6.2 Risk analysis

Risk analysis involves interactive exchange among risk assessors, risk managers, regulators, local community, news media and interest groups. A qualitative type of risk analysis should be performed at the early stages of a project to help site screening, site selection, communicating project aspects to the public, and aiding regulators in permitting projects. Subsequent to more detailed site characterisation and modelling efforts, quantitative risk analysis may be performed to estimate the likelihood of human health and environmental risks. Furthermore, stakeholders such as regulators and insurers may require risk analysis to support incentives, such as loan guarantees to large projects. A successful risk analysis will always be linked to monitoring and modelling plans for a given storage site. The risk analysis results can be used by industry, investors, and insurers to understand the potential liability associated with projects and build that into the cost of developing a CCS project (NETL, 2011). Once risks are understood, a project developer must take steps to avoid or manage the risks that are not judged acceptable. In the risk management step (Fig. 6-3), inputs from the risk and characterization processes, and a variety of social, political, and techno-economic parameters are used to prioritize, monitor, control and mitigate risks (NETL, 2011).

E. Fig . 6-3

Fig. 6-3. Risk management workflow diagram for a commercial-scale storage deployment program. Adapted from Korre and Durucan, 2009; NETL (2011) - modified.

It is important to distinguish between risk and uncertainty, although they may be related (DNV, 2010). Uncertainty is a critical factor to assess in the context of risk/performance assessment (NETL, 2011) that can be related to different features:

  1. Parameter uncertainty, associated with input parameters, is commonly recognized and addressed in modelling approaches, via e.g. numbers of simulations based on a randomly sampling of uncertain parameters (Monte Carlo approach).

  2. Conceptual model uncertainty, concerning how the real world (geological cross-sections, faults or fractures zones, etc.) is represented and abstracted.

  3. Modelling uncertainty, concerning the underlying mathematical modelling and its inherent assumptions, e.g. boundary conditions. Modelling uncertainty can be assessed qualitatively by comparison of results from different mathematical models, via benchmarking exercises, which are recommended to enhance modelling credibility and confidence.

  4. Scenario/event uncertainty, relating to whether scenarios/events representing all potential hazards have been identified and analysed (Stenhouse et al., 2009).

A core part of qualitative or quantitative ranking of risks for CGS involves assessing the level of knowledge available, and the subsequent implications on the level of risk. Proper management of uncertainty helps to manage down the assessed level of risk throughout the life of a CGS project. In particular, if risks are ranked conservatively, reducing uncertainty will generally result in a lowering of the assessed risk (DNV, 2010).

An important step of a detailed risk is the qualitative or semi-quantitative prioritisation of the risks, where risks are categorised and ranked in terms of likelihood and magnitude of consequence. The ranking allows high-priority risks to be identified and plans for mitigating or controlling them to be developed, while lower-priority risks can be placed on a watch list. Other risks, with mid- or unknown-priorities, may undergo further analysis or investigation. As more information is obtained from site characterisation, modelling, and monitoring, the risk priorities can be updated. Later stages may also include model simulations to assess the probabilities and impacts of selected scenarios. Such plans will heavily rely on monitoring data and will generally stipulate an "if-then" process: if the monitoring system detects a problem, then specific actions will be performed to address the problem, either immediate action or need for an additional, focused monitoring. A good monitoring and mitigation plan will decrease the risk and uncertainty associated with many potential consequences (NETL, 2011).

 

in depth

6.2.1 Probabilistic risk assessment

In probabilistic risk assessments, explicit probability distributions are used for some (or all) parameters. ...

6.2.2 Risk assessment methodologies

A number of different assessment methodologies have been and are being applied to CCS-related projects....

6.2.3 Features, Events and Processes methodology as an approach to risk assessment for CO2 storage

Many of the ongoing risk assessment efforts are now cooperating to identify, classify and screen all factors that may in...