Interested in linking to "Power to the process"?
You may use the Headline, Deck, Byline and URL of this article on your Web site. To link to this article, select and copy the HTML code below and paste it on your own Web site.
By Larry O’Brien, Fieldbus Foundation; John Rezabek, ISP; and Herman Storey, Herman Storey Consulting
Intelligent field devices, whether they are based on Foundation Fieldbus, Profibus PA, HART, or other protocols, are pervasive in modern process manufacturing. The original promise of having microprocessor-based devices was that they would transform the way we see the information related to these devices and to the processes they control. Maintenance practices could be transformed so that devices with impending problems could be identified sooner. No longer would field technicians have to go to the device itself to get relevant information. Instead, information would be provided directly to the process automation system, plant asset management system, or indeed any other systems or software in the plant that required it. The technology offers the promise of significantly lowering risk while lowering maintenance costs.
In many cases, the promise of intelligent field devices in the plant remains unrealized. This is not so much a technology issue as a people or work process issue. Too many users are employing old maintenance work processes with new technology. The new devices and applications are installed, but the operators and technicians stick to their old way of doing things, their old preventive or routine maintenance practices, and never really take advantage of the huge amount of information that is available to them. It seems clear that the process industries would benefit from a standard set of work processes and best practices for intelligent device management (IDM). This would give end users an effective blueprint for achieving the significant economic lifecycle benefits associated with intelligent devices.
The process industries are notoriously conservative and reluctant to change. In many cases, this reluctance is justified. The processes under control are volatile and dangerous, and any misjudgment can result in serious consequences. If you look around today, however, you will see that the process industries have enthusiastically adopted intelligent devices and IDM applications. Purely analog 4-20 mA field devices are being sold less and less. Why then have we not embraced this technology to its full potential?
Current work processes and current standards for maintenance management are built around time honored traditions including run to failure and periodic inspection and testing programs. These work processes were developed for equipment that doesn’t have built-in diagnostics. Pressure vessels and piping can be expected to fail very slowly over decades if operated within their design limits. Inspect and test programs are good work processes for such equipment, especially if they are optimized by risk-based techniques.
Instruments can also be pressure-retaining devices, but their primary purpose is for measurement and control. The measurement and control functions are in direct contact with the process and therefore subject to wear and tear that can degrade very quickly or suddenly compared to typical pressure retaining equipment. Furthermore, microprocessors embedded in these devices can do diagnostic work to identify when components are malfunctioning or degrading. In many cases, testing is unnecessary for these devices, and inspection can be simplified. Instead of test procedures, all that is needed are tools and processes that utilize the built-in diagnostics.
In the real world, however, we are finding that these tools and processes are often implemented incorrectly, and many times not at all. New technologies require new ways of doing things. Look at cars today for example. Any new vehicle sold today is going to contain a sophisticated onboard diagnostic computer that will be able to tell you what’s wrong with your vehicle based on specific diagnostic codes. Mechanics no longer use trial and error to diagnose problems. Just plug in the diagnostics unit and you can find out what the problem is. But this didn’t happen overnight. Mechanics had to be trained in the new way of doing things. New processes were instituted and taught at trade schools.
The argument for using intelligent device management has a strong economic value proposition. It’s not just technology for technology’s sake. Detecting problems before they happen reduces unplanned shutdowns. Some in the industry call this a “save.” Depending on the application, one save can pay for your entire automation investment. The potential to cut maintenance costs is also significant. Intelligent devices allow you to implement a more predictive and proactive maintenance strategy.
In addition to saves, routine maintenance tasks can be eliminated because you no longer have to guess if there is a problem with a device. Many maintenance activities for instrumentation result in “no problem found.” These activities traditionally require personnel to enter hazardous areas, climb to areas with poor access, and spend time on unnecessary tests. Diagnostics can confirm proper operation without all of this expense and risk and can lead to a quicker resolution of operational issues.
Perhaps using the analogy of the automobile is a little too simplistic for the process industries. In today’s process plants, operators, engineers and technicians must navigate an already complex landscape of applications and communication pathways that can often make the problem of using diagnostic data from intelligent devices a herculean task. Multiple databases and interfaces permeate the industry. Digital field devices must interface with a host system, which in turn communicates with a data historian, configuration database, alarm management systems, plant asset management systems, and higher-level computerized maintenance management systems. Enterprise connectivity is a must, since work orders are typically generated through systems like Maximo or SAP. Many users get stymied when it comes to integrating all these disparate elements together. Questions like who gets access to information from intelligent devices and when must be answered, and corresponding data flows must be planned. Workers at multiple levels in the enterprise (in multiple disciplines) must be informed, and coordinating communication between these workers creates a problem in itself.