Oil analysis has become commonplace in many commercial and military organizations, and efforts have been put forth to bring traditional oil analysis procedures into the real time realm. Real time monitoring is a vital tool, which can allow lubricants to be used to their fullest potential while minimizing machinery downtime, resulting in increased savings and productivity.
The Applied Research Laboratory, or ARL, team in collaboration with its program partners established the following working definition of “Smart Equipment” for the purpose of a current program sponsored by the United States Department of Energy (U.S. DOE) under a Nuclear Energy Research Initiative (NERI):
“Smart equipment embodies elemental components (e.g., sensors, data transmission devices, computer hardware and software, MMI devices) that continuously monitor the state of health of the equipment in terms of failure modes and remaining useful life, to predict degradation and potential failure and inform end-users of the need for maintenance or system-level operational adjustments.”
Implementing smart equipment, as suggested by this definition, will require on-line sensor capability, signal processing, data fusion and potentially automated reasoning installed at the equipment level. Several challenges exist to direct implementation.
One impediment to introducing smart equipment to existing nuclear power plants is the difficulty associated with wiring. It is often impossible to obtain the required penetration to add a sensor to a system inside containment. Furthermore, the expense of routing and installing cabled systems is often prohibitive.
A solution to this problem is to follow the current trend in industry of placing the “smarts” at the sensor level and transmitting low bandwidth information to higher levels via a wireless communication. This approach is being developed currently for smart vibration sensors or nodes. The required accuracy and issues of self-calibration and self-diagnosis are also very important.
In many ways, lubrication systems are an ideal demonstration platform to evaluate such concepts with real time sensor development. Lubrication systems are endemic in applications that require mechanical equipment with relative motion between its parts. In this case, the lubrication system can be effectively modeled with a similar laboratory setup.
From a larger perspective, equipment maintenance personnel are increasingly using information on lubricant condition in their decisions concerning maintenance. In the past, oil monitoring has predominantly been used to determine when the oil should be changed; however, it is now known that oil monitoring can also provide important information on the condition of the machine itself. Oil analysis has proven to be an effective tool for determining failure modes for both equipment and the lubricant.
The foremost goal of an oil analysis system is the early detection of oil degradation, contamination and machinery wear. This early detection can bring about several important benefits. Improved safety, early detection and warning of machinery degradation can ensure a safer work environment.
Additionally, early detection and control of the causes of lubricant degradation, contamination and wear can substantially reduce the occurrence of damage to machinery. And, early detection can result in increased equipment availability or effectiveness.1
There are a variety of methods for determining the condition of lubricants, but most of these techniques fall under one of three main categories: quality, debris or elemental. A direct result of indirect measurement of an oil’s quality can be based on additive depletion, oxidation, thermal breakdown or other physical or chemical properties.
The goal of debris monitoring is to determine the presence, size and possible origin of both metallic and nonmetallic oil debris. Elemental monitoring uses precision equipment to determine the presence of foreign elements in the fluid system. It can also be used to measure the amount of desirable elements present (i.e., additives).
In addition to the nature of oil analysis techniques, one must be aware of the various positions in which oil monitoring can occur. Oil monitoring can occur in three possible positions: off-line, on-line or in-line (Figure 1). In off-line monitoring a portion of the oil is sampled and analyzed away from the machine.
A disadvantage of off-line monitoring is that it may be affected by a variety of influences during sampling, transport and/or testing. On-line monitoring is where a portion of the oil is sampled and analyzed by direct connection to the lubrication system. It has little impact on system flow, provides direct results and has little outside influence.
However, like off-line monitoring, on-line monitoring can be misrepresentative of the system if the portion sampled is small relative to the system flow. In-line monitoring is where all the oil that passes is analyzed giving immediate results, with no outside influence. In-line monitoring can be difficult to implement and can influence the system. Note that on-line or in-line monitoring are required for real time oil analysis.2
To realize the need for real time oil sensors, one must examine the deficiencies of off-line methods. Laboratory analysis takes time, and in the time it takes to process the sample, machinery can be damaged from poor lubricant quality. One author reported that “50% of off-line analyses find no problems, and only 5% detect serious problems, the remaining 45% of analysis show imminent failure.”3
This suggests that for many applications full time monitoring is imperative in determining lubricant faults in a timely manner. Additionally, one can never be sure that the oil sampled is representative of the entire lubricating system. Various sampling techniques are used in an attempt to acquire the best sample, but there are still possibilities that the sample collected is not the most representative of the system.
Also, when a sample is taken, it is difficult to ensure no outside contamination from the sampling procedure, container or laboratory has been introduced. Finally, off-line oil sampling and analysis can be costly.
Real time sensors provide the ability to conduct continuous monitoring. This is beneficial on many levels, especially in responding to suddenly occurring faults and condition trending.
Real time sensors enable the integration of diagnostic and prognostic maintenance systems. Moreover, real time sensors eliminate the burdensome cost of oil sampling and laboratory analysis.
Developments in real time sensors have divided into two paths: true real time sensors and near real time systems. True real time systems are placed either directly in the system flow, or in a rerouted flow branch.
They can be connected directly to monitoring systems to allow for continuous real time monitoring and diagnostics. Near real time sensors bring laboratory procedures onsite to allow for quicker response time.
Some near real time systems still require oil sampling, but the tests only take a matter of minutes to complete. Alternatively, some near real time systems can be connected to monitoring systems for lubricant diagnostics and prognostics.
True Real Time Sensors
True real time sensors continually monitor the state of lubricants. They take many shapes and use a variety of techniques to determine lubricant condition. It should be noted that since many of these sensors use different techniques to determine lubricant condition, they have differing capacities to detect specific lubricant faults.
Kavlico Oil Quality Sensor®
The Kavlico Oil Quality Sensor works on the principle that the dielectric constant of oil changes as it degrades or becomes contaminated. Dielectric constant is a measurement of a substance’s ability to resist the formation of an electric field within it.
As the quality of oil deteriorates, the sensor measures its dielectric constant and outputs this information in the form of a voltage that is correlated to the quality of the oil. Dielectric sensors have been shown to be adept at detecting the presence of water in lubricating oils due to the large difference in dielectric constant between water and oil.
In the development of dielectric sensors it was noted that the dielectric constant is dependent on temperature. Kavlico bypassed this issue by installing an automatic correction into their sensor.
The lubricant temperature is measured and the dielectric signal is compensated to make up for the temperature changes. The sensor can provide real time qualitative analysis of oil condition and some quantitative analysis of water content.4
Lubrigard® Dielectric Sensor
The Lubrigard Dielectric Sensor works on a similar principle as the Kavlico sensor. It uses high frequencies to measure the dielectric constant of the lubricant and provides real time qualitative analysis of oil quality.5
The sensor contains a built-in microchip, which interprets measurements and then relays information to outside systems. The Lubrigard sensor also contains a built-in temperature device to account for changes in operating temperature.
Oil Condition Monitor® by Foster-Miller
The Oil Condition Monitor (OCM) is a real time optical system for the measurement of oil condition acidity (TAN), water content, thermo-oxidative degradation, fuel/coolant dilution and antioxidant depletion. The sensor functions on the same principles as laboratory FTIR systems.
The OCM features a miniaturized infrared spectrometer that tracks pre-established wavelength regions. The regions of interest correspond to particular oil characteristics classified by JOAP. In testing, the OCM was able to provide real time data on degradation, contaminants, water, glycol and incorrect oil.
A standard ribbon cable and PCM CIA card connect the OCM to a PC where data acquisition software converts output voltages into absorption values associated with oil quality.6
GasTOPS MetalSCAN® Debris Monitor
The MetalSCAN unit is a through-flow sensor that is installed on lubricant lines upstream of the oil filter. The sensor uses a magnetic coil assembly to detect and categorize metallic particles by size and type (ferrous or nonferrous).
The minimum detectable particle size is determined by the bore size of the sensor. Currently, systems are designed to detect particles as small as 50 microns. The sensor consists of three coils surrounding the inside bore. Two coils create a magnetic field, and the third coil detects any disturbances in the field.
Depending on the type and magnitude of the disturbance, the control unit determines the type of particle and the particle size. The control unit also reports the total mass of ferrous material that has passed through the sensor. GasTOPS claims that the MetalSCAN unit has 100% detection efficiency. This allows users to track the debris progression over time and to trend the information to determine the current state of the lubrication system.7
Inductive Debris Monitor® by Smiths Industries
The Smiths Industries Inductive Debris Monitor (IDM) provides real time monitoring of lubricant debris. The IDM measures ferrous and nonferrous metallic debris in a similar manner as the GasTOPS MetalSCAN system. It is a non-intrusive system which threads directly in lubricant flow lines.
Temperature, flow rate, air entrainment or bubbles have no effect on its measurements. Smiths Industries says the IDM can detect particles with 100% efficiency. The IDM can detect ferrous particles as small as 50 microns depending on bore size. As with the MetalSCAN unit, the IDM allows the trending of wear particle production over time.8
Quantitative Debris Monitoring by Vickers Tedeco
Vickers Tedeco produces two quantitative debris monitoring systems. The first system is the Quantitative Debris Monitor (QDM). The QDM creates a magnetic flux field, which detects the presence of debris. When the debris disturbs the field, a voltage output is produced. This voltage is correlated to the mass of the debris.
The debris is also retained for further inspection. The other Vickers Tedeco system is the IQ Debris Monitor. The IQ system provides a continuous analog output proportional to the amount of debris accumulated and can sense a change in debris accumulated as small as 0.1 mg of magnetic material. As with the QDM, the IQ system retains the debris for further inspection. Additionally, these systems can interface with consumer software for real time monitoring.9
Electric Chip Detectors by Vickers Tedeco
Vickers Tedeco manufactures two types of electric chip detectors as well: the Electric Chip Detector and the Pulsed Electric Chip Detector. The Electric Chip Detector creates a magnetic field that attracts ferromagnetic debris particles. The debris bridges a gap between two electrodes, which act as a switch closure for an alarm output.
The device can be threaded directly into a lubrication system. Debris particles captured by the system can be removed for further inspection to determine wear mechanism and material type. The switch nature of the Electric Chip Detector does not allow for debris trending and can cause false alarms by detecting insignificant debris build-up.
The Pulsed Electric Chip detector attempts to alleviate this problem. A low energy current pulse clears fine debris away to prevent false alarms. The number of pulses the unit outputs can be programmed into its memory. The pulses clear away fine debris, but the larger debris particles are still detected. As with the conventional unit, the debris is retained for further inspection.10
Electromesh® Indicating Screen by Vickers Tedeco
The Electromesh screen provides a warning of ferrous and nonferrous conducting debris particles. Debris particles bridge the gap between strands to close a circuit. Oil flow passes through the screen with minimal pressure drop. The Electromesh screen is effective at detecting non-magnetic conducting particles i.e., aluminum, bronze, magnesium and Babbitt, but is only capable of detecting large particles in an on/off alarm fashion.11
In-Line X-Ray Fluorescence Spectroscopy (in-line XRF)
X-Ray fluorescence spectroscopy is now being developed as an in-line sensor for real time quantitative analysis of wear and additive metals. XRF systems have been extensively used in laboratories for elemental analysis. It was a significant step to try to adapt these systems for real time analysis. In-line XRF systems work on the same principles as their laboratory counterparts.
An x-ray source is imparted on the oil flow and a detector reads the x-ray emissions from the flow. The resulting data is analyzed and the elemental content is displayed. This system performs a quantitative detection of 12 elements in the parts per million range. EDAX PPD created the original portable XRF system. The In-line XRF system is being further developed at the Pacific Northwest National Laboratory (PNNL) for additional applications.12
Pall Water Sensor
The Pall Water Sensor allows users to obtain real time information on lubricant temperature and relative water saturation. The systems sensing element is a capacitive device that measures the capacitance change of the lubricant with changes in water saturation. The unit’s circuitry converts the capacitance into a linear, amplified voltage output.
The control unit interprets the voltage and determines the percent saturation. What makes the Pall different from other water content measuring devices is that it returns the water content information in percent saturation not parts per million (ppm). This is important due to the fact that that the acceptable water content measurement is not absolute.
A 200 ppm water content in one oil might be well within limits, while in another it might be catastrophic. Measurement of percent saturation overcomes this issue because a fluid’s saturation point is affected by many factors including temperature, aging, additive depletion, degradation and contamination.
An oil’s saturation point is the point when the maximum amount of water is absorbed and a phase separation takes place and free water is formed. This system can provide an accurate measure of an oil’s percent saturation of water over a variety of conditions and qualities.13
A large amount of time and money have been invested in advancing oil analysis into the real time realm. Great strides have been made to develop sensors that would enable complete real time monitoring of lubricant condition.
Although there are many challenges remaining before an all-encompassing real time oil analysis device exists. First of all, the in-line XRF and the Foster-Miller OCM need to be further tested to determine their true operating capabilities. Additionally, the near real time systems need to be advanced into true real time.
Finally, research needs to be conducted to establish the feasibility of combining several real time systems to provide an all-encompassing real time oil analysis module. A complete oil analysis system when combined with data fusion and automated reasoning could provide a huge boost forward for lubrication system conditioned-based maintenance.
The original work related to lubrication system sensing, modeling and test bench development was funded by Dr. Phillip Abraham from the Office of Naval Research on Grant Numbers N00014-96-1-0271 and N00014-97-1-1008. Subsequent effort in assessing real time methods as could be applied to an LSTB demonstration has been supported by the NERI Smart Equipment Contract.
1 Toms, Larry A., Machinery Oil Analysis: Methods, Automation and Benefits, 1995.
2 Hunt, Trevor M., Condition Monitoring of Mechanical and Hydraulic Plant, 1996.
3 Wilson, Bary W., et. al., Development of a Modular In-Situ Oil Analysis Prognostic System, International
Society of Logistics (SOLE) 1999 Symposium, Las Vegas, Nevada, August 30 - September 2, 1999.
4 Kavlico Oil Quality Sensor, http://www.kavlico.com/ksensor/oil_sensor.html
5 Lubrigard Sensor, http://www.lubrigard.com
6 On-Line Oil Condition Monitor, Foster-Miller Product Notice.
7 In-Line Oil Debris Monitor, Aerospace Engineering, October 1996.
8 Inductive Debris Monitor, http://www.smithsind-aerospace.com/prods/fvms/idm.html
9 Quantitative Debris Monitoring, http://www.tedeco.com/p_qdm.html
10 Chip Collectors and Detectors, http://www.tedeco.com/p_zappers.html
11 Electromesh Indicating Screen, http://www.tedeco.com/p_electromesh.html
12 Wilson, Bary W., Price, Stephen, In-Line X-Ray Fluorescence Spectroscopy, Lubrication & Fluid Power,
August 2000, pp. 16-19.
13 Are You Sensing Moisture Problems? Pall’s New Water Sensor Reports Dissolved Water Concentration,
Practicing Oil Analysis, July 1999.
14 Byington, Carl S., Lubrication System Test Bench for Gas Turbine Engine Diagnostics, ARL Technical
Memorandum, File No. 99-015, February 1999.