It is no secret that each year petroleum refiners are spending less to process a barrel of crude oil into marketable fuel. Corporations realize they have to go beyond working harder to stay competitive. Forward-thinking companies continuously find more cost effective ways to streamline their processes by leveraging maintenance technologies to improve the bottom line. It is a survival of the fittest for the refining industry—fail to perform optimally over a substantial period and chances are you will end up as the slowest in the herd, eventually falling out of the pack, or worse, a more efficient organization consumes you.
Progressive organizations are using mobile, wireless workflow technology to streamline their process and maximize equipment reliability to stay ahead of the competition. This article will present a review of the reliability goals (business drivers) and the results of implementing an operator driven reliability (ODR) process using mobile, wireless workflow technology at a Gulf Coast refinery.
The first step
Valero Energy Corp. purchased a Gulf Coast, lower-quartile performing refinery in 1997 from a crude oil trading company. After an infusion of refining management, a reliability process was initiated using consultants to facilitate teams that were formed to yield reliability and quartile ranking improvements. Although several teams were successful, two became paramount: the reliability measurements and the ODR teams. Reliability measurements made sense and became a way of life--you cannot manage what you cannot measure. The metrics for reliability measurements were completed and implemented in early 1999, while ODR blossomed in early 2000.
Both teams identified a need to streamline operator rounds and obtain technology to facilitate the review of all data collected. These needs became key requirements for the ODR process. It was essential to develop trending capability of field-collected data to identify equipment in early phases of failure. If problems were detected early, the equipment could be halted and corrective action taken to minimize the cost to repair.
Profits can be increased by producing more or spending less. Management hoped the process of increasing reliability without adding personnel or workflow would result in less spending. To gain data consistency through the implementation of a best practice, the team identified the need to eliminate paper and implement electronic devices where data would be entered and seamlessly transferred to a system that was available for analysis by interested parties. The goal was for operators to accept ownership of their equipment and to accept a new system, one that enforced a best practice workflow process.
After evaluating available technologies, IntelaTrac from SAT Corp., Houston, TX, was chosen. IntelaTrac is industry workflow automation software used on rugged mobile handheld devices. The software aligned with management's vision for a tool that crosses all disciplines and collects data for analysis—from operations to maintenance, environmental, and turnarounds. As a bonus, the software leveraged existing legacy software systems including the Aspen process historian and SAP.
While in front of equipment, operators can take immediate, preventive actions on their routine rounds. Asset status can be documented with bar code technology or radio identification tags. Audit ability can also be preserved and the data updates into back-end systems for additional analysis in other departments.
During the ODR implementation, a review was performed of what field data should be collected. Input from several departments ensured that all collected field data would be of value and analyzed. The golden rule was, "if data is not going to be analyzed then it will not be collected." The final result provided key visual, vibration, and temperature readings that provide early detection of equipment degradation, and include several preventive maintenance activities that previously were performed by maintenance personnel. This is beneficial to the maintenance craft as they have more time to perform craft skills activities.
Results in phases
The first phase was a pilot installation of the software and the mobile technology in three process units within the complex. Several specific assets in the early stages of failure were detected and failure was prevented. The return on investment (ROI) for the pilot from that failure prevention indicated a three-month payback. Satisfied with the results, the site rolled out ODR to the remainder of the complex. Similar results were observed and a two-month ROI was determined.
Over a 12-month period, the early detection of equipment failures was recorded and a dollar value assigned. To define the value, historical information from SAP provided the average cost of the specific equipment repair that was used as a baseline against any items identified using the software. The cost to repair equipment identified through early detection was then subtracted from the historical baseline average. Items found included worn bearings and seals, equipment out of alignment, unexpected process conditions (plugged strainers or seal pots), and coupling adjustments. Each item found to have a problem was taken out of service before having a catastrophic failure event. More than $558,000 in savings and maintenance avoidance was identified.
Measuring the result
To align with the reliability measurements effort, total work order costs were reviewed. Fig. 1 illustrates total monthly work order cost for the complex implementing ODR.
An equal period pre- and post-IntelaTrac implementation is illustrated. Total work order costs during the period support the findings in specific equipment saves.
Total work order cost was analyzed for the months prior to the ODR process and software implementation. Comparing the 44 months prior to ODR implementation to the 12 months that followed pointed out the obvious: the complex monthly work order costs had been reduced by more than $87,000 per month, a 33 percent improvement. With this result, a complete ODR site rollout was justified and is currently in progress.
Fig. 2 is a Crow/AMSAA reliability growth diagram illustrating a positive step change in work order cost over time, and a prediction of the amount to be saved in avoided maintenance at a future referenced date.
Crow/AMSAA charts are the log-versus-log of a cumulative-sum graph. Initially used by J. T. Duane of General Electric to plot cumulative failures over time and later mathematically proven by Dr. L. Crow of the U.S. Army Materiel Systems Analysis Activity (AMSAA), the plot is a fairly accurate model and predictor of maintenance costs.
Each point in the graph represents maintenance costs for 1 mo. After a few points are plotted, a best-fit line can be drawn through them. The Beta, or slope of the line, shows if improvements are taking place. A slope greater than 1.0 indicates that reliability efforts are failing while a slope less than 1.0 indicates that reliability efforts are succeeding. A slope of 1.0 is simply treading water. Lambda is the Y intercept of the work order cost line. R2 is the mathematical fit of the work order cost line to the monthly cost points (the closer to 1.000, the more accurate the line is drawn, i.e., the better the slope is).
The figure illustrates the total work order cost each month for the complex implementing ODR. An equal period pre- and post-software implementation is illustrated. A best fit line was drawn for months prior to ODR and the slope calculated. A 1.24 slope was determined indicating that a step change, almost a paradigm shift, was warranted. A second best fit line was drawn for months after ODR implementation and the slope calculated. A 0.75 slope was determined indicating that a paradigm shift in maintenance avoidance had taken place; hence the $87,000 per month cost savings in the complex.
Could all of these savings be attributed to the implementation of the ODR process? Probably not. A refinery is a dynamic place with many influencing factors. During this period, operations management took a strong reliability stance, seasoned reliability engineers were assigned to the complexes, and a turnaround was conducted during the period. Maintenance activities likely took place during the outages that may not have been true turnaround items. Operations also took an equipment ownership position during this period that continues today. However, it is difficult to dispute the data. The step change only occurred in the complex that implemented the ODR process, including the installation of the software, and the use of wireless, handheld data collectors.
The trickle-down effect
Operators and other personnel have begun using the software and the wireless handhelds outside of the ODR realm.
Process safety management and occupational safety inspections, such as car seal, fire extinguisher, and hose station inspections, are being conducted. Volatile organic compounds information is being collected, stored, and transferred to the environmental department for regulatory requirements. Caustic concentrations are being reviewed and calculated in the field. Even a large expansion and turnaround conducted early in 2002 was followed and reported to corporate management using the handhelds and software. This was received with enthusiastic acceptance by the turnaround management team.
Staying ahead of the pack
Realizing the need to go beyond just working harder to stay competitive and trying to be the faster animal in the herd, investigation is underway to meld ODR into reliability centered maintenance and risk-based inspection initiatives to develop a more seamless reliability process at the site. The wireless handhelds are rapidly becoming the mobile laptop for many individuals who are required to complete various tasks within the site. It will be another year before the effort to implement ODR refinery-wide is fully analyzed, reviewed, and reported. MT
James R. Cesarini is manager of reliability and turnarounds, Valero Energy Corp., San Antonio, TX
Fig. 1. Total work order costs during the period support the findings in specific equipment saves.
Fig. 2. CROW/AMSAA reliability growth diagram of work order cost after implementation shows a positive step change over time, and a prediction of the amount to be saved in avoided maintenance at a future referenced date.