Powering up older systems: a guide to upgrading legacy systems--what does and does not work.
Hazen Danforth, director of technology, Georgia-Pacific Corp., Brunswick, Georgia, USA
Taina Heimonen, product manager, machine automation RTD, Metso Automation, Tampere, Finland
Kathy Hutson, marketing manager, Honeywell Pulp, Paper and Printing, Phoenix, Arizona
Ian Journeaux, manager, process development and control, Stora Enso North America, Wisconsin Rapids, Wisconsin
Betty Naylor, director of legacy system migration program, Invensys Foxboro, Foxboro, Massachusetts
Alison Rowat, product manager, Metso Automation USA Inc., Norcross, Georgia.
SOLUTIONS! What is the biggest challenge to upgrading older "legacy" process control systems?
DANFORTH: The ability to upgrade legacy systems largely depends on the vendor with whom you are working. Some have maintained compatibility over 25 years. Others have replaced their systems every five or six years. Someone looking to have a clean, crisp, new system might look down on all the gateways and software transition programs you need to make these kinds of upgrades, but the gateways are necessary to enable communication with each different software platform. They allow you to achieve maximum value from your precious investments while obtaining modern technology. Of course, you must recognize that with an old controller running in a new platform you cannot do everything. As you mine down into the system, you reach a point where you have old hardware that simply cannot do what the fully modern system will do.
HEIMONEN: The biggest challenge is making sure all the existing information works effectively in the upgraded system. The updated system should be able to work with new systems and accept applications from older versions. To protect customers' investments, systems that provide reusability of existing system components and expandability with new components provide the best results.
HUTSON: It is vital to preserve as much of the original investment as possible while taking full advantage of the features of the new system. System advances help to justify the migration by providing improved decision support, diagnostic and asset management tools, and advanced control strategies designed to maximize business value. Users want to improve the return on capital investments continually by boosting productivity or extending the life of existing assets. When upgrading a process control system, minimizing the cost of change is key. This encompasses many aspects of the project from training of personnel to maintaining intellectual assets to preserving as much wiring as is practical. The control system vendor should have solutions in all these areas. Ultimately, reliable systems that will grow as your needs grow and offer a history of migration support will minimize this cost.
JOURNEAUX: The biggest challenge to upgrading "legacy" process control systems is capital justification since the industry's capital availability is limited and the underlying measurement technology and control strategies have not evolved greatly since the original systems were installed. This is amplified by the fact that a paper machine gauging system upgrade can cost upwards of several hundred thousand dollars.
NAYLOR: The biggest challenge is minimizing cost, risk, and process downtime. Newer systems have input/output (I/O) modules with a different signature size and shape and provide different capacity and functions than the older I/O products. Some vendors have forced their customers to do a total ripout replacement and install new wiring, new cabinets and racks, and new termination assemblies. Instead, I/O modules should be designed to be form-fit replacements for the original legacy I/O modules. These plug-in replacements should ensure that no field rewiring is necessary and the existing infrastructure of racks, cabinets, termination assemblies, and power supplies are reused. A typical migration installation should take hours rather than days or weeks.
SOLUTIONS! What are elements of a successful legacy system upgrade?
DANFORTH: The most important element is defining what the upgrade will accomplish. If you can open and close the valves smarter, save steam, or be more efficient in chemical addition, then it is worth it. If you cannot, it is not. Your ability to upgrade legacy systems depends largely on what you have done since the original system installation to keep current with new software versions. If you are only one or two generations behind, you should be able to upgrade successfully. If you are fifteen generations behind, you have a huge problem. That situation varies by mill even within the same company. Some mills install new software versions on a regular basis while others in the same company are using 20-year old software. If you are more than three to five major releases behind, you probably will not have sufficient memory to run the software. You will need to spend heavily in hardware upgrades. Our mill has up-to-date systems. We believe that this has helped us operate the mill better and smarter.
HEIMONEN: Application compatibility should ensure that control logic will operate properly after the upgrade. Also, new applications must be easy to implement through effective training. Mills should also understand the options for adding new components or subprocesses to the system and how easily they can be added to the existing system.
HUTSON: Identifying and staffing a strong migration team with expertise on the process, the system, and system upgrades will greatly reduce the project risk and ultimately the adoption of the new system by the mill team. Elements that must be addressed include:
* Minimizing or completely eliminating process downtime
* Operations, maintenance, and engineering involvement in planning and training for the "new" system
* Efficient reuse of wiring, configuration, graphics, and control strategies
* Availability of migration services that will make use of old graphics, control strategies, and I/O mappings.
The mill must decide on an approach that best meets their needs and goals whether it is a complete mill shutdown or a more phased migration plan. Many mills choose the phased approach. This spreads the costs over many years. Open standards typically enable the human interface to be replaced first followed by the rest of the system. Replacing the human interface can bring many benefits immediately including improved operator awareness, intelligent alarm management, and quick access to all process-related information.
A mill can also use a phased approach when upgrading a process controller. The key to controller replacement is minimizing disruption when moving the configuration from a legacy controller platform to the "new" system. Using services that retain field wiring and optimize the translation from the old system to the new system can do this.
JOURNEAUX: Connectivity is critical. In a gauging system upgrade at our mill, the availability of standard OPC interfaces for our gauging systems regardless of vintage ensured that the upgraded systems integrate seamlessly with the existing information infrastructure.
NAYLOR: Time and risk are the most important elements. The design of the upgrade is critical. If no field wiring or infrastructure is altered, the original system can be put back in place quickly in the event of a problem while the vendor takes the problem off-line for resolution without further affecting the customer's production schedule. If some wiring needs moving to spread out the load to match the capacity of the new system I/O modules and a problem occurs, a fix must be determined, causing the customer increased production downtime. This could be days or weeks.
ROWAT: The most important elements are little or no process downtime, user-friendly system features or operators (which makes for easy training), smooth application transformation from the old system to the new system upgrade, and minimum manpower required for programming the new system.
SOLUTIONS! How can conflicts be avoided between legacy components and new system components?
DANFORTH: The number one issue is the amount of memory in the old components. In some cases, the hardware does not have enough memory to run the new software and cannot be upgraded. If I were purchasing a new system, I would look at the long-term ability of the vendor to maintain a modern level of performance in its older systems. If they have a history of regularly making software upgrades for older platforms without having to upgrade hardware, it is more likely that the new platform will follow that same arc. In one example at our mill, the basic system processor included a chip that was no longer manufactured. The vendor designed a new board with modern components that fit in the older controller. In another case they told us that the controller was obsolete and did
not upgrade it.
HEIMONEN: Carefully designed connectivity architecture between legacy components and new system components is critical. Several performance levels are possible such as full compatibility, the new system is able to handle older processes, or no compatibility exists (requires double application).
HUTSON: The easiest way to avoid conflict between legacy components and new system components is to select a vendor that has designed evolution of the legacy components into the new components. This eliminates most if not all common problems associated with interfacing multiple systems. When you need to mix components, use OPC standards when possible and make sure the new system architecture allows for communications between multiple areas.
JOURNEAUX: In general, conflicts occur at the information access and operator interface levels. These issues can be addressed by ensuring that systems use a common industry-standard communications protocol and by moving operator monitoring to mill-wide systems. The availability or more correctly non-availability of spare parts for legacy process control systems can force an upgrade to avoid possible extended outages. In some cases, the availability of sensor technology that is not available for older architectures can be the catalyst for a new system.
NAYLOR: If the vendor repackages I/O modules to be form-fit replacements for the original legacy I/O modules, it can ensure there will be no conflicts or I/O mapping issues. The new I/O modules should line up exactly with the original system's termination assemblies. Each replacement module should meet or exceed the capacity and function of the modules it is replacing. Control database conversion and display/application conversion should be done by engineers knowledgeable in other vendors' control logic. Automated mapping tools can ensure that the conversion is done quickly and accurately. In addition, a full system factory acceptance test should be performed before installation to ensure the conversion is implemented correctly. Mills should always insist that the vendor guarantee backward compatibility between today's products and future products.
SOLUTIONS! What are the signs when upgrading is better than replacing legacy process control systems?
HEIMONEN: This depends on whether the upgrade uses proprietary hardware or third party hardware. The average lifetime for proprietary hardware is longer and in most cases better supported than commercial hardware. If the hardware is becoming obsolete, one should replace it. If the product is supported, then upgrades are a good alternative. Sometimes it is difficult to find specialists for the older systems--you are alone with your problems. Availability of spare parts is often low or declining.
HUTSON: Problems arise when older control systems become obsolete or when these older systems are asked to perform functions for which they were not designed. When a control system approaches obsolescence, maintenance and support costs rise at a rapid rate. If an evolutionary migration path to new technology is not available, the cost of ownership for the aging technology must be evaluated against the cost of the new system. Cost of ownership should include increasing spare parts and service support costs, increased risk of process downtime due to hardware failures, and the value of the knowledge base for the older system. The value of benefits from the new system should also be included in the analysis. Some advantages include:
* Benefits in the manufacturing operations through increased yield, reduced chemical and energy costs, and improved quality
* Reduced maintenance and support costs
* Improved work processes and workflow.
NAYLOR: A customer's decision to replace a legacy control system typically revolves around increased cost of ownership to continue to service and maintain the older legacy system, the need to be able to do more with their system in terms of plant performance and third party application integration, and a recent trend by vendors to declare their legacy systems "end-of-life" forcing customers no alternative but to replace their older system with a newer one.
IN THIS ARTICLE YOU WILL LEARN:
* The biggest challenges to upgrading older "legacy" process control systems
* The most important elements for legacy system upgrades
* Avoiding conflicts between legacy components and new system components
* Signs indicating when replacing is better than upgrading legacy process control systems
* "Integration of Process Control and IT," Steve Sena, Solutions!, September 2003, p. 42.
* Search the TAPPI database at www.tappi.org. Type in "Process Control" in the search engine in the upper right hand corner of the home page.
Stora Enso North America and Aracruz Celulose case studies available at www.tappi.org
Two case studies on process control upgrades are available at www.tappi.org: Stora Enso North America's upgrade of its gauging systems and control technology and Aracruz Celulose's upgrade of its Fiberline B evaporation unit controls. To access these articles, go to www.tappi.org and click on the Solutions! link. Then click on Online Exclusives for October. If you are accessing these stories after October 2003, click on the Solutions! link and then the "Online Exclusives Archives" link in the middle of the web page.
Edited by ALAN ROOKS, Editorial Director
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Process Control|
|Publication:||Solutions - for People, Processes and Paper|
|Date:||Oct 1, 2003|
|Previous Article:||Looking for a little R-E-S-P-E-C-T.|
|Next Article:||What it means to be green: how do you define "green chemistry" and what does it mean for papermaking?|