncertainty and limited data surrounding the GCS system trigger constant updates and refinement of existing RA frameworks. For example, the well integrity of legacy wells is a significant threat, as unplugged abandoned wells or poorly sealed plugs have potential leakage pathways for CO₂ in depleted oil and gas reservoirs [26]. Recently, the State of Texas introduced a bill to plug more than 150,000 inactive legacy wells in the next 15 years across the state that pose a risk of leakage, highlighting the importance of RA [27,28].

The research examines more than 20 RA tools used in GCS operations, including all four RA methodological scales: qualitative, semi-quantitative, quantitative, and integrated. Figure 2 depicts the risk levels of CCS projects, with the early time being the highest risk level [29]. The study shows that early-time RA can be accomplished through qualitative methods, which is helpful regarding limited data. However, these techniques may not identify high-risk formations and require much data, making them open to human errors. Conversely, quantitative frameworks afford more accurate assessments at the expense of requiring large datasets to gauge CO₂ leakage rates accurately. Recently, integrated models have incorporated qualitative and quantitative methods, gaining popularity due to enhanced decision-making. For the first time, this review provides a comparative analysis of two such integrated models to help stakeholders determine tools appropriate for particular CCS projects’ needs.

This review demonstrates the necessity for additional studies regarding integrated (hybrid) RA frameworks, combining qualitative and quantitative approaches for optimizing CCS deployment at a large scale. Through advancements in risk analysis models, improved uncertainty management, and more effective risk prevention protocols, stakeholders can enhance the security of CCS projects alongside operational performance while gaining public acceptance. The evolution of risk analysis methodologies is crucial to successful CCS implementation because it is a key strategy for mitigating climate change.

The paper is divided into three main sections. First, it explains the basic elements of RA for GCS, detailing its crucial components and technical and non-technical challenges. The following section depicts an extensive critical review of more than 20 RA methodologies, including two integrated frameworks. Finally, the paper discusses the results based on the analysis conducted in each section.

The novelty of critical review lies in its comprehensive evaluation of all four primary RA tools currently available, highlighting their strengths and limitations and addressing both technical and non-technical risks within a single study. Previous reviews, including those by Condor et al. [30], Li et al. [31], and others did not include all existing RA methods, particularly the recent integrated framework. Other studies, such as those by Damen et al. [32], focused solely on technical risks, while Sneddon et al. [33] only performed categorical overviews without a critical evaluation of practical tools. This review provides a more holistic understanding of RA methodologies within GCS by addressing these gaps.

2. Fundamentals of Risk Assessment in Geologic CO₂ Storage

2.1. Definition and Importance of Risk Assessment

The RA of GCS is a conceptual plan that outlines steps for recognizing, evaluating, and managing hazards linked to the geologic storage of carbon dioxide. These risks include leakage of CO₂, risks arising from induced seismicity, brine displacement, and environmental and social risks to the fauna and flora and residents living nearby [34]. Efficient RA is critical for making geologic CO₂ storage safe, reliable, and acceptable to the public as a climate change mitigation measure.

The industry’s CCS initiatives have a noticeable lack of documented RA failures. The only natural catastrophe often referred to is that which occurred at Lake Nyos, Cameroon, in August 1986. Although this event is not related to CCS, this kind of event immediately drew public concern about the risks of leakage of CO₂ and increased pressure on the appropriate safety and RA of CO₂ storage. This underlines the need for rigorous RA approaches to dent-specific issues and improves the perception of CCS technologies [35].

RA is crucial due to a reduction in uncertainty that might surround a project. Improvements to project designs, together with regulatory compliance, generate significant benefits. A rigorous approach to RA, based on the likelihood of future difficulties, allows stakeholders to build effective monitoring, mitigation, and maintenance strategies [36]. It also builds confidence among policymakers, investors, and market participants, who are key to deploying CCS on a larger scale worldwide.

Due to the safe, long-term storage of CO₂ in geological formations, employing CCS technology may exert considerable influence on the sustainable use of GCS. Therefore, it is necessary to follow systematic strategies.

2.2. Components of Risk Assessment

The long-term success of the CCS project relies on a comprehensive analysis of potential leakage risks of CO₂. This process is a step-by-step approach to assessing risks that are likely to affect the reliability, effectiveness, and sustainability of CO₂ storage [32]. RA consists of the following fundamentals: hazard identification, risk analysis, risk evaluation, and risk mitigation (Figure 3). These elements collectively assist in growing strong measures to overcome any risks and future contingencies linked to GCS operations.

Hazard identification is the first step in the risk assessment process [9]. It requires the identification of specific incidents or circumstances that may result in the release of the stored CO₂ or impact on the storage site. In this stage, one looks at the geological suitability of the area in the context of caprock to determine whether it is stable or likely to slip along the fault or area of seismic activity and other operational risks that may affect well integrity, as has been determined in some case studies [37]. Identifying hazards is the primary key to assessing the range of risks and providing the framework for the above analyses.

Risk analysis is the tool that helps improve the identification of hazards by evaluating the probability and potential impacts of identified risks [38]. This stage may use computer models, historical data, probabilistic methodologies, or other quantitative and qualitative approaches to determine the size and likelihood of undesirable incidents [39]. Considerations for CO₂ plume migration include pressure fluctuations, possible implications on ecosystems, and human health. This analysis allows stakeholders to obtain insight into the most serious hazards to the storage facility and prioritize risk mitigation activities accordingly [32].

The next phase is risk evaluation, which involves comparing the hazards that have been identified to the degree of risk acceptance. This process helps determine if the risks that the corporations have identified are within acceptable bounds for the project’s stakeholders [36], which include community organizations, regulatory bodies, and environmental non-governmental organizations [38]. In situations where specific risks are deemed unacceptable, it may be necessary to redesign aspects of the project or implement additional control measures to address those risks.

Risk mitigation in CO₂ storage involves implementing strategies to mitigate identified risks, such as CO₂ migration, leakage, and induced seismicity. This process is essential for ensuring the safety and reliability of GCS operations and optimizing monitoring plans [30].

Risk management constitutes the final framework of risk assessment. After the risks have been assessed, the crucial approach to be taken should be risk management to reduce the risks and occurrences of dangerous consequences in the CCS project [36]. To avoid experiencing the consequences of risks, risk management solutions entail developing monitoring and confirmation programs, modifying operations, and preparing contingencies. Thus, by managing identified risks, potential negative impacts on technology sustainability, the environment, and public health can be minimized [38,40].

2.3. Technical Challenges in RA for CCS

RA of CCS is accompanied by several technical challenges, mainly associated with geological variability, complexity of operations, and technical limitations. Another area of interest entails the assessment of CO₂ migration in subsurface formations (Figure 4) to avoid contamination with Underground Sources of Drinking Water (USDW). Moreover, a paramount issue is the well integrity of legacy wells, which is considered a leakage pathway for CO₂ to the subsurface [39,41]. Forecasting risks associated with leakage and induced seismicity adds another layer of complexity, considering subsurface behavior is inherently unknown. The complexity of geology makes it difficult to create effective strategies, and the lack of information about the condition of the legacy wells complicates the process of safe application of CCS projects. Addressing these concerns is critical, especially for large-scale CCS operations that require extensive risk mitigation techniques and cutting-edge technology solutions to improve project feasibility and safety [2].

Geomechanical risks during injection tests are also considered a technical challenge in RA for the CCS project [42]. Key challenges include accurately calculating injection pressure thresholds to prevent formation fracturing and sand production [43]. These risks are influenced by factors like injection pressure thresholds and temperature differences between the storage formation and injection fluid, which can significantly affect fracture initiation pressure [44].

Figure 4. U.S. Environmental Protection Agency Underground Injection Control Program: 1—Confining layer; 2—CO₂ Injection formation; 3—inactive fault; 4—Water quality tracked at surface; 5—Monitoring well to track CO₂ pressure; 6—Permitted CO₂ injection well; 7—High-quality cement preventing CO₂ migration; 8—Corrosion-resistant well material; 9—High-quality plugs; 10—Monitoring of seismic activity; 11—Isolated groundwater well; 12—Seismic surveys to study CO₂ plume movement underground; 13—CO₂ injection pressure monitored to prevent formation fracturing; 14—Testing to confirm the well integrity; 15—Monitoring of flow and pressure; 16—Response plan (adjusted from [45]).

Figure 4. U.S. Environmental Protection Agency Underground Injection Control Program: 1—Confining layer; 2—CO₂ Injection formation; 3—inactive fault; 4—Water quality tracked at surface; 5—Monitoring well to track CO₂ pressure; 6—Permitted CO₂ injection well; 7—High-quality cement preventing CO₂ migration; 8—Corrosion-resistant well material; 9—High-quality plugs; 10—Monitoring of seismic activity; 11—Isolated groundwater well; 12—Seismic surveys to study CO₂ plume movement underground; 13—CO₂ injection pressure monitored to prevent formation fracturing; 14—Testing to confirm the well integrity; 15—Monitoring of flow and pressure; 16—Response plan (adjusted from [45]).

2.4. Non-Technical Challenges as Barriers for CCS Projects

Non-technical challenges can be categorized as illustrated in

Table 2. Main non-technical challenges.

Name of Challenge	The Nature of the Challenge
Public perception	Local people oppose the implementation of the CCS project due to the risks of CO2 leakage to the environment.
Regulatory concerns	Insufficient regulatory support from governmental institutions.
Financial risks	Absence or insufficient attention from financial institutions to the CCS projects.
Political challenges	Political instability towards aligning national interests with global climate goals.

, providing a structured and academic approach to their classification. Non-technical challenges arise from external factors outside the direct scope of CCS operations but significantly impact the project, potentially leading to deviations from its intended goals and delays [46]. These challenges and concerns have the potential to create delays and overruns of funds, hence jeopardizing the proper management of CCS projects. In order to preserve project control and guarantee successful outcomes, it is crucial to identify and resolve non-technical challenges and uncertainties promptly [47].

2.4.1. Public Perception

Public perception is critical to adopting and implementing CO₂ storage projects and is pivotal to their success. Scientifically, however, very little is known about this subject to the public, with only 6.4% admitting familiarity with CO₂ storage projects and only 11.5% opposing it in the US [48]. Lack of technological knowledge, past experiences with organizations, and contestable soil information probably influence CO₂ storage initiatives among the public. Early community engagement and educational programs can improve understanding and acceptance [49].

Public perception is acknowledged as a key risk that might hinder the effective deployment of CCS programs. The use of CO₂ injection ought to be safe and free from the danger of fatalities or serious injury to people. This framework works under the assumption that risks believed to be sufficiently low are likely to be accepted in the community. This approach has gained so much popularity and has been used by many industries in a global manner [36].

The cases of successful CCS implementation in the leading countries reveal that even where a project carries economic sense, its cancellation is possible without a proper strategy of engaging the local communities. A notable example is the Barendrecht project in the Netherlands, proposed by Shell. These disagreements resulted in massive demonstrations, for which the government was forced to ban this project and other CCS projects, such as onshore storage of CO₂. This case reveals how preliminary studies should be conducted before implementing project development. Such studies should strive to improve public perception of CCS technologies while also measuring public understanding and examining social attitudes toward these technologies [50]. As some studies show, public perception is the second risk that might jeopardize the CCS project [51], popularizations methodologies of scientific achievements and new technologies presented by some researchers [52].

To eliminate any opposition from local communities, the EPA (Environmental Protection Agency) suggests engaging with the public before applying the project to the agency. The EPA’s regulatory framework for Underground Injection Control (UIC) programs emphasizes the importance of public participation, particularly for Class VI permits related to geological sequestration (GS) projects. To ensure affected communities have opportunities to engage, the EPA publishes draft permits for public review, typically allowing a minimum 30-day comment period. Notifications are disseminated through various media and mailing lists, enabling community members to submit comments, attend public hearings, or appeal final permitting decisions. Establishing early communication fosters trust and inclusivity while addressing concerns such as disproportionate or cumulative environmental impacts. Effective community engagement involves tailoring outreach efforts to meet community needs, such as offering multilingual materials or flexible meeting times, and conducting research to understand local concerns [53].

Although the EPA has outlined ways to engage with communities before implementing the project, it has no comprehensive strategies for dealing with public perception issues that might arise in the early or even later stages or during the implementation process. Nevertheless, this gap shows that public perception is still one of the major non-technical issues that should be investigated and addressed extensively.

2.4.2. Regulatory Concerns

Several regulatory challenges must be addressed for the successful implementation and global adoption of CCS projects. Current laws sometimes lack clarity, especially concerning legal issues such as legal responsibility, environmental regulations, and intellectual property rights, which might make it difficult to adopt CCS. These features raise significant challenges that limit the deployment of CCS technologies and, therefore, their effectiveness in providing sustainable energy solutions [54,55].

There is also a lack of regulatory clarity on the offshore storage of CO₂, where international collaboration is necessary. Particularly under conventions like the London Protocol, which regulate marine CO₂ storage, the necessity of global cooperation is underlined [54]. In order to ensure environmental protection and productive international collaboration in CCS projects, regulatory concerns in CO₂ storage under the London Protocol include adherence to specifications for CO₂ stream composition, assessment processes, permit issuance, and reporting to the International Maritime Organization [56].

Regulatory challenges also arise when it comes to transporting CO₂ through international borders; some states within one country lack regulatory clarity regarding the transfer of CO₂. Moreover, since geological storage of CO₂ aims at its long-term containment, the current legislation structure does not have legal liability mechanisms in cases where the CO₂ leaks after decades. It therefore suggests that mechanisms such as straightforward assignment of liabilities and the introduction of the proper forms of insurance should be developed to sustain responsibility and project future [57]. Several other significant challenges, therefore, include coordinating the laws across jurisdictions; determining exact rights to the pore space; and implementing mechanisms for long-term management and responsibility [55].

2.4.3. Financial Risks

Financial risks and some barriers to CCS project investment have been acknowledged as crucial issues. The type, source, and cost of the energy used in CCS projects have a significant influence on operational costs and, therefore, on investment decisions [58]. Financial risks are categorized (

Table 3. The categorization of financial risks at various stages throughout the lifecycle of a CCS project.

Operational Costs				Investment
Capture		Transportation	Storage	Environmental Impact	Skepticism
Factors	Plant dimensions	Distance to the injection site	Geological specifications	CO2 leakage from storage and pipelines	Technological maturity
	Fuel characteristics	Route complexity	Project dimensions
	Fuel cost	Pipeline diameter	Storage type	Consequently, high financial costs due to environmental damage
	Interest rates	Location of storage site (off or onshore)	Specifications for MVA (monitoring, verification, and accounting)
	Plant lifetime

) into two major sections: operational expenditure and investments.

In turn, operational expenses are divided between capture, transportation, and storage costs, each of which individually contains financial risks and challenges [47]. It is imperative to select economically feasible plant size, fuel characteristics, and net efficiency for CCS plant capture technologies to keep the CCS project financially competitive. Moreover, parameters of the price of fuel, interest rates, and plant lifetime play a crucial role in the costs and success of the CCS project [59]. These financial risks, primarily associated with the capture process, must be addressed early in the design phase, as the selection of the capture method should also be guided by economic feasibility.

There are mainly three factors associated with the financial risks of transportation of CO_2, namely the distance from the plant to the injection site, the route complexity, and the pipeline diameter [60]. It is well known that the diameter of a pipeline depends on the phase of CO₂, with dense-phase CO₂ requiring a smaller diameter and vice versa. Although pipeline costs are relatively consistent across countries, the costs for offshore CO₂ storage sites can vary significantly due to the additional monitoring and maintenance techniques required. Generally, CO₂ transportation to offshore storage sites is more expensive than to onshore storage sites. However, to mitigate the higher costs, transporting CO₂ using ships could be a cost-effective alternative, mitigating the financial risks of transportation [61]. The selection of cost-effective transportation methods is crucial to managing the financial risks of CCS projects.

Another critical factor in implementing CCS projects is the cost of selecting a storage site. Factors influencing these costs include geological characterization, project dimensions, the type of storage site, and specifications for monitoring, verification, and accounting (MVA). The storage site itself is characterized by geological features that significantly affect the project’s economic feasibility. Key geological characteristics include the thickness, porosity, permeability, lateral continuity, and depth of the storage formation, all of which determine its suitability for CO₂ injection and have a direct impact on storage costs [61]. Accurate evaluation of geological features, scale factors, and pressure management is essential for cost-effective CO₂ storage and the success of CCS projects [47].

Regarding investments as financial risk, two significant difficulties must be addressed, namely the reduction in environmental effects and the doubts about the efficiency of CCS technology. Some of CCS’s critics argue that the technology is still not sufficiently reliable and they need more studies to solve all the existing ambiguities before the broad application of this technology [47,62]. Others do not want to invest in their CCS projects due to the potential leakage of CO₂ into the atmosphere from storage and pipelines; the legal procedures required to address this issue may cost them considerable money.

To eliminate CCS operational and investment-related financial risks, it is necessary to look for investment and sources of financial capital for CCS projects. More efforts should be made to prove that these technologies are now mature, cost-efficient, and environmentally friendly by developing research and pilot projects. CCS should not be the primary sustainable energy solution method. It has to be a transitional solution to cut down on the carbon footprint. This approach makes CCS a valuable tool for combating climate change, though it responds to other financial gaps shown in Table 3. Adaptive independent monitoring, verification, management, and transparent legal frameworks and standards are prerequisites to achieving CCS projects’ safety, compliance, and integrity. They build the confidence of stakeholders, create goodwill among investors, and market CCS as a reliable solution to advancing the journey to a more climate-resilient economy.

2.4.4. Political Challenges

Political instability poses a substantial threat to aligning national interests with global climate goals, especially within the framework of CCS implementation. The practical implementation of CCS technology depends on the interaction of public investment, political will, and international collaboration. Based on the available literature, political challenges could be categorized (

Table 4. Categorization of political challenges.

Political Challenges
Contributing Factors	National Political Challenges	International Political Challenges	Cross-Cutting Challenges
	1. Policy Instability	1. Geopolitical Disputes	1. Investor Volatility
	2. Lack of Regulatory Frameworks	2. Weak International Agreements	2. Long-Term Commitment Issues
	3. Permitting and Licensing Delays	3. Accountability and Enforcement	3. Global Coordination Issues
	4. Weak Government–Private Sector Collaboration

) as national, global, and cross-cutting challenges.

National political challenges mainly arise from policy shifts that are the result of changes in governments. As CCS projects are long-term projects associated with a high degree of trust between the private organizations and the local government, any political changes affect the investment heavily, thus making it risky [63]. Therefore, for these projects, it is possible to state that to conduct them successfully, it is necessary to have a stable government and proper communication between the government and private businesses [64]. The absence of clear regulatory frameworks on a national level for issuing permits and licenses jeopardizes project implementation and investors’ confidence [65]. Governments, businesses, and civil society must work together to address regulatory frameworks, which are essential obstacles. Moreover, difficulties and delays in obtaining permits and licenses threaten the timely implementation of CCS projects and maintaining investors’ interest in funding the project [47]. A clear and supportive regulatory framework for obtaining and approving permits and licenses is essential to eliminate national political challenges.

International political challenges mainly arise from geopolitical disputes and conflicts, international agreements lacking enforcement mechanisms, and global regulatory frameworks to push or penalize countries for not implementing CCS projects. For instance, the Paris Climate Agreement does not have any mechanisms to sanction or penalize countries that do not comply with the international environmental agenda, even if the government must reduce CO₂ emissions by applying sustainable energy solutions, CCS in particular. Therefore, some countries continue to produce CO₂ on a large scale without facing any penalties [66,67]. The weak international agreements and the absence of a regulatory body with the authority to enforce government compliance with CCS projects pose a significant threat to the international climate agenda. Therefore, it is crucial to establish an international regulatory body with the power to hold countries accountable and enforce compliance with the global agenda through sanctions and fines [68,69]. The international community should address these gaps from international political challenges.

Some cross-cutting political challenges are formed by the contribution of national and international challenges as well. Political instability, both at the national and international levels, can negatively impact the global pace of CCS implementation, rendering investments risky and unreliable. Long-term commitment issues arise from political and financial commitments, where CCS project implementation is prolonged or postponed due to national policy changes. Here, limited enforcement mechanisms in the agreements (national and international) keep the involved parties unpunished. Lastly, limited cooperation among nations and the absence of institutional mechanisms to collaborate with countries hinder the creation of a stable global atmosphere for CCS [67].

In summary, the main political challenges can be identified in gaps such as unstable governance, ineffective regulatory frameworks, a conspicuous lack of cooperation between governments and the private sector, and low international cooperation. Addressing these issues would facilitate improvement in the investment system, clarify market regulation, and provide appropriate conditions for CCS development. In this manner, objective accomplishments regarding low-carbon emissions and global environmental goals will be achieved [47].

2.5. Historical Background of RA Tools in GCS

Those mentioned above technical and non-technical challenges exclusively identify all respective challenges and associated gaps. To provide more comprehensive information about the need for and importance of CO₂ storage, a review of the historical background is presented in this subchapter. As global interest in GCS has dramatically increased over the past few decades, the development of RA has also evolved significantly during this time. This review divides the historical development into three primary timelines: early foundations (2000–2010), middle times (2010–2015), and present times (2015–present).

The time period from 2000 to 2010 involved qualitative RA tools, primarily the application of environmental risk frameworks derived from the petroleum and nuclear waste industries. For example, one of the first comprehensive tools, the CO2QUALSTORE guideline by DNV [70], established a systematic procedure for identifying potential failure modes, especially in saline aquifers and depleted reservoirs. While the US DOE-NETL and EPA developed their qualitative RA frameworks to address risks in GCS as VEF (Vulnerability Evaluation Framework) [71].

From 2010 to 2015, probabilistic modeling and semi-quantitative matrices became the focus of RA tool development. The National Risk Assessment Partnership (NRAP) Toolset modeled pressure migration processes, brine displacement dynamics, and wellbore integrity behavior using physics-based simulations [72]. The RISCS project worked alongside other groups in Europe to create specific protocols that merged geological with ecological RA [73].

The recent literature has shifted toward integrated RA frameworks that combine site characterization, monitoring data, and real-time simulations. International standards under ISO 27914:2017 [74] play a key role in advancing common practices with standardized regulations and in promoting integrated RA frameworks [75]. The modern integrated RA tools NRAP-Open-IAM, REX-CO₂ use real-time risk updates, integrating qualitative and quantitative features of RA, validated through empirical case studies at Sleipner, In Salah, and the Illinois Basin–Decatur Project [76,77].

A more detailed and critical evaluation of the various RA tools is provided in Section 3 below.

3. Overview of over 20 RA Methods

Risk assessment (RA) methodologies are typically divided into two primary categories: qualitative and quantitative. Since the field of RA and CCS is developing semi-quantitative and integrated (hybrid) methods available on the market. This chapter will discuss over 20 RA methods available on the market and give a comparative analysis of qualitative, semi-quantitative, and quantitative models separately.

3.1. Qualitative Methods

Qualitative risk analysis aims to determine possible hazards/risks and develop mitigation solutions without giving numerical estimates of the potential or effect of specific risks or scenarios on the project. Instead, it uses expert judgment and historical data to determine which risk variables pose the biggest threat to a project, in terms of compatibility with health, safety, and environmental (HSE) regulations [78]. Qualitative risk assessment methods are extremely easy to apply in the early stages of a CCS project when the numerical data are limited or absent. As CCS projects require assessment of technical (CO₂ leakage, induced seismicity, etc.) and non-technical (public perception, regulatory, political, etc.) challenges, the qualitative RA permits quick and structured decision-making under the degree of uncertainty. Qualitative RA is considered more effective and sufficient when there is a lack of numerical data, time, knowledge, and experience [30].

Table 5. Overview of qualitative RA tools/methods studies.

Name	Company/Author(s)	Description	Year
FEP (Features, Events, Processes)	Quintessa	Qualitative	2010
VEF (Vulnerability Evaluation Framework)	U.S. EPA	Qualitative; expert estimation of vulnerable conditions	2008
SRF (Screening and Ranking Framework)	LBNL Oldenburg	Qualitative; expert-elicited probabilities	2005
SWIFT (Structured What-If Technique)	Sollie et al. and DNV (Det Norske Veritas)	Qualitative; group-based assessment	2010–2011
CASSIF (Carbon Storage Scenario Identification Framework)	TNO, Yavuz et.al	Qualitative, scenario-based	2008
DPSIR (Drivers, Pressures, State, Impact, and Responses) framework	Ecofys Netherlands	Qualitative, casual, and systematic RA framework	2010
MCA (Multicriteria Assessment)	Gough and Shackley	Qualitative, scenario-based	2006
ESL (Evidence Support Logic)	Shell (TESLA tool)	Qualitative, evidence-based	2013
RA-CO2	Paraguassú et al.	Qualitative, inductive	2013
Bow-tie diagrams	BRGM (GERICO)	Qualitative; systematic decision-making bow-tie diagrams	2011

illustrates several useful qualitative RA tools examined and compared in this chapter.

3.1.1. Features, Events, Processes (FEP) Method

The Features, Events, and Processes (FEPs) method provides a highly systematic and transparent structure to distinguish and explain the characteristics of the several systems specifically under study. In the context of CO₂ storage, the system may include the overall storage framework together with the reservoir, caprock, overburden, side burden, underburned, and wells, or it may consist of only a selected part of that framework, for example, the well and its immediate surroundings [79,80,81]. Features are site-specific qualities such as faults, reservoir permeability, porosity, etc. In the context of CO₂ storage, events that could be considered are earthquakes, induced seismicity, well blow-out, etc. Processes are physical interactions with features such as geochemical or geomechanical processes [42]. The Quintessa database presents more than 200 FEPs that are divided into eight categories and subcategories [79,82]

There are two main approaches to analyzing the FEP method: the bottom-up and top-down. The ‘top-down’ methodology first defines major high-level FEPs and classifies the identified FEPs into scenarios, further developing the scenarios into detailed conceptual models for modeling. A ‘bottom-up’ approach starts with a detailed description of the site where hazards could occur and has a lot of FEPs involved, and these are integrated to come up with conceptual models and scenarios [83,84,85,86,87]

Abid et al. utilized the FEP method, integrated it with the interaction matrix IPM (Integrated Potential Matrix), and converted qualitative risk assessment into a semi-quantitative one using cause-effect plot diagrams. The study of two fictitious wells was used to identify key “features”, “events”, and “processes” in terms of CO₂ storage safety. As per the authors, the primary safety concerns were the chemical reaction of water and subsidence with the cement and casing. The risk assessment tool aided in identifying the actions required to remediate the identified risk through multiple barriers (primary and secondary) within the wells. However, there is a need for expert judgment and testing for accuracy [88].

The FEP qualitative risk assessment method considers various factors and aids in stakeholder communication and traceability. However, the technique is time-consuming and complex for the initial stage of a CCS project, where quick decision-making is preferred.

3.1.2. Vulnerability Evaluation Framework (VEF) Method

VEF is a qualitative risk assessment method developed by the EPA. The technique identifies attributes (permeability, wells, faults, etc.) of the CCS system that can make them vulnerable to adverse effects, outlines impact categories, and offers systematic assessment by decision-making flowcharts Figure 5 [71]. Firstly, injection and confining zone-related attributes that could influence the Geological Storage System (GSS) were identified. Furthermore, the spatial area for assessing the adverse effects of migration, leakage, or unexpected pressure change is defined using the presented method. Lastly, the potential effect categories and important receptors identified include human health and welfare, the atmosphere, groundwater, surface water, and the geosphere. The final report of VEF qualitatively defines whether the project has Low or Elevated vulnerability. Low vulnerability means that adverse effects are unlikely to occur on assessed attributes of the system, while elevated vulnerability means that attributes assessed require special attention [89]. The VEF is not meant to choose sites, set performance standards, or specify data requirements. Instead, it is a helpful guide for regulators and technical experts to pinpoint areas that need design evaluation, specific risk assessment, monitoring, and management [71,90].

Although the method provides systematic, holistic decision-making tools, its complexity and data requirements make it time-consuming, and it lacks uncertainty management for elevating vulnerable projects. Also, the method does not consider operational and CO₂ transportation-related risks (attributes), which makes it more specific. Existing gaps could be addressed by integrating with other models to provide a quicker and more comprehensive risk assessment.

3.1.3. Screening and Ranking Framework (SRF) Method

The Screening and Ranking Framework (SRF), presented by Oldenburg et al. [91], aims to estimate potential geologic CO₂ storage sites utilizing the rating of health, safety, and environment (HSE) risks associated with CO₂ leakage. This framework describes risks of CO₂ leakage based on three fundamental characteristics of a geologic CO₂ storage site: (1) the capability of the target formation to provide primary containment, (2) the ability of secondary containment to mitigate leakage if the primary formation fails, and (3) the potential for the attenuation and dispersion of CO₂ if both primary and secondary containment measures fail. SRF requires general input, often relying on expert opinion based on the level of site characterization and available published information. The framework’s assessment is structured around four information classes (

Table 6. SRF methodology (adjusted after [91]).

Characteristics	Attributes	Properties	Prox For…
Potential for primary containment	Primary seal	Thickness Lithology Demonstrated sealing Lateral continuity	Likely sealing effectiveness Permeability, porosity Leakage potential Integrity and spill point
	Depth	Distance below surface	Density of CO2 in reservoir
	Reservoir	Lithology Permeability, porosity Thickness Fracture or primary porosity Pore fluid Pressure Tectonics Hydrology Deep wells Fault permeability	Likely storage effectiveness Injectivity, capacity Areal extent of injected plume Migration potential Injectivity, displacement Capacity, tendency to fracture Induced fracturing, seismicity Transport by groundwater Likelihood of well pathways Likelihood of fault pathways
Potential for secondary containment	Secondary seal	Thickness Lithology Demonstrated sealing Lateral continuity Depth	Likely sealing effectiveness Permeability, porosity Leakage potential Integrity and spill point Density of CO2
	Shallower seals	Thickness Lithology Lateral continuity Evidence of seepage	Likely sealing effectiveness Permeability, porosity Integrity and spill point Effectiveness of all seals
Attenuation Potential	Surface characteristics	Topography Wind Climate Land use Population Surface water	CO2 plume spreading Plume dispersion Plume dispersion Tendency for exposure Tendency for exposure Form of seepage
	Groundwater hydrology	Regional flow Pressure Geochemistry Salinity	Dispersion/dissolution Solubility Solubility Solubility
	Existing wells	Deep wells Shallow wells Abandoned wells. Disposal wells	Direct pathway from depth Direct pathway Direct pathway, poorly known. New fluids, disturbance
	Faults	Tectonic faults Normal faults Strike-slip faults Fault permeability	Large permeable fault zones Seal short-circuiting Permeable fault zones Travel time

): (1) site characteristics, (2) attributes defining these characteristics, (3) properties defining the attributes, and (4) user-input values defining the properties [91]. The SRF is represented with a spreadsheet where the user can input available limited data and plot the numerical points with automatically generated lines where the sealing formation’s poor, fair, or good confinement is defined. Oldenburg et al. [91] applied the method in three case studies for potential CO₂ storage. The formations were evaluated with limited data and in a short period. The author emphasizes that the presented method is valid only for qualitatively screening the storage sites; further steps require more in-depth quantitative and simulation models for the final decision.

The presented methods allow users to make rapid decisions with relatively limited data from open sources and publications, which is vital at the initial stage of any CCS project. This method is also advantageous in terms of transparency and ease of use. However, there is still a need for more in-depth evaluation of the storage site, testing on the actual storage site, and uncertainty in predictions.

3.1.4. SWIFT (Structured What-If Technique) Method

SWIFT is a group-based technique used by DNV in consultation with GE Plastics to identify risk and uncertainty within commercial operations and industrial sectors, particularly in CO₂ storage. Originally used in engineering and heavy lifting, it was transposed to CO₂ storage when DNV was helping Norway identify locations for offshore storage. The SWIFT method is advantageous in comprehensive, efficient, and proactive hazard identification.

Some aspects of applying the SWIFT method include the ability to identify risks qualitatively, generate suggestions of risk control measures, and take advantage of expert opinion. It includes brainstorming techniques, discussions, and checklists that systematically discuss hazards from a global perspective, shifting the focus from objects to systems or operations [92].

Discussions and different segments on the CO₂ storage site are available in the following categories: reservoirs, seals and overburden, wells, and atmosphere. The workshops follow checklists based on the Quintessa database of the Features, Events, and Processes developed with the IEA Greenhouse Gas R&D Program (IEAGHG). These ensure intensive identification of the risks involved, forming the root for future risk assessment. Experts participating in the discussions must be selected based on their outstanding field knowledge. Discussions focus on the causes, potential repercussions, and safeguards for each hazard, emphasizing the need to examine all possible risks, regardless of perceived likelihood [93].

This approach is constructed with a series of questions (What If?; How could?; etc.) to identify the risks or potential dangers to the project’s safety. SWIFT is considered an easy qualitative estimation method of hazards, where other complex hazard identification techniques do not work or are too complex. The process is also advantageous due to its adaptability, time-saving, clear documentation, and top-down approach (see FEP method above). However, the qualitative nature and limited details make it less reliable, and the technique highly relies on the experts participating. Here, biases based on participants’ opinions might result in subjective results. The method qualitatively identifies the hazards without giving any probabilities or likelihood of events, which is needed for advanced decision-making.

3.1.5. CASSIF (Carbon Sequestration Scenario Identification Framework) Method

CASSIF is a scenario-based analysis framework developed based on the FEP scenario analysis database. The framework assesses containment ability regarding fault and seal integrity [89]. It consists of four modules, as described in

Table 7. Modules of CASSIF and their description.

Module	Description
FEPQuest	This module represents a questionnaire (40 questions) that experts must fill in. This allows us to have an initial understanding of the storage site.
FEPMan	Based on answers received from the first module, FEP will be selected based on answers that need more attention. In this module, qualitative data deemed uncertain will be selected for further discussion during the workshop. Experts must choose the FEP numbers already provided in the FEPMan database.
FEPChain	Markov Chains are utilized to describe the response of highlighted FEPs to one another, and based on probabilistic models, the most likely state of the system is predicted.
FEPMon	Based on the above assessment, monitoring and response actions are presented to ensure the project’s safety.

.

This module requires less time to analyze risks than the FEP model from which CASSIF was derived. This efficiency is achieved because the online database is implemented in software. The method is considered comprehensive and transparent for detecting, interpreting, and mitigating leakage from storage formations. However, its application has been limited to a specific field. The tool needs to be made available as open-source software for broader adoption.

3.1.6. DPSIR (Drivers, Pressures, State, Impact, and Responses) Framework

Within the DPSIR framework, the influence of the components is assessed and defines the state of knowledge on environmental impacts that might hinder the deployment of the CCS project [94]. The methodology behind this framework contains the following definitions of each component. Drivers or driving forces are environmental forces that lead to CO₂ and highlight the importance of storing CO₂ underground. Pressures refer to CO₂ emissions due to human activities. The state describes the condition of the environment as a result of CO₂ emissions and human activities. Impact refers to the results of environmental changes on human beings and the ecosystem. Responses are actions taken to reduce the effects of the factors mentioned above [95].

The DPSIR framework evaluates every component meticulously through a pictorial representation found in Figure 6. Panels within this framework detect and resolve the potential risks involved in CCS projects. This method helps policymakers better understand environmental challenges and possible risks. However, complex data computational needs to prevent the DPSIR framework from performing effectively in analytic tasks. Despite recognizing uncertainties, the framework does not provide any resolution for them.

3.1.7. MCA (Multicriteria Assessment)

The MCA method estimates the suitability of CCS reservoirs and scenarios through weighted analysis, which matches these elements to established performance standards. Through MCA, participants conduct assessments by choosing their preferred assessment factors and ratings to create a structured framework that gathers stakeholder opinions. Participants employing the method gain full accessibility because it works for everyone, no matter their technical level, and maintains straightforward assessments that are easy to comprehend. Using this method, providers have complete flexibility for transparent evaluation procedures [97].

The Multicriteria Analysis (MCA) was carried out in two stages: first, evaluating carbon storage reservoirs, and second, assessing various scenarios, which leveraged detailed geological expertise. In the first stage, a set of criteria is defined against which the CCS reservoir will be evaluated. In the second stage, various criteria are proposed to assess scenarios. The two-stage MCA procedure enabled comprehensive reservoir assessments and quick implementation of broad scenario trade-off assessments. The process adopted two stages, which integrated specialized carbon storage criteria for expert geological analysis, combined with a general understanding for non-expert participants [98]. Based on a study conducted by Gough and Shackley [98], the criteria of CCS reservoir assessment and scenarios could be categorized as in

Table 8. Evaluation criteria for reservoir and scenarios by [98].

Reservoir Assessment Factors (Stage One)	Scenario Assessment Factors (Stage Two)
Storage timescale Potential storage capacity Proven security Monitoring Adverse impacts on human health Adverse impacts on ecosystems Planning/legal barriers CCS costs Public perception Reversibility (proposed by respondents) Robustness to uncertainty (by academic) Research and development (by NGOs) Energy penalty (by NGOs)	Reservoir performance (received a high weight) Costs Infrastructure change Lifestyle changes Security of energy supply Environmental impact Credibility International effects or distributional issues

. The evaluation has proven instrumental in discovering vital ambiguities and threshold issues affecting storage technologies, while the subsequent phase investigated carbon storage across various CO₂ reduction scenarios.

However, MCA requires intensive resources and extensive stakeholder consultations and data collection, which makes it time-consuming, costly, and site-specific. As the results depend on ratings from respondents, assigning weights to different criteria has the potential for biases in the evaluation process. Although the presenters declare that people with limited geological (technical) knowledge could participate in the evaluation, a certain level of expertise in technical fields like geology is essential, which can limit the participation of non-experts.

3.1.8. ESL (Evidence Support Logic)

The main idea of an evidence-based logical hypothesis is to identify risks and challenges that have the most significant impact on the confidence of the system [99]. ESL helps identify if deeper research and evidence of the risk and uncertainty would improve confidence in the problem that lacks information. Sensitivity analysis using an ESL model assists in determining the areas where further research and evidence collection are most likely to have the most significant influence on lowering uncertainty brought on by ignorance. ESL tools and other decision-support platforms deliver systematic ways to assess data while maintaining independent human analysis and avoiding objectification of procedures. Using logical models to decompose problems promotes full coverage of essential elements and uncertainties and allows documentation of decision processes. A standard assessment helps discover biases while showing their impact on the process.

Metcalfe et al. [99] developed a software tool, TESLA, based on the ESL approach and applied it to the Salah CCS project in Algeria. TESLA is a tool that helps users with decision-making when such issues arise. The tool divides the decision into a structure with a hierarchy, simplifies the problem, and allows users to collect and categorize the required information. Although the process does not automate the decision-making process, it provides helpful assistance [100,101]. When used in the Salah storage system, the tool revealed high confidence in complete and permanent CO₂ confinement, with limited danger of substantial release due to effective management and the absence of transmissive defects or fractures.

The proposed TESLA tool (based on ESL) allows integration of various data types like seismic surveys, reservoir models, and collective assessment. The processes are transparent and boost the confidence of stakeholders while making decisions. The tool enables us to highlight uncertainties that need to be targeted first and allows authors to apply it in practical applications in long-term storage stability assessment. However, the complexity in constructing decision trees and analyzing them, extensive data requirements, and reliance on spreadsheets that the tool is based on limit its application in initial scanning of storage sites where data, time, and knowledge are limited.

3.1.9. RA-CO₂ Method

Paraguassú et al. [102] proposed this qualitative approach of RA for CCS projects in Brazil. The authors analyzed previously described qualitative methods like VEF, SWIFT, MOSAR, FEP, MCA, and CASSIF and adopted different terms from each method (

Table 9. Terms and descriptions taken from different sources.

Name of the Parts of the Tool	Source Method	Description
Compartment	CASSIF	This refers to the predominant leakage paths, i.e., seal, fault, and well.
Attenuating	SWIFT	A set of measures that might lower the risk.
Aggravating	FEP	Potential risks that could trigger the leakage of CO2.
Frequency	MOSAR	The possibility of CO2 leakage happening.
Severity	MCA	The impact of CO2 leakage on the environment.
Cause or initiator event	VEF	A series of causes that oversee the occurrence of CO2 leakage.
Uncertainty	SRF	Refers to the degree of skepticism regarding the analysis results.

). RA-CO₂ qualitatively assesses the risks and categorizes them, evaluating the level of damage to the environment, the reputation of the operating company, public perception, and the confinement integrity of the storage reservoir [102,103].

RA-CO₂ is considered a beneficial qualitative RA tool due to its advantages: low cost, ease of use, transparency, comprehensiveness, systematicity, and flexibility. The authors proposed a worksheet that consists of nine columns where participants can indicate a range of aspects like a migration of CO₂, things that trigger CO₂ leakage, consequences, mitigation actions, frequency, the impact if leakage occurs, risk, monitoring measures, and a column for an identical number for each scenario of an accident. The column “risk” classifies the risks as low, medium, or high using a risk matrix (Figure 7).

Paraguassú et al. [102] developed this tool because the methods listed in Table 8 could not address such gaps as frequency and uncertainty in the sense of limited data and experience. However, the authors consider the SWIFT tool to be more fit for these purposes, but the tool is not user-friendly. It does not systematically tackle the challenges defined during the assessment. The RA-CO₂ method is capable of identifying risks earlier, which might aid the operating company in cost-saving and resource distribution based on the level of risk and uncertainty [102].

Despite the advantages mentioned in the method, there is no evidence of economic feasibility as the technique was not applied in real-world applications. The authors state that the process is cost-effective compared to other qualitative tools. Still, it is not proven, and success heavily depends on expertise, data availability, and organizational resources. If successfully proved in practice, the method’s qualitative nature limits its application only to the initial stage of the project.

3.1.10. Bow-Tie Diagram (GERICO Database)

Guénan et al. [104] concluded that the main gaps in RAs are due to a lack of an exact mitigation structure for identified potential risk scenarios and the actual data the operating company needs in the decision-making process. To address these gaps, authors came up with the GERICO database. GERICO database addresses the first gap by developing bow-tie diagrams with safety barriers, and the second gap was filled by constructing sheets with descriptions and pertinent metrics of each measure studied [104].

The authors used expert judgment to define eleven possible risk events from workshops and developed bow-tie diagrams based on them. Each of these defined risk events was further evaluated at three stages of project development: the planning phase, operational phase, and long-term storage phase. Later, one bow-tie diagram was developed per risk event for each stage (Figure 8).

The proposed method is beneficial in identifying mitigation measures specific to risk scenarios and developing bow-tie diagrams for each phase. The tool is specifically helpful in preparing preventive measures to ensure they comply with regulatory requirements. The bow-tie diagrams enable stakeholders to enhance decision-making by ranking measures based on relevance to a particular scenario. However, the tool lacks clarity of preventive and corrective measures while classification of measures. Some risk scenarios do not present clear and time frame management measures, such as “ignorance of legacy wells” (abandoned wells), making it challenging to relate to one of the three phases. The diagrams cannot address post-closure risk management because the well becomes inaccessible. It is doubtful that the tool might be immediately practical, as the authors point out that it is still under development. There are plans to transfer to another online web database that will require additional expenses.

3.1.11. Risk Matrix for RA of Legacy Wells

Arbad et al. [105] developed a Risk Matrix to qualitatively assess risks from legacy wells within the Area of Review (AoR), penetrating sealing rocks, and storage formations. The method allows the development of datasets for wells that need remediation. The proposed methodology categorizes legacy wells, facilitating a fast and more straightforward comparison of various CCS projects based on AoR. This method benefits operators in obtaining permits for Class VI CO₂ injection wells, as they must present corrective and monitoring actions for poorly plugged or abandoned wells. Although most papers consider wells with sustained casing pressure (SCP) and cement bond logs (CBL) in RA, this methodology aims to focus on different well types, from producers and injectors to orphaned, abandoned, and plugged wells, etc., which are categorized in the “Well Type” column of the risk matrix (Figure 9). The classification of wells ranges from Type 1, which lacks documentation, to Type 9, which does not penetrate the primary geologic seal or storage reservoir, with varying levels of protection for USDWs, the primary geologic seal, and the storage reservoir, depending on the type [40,41,105].

The proposed method is practical in the early stages of storage site screening and selection, aiding operating companies in identifying faulty wells and those requiring immediate remediation actions. The tool was applied in a real-world experience in an Illinois basin case study, which proves its applicability. However, the tool is limited to the legacy wells RA and an early-stage application of the CCS project. More comprehensive and integrated models must be applied for a more detailed scenario-based and numerical estimation of all leakage pathways to give a numerical analysis of leakage amounts.

3.2. Comparative Analysis of Qualitative RA Methods

The 10 qualitative RA methods undergo comparative analysis, which is displayed in

Table 10. Ranking of qualitative RA methods (Scoring criteria are detailed in Appendix A).

Method	Ease of Use	Time Consumption	Data Requirements	Expert Dependency	Applicability to CCS
FEP	3	2	3	5	5
VEF	3	2	4	5	4
SRF	5	5	2	4	5
SWIFT	5	5	2	5	5
CASSIF	4	4	3	5	5
DPSIR	3	2	4	4	4
MCA	3	2	4	4	5
ESL (TESLA)	3	2	2	5	5
RA-CO2	5	5	4	4	5
Bow-tie	4	2	2	5	5
Risk Matrix	5	5	2	4	5

in a tabular presentation. The comparison uses key factors such as user friendliness, time usage requirements, data collection needs, expert assistance needs, CCS project suitability, and strengths and limitations.

Table 11. Comparison of eleven qualitative RA methods: Top strengths and limitations.

Method	Top Strength	Top Limitation
FEP	Systematic and comprehensive for CCS projects	Time-consuming and complex
VEF	Holistic and systematic decision-making	Fails to identify operational risks as well as correctly manage uncertainty
SRF	Quick and transparent screening of storage sites	Limited to qualitative screening
SWIFT	Adaptable and proactive hazard identification	Relies heavily on expert opinion, subjective
CASSIF	Offers an efficient scenario-based analysis through transparent methods	Limited application, needs software implementation for execution
DPSIR	Suitable for policymakers and environmental risk identification	Lacks clear solutions for data uncertainties while requiring sophisticated inputs
MCA	Allows stakeholders to be involved while remaining flexible	Time-consuming and costly
ESL (TESLA)	Systematic and boosts stakeholder confidence	Complex and data-intensive
RA-CO2	Low cost and easy to use	Not proven in real-world applications
Bow-tie	Specific mitigation measures for risk scenarios	Lack of clarity in measures under development
Risk Matrix	Quick comparison and screening of wells to obtain permits.	Limited to legacy wells

compares eleven qualitative RA methods based on their top strengths and weaknesses. SWIFT, SRF, and RA-CO₂ serve as cost-efficient tools during the CCS project’s beginning, but the detailed risk analysis benefits from FEP, CASSIF, or ESL. DPSIR and MCA evaluate policies and environments, and the Risk Matrix developed by Arbad et al. [105] functions for quick qualitative evaluations on legacy wells and speeds up permit obtaining for Class VI CO₂ injection wells.

This analysis allows stakeholders to select their preferred RA tool. It shows the advantages and limitations of each method regarding project demands and resource availability.

3.3. Semi-Quantitative Methods

Semi-quantitative tools utilize both qualitative and quantitative RA methods: qualitative methods to estimate likelihood and impact levels, and quantitative techniques to prioritize and analyze the results for detailed risk identification and assessment. Plenty of semi-quantitative tools have been developed over the years for CO₂ leakage RA. In this chapter, semi-quantitative methods that have been applied in industry were introduced and compared (

Table 12. Overview of evaluated semi-quantitative tools.

Tool Name	Author/Company	Short Description	Year
CarbonWorkflow™	Hnottavange–Telleen K./Schlumberger	Experts determine FPE’s from a database, and risks are ranked on the risk matrix (Figure 10).	2008
CO2QUALSTORE Guideline	DNV	The risk-based approach provides assurance based on the quality and scope of input data. Site-specific.	2011
RISQUE (Risk Identification and Strategy using Quantitative Evaluation)	URS	Expert-elicited, works by developing probability and consequence matrices.	2001
ICARAS (Integrated Carbon Risk Assessment)	TNO	Provides RA from scenario definition to probabilistic impact assessment, covering the entire life span of the CCS project.	2013
BN (Bayesian Networks) in CCS	Gerstenberger et al.	Expert elicitation and uncertainty in CCS risk assessment.	2015
MOSAR (Method Organized for a Systematic Analysis of Risk)	A. Cherkaoui et al.	A typology grid of hazard sources for CO2 geological storage sites organizes risk scenarios hierarchically based on probability and severity.	2009

).

3.3.1. CarbonWorkflow™ Tool. (Schlumberger)

The CarbonWorkflow™ Process from Schlumberger delivers a structured semi-quantitative RA method for resource allocation in risk reduction and documentation alongside tracking how risks are mitigated. Coherent project criteria help experts evaluate potential FEPs in a risk matrix with five severity and likelihood assessment category scales. The evaluation tracks health, safety, financial, environmental, research, and industry viability effects.

Expert panels work in two cohorts, which comprise six groups dedicated to examining different project aspects, like air, surface, near surface, and subsurface, to review pre-screened FEPs. Trained experts learn about the project-specific data and RA protocols as a prerequisite to performing the ranking process. The experts supply three quantitative measurements of lower, best, and upper bounds for negative impact estimation as well as likelihood probabilities that relate to the baseline standard of “100 similar projects over 100 years” to establish approximate confidence ranges. The RA for each FEP is derived from producing severity scores with likelihood ratings. The assessment results enable managers to develop key risk scenarios that lead to the creation of risk response actions (RRAs), which form part of risk response action groups (RRAGs) that receive implementation assignments and documentation alongside periodic review processes [106].

The presented methodology was applied to RA in the Illinois Basin–Decatur Project (IBDP), where evaluation was conducted on five parameters: financial stability, impact on the environment, industry responsibility, and research activity on the project. IBDP experts addressed questions by utilizing technical and non-technical FEPs on the severity vs. likelihood (S × L) scale of the risk matrix [106,107,108]. The Kimberlina site CCS project also utilized the CarbonWorkflow™ tool for RA, which depicted risk scenarios and ranked them utilizing the risk matrix on Figure 10.

While the CarbonWorkflow™ tool is comprehensive, involves cross-disciplinary collaboration during workshops, and has proven its practical implementation on real-world challenges, the significance of time and resources and data extensiveness might limit the application of the methodology. The S × L matrix approach used semi-quantitative methodology despite potential subjectivity from absent probabilistic risk modeling, which future CCS projects should consider implementing quantitative or hybrid methods for better accuracy.

3.3.2. CO2QUALSTORE Guideline

The CO2QUALSTORE guideline serves as a tool to help operators, authorities, verifiers, and other stakeholders through its development by DNV and its industrial partners with support from national regulators. The main function of the CO2QUALSTORE guideline focuses on establishing a transparent and uniformly applied method to qualify storage sites at reasonable costs. The workflow of the evaluation is as follows: screening stage (site screening, shortlisting), assessing and selecting stage (selection of the best sites and permit application), operation and review of permit stage (commence CO₂ injection, site closure), and lastly, transfer of responsibility stage (final milestone decommissioning) [109]. The proposed guidelines emphasize the importance of early communication between project operating teams and regulators to speed up permit application [110].

The guideline is capable of risk management of major steps within the CCS project. The tool is comprehensive and allows stakeholders to assess risks from the screening to the closure phase. The method boosts stakeholder confidence and transparency and aids in harmonizing engineering practices with regulations. However, the framework heavily relies on extensive data collection, and risk documentation may pose a challenge, particularly in the initial stages of the project.

3.3.3. RISQUE Method

Bowden et al. developed this method that is based on the judgment of experts providing stakeholders transparency in understanding the leakage risks throughout the CCS project. The method was successfully applied in the major CCS projects including GCS site in Salah (Algeria), the CO2CRC Otway Stage 1, various sites in Australia and Weyburn–Midale Project [111,112,113,114]. RISQUE identified leakage pathways like faults, legacy wellbores etc., and uncertainties linked to them, and later they were compared to the dynamic changes at various leakage scenarios. The quantitative part of RA pertains to the leakage rate to be less than 1 percent over 1000 years. The analysis was conducted on such variables as; the potential for leakage on log scale from 0 to 1, amount of CO₂ leak in tons, and the likely time when the scenario would happen [112]

Although the tool has been used in non-technical spheres for a long time, it has also proven its applicability and effectiveness in a real-world CCS project. RISQUE allows diverse technical data to be integrated into the site characterization process, helps regulators gain confidence through an evaluation structure, and smooths the public perception process with transparency. However, in terms of leakage quantification, the tool is only able to estimate the amount of leakage through a particular containment lease, while the tool is not capable to estimate the leakage and impact on other leases around the geological trap—which limits the tool only in short- to mid-term risk evaluation. Furthermore, the method is limited in consideration of emerging risks like induced seismicity, fault reactivation, and geomechanical risks that may cause well integrity issues.

3.3.4. ICARAS Method

According to Wollenweber et al. [114], the ICARAS framework is a holistic modular approach to monitoring geological CO₂ storage-related risks (Figure 11).

The methodology connects pre-existing qualitative and quantitative tools, starting from CASSIF scenario definition through detailed models of reservoirs and geomechanics to migration analysis, thus providing step-by-step procedures for the complete storage period. The integration of preliminary screening systems that employ fast models like COSE enables efficient operations through a speed and accuracy balance [115]. Integrating response surface methods through Cougar™ strengthens risk quantification by improving its robustness.

The framework complexity creates two main practical difficulties: it requires the integration of diverse software programs, and it is essential to confirm that easy-to-use models still accurately represent natural geological systems. Well integrity assessment aspects remain in development, which may impact the reliability of final RA results. ICARAS enables substantial progress toward methodical system-wide RA for GCS; however, the system requires advanced improvements in tool matching along with model testing for comprehensive regulatory backing.

3.3.5. Bayesian Network (BN) in the Bi-Directional Approach of RA

BN was applied as an RA tool in CCS projects with a focus on bi-directionality (predictive and diagnostic reasoning) [116,117,118]. The method is extensively used in diverse technical and non-technical industries because it is a flexible and probabilistic graphical modeling tool where variables are depicted by nodes and arrows connecting nodes. The whole idea of the mechanism and spread of influence due to a change in the state of a node in BN comes from Bayes’ theorem [116].

Creating a BN model can be complex and time-intensive, particularly if the system that has to be modeled has not yet been defined through variables that constitute the nodes. There are two main components to the process: first, the design of the BN structure, and second, the population of the Conditional Probability Tables (CPTs) for each node within the BN (Figure 12). As the RA process depends on time, the framework needs to be flexible. BN incorporates RA in the initial site screening process with limited data available (corresponds to the initial time) and the time frame that covers later RA, such as injection and post closure, making it a flexible model in terms of time of RA and amount of input data [116].

The BN method enables us to conduct predictive analysis of adverse outcome risk and diagnostic analysis of possible cause identification through its inherent bidirectional capabilities. The BN framework establishes an explicit graphical structure that displays relationships among system elements (such as CO₂ input, pore space, injectivity, and migration). This approach allows stakeholders, particularly regulators and the public, to better understand the RA process since it lacks closed “black box” behavior. Expert-elaborated risk assessment processes occur during network model development and CPT population activities. The model relies heavily on expert judgment for its development because CCS data remains insufficient; thus, integrating multiple professional insights helps improve its accuracy [116].

However, the expert-based extraction of thousands of probabilities requires extensive time and introduces inconsistent results, particularly involving multiple disciplinary backgrounds. Additionally, the model operates statically for one time period, while the CCS system shows dynamic patterns extending throughout its entire lifecycle (e.g., during injection, post-injection, and thousands of years post-closure). This simplified model might restrict the model’s predictive power for long-term RA. Expert elicitation represents a strength, yet it brings subjectivity and potential bias. Implementing this approach requires turning continuous parameters into discretized versions through a process that may reduce precise detail recovery.

3.3.6. MOSAR Framework

MOSAR provides an organized solution for technical risk analysis of human installation infrastructure and existing facilities that help detect and eliminate risky conditions through preventive actions [119]. The method consists of two stages. In stage A, the significant risks, sources of risks, and consequences are identified. In stage B, a detailed evaluation of identified risks is conducted (Figure 13). The whole process of RA with MOSAR takes 10 subcomponents. This method is solely based on site observations. Although all the subcomponents interact with each other and no step can be neglected, the system still provides a means of flexibility by the ability to include new risks into the subsystems without changing the RA results obtained initially [120].

Through its RA methodology, MOSAR structures, the evaluation process divides storage systems into key parts like injection wells and aquifers. Any risk falls under the classification of mechanical, chemical, or biological aspects in the typology grid, which delivers comprehensive coverage across different RA factors. This method enables teamwork between operators, policymakers, and community members, enhancing RA transparency and shared responsibility. Through its interconnected subsystems approach, MOSAR enables risk detection of sophisticated leakages when fluid faults intersect with aquifers.

However, the authors state that this method requires large amounts of time and resource expenditure. Even though MOSAR needs extensive time to design hazard grids and scenarios, its reliability becomes unstable when used for storage site RA during initial development phases. The authors claim the superiority of MOSAR over complex FEP. Still, its application at the Ketzin site fails to prove practical worth since it lacks quantitative evidence to support the system. Since it lacks time-dependent probability data for these elements, the method experiences substantial limitations when assessing long-term slow risks linked to geological shifts, legacy wells problems, and induced seismicity.

3.4. Comparative Analysis of Semi-Quantitative Methods

Table 13. Ranking of semi-quantitative RA Tools (Scoring criteria detailed in Appendix A).

Method	Ease of Use	Flexibility	Handling Uncertainty	Stakeholder Transparency	Data Needs	Long-Term RA
CarbonWorkflow™	3	3	3	4	5	2
CO2QUALSTORE	3	2	2	5	5	3
RISQUE	3	2	3	5	4	2
ICARAS	2	5	5	4	5	3
Bayesian Network	2	5	5	5	4	2
MOSAR	3	3	2	5	5	2

represents ranked and assessed standard parameters and features across the six semi-quantitative RA tools for CO₂ storage. This enables users to compare their capabilities, practicality, and suitability directly.

Each RA technique provides exclusive coverage for different CCS risk factors. The selection of RA tools should match project phase requirements while using existing data sources to fulfill stakeholder needs through multiple methods.

Table 14. Summary of comparative analysis of the tools studied.

Tool	Methodology	Strengths	Limitations	Project Phase Suitability	Key Differentiators
CarbonWorkflow™	Semi-quantitative risk matrix (S × L); expert panels; three-bound estimates	Cross-disciplinary collaboration, proven real-world use; structured workflows	Subjective due to lack of probabilistic modeling; time/resource-intensive	Screening, Implementation	Focus on financial/environmental impacts; integrates RRAs/RRAGs
CO2QUALSTORE	Phased workflow (screening, selection, operation, closure)	Transparency; harmonizes regulations; stakeholder confidence	High data/documentation demands; challenges in early stages	All phases	Emphasizes regulator engagement; standardized for permit applications
RISQUE	Expert judgment; leakage pathway identification	Transparency. Integrates technical data and public perception, smoothing	Ignores cross-lease leakage; limited for emerging risks (e.g., seismicity)	Short- to mid-term	Focus on leakage pathways; quantitative leakage rate thresholds
ICARAS	Modular integration (CASSIF, COSE, CougarTM); combines qualitative/quantitative	Step-by-step procedures: balances speed/accuracy; robust risk quantification	Software integration challenges; well integrity assessment under development	Detailed planning	Integrates reservoir/geomechanical models; modular flexibility
Bayesian Network	Probabilistic bidirectional reasoning; nodes/CPTs; expert judgment	Flexibility; handles uncertainty; transparent “white box” structure	Time-intensive CPT population; static timeframe; discretization reduces detail	Early screening	Bidirectional reasoning; integrates multidisciplinary expertise
MOSAR	Two-stage process (risk ID/evaluation); subsystem division; typology grids	Structured teamwork; detects complex leakage scenarios	Time/resource-heavy; unreliable in early phases; lacks long-term risk assessment	Operational phases	Focus on mechanical/chemical/biological risks; subsystem interaction analysis

summarizes a comparative analysis of the semi-quantitative RA frameworks.

The critical review indicates that CarbonWorkflow™ facilitates cross-disciplinary collaboration but is limited by a subjective SxL matrix and short-term focus. CO2QUALSTORE supports regulatory compliance, although it imposes a high documentation burden. RISQUE is effective for leakage pathway analysis but does not account for cross-lease impacts. ICARAS enables holistic system-wide modeling, though it is hindered by software complexity. Bayesian Networks provide predictive and diagnostic analysis but are constrained by static timeframes and potential expert bias. Lastly, MOSAR is suitable for infrastructure risk detection but performs poorly in managing long-term risks.

In conclusion, the comparative analysis notes that none of the tools can assess long-term geomechanical risks (e.g., induced seismicity risk), which is a key critical gap among them.

3.5. Quantitative Methods

Various RA methods have been introduced over the past few decades to assess leakage of GCS; however, a significant shift towards quantitative tools is very noticeable. Quantitative models can predict and quantify leakage through wells, faults, and other mechanisms by relying on physics-based models to forecast long-term CO₂ plume behavior [72,121,122,123]. Quantitative tools are considered accurate RA methods; however, the accuracy highly relies on data quality, in the sense of higher uncertainty, requiring running sensitivity analysis and stochastic models, the computation is costly, time-consuming, and unreliable [122,123,124,125,126,127,128]. Among quantitative tools, Reduced Order Models (ROM) were used to reduce the time of computation by incorporating factors that affect well leakage and reservoir parameters [129,130,131,132]. ROM could be developed from surrogate, proxy, meta, and surface response models. Among these subtypes, surrogate models are found helpful in conjunction with numerical models, as surrogate models handle complex computations, keeping accuracy high [133,134]. Moreover, surrogate models with emerging deep-learning techniques were used to predict CO₂ flow and the long-term geomechanical response of the formation, which semi-quantitative models were not able to predict [135,136]. Southwest Regional Partnership on Carbon Sequestration (SWP) developed ROM with Monte-Carlo simulation for RA and uncertainty quantification in the late Pennsylvanian classis reservoir, Morrow B sandstone formation (initially under CO₂ EOR) at the Farnsworth Unit (FWU) [137,138,139]. ROMs assessed relationship between uncertain input values (porosity, permeability, anisotropy, initial oil saturation, cation exchange capacity (CEC), absorbent specific surface area (SSA), CO₂ leakage rate), and dependent model outputs (cumulative oil production, total dissolved solids (TDS), volume storable CO₂ in different trapping mechanisms, pH of water, etc.) [140,141,142].

Table 15. Quantitative RA frameworks studied.

Tool Name	Organization/Author	Short Description	Year
PA (Performance Assessment)	Quintessa	Integrated with TESLA, QPAC-CO2, and Quintessa FEP database to conduct Performance Assessment of GCS.	2009
P&R™ (Performance and Risk Methodology)	OXAND	Simulation tool SIMEO™-STOR enables quantifying leakage through wells and RA of well integrity issues (casing corrosion and cement carbonation).	2009
CF (Certification Framework)	LBNL C.M. Oldenburg	Leakage assessment through probabilities. Quantifies leakage by dividing it into small components. Total probability is the sum of the probabilities of all components.	2009
GERAS-CO2GS (Geo-environmental Risk Assessment System, CO2 Geological Storage RA)	AIST	Assessment conducted by quantifying prior probability. Analyzes the risk of leakage of injected CO2.	2012
CO2-PENS (Predicting Engineered Natural Systems)	LANL	A tool that facilitates organizing site-specific data into a system-level framework for quantitatively assessing risks and uncertainties.	2006–2013

demonstrates the Quantitative RA methods studied in this paper.

3.5.1. Performance Assessment (PA) Framework

Metcalfe et al. [143] from Quintessa introduced an integrated framework to perform PA of GCS by connecting qualitative data with quantitative models and expert knowledge. The methodology implements three essential instruments: (1) TESLA, which utilizes Evidence Support Logic (ESL) on its decision-support system to build hypothesis-based decision trees, (2) a generic FEP database that functions as a system model audit scenarios, and (3) QPAC-CO₂, which works as a numerical modeling code for physical-chemical process simulation (Figure 14). The methodology addresses epistemic and aleatory uncertainties as well as system couplings, to support stakeholder decisions through all project steps [142,143,144].

The framework creates a synergistic evaluation method that utilizes computer output, expert input, and observed data to understand the system’s functioning as a whole. The interval probability approach of TESLA defines separate categories of evidence for hypothesis testing and remaining unawareness while offering improved alternatives to standard probability statistics. The FEP database facilitates thorough scenario development through systematic system components and process cataloging, which helps prevent missed RA areas. QPAC-CO₂ demonstrates the flexibility that lets researchers join different processes at different measurement scales to conduct comprehensive system-level and detailed process analyses simultaneously. The complete audit trail between hypotheses and decision points and their models helps build trust among stakeholders while ensuring regulatory compliance [142,143].

However, among the advantages mentioned, the method has several drawbacks. Numerical evidence derived from qualitative data (for example, TESLA’s 0–1 scale) becomes susceptible to biased interpretation because expert opinions may differ without sufficient empirical evidence. The probabilistic capabilities of QPAC-CO₂ face difficulties when integrating detailed and system-level models because it requires substantial computational resources. The restricted example of caprock leakage demonstrates the system but fails to provide experimental proof. The predictions spanning 10³ years exist in a speculative state because geological and chemical processes produce significant uncertainties. The adoption of TESLA and QPAC-CO₂ depends on developing user-friendly interfaces and training users since their usability and learning preferences remain unexplored.

3.5.2. Performance and Risk (P&R™) Methodology

Oxand S.A. developed a quantitative RA framework, known as the Performance and Risk (P&R™), for RA of leakage in CO₂ geological storage systems [145]. The approach integrates:

• The system storage is described as a network of components (i.e., casings, cement sheaths, geological layers) and processes (i.e., CO₂ flow, cement degradation, casing corrosion).
• Kinetic models are used to simulate key aging mechanisms such as cement leaching/carbonation and casing corrosion, while the development of future technologies related to abdication is investigated. These processes cause impedance changes over time, altering material properties (e.g., permeability-thickness).
• Scenarios are generated probabilistically for uncertainties in parameters such as cement permeability and corrosion rates. For instance, the synthetic case was run with 243 scenarios, and random vertical cement permeability values were tested.
• Risk expressed as: Risk = Probability (P) × Severity (S). Failure likelihood (e.g., degraded cement pathways) is used to derive probability, and severity impacts on stakes like economic health and the environment. The result is mapped onto critical grids (e.g., F-G matrices) to rank the risks.
• In the end, a risk remediation strategy is recommended based on cost–benefit (C/B) ratios and risk reduction potential.

This methodology is operationalized in the simulation platform SIMEO™-STOR, which couples two-phase flow models (Darcy’s law) with degradation kinetics. It predicts CO₂ leakage pathways and rates, as well as cumulative leakages on millennial timescales (i.e., 1000 yrs) [145,146,147].

The P&R™ methodology delivers an effective method to evaluate CO₂ storage risks that combines detailed engineering analysis with outcomes that satisfy stakeholders. The methodology incorporates dynamic degradation modeling and scenario analysis to handle key uncertainties within long-term well integrity assessments. However, additional improvements should be made because the method depends on ideal input assumptions, high computational requirements, and insufficient testing of forecasts beyond short periods. The predictive accuracy and adoption rate of the tool will increase when field monitoring programs (such as Sleipner and Weyburn) develop open-source tools alongside it. Regardless of its restrictions, the method is a notable progress in performance-based safety evidence development for CCS projects.

3.5.3. Certification Framework (CF)

Proposed by Oldenburg et al. [148], the CF is an RA method to evaluate GCS site safety and effectiveness through analysis of potential CO₂ and brine leakages. Effective trapping refers to CO₂ trapping within specific storage areas according to a framework that determines leakage risk factors via probabilistic pathways (wells and faults) and environmental compartments (e.g., USDW and atmosphere) analysis. The CF incorporates reservoir simulation models with probabilistic RA and simplified leakage models to predict impacts using proxy indicators like fluxes and concentrations (

Table 16. Key aspects and their definitions in CF.

Key Aspect	Definition of Aspect
Effective trapping	CO2 containment provides the basis for safe operations since acceptable amounts of leakage remain within regulatory limits.
Conduits and compartments	The assessment examines wells and faults as main leak channels while performing five different impact checks (USDW, health/safety, etc.).
Probabilistic intersection	The RA relates to two fundamental elements that include the probability of CO2 plume interactions with conduits and conduit interactions with compartments.
Modeling	The system uses predefined reservoir simulation results (plume migration) along with simplified leakage models that allow anyone without specialist knowledge to access the system.
Flexibility	Adaptable to diverse regulatory contexts through user-defined impact thresholds and risk acceptability criteria.

). A theoretical Texas Gulf Coast model shows the framework for application [148,149,150].

The methodology offers simplicity and transparency in RA by dividing complex systems into components that benefit stakeholders and regulators. The framework incorporates regulatory flexibility, thus accommodating various national standards with those from the framework itself. Also, CF prioritizes leakage pathways by focusing on legacy wells and faults, which are the main pathways in sedimentary basins, reducing unwanted complexity. However, reliance on simplified models (e.g., one-dimensional, homogeneous reservoir assumption) limits their accuracy. CF is impractical for early-stage RA as probabilistic inputs (e.g., well integrity) require extensive site-specific data, which accounts for data gaps. Addressing these gaps is critical for future models for risk and leakage quantification.

3.5.4. GERAS-CO₂GS Framework

Tanaka et al. [151] created GERAS-CO₂GS as an RA platform to determine safety risks from CCS operations. The framework combines CO₂ migration simulation models with atmospheric dispersion computation and industrial accident frequency calculations to predict risks in geological formations, marine settings, soil surfaces, and air environments. Visualization tools, modular risk scenarios, and probabilistic hazard assessment (PHA) are used to ensure stakeholder engagement. The software tool meets criteria from both ISO/TC 265 and CDM requirements to improve operational safety while increasing public acceptance [151,152,153].

It includes the following essential characteristics that GERAS-CO₂GS contains (Figure 15):

• GERAS-CO₂GS offers RA for underground reservoirs, marine facilities, surface equipment, and atmospheric emission monitoring.
• Combines numerical models (e.g., CO₂ migration via CO₂/PENS and atmospheric dispersion via ADMER 2.5) with empirical accident data (e.g., high-pressure gas leaks).
• The system allows risk scenarios with diverse events, such as well blowouts and caprock failures, by storing categorized data about endpoints (human health and ecosystems).
• The system integrates Google Earth functionality, which creates transparent spatial risk maps for public perception.

GERAS-CO₂GS provides a novel and original approach to CCS RA on the multiscale, merging simulation rigor and empirical data. The methodology leverages assumptions on industrial accident statistics (e.g., Japanese high-pressure gas incidents) to estimate surface facility risks. It supports compliance with international standards and allows project approvals by schemes such as CDM.

However, its prototype status and reliance on analog data limit its applicability to novel CCS contexts. A limited scope and partial application exist for marine impact analysis and acute event modeling (e.g., blowouts). CO₂ production facility surface risk estimates could cause CCS-specific situations (e.g., offshore storage) to be misrepresented. The model relies on steady-state leakage models, which might limit transient ones when sudden well failures due to wellbore integrity issues or Blow Out Prevention (BOP) equipment failures occur.

Figure 15. The workflow of GERAS-CO₂GS methodology (adjusted from [150]).

Figure 15. The workflow of GERAS-CO₂GS methodology (adjusted from [150]).

3.5.5. CO₂-PENS (Predicting Engineered Natural Systems)

The CO₂-PENS model functions as an RA system for GCS that combines system-level and process-level assessment components for CO₂ capture, transport, as well as storage and leakage potential evaluation [154]. CO₂-PENS utilizes the GoldSim^® platform to build its RA model through probabilistic modules that handle uncertainty in GCS [155,156,157]. The methodology utilizes physical models that analyze CO₂ plume migration, wellbore leakage patterns, and economic RA elements. Latin Hypercube sampling generates probabilistic results for uncertainty propagation for reservoir permeability, wellbore integrity, and energy costs. The methodology uses 3D geologic information about reservoir topography and fault patterns to evaluate spatially based leakage risks from abandoned wells and faults [158]. The approach also analyzes the production expenses for treating extracted saline water (e.g., reverse osmosis vs. thermal desalination) and estimates their relationship to the sequestration advantages.

The CO₂-PENS model was used to assess storage capacity and leakage risk potential (through faults and wells) and estimate brine disposal expenses at the Rock Springs Uplift in Wyoming [159,160]. The framework was also used to evaluate the brine extraction process and select optimal water treatment options under different water salinity and temperature parameters [156]. Beyond these, CO₂-PENS helps conduct initial site screening through CO₂ plume migration simulation, and pressure management optimization, which aids in selecting the safest sequestration formation [158]. The method proves suitable for various CCS projects based on these applications.

CO₂-PENS establishes itself as an effective method for initial site screening RA and functions across multiple geological environments. However, it requires excellent input data and suitable computational facilities. High-resolution 3D simulations such as FEHM for multiphase flow demand substantial computational resources, making them impractical for storage sites lacking such data. Moreover, using literature-derived data for cost and permeability assessment produces potential biases or insufficient simplifications within the system.

3.6. Comparative Analysis of Quantitative RA Methods Studied

The critical review shows specific strengths and weaknesses that match different phases and goals within CCS projects. The PA Framework successfully merges expertise-generated information with modeling initiatives to build organized scenarios that address various uncertainties effectively. The method’s performance suffers from qualitative data requiring expert interpretation, high computational needs, and minimal experimental evidence. The P&R™ Methodology performs dynamic model estimations for degradation alongside probabilistic modeling and offers cost–benefit analysis to evaluate remediation plans. This approach contains substantial advantages, yet it depends on simplified assumptions, needs complex calculations, and shows constrained proof during long monitoring periods. The CF method receives strong regulatory approval due to its clear compliance standards, which also allow proper adjustments to changes in governing laws. This method lets users focus on key leakage pathways, making it more suitable for regulatory compliance needs. The simplified model approach in this method compromises accuracy while needing plenty of site-specific data and limitations during early-stage evaluation. The GERAS-CO₂GS framework offers broad RA capability, incorporating experimental data and visualization elements designed for stakeholder involvement. This technology falls short because it only works with prototype versions, needs analog data, and has limited skills for handling transient occurrences. The mixed set of strengths and weaknesses across models allows stakeholders to pick particular modeling systems based on individual project requirements or create multi-model systems that address all CCS project stages.

Table 17. A comparative study based on key criteria (the scoring criteria are detailed in Appendix A).

Criteria	PA	P&R™	CF	GERAS-CO2GS	CO2-PENS
Integration of Data	5	4	3	5	5
Uncertainty Handling	5	5	3	5	5
Computational Efficiency	2	2	5	3	2
Stakeholder Engagement	5	5	5	5	4
Regulatory Compliance	5	5	5	5	5
Data Requirements	3	3	5	2	5
Long-Term Predictions	2	3	3	3	3
Usability	2	3	5	2	2
Validation	2	2	3	2	5
Flexibility	3	3	5	3	5
Risk Remediation	3	5	3	3	5

shows a comparative study of five quantitative methods. As can be seen, CO₂-PENS is the most versatile method among all in terms of technical data handling and initial and long-term RA, while CF is the best fit for regulatory compliance. PA Framework and GERAS-CO₂GS show excellent results in stakeholder communication.

3.7. Emerging Integrated or Hybrid Methods

Among emerging methods, integrated ones, which combine features of qualitative and quantitative methods in one model, have shown great applicability in real-world applications. NRAP-Open-IAM and REX-CO₂ are the most well-known integrated models, and this chapter will describe and compare them.

3.7.1. NRAP-Open-IAM

In 2016, NRAP released its first leakage risk forecasting tool for GCS site operators, the NRAP Integrated Assessment Model for Carbon Storage (NRAP-IAM-CS) [159]. The model framework divides the GCS systems into four components. Each component simulates physical processes in GCS (

Table 18. Concise summary of NRAP-Open-IAM components.

Component	Key Method Description
Geologic Stratigraphy	Defines geological layers (reservoir, caprock shale, aquifers) with depth/thickness.
Reservoir Models	Simulate dynamic pressure and CO2 saturation via - Simple Reservoir (homogeneous, analytical) - Lookup Table (precomputed high-fidelity data).
Leakage Pathways	Model CO2/brine upward migration through - Open/cemented/multisegmented wellbores or faults (physics-based equations).
Receptors	Assess leakage impacts - Aquifer: Geochemical interactions (pH, TDS) - Atmospheric: CO2 dispersion (wind speed, leak rate).

).

The Geologic Stratigraphy component aims to provide foundational data for other models and integrate all components through defined stratigraphic data [123]. A simple reservoir model (with homogeneous permeability and constant thickness) is used to quickly estimate the storage site screening scenario. Lookup Table models are employed for detailed, complex, and site-specific reservoir simulations. Lookup Table reservoir models are created in CMG, ECLIPSE, or TAUGH2 simulators. The Leakage Pathways model utilizes a completely uncemented legacy wells scenario [124,125,161], a cemented legacy wellbore model [130], and a multisegmented wellbore model where the leakage is modeled along the wellbore with varying properties and multiple segments [123,162,163]. The main goal of the Aquifer Receptor model is to simulate the geochemical reactions caused by CO₂ and/or brine leakage into the aquifer and evaluate the change in groundwater quality, such as pH and TDS. [134,164]. The atmospheric Receptor model is the leakage into the atmosphere and dispersion of CO₂ in the atmosphere, depending on the leakage location, wind speed, and leakage magnitude [165,166].

However, NRAP-Open-IAM does not directly address essential parameters such as casing corrosion rates, the number and location of plugs, elastomer/seal mechanical properties, and the operational effects on in situ stress levels. The latter can significantly influence well integrity by inducing fractures and altering the stress distribution around the wellbore. Incorporating these elements into the assessment tool is crucial to achieving a more accurate risk assessment of wellbores.

3.7.2. REX-CO₂ (Re-Using EXisting Wells for CO₂ Storage Operations)

The REX-CO₂ method is structured to assess risks associated with storing CO₂ in underground reservoirs. It evaluates well design, material compatibility, and potential leakage pathways to ensure safe and long-term CO₂ containment. The tool is designed to evolve based on feedback from industry partners and real-world case studies [76,167,168]. Below is a breakdown of its key components:

The process starts by gathering basic data about the storage site and wells:

⚬

Field Information: Location of the oil/gas field, type of storage reservoir (e.g., depleted oil field, saline aquifer).

⚬

Well Data: Details like well purpose (injection/monitoring), operational history, and current status.

⚬

Storage Parameters: Reservoir depth, pressure, temperature, and injection rates.

The next step will be the Well Design Assessment. This step checks if the well can safely contain CO₂ under operational conditions. It includes four assessments:

A.

Well Integrity Assessment.

Evaluates barriers that prevent CO₂ from leaking out of the well.

Table 19. Two types of barriers are considered in the Well Integrity assessment.

Primary Barriers	Secondary Barriers
Caprock (impermeable rock layer)	Impermeable geological formations
Production casing/liner cement	Casing cement
Production casing/liner	Casing with hanger and seal assembly
Production packers	Annuli fluids (sealing fluids)
Tubing	Wellhead with valves
Downhole safety valve (SSV)	X-mas tree (surface valves)
Plugs inside the wellbore are NOT considered	Tubing hanger with seals

depicts the types of obstacles that are assessed.

These checks include corrosion, cement bond quality, pressure testing details, and Primary and Secondary well barriers.

B.

Risk of Out-of-Zone Injection.

This section assesses whether CO₂ might leak outside the intended storage zone. Checks include cement integrity around the casing, integrity of the casing shoe, and proper sealing of overlaps between casings. Also, it ensures corrosion resistance (material compatibility assessment) of the well (e.g., to CO₂ or hydrogen sulfide) and compatibility of elastomer/seals to the change in temperature and pressure within the wellbore (Structural Integrity Assessment) [76,166,167,168,169].

Moreover, the assessment tool includes cement integrity prediction. This predicts risks of cement failure and CO₂ leakage using two simplified models. The first model, ROM 1 (Geomechanical), is employed by industrial designers to forecast the likelihood of cement failure and the formation of fractures. This model’s predictions hinge on three crucial parameters: Young’s Modulus of the cement’s stiffness, the bond strength between the cement and rock, and the temperature at which the cement is injected.

The second model, ROM 2, assesses leakage and calculates potential CO₂ escape rates. This model requires data on three key factors: cement permeability, reservoir depth and pressure, and CO₂ concentration values. The REX-CO₂ tool includes essential features for comprehensive risk assessment. It collects vital site and well information during the initialization phase. The Well Design Assessment module evaluates barrier integrity and material suitability while identifying potential leakage sources. Cement Integrity Models analyze failures, predicting leaks, and determining failure rates. A Programming Framework ensures a user-friendly experience and efficient risk calculations, while a Feedback Loop mechanism continuously updates predictive models based on real-world data.

The REX-CO₂ assessment tool utilized ISO 27914 [74] and US EPA Class VI [45] standards for CO₂ storage and injection, as well as ISO 16530 [170] and NORSOK D-010 [171] standards for injection well design and the development of a decision tree, which emphasizes strong regulatory compliance. The tool successfully leverages ROM-based geomechanical cement debonding and well leakage models, which are lacking in most quantitative tools. Moreover, the tool demonstrates flexibility in adapting to changing regulatory requirements across different states and countries. However, the assessment tool does not consider complications like junction integrity of lateral wells and plugs, durability of elastomers, and limited data scenarios.

3.8. Comparative Analysis of Integrated Assessment Tools (NRAP-Open-IAM and REX-CO₂)

Table 20. Comparative study of NRAP-Open-IAM and REX-CO₂ assessment tools.

Aspect	NRAP-Open-IAM	Rex-CO2
Primary Purpose	Leakage risk forecasting for CO2 storage systems, integrating geological, reservoir, and receptor models.	Well-designed integrity assessment and leakage risk prediction, focusing on wellbore and cement failures.
Core Functionality	Simulates CO2/brine migration through leakage pathways and impacts on receptors.	Evaluates well integrity barriers, material compatibility, and cement debonding risks.
Key Components	1. Geologic stratigraphy.	1. Well-designed assessment.
	2. Reservoir models.	2. Cement integrity prediction.
	3. Leakage pathways.	3. Tool programming framework.
	4. Receptors.
Model Types	- Reduced Order Models (ROMs) for leakage pathways.	- Decision trees for well integrity.
	- Integration of reservoir and receptor models.	- ROMs for cement debonding and leakage rate.
Well Integrity Focus	Primarily focuses on wellbore leakage (e.g., open/cemented models) without explicitly addressing corrosion, plug placement, or operational stress effects.	Explicit focus on well barriers (primary/secondary), corrosion, cement bond quality, and material compatibility.
Cement Integrity	The cemented wellbore model assumes attenuation by formations; no explicit debonding prediction.	Utilizes two ROMs to predict cement debonding and leakage rates from geomechanical and permeability inputs.
Decision Support	- Area of Review (AoR) delineation.	- Well-integrity decision trees.
	- Risk assessment.	- Quantitative leakage probability estimates.
	- Sensitivity/scenario analysis.	- Mitigation guidance.
	- Post-injection care.
Operational Parameters	- Reservoir pressure/saturation.	- Well material properties.
	- Leakage pathway properties.	- Cement mechanical properties.
	- Receptor impacts.	- Operational temperature/pressure.
Tool Framework	Open-source framework (MATK-based), customizable with external models.	Built on NRAP-Open-IAM framework, enhanced with GUIs and user-friendly interfaces.
Feedback Mechanism	Updates via site monitoring data integration (e.g., pressure/plume observations).	Continuous updates based on industry partner feedback and national case studies.
Strengths	- Holistic system integration.	- Detailed wellbore/cement failure analysis.
	- Regulatory compliance (AoR, post-injection care).	- Practical for well-design optimization.
	-Open-source.
Limitations	- No explicit modeling of casing corrosion, plugs, or operational stresses.	- Does not account for plugs in the wellbore.
	- Limited focus on well-designed.	- Relies on ROMs with limited validation.

depicts a comparative analysis and summary of each assessment tool’s capabilities. This comparison could be a good starting point for stakeholders in choosing the right integrated assessment tool based on project needs.

The comparison shows that NRAP-Open-IAM is best suited for in-depth leakage analysis, regulatory compliance, AoR delineation, and long-term storage site monitoring. At the same time, it lacks some detailed well integrity assessments. On the other hand, REX-CO₂ is most effective for evaluating injection well design, cement debonding analysis, and mitigating certain operational risks; however, the assessment tool has a narrow scope, primarily focused on wells, and has gaps in plug modeling (e.g., cement plug failure). NRAP-Open-IAM provides a macroscopic assessment of the field (from the reservoir to receptors), while REX-CO₂ offers a microscopic evaluation of the well. By coupling both assessment tools, stakeholders could significantly enhance the quality of RA in GCS systems.

4. Cross-Category Comparison of RA Frameworks and Discussions

Our evaluation integrated the relevant criteria from four primary RA frameworks: Qualitative, Semi-Quantitative, Quantitative, and Integrated. This would help projects select a suitable RA framework based on their needs. Each category was assessed against seven standardized criteria using a normalized 1–5 scoring scale (5 = excellent, 1 = poor), based on usability, technical demands, and applicability across CCS project phases (Figure 16).

The qualitative approaches obtained the top scores in Ease of Use (5) and Time Requirements (5) because they maintain a simple structure requiring expert assessment. However, this method is limited in handling uncertainty and needs better data acquisition methods. The evaluation tools work well for early-stage screening but demonstrate restricted capability for leakage quantification and RA in advanced phases.

The semi-quantitative techniques performed strongly overall, earning 4 points in both Ease of Use and Flexibility categories. The approach quantifies risk to a certain extent but remains easy to handle with limited data and expert resources. These models demonstrate excellent suitability for the period, which bridges site selection with operational monitoring phases.

Quantitative methods achieved strong performance in Data Requirements (5) and Uncertainty Handling (5) since they excel at providing measurable risk predictions through scenario-based modeling. These methods failed to perform well in terms of Ease of Use (2) alongside Time Requirements (2) because they required complicated setup periods and extensive computational resources. These methods are best reserved for mature project stages where comprehensive data are available.

The integrated frameworks NRAP-Open-IAM and REX-CO₂ demonstrated a steady performance pattern in key operational areas. The system displayed high scores of 5 in all categories of Uncertainty Handling, Real-World Application, and Project Phase Flexibility, showing its ability to maintain precision and adaptability. The hybrid characteristics of these assessment methods make them the optimal selection for complete CCS risk management despite their moderate requirements regarding data and time usage.

5. Conclusions and Recommendations

Comparative studies for CCS RA show distinct benefits and drawbacks between qualitative, semi-quantitative, quantitative, and integrated assessment approaches. Qualitative RA methods like Risk Matrices by Arbad et al. [104] and SRF deliver superior site screening results in the early stages through expert hazard identification despite the limited data availability. However, these tools struggle to provide accurate assessments because they use subjective assessments instead of probabilistic models when used in later phases of the CCS project. The RISQUE and Bayesian Networks semi-quantitative tools merge engineering data and scenario RA but fail to detect long-term geomechanical hazards, including fault reactivation. Although some quantitative tools and integrated tools (CO₂-PENS and NRAP-Open-IAM) offer high accuracy in leakage quantification, these tools require extensive datasets and computational power, making their initial application impractical.

NRAP-Open-IAM and REX-CO₂ show effective results through their fusion of reservoir-scale geologic models with wellbore integrity assessment. For instance, NRAP-Open-IAM’s modular design allows for dynamic risk simulations across reservoir, leakage pathway, and receptor components, while REX-CO₂ focuses on cement degradation and material compatibility in legacy wells. The open-source design of NRAP-Open-IAM enables users to conduct dynamic risk analysis across reservoirs, leakage pathways, and receptors, but REX-CO₂ specifically analyzes cement degradation in legacy wells. However, current frameworks still exhibit gaps in addressing uncertainties regarding the induced seismicity in the storage sites and the integrity of legacy wells (existence of plugs, integrity of junctions in the lateral section).

Based on the comparative analysis of this review paper, the following points are recommended during RA in GCS projects:

• For comprehensive RA, utilize qualitative tools in the beginning and quantitative tools at later stages.
• For site-specific RA, employ an integrated RA tool to ensure wellbore integrity assessment and reservoir modeling.
• Improve current RA tools (specifically quantitative and integrated) to capture uncertainties related to induced seismicity and legacy wells configurations, especially the integrity of plugs and lateral well junctions.
• Promote open-source risk modeling platforms to increase transparency and reproducibility in the RA process.
• Strengthen regulatory frameworks and protocols to improve public trust and investor confidence in CCS.
• Encourage interdisciplinary collaboration between geoscientists, engineers, and policymakers to bridge technical assessment with non-technical considerations.

In conclusion, this review demonstrates that employing various RA techniques—qualitative, semi-quantitative, quantitative, and integrated—throughout the lifespan of a CCS project is essential for developing a comprehensive risk management strategy. Future research should focus on creating dynamic hybrid RA models that integrate both technical and non-technical aspects to improve stakeholder confidence and compliance while ensuring the long-term reliability of CO₂ storage systems.