Claims:1. A method (200) for monitoring contaminant dilution in lubricating oil (14) used in a reciprocating engine (12), the method comprising:
generating (210) a tunable predictor model based on relationships between a plurality of oil integrity parameters of the lubricating oil (14) and a plurality of impact parameters having an impact on the plurality of oil integrity parameters, wherein at least one of the impact parameters is contaminant dilution;
receiving (220) a plurality of sensor measurements for the plurality of oil integrity parameters for the lubricating oil;
retuning (230) the tunable predictor model using time series data for at least one of the purality of oil integrity parameters and corresponding measured contaminant dilution values;
predicting (240) an amount of contaminant dilution for one or more contaminants in the lubricating oil using the tunable predictor model and the plurality of sensor measurements; and
generating (250) a control action for an optimal operation of the reciprocating engine based on the amount of contaminant dilution.
2. The method (200) of claim 1 wherein the one or more contaminant comprise at least one of fuel or water or combination of fuel and water.
3. The method (200) of claim 1 wherein the plurality of oil integrity parameters are selected from kinematic viscosity, dielectric constant, and temperature.
4. The method (200) of claim 1 wherein the plurality of sensor measurements for the plurality of oil integrity parameters are generated for the reciprocating engine under operation.
5. The method (200) of claim 1 wherein the retuning of the tunable predictor model is a periodic retuning.
6. A lubricating oil system (10) for a reciprocating engine (12), the lubricating oil system (10) comprising:
a lubricating oil circuit (28) configured for circulating lubricating oil (14) in the reciprocating engine (12);
a plurality of sensors (30) disposed along the lubricating oil circuit (28) configured for obtaining a plurality of sensor measurements corresponding to a plurality of oil integrity parameters of the lubricating oil (14);
a tunable predictor model module (32) comprising a tunable predictor model (60) derived based on relationships between a plurality of oil integrity parameters of the lubricating oil and a plurality of impact parameters having an impact on the oil integrity parameters, wherein at least one of the impact parameters is contaminant dilution;
a retuning model module (34) configured for retuning the tunable predictor model using time series data for at least one of the plurality of oil integrity parameters and corresponding measured contaminant dilution values;
a contaminant determinant module (36) configured for determining an amount of contaminant dilution for one or more contaminants in the lubricating oil (14) using the tunable predictor model (60) and the plurality of sensor measurements; and
a controller (40) configured for generating one or more control actions based on the contaminant dilution for optimal operation of the reciprocating engine.
7. The lubricating oil system (10) of claim 6 further comprising a communication module (38) configured for communicating the amount of contaminant dilution to the controller.
8. The lubricating oil system (10) of claim 6 wherein the one or more contaminants comprise at least one of fuel or water or combination of fuel and water.
9. The lubricating oil system (10) of claim 6 wherein the plurality of oil integrity parameters comprise kinematic viscosity, dielectric constant, and temperature.
10. The lubricating oil system (10) of claim 6 wherein the one or more control actions comprise generation of an alarm as an indication for oil change for the reciprocating engine (12) based on the amount of contaminant dilution.
, Description:BACKGROUND
[0001] This invention relates generally to lubricating oil contamination due to leakage of fuel and water into the lubricating oil in the reciprocating engines, and system and method for determining amount of contaminant dilution to improve engine (reciprocating engine) performance.
[0002] The lubrication system of a reciprocating engine used for example in different vehicles such as locomotives, automobiles etc., for providing a supply of lubricating oil to the various moving parts of the reciprocating engine. The main function of the lubricating oil system is to enable the formation of a film of oil between the moving parts, which reduces friction and wear. The lubricating oil is also used as a cleaner and in some reciprocating engines as a coolant.
[0003] Due to the severe operating load, speed, temperature and the introduction of contaminants into the lubrication oil system, there are chances that the system gradually lowers the lubrication oil properties and hence it becomes harmful for the engine operation. Therefore, analysis of lubricating oil is done periodically, as a part of engine maintenance, to identify and analyze the possible sources of contamination and monitor them from time to time.
[0004] Some of the common lubricating oil contamination reasons include leakage of fuel in the lubricating oil system, because of which there is a reduction in flashpoint, viscosity, and load carrying capacity of the lubricating oil. Another reason is leakage of water in the lubricating oil due to which there are chances of formation of emulsion and reduction of load carrying capacity of the lubricating oil.
[0005] Analysis of lubricating oil is typically done by sending a sample of the lubricating oil to a laboratory for analysis. The oil may be subjected to tests such as viscosity, fuel oil dilution, water, solid impurities, etc. Such analysis usually takes some days or weeks for completion, and the lubricating oil may deteriorate further in this time period, exposing greater risk for engine operation.
BRIEF DESCRIPTION
[0006] In one aspect, a method for monitoring contaminant dilution in lubricating oil used in a reciprocating engine is described. The method includes a step for generating a tunable predictor model derived from relationships between oil integrity parameters of the lubricating oil and different impact parameters having an impact on the oil integrity parameters, where at least one of the impact parameters is contaminant dilution. The method includes a step for receiving sensor measurements for the one or more oil integrity parameters for the lubricating oil. The method also includes a step for retuning the tunable predictor model using time series data for at least one of the oil integrity parameters and corresponding measured contaminant dilution values. Next, the method includes predicting an amount of contaminant dilution for one or more contaminants in the lubricating oil using the tunable predictor model and the sensor measurements; and generating a control action for an optimal operation of the reciprocating engine based on the amount of contaminant dilution.
[0007] In another aspect, a lubricating oil system for a reciprocating engine is described. The system includes a lubricating oil circuit configured for circulating lubricating oil for lubricating components of the reciprocating engine and sensors disposed along the lubricating oil circuit configured for obtaining sensor measurements corresponding to oil integrity parameters of the lubricating oil. The system includes a tunable predictor model module having a tunable predictor model derived from relationships between oil integrity parameters of the lubricating oil and different impact parameters having an impact on the oil integrity parameters, where at least one of the impact parameters is contaminant dilution. The system also includes a retuning model module configured for retuning the tunable predictor model using time series data for at least one of the oil integrity parameters and corresponding measured contaminant dilution values. The system further includes a contaminant determinant module configured for determining an amount of contaminant dilution for one or more contaminants in the lubricating oil using the tunable predictor model and the sensor measurements; and a controller configured for generating one or more control actions based on the contaminant dilution for optimal operation of the reciprocating engine.

DRAWINGS
[0008] These and other features, aspects, and advantages of the present invention will become better understood when the following detailed description is read with reference to the accompanying drawings in which like characters represent like parts throughout the drawings, wherein:
[0009] FIG. 1 is a diagrammatic representation of a lubricating oil system for a reciprocating engine;
[0010] FIG. 2 is a flowchart representation of select modules of the lubricating oil system of FIG. 1, for determining an amount of contaminant dilution in the lubricating oil system; and
[0011] FIG. 3 is a flowchart representation of a method for monitoring contaminant dilution in lubricating oil used in a reciprocating engine.
DETAILED DESCRIPTION
[0012] The foregoing description is directed to particular embodiments described herein is for the purpose of illustration and explanation. It will be apparent, however, to one skilled in the art that many modifications and changes to the embodiments set forth above are possible without departing from the scope and the spirit of the disclosure. It is intended that the following claims be interpreted to embrace all such modifications and changes.
[0013] As explained herein above, an accurate and timely analysis of the level of contaminants in the lubricating oil is one of the valuable preventive maintenance tools for the reciprocating engine(s) (also referred as “engine(s)” herein), that enables optimal control actions for engine operation, to avoid a major repair. The contaminants described herein include fuel, water, or combination of fuel and water, as non-limiting exemplary contaminants, however the aspects described herein may be applied to other contaminants as well.
[0014] It would be appreciated by those skilled in the art that the lubricating oil analysis can be advantageously used for increasing the engine's life, decreasing failure rate of engines, and reduce repair costs. The aspects described herein relate to using sensor measurements from the sensors deployed in the lubricating oil system, and correcting these sensor measurements, based on data parameterization approach, to derive percent contaminant dilution in the lubricating oil.
[0015] FIG. 1 is a diagrammatic representation of a lubricating oil system 10 for a reciprocating engine 12. Typically, the lubricating oil 14 for the engine 12 is stored in the bottom of a crankcase, known as “sump” 16, or in a drain tank located beneath the engine 12. The lubricating oil 14 is drawn from this tank through a strainer 18, one of a pair of pumps 20, into one of a pair of fine filters 22. The lubricating oil 14 is then passed through a cooler 24 before entering the engine 12 and being distributed to the various branch pipes to reach the different components of the engine 12, referred generally by distribution manifold 26. Pumps 20 and fine filters 22 are arranged in duplicate with one as standby. The fine filters 22 are arranged so that one can be cleaned while the other is operating. After use in the engine 12, the lubricating oil 14 drains back to the sump 16 or drain tank for re-use. The sump 16, strainer 18, pair of pumps 20, pair of fine filters 22, cooler 24, distribution manifold 26 form a lubricating oil circuit 28, for the flow of lubricating oil 14 for lubricating different components of the reciprocating engine. The different components may typically include different bearings, cylinder liners etc.
[0016] The lubricating oil system 10 also includes some sensors 30 disposed along the lubricating oil circuit for measuring different oil integrity parameters. The sensors 30 are typically rugged to withstand the temperatures and pressures during the operation of the reciprocating engine. In one example, some sensors are based on tuning fork resonance for measuring kinematic viscosity and density, some other sensors are capacitve element based for measuring dielectric constant, and still some other sensors are thermocouple based for measuring temperature. Oil pressure sensors also utilized on the engine inlet and outlet so as to measure the pressure drop across the engine
[0017] The lubricating oil system 10 further includes a tunable predictor model module 32 that incorporates a tunable predictor model (described in more detail herein below), a retuning model module 34 that incorporates functionalities to retune the tunable predictor model so that it represents an operating environment of the reciprocating engine, and a contaminant determinant module 36 for determining an amount of contaminant dilution for one or more contaminants using a predictor-corrector approach, that is described herein below in reference to FIG. 2 in more detail. The one or more contaminants referred herein include at least one of fuel or water or combination of fuel and water.
[0018] The lubricating oil system further includes a communication module 38 configured for communicating the amount of contaminant dilution to a controller 40. The controller 40 is configured for generating one or more control actions 42 based on the contaminant dilution for optimal operation of the reciprocating engine 12. The control actions include, for example, generation of an alarm as an indication for oil change for the reciprocating engines based on the amount of contaminant dilution. The alarm may be in a form of a visual display, an audio alarm, a text message, or other such modes of alarm.
[0019] It would be understood by those skilled in the art, that the different modules described herein above are configured using a processor 44 and a memory 46. The processor 44 may include at least one arithmetic logic unit, microprocessor, general purpose controller or other processor arrays to perform computations, and/or retrieve data stored on the memory. In one embodiment, the processor may be a multiple core processor. The processor processes data signals and may include various computing architectures including a complex instruction set computer (CISC) architecture, a reduced instruction set computer (RISC) architecture, or an architecture implementing a combination of instruction sets. In one embodiment, the processing capability of the processor may be limited to supporting the retrieval of data and transmission of data. In another embodiment, the processing capability of the processor may also perform more complex tasks, including various types of feature extraction, modulating, encoding, multiplexing, and the like. Other type of processors, operating systems, and physical configurations are also envisioned.
[0020] In one embodiment, the memory 46 described herein above may be a non-transitory storage medium. For example, the memory may be a dynamic random access memory (DRAM) device, a static random access memory (SRAM) device, flash memory or other memory devices. The memory may also include a non-volatile memory or similar permanent storage device, and media such as a hard disk drive, a floppy disk drive, a compact disc read only memory (CD-ROM) device, a digital versatile disc read only memory (DVD-ROM) device, a digital versatile disc random access memory (DVD-RAM) device, a digital versatile disc rewritable (DVD-RW) device, a flash memory device, or other non-volatile storage devices.
[0021] Now turning to FIG. 2, a flowchart representation 50 for showing flow of steps in the different modules, namely a tunable predictor model module, retuning module and contaminant determinant module referred in FIG. 1, is illustrated. A tunable predictor model shown as 60 is derived from a relationship between different oil integrity parameters of the lubricating oil and different impact parameters having an impact on the oil integrity parameters. The oil integrity parameters include, for example, kinematic viscosity (required for measuring percent fuel dilution in the lubricating oil), dielectric constant (required for measuring percent water dilution in the lubricating oil), and temperature, and the impact parameters include contaminant dilution like water dilution and fuel dilution.
[0022] As shown in FIG. 2, the tunable predictor model receives training data shown by step 70, which includes the measurements (from the sensors) of oil integrity parameters 80 and impact parameters 90 (measured values from a laboratory) for a given period of time (the time period may be for example 3 months, 6 months or other such time period that is user determined). The tunable predictor model 60 is derived using correlations between the oil integrity parameters and the impact parameters, based on the training data, and generates statistical prediction values for contaminant dilution corresponding to sensor measurements of oil integrity parameters. The correlations are established, using for example, but not limited to neural network techniques in one example, and regression based polynomial techniques, in another example.
[0023] For example, in one embodiment a neural network (NN) based technique is used for determining percent water dilution (% WaterNN), i.e. network output which uses the following equation:
..Equation (1)
where ? represents kinematic viscosity
? represnets dielectric constant
T represents temperature
X represents network input
F represents hidden layer neurons
V represents input layer synaptic weights
W represents output layer neuron weights
b1 is input bias
bo is output bias
[0024] Similarly, in one embodiment where regression based technique is used that uses the following equation for determining percent water dilution:
Equation 2
[0025] Similarly for percent fuel dilution using regression based technique, the following equation is used in one embodimentf
..Equation 3
where a0, a1 , a2 are regression coefficients, ? represents kinematic viscosity, T represents temperature.
[0026] Thereafter in one embodiment, physics based modelling is also used for deriving the tunable predictor model 60 in conjunction with regression/neural network based model, for example,a Maxwell-Garnet mixture model framework known in analytical modelling, is used to obtain physics based predictions for contaminant dilution. Physics based modelling may contain certain biases since environment of each reciprocating engine is different, these biases are corrected using statistical prediction values from the regression/neural network based model, to finally generate predictions for amount of contaminant dilution shown at step 120. After building the tunable predictor model using training data, the tunable predictor model is tested using test data from laboratory measurements for the contaminant dilution.
[0027] After deployment of the tunable predictor model, input data as shown at step 100 is received by the tunable predictor model in a pre-defined time interval, such as thirty seconds that includes sensor measurements for kinematic viscosity, dielectric constant and temperature. The tunable predictor model processes the input data referred herein above and shown at step 100 using the regression or neural network based and physics based modelling incorporated in the tunable predictor model 60, and is able to generate the amount of contaminant dilution at same periodicity as the pre-defined time interval, for example every thirty second, as shown at step 120. This provides a great advantage over existing techniques, where there is a time lag of several days in obtaining the amount of contaminant dilution which continues to detiorate the performance of the reciprocating engine in this time interval.
[0028] It would be appreciated by those skilled in the art that the tunable predictor model 60 described herein would require periodic retuning which is shown at step 110, and is advantageously used, so the tunable predictor model continues to correspond accurately to the operating environment of the reciprocating engine, which tends to change due to several factors including periodic control actions, operator interventions, and use environment. For retuning, a correction factor is obtained separately using the time series based correction methods for retuning the predictor model to ensure alignment of the predictor model with a field environment (also referred as “onboard” environment or “in-use” environment, or “under operation” environment or “during operation” environment, also the word “environment” may not be used every time for retaining the flow of description). In one example, the model retuning is done every week or five days. However, the periodicity may be user or operator determined, and other time intervals for retuning will be equally applicable.
[0029] In one example, for retuning, time series data for at least one of the oil integrity parameters and corresponding sensor measurements, for example for kinematic viscosity is received for the given time interval. For the same time interval, predicted contaminant values corresponding to the same oil integrity parameter are obtained from the tunable model. A comparison is done between the actual measurements for the contaminant that are derived in one example using Karl Fischer laboratory analysis, a standard analytical chemistrty analysis, and predicted values from the tunable predictor model, and a correction factor is obtained based on the comparison. The correction factor is applied on the regression based and physics based correlations, and the tunable predictor model is updated. Model update referred herein implies that the correlations in the regression based or neural network based model and the physics based model are updated. In one example the retuning is achieved using Bayesian filter techniques (this will be well understood by those skilled in the art, as a general probabilistic approach for estimating an unknown probability density function recursively over time using incoming measurements and a mathematical process model), such as Kalman filter technique or particle filter techniques.
[0030] It would be appreciated by those skilled in the art that using the regression based or neural network based modelling techniques allows representation of behavior of an in-use reciprocating engine, and the physics based modelling allows to include the physics behind the behavior of the in-use reciprocating engine, and any changes in an ideal operation of the in-use reciprocating engine due to the operating environment. Thus the merging of these two approaches provides a unique ability to emperically estmate amount of contaminant dilution such as percent water dilution or percent fuel dilution, removing inaccuracies that arise due to use of only one of the appraches, and provides faster response time to determine the amount of contaminant dilution, as compared to the existing techniques. Further the retuning aspect, further provides a capability to keep the tunable predictor model relevant to changing operating environment of the reciprocating engine, ensure reliability of the predicted values for the contaminant dilution over a long period of time.
[0031] Turning now to FIG. 3, a summary for a method for monitoring contaminant dilution (fuel, water or a combination of fuel and water) in lubricating oil for a reciprocating engine is shown in flowchart 200. The method includes a step 210 for generating a tunable predictor model based on relationship between oil integrity parameters of the lubricating oil and impact parameters having an impact on the oil integrity parameters, where at least one of the impact parameters is contaminant dilution. The method further inlcudes a step 220 for receiving sensor measurements for the oil integrity parameters for the lubricating oil. The method inlcudes a step 230 for retuning the tunable predictor model using time series data for at least one of the oil integrity parameter and corresponding measured contaminant dilution values. The method then includes a step 240 for predicting an amount of contaminant dilution for the one or more contaminants in the lubricating oil using the tunable predictor model and the sensor measurements. As a final step 250, the method includes generating a control action for an optimal operation of the reciprocating engine based on the amount of contaminant dilution.
[0032] It would be appreciated by those skilled in the art that the system and system described herein above address the need for knowing more accurately, and timely, the percent dilution of the lubricating oil in the reciprocating engines for optimizing the engine life, thus allowing for an overall safer and more productive operation of the reciprocating engine, that is an important requirement for locomotives and other general purpose vehicles as well.
[0033] While only certain features of the invention have been illustrated and described herein, many modifications and changes will occur to those skilled in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention.