Martin Flatley
Sunoco Inc.
Sarnia, Ont.
The Sunoco Sarnia, Ont., refinery has a crude-distillation capacity of 80,000 b/d. A new asset-management software system allows information from intelligent devices to be seen on a personal computer in the maintenance shop as well as on a screen in the control room for fast troubleshooting (Fig. 1).
A relatively inexpensive asset-management software package has produced immediate benefits for the Sunoco refinery in Sarnia, Ont. (Fig. 1).
The package has sped up the diagnosis of production problems, reduced maintenance costs, prevented process upsets, and eliminated the need for maintenance personnel to enter hazardous areas.
Sunoco needed to reduce instrument maintenance costs by 25% over a 3-year period. One alternative was to cut personnel-a move that could eventually cost money in equipment deterioration and downtime.
The other option was to solve the problem with advanced technology. Sunoco chose Fisher-Rosemount's newly developed Asset Management Solutions (AMS) to achieve the projected savings.
A few years ago, the refinery began installing "smart" digital instrumentation to replace existing standard analog (4-20 ma) transmitters, especially in critical process areas.
In addition to performing the normal input/output functions of the distributed control systems, these instruments can perform self-diagnostics, store information on their status and condition, and communicate when queried. The large amounts of useful information that these instruments generated, however, were not being fully utilized until the asset-management program was implemented.
Product application
In mid-1997, Fisher-Rosemount's newly developed AMS software was beta tested in the refinery to evaluate its ability to access information from smart field devices, organize the data, and present it in a useful manner.Sunoco was pleased with the results of the 60-day test. The product started up easily and was configured easily. Most importantly, it supplied the information that was promised.
AMS allowed the operators to receive information from pH meters in real time. Technicians no longer had to go to the field to check pH cells.
For these reasons, Sunoco decided to install the software in a file server in one of its three refinery control rooms. General illustrations of the hardware configuration are shown in Fig. 2 [61,109 bytes] and Fig. 3 [81,715 bytes]. The final project came in on time and on budget.
Since its original installation, AMS has been used primarily as a diagnostic aid, providing information for the instrumentation department to evaluate the condition of field equipment and determine what repairs, if any, are needed, and how soon they should be made.
For example, two valves handling hydrofluoric acid in an alkylation unit have a history of sticking. Because of the presence of the acid, personnel must wear heavy rubber suits when entering the unit to check the valves or perform maintenance on them. The same is true of checking and maintaining pH analyzers in that area.
AMS enables the operators to remotely check the packing friction and other operating parameters of these valves. Operators can frequently obtain important information about the valves' conditions without sending personnel into a potentially hazardous atmosphere.
The result is a saving of maintenance dollars by avoiding time-consuming trips into the acid unit by technicians.
In addition to using the system as a diagnostic aid, Sunoco can use the system to collect historical data. The Audit Trail option provides an historic record of maintenance activity with connected field devices.
For example, configuration and calibration information is recorded for each device. If there is an alarm, the Audit Trail provides a transmitter status record that engineers can use to reconstruct what happened in order to determine the cause.
Economics
While a thorough financial evaluation has not yet been done, it appears that savings of 20-30% are being achieved vs. the cost of maintaining problem transmitters. As AMS is more fully implemented, Sunoco expects the savings to reach 40-50%.The software is currently being used to configure new smart field devices as they are wired into the system. In the future, the software will also be used to calibrate those units. Other process industry companies have reported saving up to $100/device in configuration costs and $50/device per year in calibration costs.1 2 In addition, these activities are automatically documented for future reference.
Asset management also saves the refinery money by helping to avoid upsets caused by sudden cold snaps, which usually occur in the fall season.
Steam tracing is used during the winter to keep lines from freezing, but unexpected hard freezes occasionally occur before the steam is turned on. The resulting upset can be costly, especially if the transmitter is for a reactor pressure or some other critical operating parameter.
If control room operators do not get accurate information from the transmitters, the unit may be in an upset condition by the time the problem is resolved. The last time this happened, it was an expensive situation.
By monitoring temperatures of exposed field device bodies, Sunoco knows when those lines become dangerously cold. When this occurs, operators check the steam tracing to make sure it is on. This protection alone is worth the asset-management system, which cost more than $55,000. The cost includes wiring the smart devices to Hart (highway addressable remote transducer) interface units, the computer, software, and associated equipment.
Currently, there are about 50 smart transmitters wired into the AMS system. Worn out transmitters throughout the 700 I/O (input/output) control system are being replaced with smart pressure and temperature transmitters, radar-level gauges, mass-flow meters, NIR (near infrared) analyzers, and instruments compatible with the Fieldbus communications protocol. Smart positioners will be installed selectively on valves in critical operations.
The gradual replacement of failed transmitters with smart transmitters and the use of information generated by these devices could save at least 20% of total maintenance dollars annually, enabling Sunoco to reach its cost reduction goal.
Future plans
The results have been favorable, both in problem identification and acceptance of the technology by the maintenance group. Consequently, use of the software is being expanded. The asset-management information will soon be available on Sunoco's plant wide network, permitting direct access by technicians in the instrument maintenance shop as well as by the instrument engineering and design groups.Eventually, asset-management information will be integrated with Sunoco's computerized maintenance-management system (CMMS) to assist in maintenance planning and scheduling.
This will provide an online trigger based on monitoring the status of field equipment and device alerts when a piece of equipment begins to exhibit signs of stress. This accurate, early warning can be used to initiate work orders and spare parts. Maintenance personnel can then be directed to precisely the right piece of equipment to handle a known problem without unnecessary expense.
While the initial application of AMS has been promising, Sunoco does not expect to realize all the potential benefits until there is a greater population of field instrumentation connected to the system. Based on the early results, however, the validity of using information generated by smart field devices to reduce maintenance costs has been proven.
References
- Giovannelli, Steve, "Effective Asset Management Begins with Accurate Field-based Data," InTech, May 1998, pp. 92-96.
- Giovannelli, Steve, "Controlling Maintenance Costs," Chemical Processing, May 1998, pp. 55-58.
The Author
Martin Flatley is senior instrument engineer at the Sunoco refinery in Sarnia, Ont. Prior to this position, he was an instrument technician and technologist for about 10 years. Before joining Sunoco, he was an instrument technician at other companies. Flatley holds bachelors and masters degrees in electrical engineering from the University of Western Ontario, London, Ont. He is currently working on a PhD at the same university.
Copyright 1998 Oil & Gas Journal. All Rights Reserved.