By Michael Marvin, P.Eng., Senior Applications Engineer
WHETHER YOU ARE establishing alarm limits within a control system, acknowledging alarms, reacting to them or even at times, ignoring them, the notion of alarm management is not new to the process control industry. Wherever you look in chemicals, oil and gas, power and utilities, pulp and paper, mining or other industry verticals, youíll find control systems generating large quantities of alarms and events, and typically, more than operators can effectively handle.
A White Paper released by ARC in October of 2004, on Unified High Performance Manufacturing Using Operational Excellence, reports that reducing unscheduled downtime by 2-5% can realize a 40-50% benefit. This benefit includes increased productivity without compromising quality or safety. Plant personnel are keenly aware of the problem; anyone who has spent more than 15 minutes in a control room can see it clearly. So why are there so many alarms? The problem isnít that operators, engineers, managers or supervisors donít want to deal with nuisance alarms. They do. They want to take control of this overwhelming situation. The real problem is that they donít have the methodology and tools they need to systematically integrate and successfully administer an effective alarm management solution.
This article will outline five practical steps you can take to reduce nuisance alarms and improve alarm quality for control room operators. This includes the ability to identify deviation from normal operating ranges as the process or equipment approaches equipment design thresholds, unsafe, or undesirable operating conditions. The result is improved day-to-day operations, reduced incidents and downtime, and in those cases where an incident does occur, dependable and safe management of the situation.
The steps are as follows:
- Benchmark, Assess, and Identify
- Create a Master Alarm Philosophy Document
- Acquire the Right Tools
- Engineer Alarm Settings
- Monitor and Maintain Improvements
Letís take look at each in more detail:
Step 1: Benchmark, Assess, and Identify
The first step of any serious improvement initiative is benchmarking. If you canít measure your present situation, you canít improve it. Accordingly, the first step is to keep track of alarm rates for several weeks in order to get a baseline measurement. Once thatís done, the next step is to assess how your plantís current alarm levels measure up to industry standards. The Engineering Equipment and Materials Users Association (EEMUA) recommends a wide range of best practices for alarm management and has published a document that outlines industry standards: EEMUA #191 Alarm Management Guidelines (visit eemua.co.uk to obtain your copy).
When you have finished benchmarking and assessing your current performance, you can start identifying opportunities for improvement. Below are the key questions one needs to answer when making an assessment. Note the checklist in descending order of importance:
- Is dynamic (real-time) alarm loading acceptable for all operators?
- Does the dynamic alarm prioritization meet industry standards?
- What are the troublesome tags on the system during steady-state operation?
- How does the configured DCS alarm count measure up to standards (alarms per tag)?
- What does the configured alarm distribution look like in comparison to standards?
To get a very quick snapshot of where your plant ranks according to EEMUA standards, Matrikon has posted a calculator on its website (www.matrikon.com/plantperformance). Upon benchmarking your plantís current state and assessing the areas for improvement you can move on to stage two and begin to develop a strategy.
Step 2: Create an Alarm Philosophy Document
Develop a corporate standard that clearly defines your siteís alarm management strategy. This is known as an Alarm Philosophy Document, and it should answer a number of key questions:
- When should an alarm be configured? It is important to note that alarms should not be configured so that they only indicate when a value has ďdeviated from the norm.Ē Instead, alarms should only indicate the need for operator intervention.
- What rules should govern alarm prioritization?
- What are the alarm system tasks and who will be responsible for each task?
- What are the change control procedures for alarm settings?
- What are your performance targets and how will you measure them?
With an alarm philosophy defined the challenge many sites face is effectively measuring and managing the goals outlined in the philosophy. Many tools that exist on the market today claim to manage and reduce process alarms. I have found, however, that most of them fail to first define the objectives of the alarm system and the ground rules for implementation. Without this information, it is difficult to create an effective solution.
Step 3: Acquire the Right Tools
Buy software that enables you to monitor and analyze the performance of your plantís alarm system. A good software package will pay for itself quickly by reducing or even avoiding downtime and by increasing operator productivity and thus plant productivity.
Be cautious not to confuse simple printer replacement software solutions with true alarm management packages. Carefully scrutinize a vendorís solution before committing to it.
The following is a list of functional requirements that any good alarm management software should satisfy. Although it is not an exhaustive list, it gives an outline of the most common tasks that the tool should perform:
- Provides an alarm and event historian that constructs an audit trail of alarms and events, facilitating incident reviews, nuisance alarm analysis, alarm flood analysis, and operator action analysis. These are the reports you need to effectively diagnose and improve all plant states from normal operations to upset conditions.
- Easily calculates alarm-related Key Performance Indicators (KPIs) as outlined in the EEMUA #191 standard.
- Automates the Management of Change (MOC) process and reconciles the Distributed Control System (DCS) against an engineered, Master Alarm Database.
- Provides an audit trail of changes to DCS configurations.