Examining oil analysis programs – both off-site and onsite programs across various industries- it is observed frequently that oil analysis abnormal alarm observations and recommendations are not being acted upon promptly.

The ability of reliability teams to ignore the results is primarily due to a lack of trust in the triggered alarms. This behavior is known as Alarm fatigue. First identified in the medical community, Alarm fatigue is a sensory overload condition where technicians are exposed to excessive alarms, which can result in desensitization and missed alarms.

An immediate issue is the identification of the systems needing the most attention. If everything is in alarm, how do you prioritize? In facilities where workers multitask and already have significant task backlogs, ignoring or failing to respond appropriately to such warnings can lead to unplanned downtime.

This article outlines why the condition occurs and a process to intelligently set alarms using new TruVu 360 Fluid Intelligence Platform features for onsite MiniLab and FieldLab analysis.

Understanding Alarm Settings for Oil Analysis

Oil analysis alarms may be set in several ways: rules of thumb, rate of change, or OEM / Lubricant supplier guidance. However, when starting up a new program, most reliability engineers opt to focus on OEM limits. OEM guidance is focused on best practices, usually on optimum equipment design duty cycles.

The challenge arises when the equipment duty cycle, lubricant type, or ability to change results from lack of repair ability (e.g., no filtration systems) leads to “normal” conditions with oil analysis measurements considered abnormal with these OEM values.

False alarms can be very problematic. False positive alarms may be triggered when severe level alarms are set too low. A conservative approach to setting alarms at the start of a condition monitoring program (e.g., OEM limits or generic industry values) can create false positive alarms that, when presented to equipment operators, lead to a lack of confidence in the capability of the oil monitoring program.

Today, oil analysis programs have more data available than ever to inform lubrication and reliability strategies. An effective lubrication program can return a 40:1 return on investment; however, this is only possible when oil analysis programs are trusted, and more specifically, the alarms and diagnostics are appropriate for the facility and machine being monitored.

Keeping Control

In any organization, the pareto based principle is assumed in a controlled facility. For oil analysis programs, the case is similar, whereby a majority of equipment should be in “normal” operating mode, and a percentage (typically around 20%) is what requires attention. Usually, sample results are color-coded yellow for caution and red for severe so the owner can manage the assets that need the most attention. This has a practical benefit in any facility where assets are working and resources are limited.

When a facility begins to have a significant amount of alarms greater than 20%, it becomes challenging to manage, and this is where alarm fatigue develops. The reliability engineer has more assets that need attention than resources available. So now what?

Figure 4. An oil analysis program where alarms are not set for practicality

Looking at the example in Figure 4, it is nearly impossible for a reliability engineer to manage a program when ~ 84% of the equipment is in severe alarm. If addressed at all, these alarms will likely be ignored until the next sampling.

Determining Better Alarms Using Historical Data

Determining better alarms and avoiding alarm fatigue can be achieved by performing a periodic statistical analysis, evaluating and adjusting alarms based on ASTM D7720-21 Standard Guide for Statistically Evaluating Measurand Alarm Limits when Using Oil Analysis to Monitor Equipment and Oil for Fitness and Contamination. ASTM D7720-21 defines statistical techniques for evaluating whether alarm limits are appropriate for flagging problems requiring immediate action.

The methodology and mathematics can be developed with Excel or other software; however, it involves exporting data and some effort to generate information. A new capability called Condition Based Alarms (CBA) based on ASTM D7720 is now available in TruVu 360, the fluid intelligence platform for onsite MiniLab and FieldLab systems at Ametek Spectro Scientific

The statistical evaluation using D7720 can be applied to all existing components within TruVu 360. A historical sample set of at least 80 samples within a component is required to complete the calculation. Using the guidelines outlined in the ASTM D7720, TruVu 360’s CBA feature produces an output that helps the user:

• Evaluate current alarm limits in each limit set to determine if current alarm limits are effective.
• Adjust alarm limits per limit set based on historical data (>80 samples needed).
• Evaluate the effectiveness of current alarms vs recommended alarm limits.
• Carefully and systematically develop an alarm strategy that is achievable and sustainable.

Using Condition-Based Alarms to Manage Steam Turbines in a Refinery

Steam turbines are widely used to drive process trains in refineries. It’s not unusual to find older turbines that can tolerate high levels of water and particulate. In many cases, these same turbines are assigned alarm profiles with strict limits for particle counting and water. Routine monitoring can result in a series of false positives that maintenance personnel ignore.

A recent review of an oil analysis program at a Gulf Coast refinery showed that many of its steam turbines (93 of the 83 assets) were classified as abnormal or severe. Furthermore, the TriVector distribution indicates that 70% of the abnormal or severe alarms are a result of contamination and wear present in the oil.

Table 1 shows an example of evaluating the existing alarm set vs. the proposed new alarm set to help understand how the CBA statistical analysis of alarm limits is used within TruVu360.

Table 1. Statistical analysis (per D7720) of refinery oil sample data from steam turbine.

Particle Contamination

Most reports for the steam turbines were alarmed due to the high particle count. CBA analysis confirms this and provides a new suggested alarm limit that, if implemented, will only alarm the top 3% of samples.

The graph in Figure 6 presents CDF (Cumulative distribution function) plot of ISO 4406 Code >6 measurements. The steam turbine’s current severe limit for ISO Code >6 is 15, and the suggested recommendation based on the CBA analysis severe limit is 27.

With the CBA approach, it is easier to prioritize the effort based on the worst samples in the refinery first, which are identified by the top 3%.

It will take a consistent long-term contamination control program initiative to achieve the refinery’s current ISO 4406 Code limits. But, with step-level initiatives and [i]focusing attention on a portion of the assets to demonstrate success, The expected payback will be improved machinery health, asset life extension, overall reliability improvement, and substantial cost avoidance. [ii]

Mechanical Wear Parameters and Alarm Limits

The above example is easy to understand, as there is widespread awareness and publications of ISO codes for contamination control, and most reliability engineers can reset values based on their comfort zone.

However, many powerful wear debris analysis parameters have been introduced in the last ten years for condition monitoring that do not have much OEM detail and experience. Figures 7 and 8 present a distribution of refinery steam turbine measurements using two wear parameters: Ferrous Wear Severity Index (FSWI) and Total Ferrous. Both measure severe ferrous wear.

The steam turbine’s current severe limit for FSWI is 4, and the recommended severe limit is 76. The steam turbine’s current severe limit for total ferrous is 20, and the recommended severe limit is 42. Samples exceeding severe alarm limits for FSWI and Total Ferrous must be explained and never ignored. Consider increasing these severe wear limits from 85th to 97th percentiles to prioritize limited resources and focus on the most serious wear indications.

 Alarm fatigue can prevent lubrication programs from making small improvements over time. However, the concept of condition-based alarms can help users avoid this. By carefully and systematically using data to drive alarm limits, the user can set more effective and practical alarms that can sustain maintenance efforts year after year.

For more information, please visit https://www.spectrosci.com/product/truvu-360.

References

[i] “Challenges and Solutions in Implementing a World-Class Lubrication Program,” Machinery Lubrication

[ii] Annual Cost Savings From Effective Lubrication Programs With Onsite Oil Analysis, Spectro Scientific, https://blog.spectrosci.com/annual-cost-savings-from-effective-onsite-oil-analysis

For years, rig operators have used oil analysis as an essential tool for routine maintenance and future cost avoidance. On oil rigs, equipment failures risk employee safety, and missed production targets are realized very quickly. The remote locations of offshore rigs make routine maintenance very expensive.

In most cases, skilled people and supplies can only reach the platform by ship or helicopter, so the cost of bringing technical specialists, replacement equipment, spare parts, and tools is high.

Oil analysis is critical in alerting the maintenance team to problems that may damage a vital system. An effective oil analysis program also helps to allocate scarce resources efficiently by planning maintenance based on equipment conditions rather than time intervals.

Currently, most offshore platforms take oil samples and ship them by helicopter to onshore labs for analysis. Once the results are available, they are sent back to the platform.

It costs more than $1 billion to operate a typical production platform over its 10 to 20-year life cycle, so operating costs per day can be estimated at $100,000 to $300,000.

A typical offshore platform contains millions of dollars of machinery critical to the crew’s safety and whose failure can easily put the platform out of operation. Oil analysis determines the amounts of various metals in the oil, providing a fast and inexpensive way to gauge the machinery’s wear.

Oil analysis also helps determine the condition of the oil by measuring byproducts formed by oxidation and by measuring the viscosity. Monitoring oil condition reduces the risk of catastrophic failure and the high cost of changing and disposing of oil in heavy machinery.

Typical equipment sampled on an offshore rig:

• Turbo gas-powered generators
• Fire water pumps
• Diesel engines
• Gearboxes
• Pumps
• Crane engines
• Hydraulic systems

Value Of Oil Analysis on Offshore Platforms

Recently, an offshore rig in the Gulf of Mexico contacted a reliability service provider in Louisiana to seek a solution to improve their oil analysis program. The long turnaround time from their onshore lab compounded the weaknesses of their current program.

Key Issues to Address:

• Frequent resamples due to mislabeling. Collected samples from the various rotating equipment on the platform were misidentified, leading to erroneous results by the contract lab. Policy demands action to CAUTION or ALERT status samples, so long delays caused operators to discard results.
• Missed Compressor failure A genuine problem detected by onshore oil analysis on a failing air compressor was received one month after the compressor had failed without warning. The rig operator then incurred extra costs replacing it. It can cost $250K for expedited freight and installation charges to replace a critical system ASAP.
• Resources for onsite oil analysis. The rig operator had submitted their operating budget for 24 months to the oilfield license holder. They could not dedicate rig personnel to operating onsite oil analysis equipment in addition to sampling.

The Solution: FieldLab C

The oil producer asked the reliability service provider to travel to its platforms with a portable oil analyzer that could provide comprehensive analysis, deliver immediate results, and generate a full report with recommendations throughout the shift.

The service provider proposed sending a Level III Vibration Analyst/Tribologist Service Engineer, who had all the safety certifications for transport and was familiar with rotating equipment and maintenance strategies on board the rig.

The major challenge was to select a portable, solvent-free oil testing device that could measure abnormal metals, viscosity, oil chemistry, and particles with the ability to provide a report. The solution chosen was the FieldLab system.

When deploying field solutions, careful consideration must be given to turnaround time, technique, and data integrity. The FieldLab onsite oil analysis system provides fast results with the data integrity of a commercial oil analysis lab.

The FieldLab is a rugged, portable expeditionary fluid analysis system that allows operators in the field to perform comprehensive, mobile lubricant sampling. The battery-powered device enables complete lubricant assessment for condition monitoring and rapid results that permit informed maintenance decisions.

It requires only a small sample volume to perform a comprehensive analysis, and no chemicals or solvents are needed for sample preparation or cleaning. With all four tests, the complete analysis only takes about 5 – 7 minutes.

FieldLab is a unique system design where machine failure and root cause analysis are interpreted using a comprehensive process combining a modified pore blockage particle count technique with an XRF Analyzer.

This unique system design allows for a broader range of particle size detection, allowing the user to identify changing wear rates and isolate potential root causes of problems in lube systems. Additionally, viscosity and oil chemistry can be run simultaneously.

Initial Examples

There have already been several cases where the savings from onsite oil analysis exceeded the entire year’s service cost.

For example, technicians on one platform replaced the diesel engine on a crane. When the technician visited the platform and tested the oil, the viscosity was 70 cSt when it should have been 120 cSt. The technician ran additional tests and discovered the presence of diesel fuel in the oil.

A mechanic put dye in the fuel supply and found a broken injector line was leaking diesel fuel into the oil sump. This leak could damage the engine or cause a fire. The oil analysis results made it possible to fix the problem with a low-cost solution – replacement of the injector line.

In another case, the oil analysis results on a large gas turbine compressor showed a high particle count. The technician queried the platform’s maintenance team and discovered they had recently replaced a valve in the lube oil system.

The technician wondered whether the oil particle count might have spiked in response to this maintenance, so he flushed the lube system and ran another test. This time, the test showed a much lower particle count, although still above normal values.

After discussing the situation with the maintenance foreman, the decision was made to do nothing immediately but to retest the equipment the following month. The particle count had returned to normal levels when the equipment was retested.

According to the maintenance foreman, if an onshore lab had tested the oil, there would have been no chance to do an immediate follow-up study. It would have been necessary to, at the minimum, perform vibration testing and possibly perform even more expensive repairs.

Data Management and Reporting

The results of offshore oil analysis can be uploaded to the TruVu 360 Fluid Intelligence System. The results are available not only to the maintenance team on the platform but also to onshore managers and analysts who track trends and provide recommendations on whether to invest in a particular piece of equipment.

The oil analysis method used on most offshore platforms takes up to a month to send samples to a lab and receive the results.

Onsite oil analysis has the potential to provide significant value and benefits by allowing testing on the rig and delivering immediate answers to the maintenance team. Faster results can prevent breakdowns and avoid unnecessary maintenance.

The savings are realized by getting answers quickly and better managing equipment uptime on the offshore rig.