The Role and Challenges of Alarm Management in Process Industries
Larry O'Brien
Vice President
ARC Advisory Group

Alarm management and process safety are closely intertwined. Your alarm man-agement system and associated operator intervention are often the last line of defense before a safety system trip. However, the role of alarm management is also important when you consider other layers of protection, such as the process control system. The ISA 18 and IEC 62682 standards govern alarm management for the process industries, while the ISA 84 and ISA 61511 standards cover safety instru-mented functions (SIF). A good understanding of the role of alarms in process safety is necessary both for optimum operator effectiveness and plant safety.

Many end users are undergoing alarm management projects at their respective plants or across their companies right now. In some cases, they are migrating from an old alarm management system. In others, they are initiating a new alarm management program. In either case, if you are under taking such a project, it's a good idea to have a comprehensive understanding of where process alarms and safety intersect. In alarm management, there are lots of different classes of alarms, all with their own requirements. The same can be said for process interlocks. Process interlocks, for example, prevent incorrect operation and possible damage to process equipment, while safety interlocks prevent hazards presented to humans and are designed to prevent death, injury, or a major process incident. A third classification of interlocks is also emerging that can be called Independent Protection Layer (IPL) interlocks. IPLs protect against abnormal situations, while safety interlocks respond to abnormal situations and take the process to a safe state. The same logic can be applied to alarms. All alarms should require an operator response per ISA 18, but not every operator response will be a critical safety-related response. A disruption in the process for a food plant, for example, must be documented in a cer tain way to satisfy regulator y requirements as a food safety alarm. This kind of an alarm is not an everyday process alarm that you would find in a downstream petrochemical plant, nor is it a safety alarm. Instead, it falls into a class called "highly managed" alarms. So, we now have multiple classes of alarms with multiple requirements.

ARC Advisory Group recently conducted a survey on current practices and trends in alarm management in the process industries. We also wanted to learn how end users, suppliers, consultants, and system integrators are approaching the of ten challenging issue of migrating existing alarm management appl i cat ions. A larm management in general continues to be a big issue in process plants, driven largely by the need to conform to current standards and best prac tices like ISA 18.2, EEMUA 191, and IEC 62682. The primar y goal of these standards and prac tices is to develop a continuous improvement approach to alarm management and ease the alarm burden on operators so they only see the information they need to see, par ticularly during process upsets or other abnormal situations.

However, adherence to standards and best practices is not the only issue end users face today. Advanced alarm management solutions have been available on the marketplace for many years now. As a result, many end users now need to migrate to a new alarm management application as their older ones become obsolete. In many cases, users are using this as an oppor tunity to improve their alarm management philosophy and implement some of the newer aspec ts of these solutions, such as dynamic alarms that can change in lock step with the dynamically changing state of the plant.

ARC received over 170 responses to the sur vey. Close to half of the total respondents were end users, while consultants represented over 19 percent of respondents. Suppliers represented a relatively small por tion of total respondents, at just 17.5 per- cent. Other respondents included OEM and skid-mounted equipment manufacturers, and system integrators. While not all respondents answered all survey questions, we did note general alignment in the responses, whether from end users, suppliers, or third-par ties. We believe that this is significant.

Figure 1: Alarm Management System

On an industry basis, the bulk of responses came from the oil & gas sector (over 24 per cent), while petrochemicals and bulk chemicals accounted for 19 percent of re-sponses. Regionally, most respondents were from North America and Western Europe. On a whole, these regions have the largest installed base of advanced alarm management applications.

Most sur ve y respondents had recently implemented a new alarm management and rationalization project at their company or facility; many on a companywide basis. Clearly, there is still a lot of activity in alarm management and rationalization and ARC believes this effor t will only escalate over the next few years as many end users face the need to migrate from older alarm management platforms, while other users who have not yet implemented advanced alarm management solutions will embark on new projects.

Over 35 per cent of respondents were taking existing applications and applying minor upgrades. The rest were fairly evenly distributed among those who were implementing brand new projects, those who were migrating to a new solution from a new supplier, and those who were migrating to a new solution from the same supplier.

Most respondents (72 per cent) indicated they follow the ISA 18.2 standard. This is consistent with the sur vey demographics, since this standard is very popular in Nor th America. Clearly, however, many users outside of Nor th America also follow the standard. Close to 20 per cent of respondents indicated that they follow the IEC 62682 standard, which closely follows ISA 18.2.

ARC asked respondents to describe briefly the three primary challenges they faced regarding alarm management project implementation. While we received quite a few different responses, we were able to identify three main challenges in the responses to this open-ended question.

Figure 2

Challenge #1: Getting Management Buy In and Allocating Resources

Many of the primary challenges listed deal with human issues, such as getting "buy in" from operators and management, and finding the appropriate amount of time, resources, and training to do the project effectively. Cost and funding issues were also prevalent. Ac tually performing the alarm rationalization aspect of the project was also listed as a primary challenge.

Challenge #2: Lack of Subject Matter Experts

Secondary challenges also included human issues such as buy in, ease of use, and basic issues such as time and resources. However, we star t to see more specific technical and implementation challenges as secondary issues, such as alarm philosophy development challenges, configuration issues, developing or redefining KPIs, ease of use, and database issues. People issues also start to become more specific, such as finding sufficient subject matter exper ts (SMEs), developing common work processes and procedures across the enterprise, and resistance to change by operator and other personnel.

Challenge #3: Alarm Rationalization and Consistency

We still see some cost and resource concerns repeated as tertiary challenges, but there are also more specific people - oriented and technology issues listed, such as keeping alarm rationalization up to date, management of change, and implementing dynamic alarming. Achieving consistency in alarm management while dealing with disparate sources of data was also pointed out as a key challenge.