Lifecycle planning for your DCS

Lifecycle planning for your DCS

By Mitch Cochran, Process Control Solutions LLC

The advent of the distributed control system (DCS) in the 1980s brought powerplant control into the digital age, combined the formerly separate functions of control and data acquisition, and enabled process data to be integrated with enterprise-wide business-management data. In the 1990s, DCS suppliers further enhanced their systems, moving away from proprietary hardware and software, and toward more inter-operable, standards-based networks. Net result: Today’s DCS is a reliable, user-friendly product that dominates the controls sector of the electric-power industry.

But all products, no matter how dominant, have a finite lifecycle. This is particularly true for microprocessor- based products. Owner/operators of DCSs, therefore, eventually must address such lifecycle planning problems as lack of manufacturer support, shortage of spare parts, and processor capacity limitations. Lifecycle planning also involves evaluation of budgetary constraints as well as corporate strategies regarding future expansion.

This article discusses lifecycle planning for one of the most popular DCSs in North America—the WDPF (for Westinghouse Distributed Processing Family), manufactured by Emerson Process Management’s Power & Water Solutions unit, Pittsburgh, Pa. Note that Emerson purchased the former Westinghouse Process Control Div when Siemens AG acquired that company’s power generation business. Intent here is to help owners of any make and model DCS decide how long they should retain their existing control system, and at what point they should make the leap to a new system—using the WDPF as an example.

WDPF maintenance

The WDPF control system has been on the market since the mid 1980s, and is approaching the end of its product lifecycle (Fig 1). That’s not a knock against the product. That’s a fact established by the manufacturer’s product-support policy. Emerson’s policy is to provide a minimum of 10 years of product support. The following definitions from the original equipment manufacturer (OEM) apply:

  • “Current products” are the most up-to-date products offered with published pricing, normal lead times, and complete support. Current products are recommended for new installations.
  • “Active products” have been functionally replaced by “current products,” but remain available with published pricing, normal lead times, and complete support. Active products are intended for expansion of existing systems. Transition to “active product” status marks the start of the 10-yr product-support commitment.
  • “Maintained products” generally are not available for purchase. Support is available, but is focused strictly on maintenance—replacement parts, repairs, and field service. Emerson encourages owners of products in this category to initiate their lifecycle planning.
  • “Retired products” are no longer available for purchase. Support is limited, slower, more costly, and subject to material availability. Transition to “retired product” status marks the end of the 10-yr product-support commitment.

The WDPF control system is classified as a “maintained product,” and is quickly approaching “retired product” status. At the management level, therefore, WDPF users should be asking themselves: How long— and at what O&M cost—can I maintain my DCS? How much—and in what year— should I budget for a system upgrade?

At the technician level, users should be asking: Are my system backups complete and up to date? What can I do to be prepared for a software server crash? Where can I get refurbished WDPF parts? Could I benefit from having a plant simulator? What can I do to alleviate memory capacity limitations?

System backups. To prevent the loss of critical configuration software during a hardware or software failure, WDPF systems must be backed up properly. Recovering from a harddisk crash on an operator WEStation is routine maintenance, requiring only a blank hard drive and the “boot net–install” procedure from the Software Load Kit U0-8001-2.x.

The software server, however, is the most critical drop in the WDPF system, since all other drops are loaded from the software server. So recovering from a hard-disk crash on the software server is a much bigger deal. This requires Solaris and WDPF Release CD-ROMs, back-up tapes, third-party licenses, various custom files and drivers, and more. Many times, if proper back-ups have not been maintained, recovery from a hard-disk crash on the software server cannot be accomplished completely.

The solution: Maintain a “hotspare” backup hard drive for the software server. With this action, even the worst-case scenario—recovery from a software server hard-drive failure—becomes routine maintenance.

Serial interface QLC cards contain configuration files in battery-backed RAM memory. These files also must be backed up to storage media, in case a QLC card fails. The WDPF is not self-documenting. Therefore, all logic drawings— known as “SAMA” drawings because they follow the Scientific Apparatus Makers Assn standards—must be maintained. An “as-built” drawing review may be needed, in order to bring your drawings up-to-date.

All of these maintenance tasks should be performed, at a minimum, during an annual inspection of the WDPF system. At this time, preventive maintenance also should be conducted on the WDPF hardware, to prevent dust, soiling, or normal wear-and-tear from impairing system reliability. The annual inspection also should include housekeeping tasks for the Unix file system, and documentation of processor memory.

Capacity limitations of the WDPF system can hamper system operation and maintenance by making it impossible to add or modify DCS control logic, alarming, or monitoring functions. Lack of available database space in the distributed processing units (DPUs) and in the data highway system IDs—referred to as SIDs—are the most common capacity problems encountered with a WDPF control system.

One action that frees up database memory is cleanup of the DPUs. This work does not require any additional hardware, and most of it can be done offsite so that plant downtime is minimized. The cleanup procedure identifies and deletes unused points and obsolete logic. It also identifies alarmable points that could be moved to an empty DPU. A total plant outage is not required to accomplish this task.

Another option to alleviate capacity limitations is to add another DPU processor to the system. Each DPU comprises 120 kilobytes (kB) of database space to which alarms from other DPUs could be re-located. Adding a DPU with cabinet and QCrates makes it possible to relocate I/O points and control functions from other DPUs, as well as moving alarms.

Rebuilding the system point directory can alleviate the problem of SIDs being “maxed out.” In addition to identifying and deleting unused points and obsolete logic, point broadcast frequencies will be optimized. Preliminary steps can be performed offsite on the DPU source code, but a total plant outage of one to two days is required to load the modified DPU source code and rebuild the system point directory.

Hardware maintenance. Several options are available to users, when it comes to hardware maintenance of the WDPF system. Support from the OEM includes:

  • Emerson Sure-Serve Contracts. The OEM of fers repai r and exchange services designed to accommodate varying needs and delivery requirements.
  • Standard Repair and Return. Under this traditional scenario, customers return items for inspection, test, and repair. It is Emerson’s objective to service and return all company-manufactured items within 30 days of receipt.
  • 24-Hour Exchange. This option is designed to address customer needs in the event of an emergency or other critical situation. A replacement item will be shipped within 24 hours of receiving the malfunctioning part. This level of service can be provided only for instock items.
  • Advance Replacement allows for critical situations where you must have a replacement item dispatched immediately. Once you call for this service and your purchase order or credit card is received, a refurbished part, if available, will be shipped to you. You will be responsible to ship your serviceable item back to Emerson within 30 days.

Users also can tap several non- OEM suppliers of WDPF spare parts. For example, Sun Microsystems and HP parts—including RAM (random access memory), hard drives, HP optical drives and optical discs, Ultra-5, and Sparc-5 machines—can be purchased from Sun resellers. These parts also can frequently be found on E-Bay. Used hardware—including data-highway interface (DT Box) and DPU hardware—is available from resellers or customers who have recently upgraded from WDPF. For instance, DeepSouth Hardware Solutions LLC, Osyka, Miss, buys and sells surplus WDPF parts for powerplant and turbine control systems.

Advanced simulators are invaluable tools for plant operator training. In addition to learning the basic DCS windowing functions, trainees can gain familiarity with normal operating procedures—such as plant startup and shutdown—and abnormal plant situations—such as feedpump trips, gas-turbine runbacks, and so on.

Simulators also can serve as a valuable maintenance tool. For example, I&C technicians can use them to build or maintain competency in such tasks as loop checks, instrument calibrations, and testing logic modifications prior to actual implementation on the DCS. Finally, because the simulator runs on standard control-system hardware, it can double as a “hot-spare” parts bin, thus improving DCS availability when critical—and increasingly difficult to obtain—parts fail.

Time to upgrade

Despite the best efforts of the maintenance crew, a computer-based control system eventually becomes obsolete, and must be replaced. For WDPF users, replacement options include migration to Ovation retaining the QLine I/O, or a total system retrofit to another DCS technology—such as the Siemens T3000, Foxboro, Bailey, and Honeywell. A comprehensive controlsystem assessment will help users decide when it’s time to upgrade. The assessment might include:

  • System capacity evaluation—an evaluation of DPU memory and system point directory capacity.
  • Interviews with O&M personnel. A questionnaire is a useful tool to quickly assess the system hardware configuration. Interviews with operators and maintenance staff can help identify specific areas that need improvement.
  • Analysis of historical data. Operations logs and historical data can be used to identify areas where the control system should be optimized or upgraded to improve plant availability, efficiency, or emissions.
  • The Emerson System Assessment Tool. End-users and owners of WDPF systems can access Emerson’s automated assessment tool, available on the web at www.wdpfusers. com, for help with their DCS lifecycle planning.

Ovation option. Some power producers are doing away with the threeletter acronym that has for decades dominated process control (DCS), in favor of the two-letter acronym that dominates the retail and residential world—PC. Emerson seems to have carried this concept the farthest, with its Ovation product.

First released in 1997, Ovation is a fully PC-based powerplant control system that eliminates proprietary operating schemes and vendor-specific hardware. According to Emerson, Ovation reduces the risk of obsolescence often associated with proprietary control systems, and provides better and more accessible information than is available through the typical DCS.

For users of existing WDPF systems, an upgrade to Ovation can be accomplished via a complete redraw, or by using Emerson’s “migration tool” which can simplify and speed up the upgrade. The 36-MW San Gabriel Mill Plant, Pomona, Calif, for example, completed its conversion from WDPF to Ovation during a shutdown lasting only one weekend.

The automated migration tool enables owners of WDPF systems to upgrade to the Ovation system with minimal re-engineering and system retuning (Fig 2). The tool quickly converts WDPF graphics and logic codes to Ovation’s open architecture, avoiding the lengthy outages needed for the typical modernization project.

Migration to the Ovation system also can provide enhanced system functionality and process efficiency while reducing risks related to component obsolescence. If you are considering upgrading your WDPF system to Ovation, here are general answers to a couple of the questions you’re likely to ask:

Question 1: What are the pros and cons of the migration tool, versus a complete redraw? Answer: The migration tool converts the WDPF DPU source code directly into Control Builder Sheets and database, for import into Ovation. A one-to-one change-out of DPUs is required. Also, some upfront work may be required to get the DPU source code into a format that the migration tool will accept. The migration tool offers a well-proved migration path, and is a good option if documentation is lacking, or if schedule or budgetary constraints dictate the pace of the DCS upgrade.

By contrast, the “redraw” is exactly what its name implies. The point database is generated from the WDPF as-built termination lists and Q-Line I/O documentation, and the Control Builder Sheets are redrawn to match the as-built WDPF SAMA logic drawings (Fig 3).

Performing a complete redraw is more expensive because of the additional software engineering that’s required. Two DPUs—which are located side-by-side and perform the same function—can be combined into one controller both to help offset the higher overall cost of a redraw project and to improve control-system reliability. A redraw project also offers several advantages over a migration project, which should be considered in your evaluation. These include:

  • More readable SAMA logic. With the migration tool, logic is broken up into small bites in order to match WDPF execution order, and signal lines criss-cross the sheet in a visually confusing way.
  • Better control-system logic review. With the migration tool, a comprehensive design review is not performed.
  • Design modifications easily can be integrated as part of the upgrade process.
  • Simulation testing can be performed.

To compare the costs of a migration versus a redraw, users may request separate quotes from Emerson— one for a straight migration and the other for a re-draw. An additional quote from a third-party DCS consultant may be required for the as-built documentation and/or simulation development.

Question 2: Which operating system should I choose—Solaris or Windows? Answer: Solaris users often are familiar with Unix and concerned about network security, while Windows users want easy connectivity to the plant local area network (LAN) and inexpensive, off-the-shelf computer parts. Other benefits of Solaris include easy recovery from hard-drive failures—with the “boot net–install” option—and X-Windowing capability—via Hummingbird Exceed. On the other hand, Windows has Expanded Engineering Toolbox functions, which are excellent.

Ovation redraw project

A typical WDPF-to-Ovation redraw project includes the following elements:

  • Plant walk-down and drawing collection. A thorough review of the DCS process graphics is performed, in conjunction with a plant walk-down, to gain familiarity with the plant systems. Master red-line drawings (SAMA, P&IDs, electrical schematics) are collected. Also, the WDPF software server configuration, along with custom trend groups and any third-party software licenses—such as for Applix, optical drives, and printers— is documented and backed up for future reference.
  • As-built SAMA logic drawings are the starting point for the software design. Logic drawings are extracted from DPU source code.
  • As-built termination lists and QLine Card slot assignments and addressing documentation will be generated and validated. This information is entered into the Ovation DBID Tool (Database Initial Definition) to generate the initial Ovation point database.
  • Network layout/Ovation system architecture design. Fiberoptic runs between buildings and turbine enclosures provide isolation from electromagnetic interference and eliminate grounding issues. The 10-BaseT connections make for flexible system architecture. The number and physical locations of fiber runs and network switches must be determined.
  • Ovation hardware order. Based on the WDPF system hardware, software design requirements, and Emerson’s recommended network layout, the Ovation system hardware order can be placed. The Q-Line I/O hardware need not be replaced, only the human-machine interfaces and the DPU multi-bus chassis and power supplies.
  • Spare parts are more economical if complete controller upgrade kits are specified, rather than individual parts. Also, a simulator using Ovation hardware could double as a “hot-spare” parts bin.
  • WEStation graphics. Newer WDPF systems with WEStation graphics require no additional graphics work. Older WDPF systems with “classic graphics” require a redraw of the process graphics.
  • Simulator design. A simulator model will be developed in order to debug the Ovation software.
  • Final design review. Based on the interviews with O&M personnel, review of historical data, and logic review, control-logic modifications or additions to improve plant reliability, efficiency, or emissions are reviewed and included in the software design. Any additions or modifications to the logic may need to be reflected in the I/O termination lists and process graphics as well.
  • Software submittal to Emerson. All documentation, along with the Software Server backup tape, will be submitted to Emerson and used to implement the Ovation software.
  • Software validation testing. A total plant simulator model facilitates debugging of the software design and implementation. The “loopback stimulator” reads the outputs from the control system, simulates the process response, and writes the inputs to the control system. A detailed test procedure is an invaluable tool that helps maximize simulation time and ensures that nothing is missed. While the purpose of the simulation test is to validate the software design, maintaining the simulator at the plant for operator training and as an engineering tool may be an attractive option. This could be done with “virtual controllers” or using actual Ovation hardware.

Installation, commissioning

For both redraw and migration projects, the installation of Ovation’s fiberoptic network can be completed prior to the plant outage in which the total system installation occurs. After all systems are locked out and tagged out, the DPU processors, power supplies, WEStations, and WDPF Highway Interface Boxes are removed and the Ovation controllers and WEStations are installed and downloaded. With proper planning, a WDPF-to- Ovation upgrade can be completed during a normal gas-turbine combustor inspection outage—even on a large 2 × 1 combined-cycle plant.

Commissioning activities are subject to many factors—such as price of fuel and dispatch requirements. Some “hot” commissioning activities include: pre-start simulations, false fire, first sync, load-ramp tuning, and load rejections. Documentation and closeout of the upgrade project includes point database backup, software server backup to tape, and printing the as-built control builder logic. ccj oh