Barrier management evolves
Barrier failures tend to be primarily human, technological, organizational or a combination. At the Chernobyl power plant, Ukraine, unauthorized tests led to loss of nuclear containment in 1986. A software design flaw caused a drilling rig under construction in Singapore to fail in one leg in 2012. The immediate causes of the 2005 Texas City Refinery were human and technical faults, but the root cause was organizational: a poor safety culture.
Safety barrier management is well established in the industry. A succession of barriers can be applied to contain the risk of each foreseeable load or threat below a target level (figure 1). Barrier status is monitored by audits and inspections whose frequency is determined, in the RBI approach, by the additional risk that would be created if a particular barrier failed.
Barriers degrade if not actively managed. Different barriers also provide different risk reduction and degrade at different rates, so standard practice is to measure them at different intervals. If barrier failure or weakening goes undetected, unreported or unresolved, overall risk may exceed target (figure 2). It is therefore important to: know the actual status of barriers; easily monitor and analyse barrier performance; ensure that degradations are managed efficiently; and optimize operational risk levels while accounting for degraded barriers.
DBM meets these challenges by continuously assessing barrier status so maintenance, repair or replacement can restore risk to target levels (figures 3 and 4). It allows avoidance of activities that might require a failed barrier to function. Frequency and thoroughness of assessment remain risk-based, and occur at different levels of the organization.
Designing a DBM approach involves specifying an appropriate number of barriers of suitable reliability to meet the risk target. Information is visualized and accessible to relevant stakeholders through decision support tools
BG Group is similar to many operators in that its barrier management approach has evolved, and continues to, through its people’s experience and insights into risk management. The legacy is a mix of manual, semi-automated and automated tools and processes to help monitor and manage barriers.
The company’s local assets make their own risk management decisions. Barrier management processes vary in the degree to which they can be described as dynamic. Some are web-enabled and can provide a semi-live picture, Nishapati said.
“We are on a journey and nowhere near the end of it, but the benefits of DBM that I have described are already clear,” he stressed.
DBM also adds to BG Group’s understanding of the potential of Big Data management, he added. “It combines a lot of different data in a format that facilitates decision makers being able to digest it and extract pertinent correlations that help to keep operations safe in the field.”
While the industry measures barriers, it does so sub-optimally, and status updates and decisions are not always based on risk contribution, Tørstad observed. “We need the equivalent of RBI for barriers: a risk-based barrier management approach. DBM meets that need.”
The case for DBM becoming the prevailing model of barrier management is strong, Tørstad added. It reduces ignorance through more frequent sampling of barriers and smarter use of more relevant, richer and up-to-date monitoring data. It increases and improves decision support for prioritizing critical maintenance, implementing compensating measures, and for day-to-day risk analysis.
By ensuring more regularity in operations, DBM supports both asset safety and operational efficiency by avoiding process upsets. Early indications of barrier degradation provide flexibility for the scheduling of small, multiple corrective actions rather than intrusive and large ones.
“DBM reduces the cost of barrier management as prevention is cheaper than correction,” Tørstad concluded. “It can provide a better balance between safety and cost.”
Human and organizational benefits
Even if data sources for DBM are initially manual and based on engineering judgement, the value is in the process of evaluating barriers, say weekly, then planning prompt action, where needed, to restore risk to target levels, Nishapati added. Experienced operational managers do this anyhow, he noted. “People can do it differently or reach different conclusions. Everyone in BG Group now works to the same process, asks the same questions, and hopefully holds, as much as possible, the same world view from which to make their particular decisions.”
Consistency and a common approach across assets allows within-plant and wider-resource prioritization, not just benchmarking, he added.
BG Group’s experience underlines how others can begin the journey towards implementing DBM without needing vast amounts of data or a full array of real-time monitoring and response solutions.
“One big learning for us is that there is huge value in the process and discipline of people bringing the intellectual rigour of barrier management to looking at the entirety of their barrier models,” Nishapati emphasized.
“I expect this [DBM] approach to safety management to become the industry norm within five to ten years. A lot of other industries are ahead of us or are at least on the same journey, and we can learn from each other. The technical issues are not insurmountable.”
Potential challenges in designing, implementing and maintaining DBM include knowing the continuous status of human, technology and organizational barriers, and may involve analysis of large volumes of data.
A DNV GL-led joint industry project (JIP) on decision support tools (DSTs) for DBM will address a lack of common terms and language for communication in risk management. This will enable improvement of practical, real-time DSTs and risk management and deliver continuous knowledge of barrier status.
Participants will test and develop best practices, data sources, and tools for standardized bow-tie diagrams, response trees, decision protocols and pilot-scale decision support systems.
Bill Nelson, principal consultant with DNV GL - Oil & Gas said: “We believe that decision support tools for DBM will help the industry move toward the long-range vision to reduce operational costs and decrease downtime while increasing safety for offshore operations.”
 ‘Investigation report no. 005-04-I-TX: refinery explosion and fire’, US Chemical Safety and Hazard Investigation Board, March 2007