EP4733568A1

EP4733568A1 - Diagnostic method and device for a fuel supply system of an internal combustion engine methane fuelled

Info

Publication number: EP4733568A1
Application number: EP25209119.4A
Authority: EP
Inventors: Tiziano D'AMBROSIO; Mario CASTALDI
Original assignee: FPT Industrial SpA
Current assignee: FPT Industrial SpA
Priority date: 2024-10-25
Filing date: 2025-10-16
Publication date: 2026-04-29

Abstract

Computer-controlled diagnostic method for a natural gas fuel system of an internal combustion engine, wherein the fuel system comprises a rail, the method comprising the following sequential steps: acquisition of a succession of rail pressure measurements over a predetermined observation time interval (Total Engine Hours) while the engine is operating; comparison of each measurement with a pressure threshold (Th2); calculation of a first indicator (RI) representing a fraction of said predetermined observation time interval, indicative of "normal" fuel system operation, wherein said indicator is calculated based on all comparisons relating to each measurement in the sequence of measurements; calculation of a second indicator (IF) as a gradient of the first indicator over two distinct predetermined observation time intervals; calculation of a key performance indicator (KPI) as a linear combination of the first and second indicators, wherein said key performance indicator indicates a current condition and a forecast of a change from the current condition of the fuel system.

Description

Field of the invention

The present invention relates to the field of diagnostic methods and devices for internal combustion engine methane fuel systems.

State of the art

The automotive sector, focusing on natural gas (CNG), compressed natural gas (CNG), and liquefied natural gas (LNG), also known as automotive methane, is a growing industry that aims to provide more environmentally friendly and sustainable vehicles to address environmental challenges and greenhouse gas emissions.
The growing importance of natural gas vehicles in the automotive sector is due to several factors, including:

reduction of emissions,
diversification of energy sources,
the pursuit of lower operating costs.

In fact, natural gas vehicles offer significant advantages in terms of pollutant emissions compared to gasoline or diesel vehicles. Combustion of natural gas produces fewer CO2 and particulate emissions, thus helping to reduce air pollution and greenhouse gas emissions. Furthermore, natural gas can be produced from a variety of sources, including organic waste, natural gas, and biogas, reducing dependence on petroleum-based fuels.
Overall, natural gas-powered vehicles offer lower operating costs than conventional gasoline or diesel vehicles, as the cost of natural gas is often lower than that of other fuels. The automotive industry therefore continues to invest in natural gas propulsion technologies, seeking to offer more efficient engines and advanced fuel systems to improve the performance and reliability of natural gas vehicles.
Rail pressure is a key parameter, playing a crucial role in optimizing overall vehicle performance.
Optimal rail pressure is essential for optimal combustion efficiency, rapid throttle response, and reliable vehicle performance.
Rail pressure is critical in the fuel system of natural gas vehicles, as it is responsible for the proper delivery of fuel to the engine through the corresponding natural gas injectors, which are operationally connected to the rail.
A limited reduction in rail pressure leads to a reduction in engine efficiency, as the quantity of natural gas injected into the engine cylinders is insufficient for complete combustion. This results in a loss of power and reduced engine energy efficiency, as not all the methane is burned optimally to produce mechanical work.
Engine behavior may become erratic. This can manifest itself through knocking, stalling, or even engine shutdown, compromising the safety and reliability of the vehicle.
A significant reduction in rail pressure leads to malfunctions in the fuel supply system, with the risk of damaging pumps and injectors. This can cause system malfunctions such as leaks, blockages, or failures, compromising the proper operation of the engine.
Furthermore, an increase in pollutant emissions may occur. Since methane is a clean fuel, it is important to ensure efficient combustion to minimize emissions of harmful substances such as nitrogen oxides (NOx) and unburned hydrocarbons (HC), which are normally generated under lean burn conditions.
Lean burn also causes an increase in cylinder temperatures, with the risk of damage to valves and seats.
Clearly, to address these issues, it's essential to regularly monitor fuel pressure in the fuel system and perform preventative maintenance to ensure all fuel system components are in optimal condition.
There are several known methods for monitoring rail pressure in vehicles, but the most common involves monitoring rail pressure using pressure sensors directly on the fuel rail.
Refueling under critical conditions, i.e., with cold CNG and low pressure, can be the main cause of rail pressure loss.
In detail, the link between refueling with cold CNG and rail pressure loss can be complex and depend on several factors, including an increase in the viscosity of the CNG and a contraction in volume with a general reduction in pressure. Therefore, simply monitoring rail pressure is not meaningful.
Unless specifically excluded in the detailed description below, the information provided in this chapter is to be considered an integral part of the detailed description.

Summary of the invention

The purpose of this invention is to propose a method for assessing the wear status of a natural gas fuel system.
The basic idea of this invention is to acquire pressure measurements in the rail of the natural gas fuel system and to process these measurements to obtain a performance indicator indicative of the wear status of the fuel system of a vehicle engine.
Specifically, the idea is to acquire said pressure measurements and obtain a frequency distribution based on predefined threshold values.
The pressure measurements are acquired for observation intervals while the engine is operating, i.e., while it is consuming natural gas.
A first indicator is then calculated that represents the time period in which the fuel system is in critical operating condition, starting from a starting instant of observation, for example, engine start, up to the current instant.
Preferably, the time period is averaged over a longer observation period, for example, over a week or a month.
This indicator allows for the prediction of a vehicle's failure condition, allowing for timely planning of vehicle maintenance.
Furthermore, a second indicator is calculated, representing the first derivative of the first indicator over the observation period.
The second indicator helps understand the rate at which the first indicator improves or deteriorates.
The second indicator is used, according to the invention, to modify the output of the first indicator.
Preferably, a third indicator is further calculated, representing the second derivative of the first indicator over the observation period.
The third indicator helps understand whether the improvement or deterioration represented by the second indicator is constant or not.
Thanks to the present invention, the driver or manager of a vehicle fleet can promptly schedule vehicle maintenance.
The dependent claims describe preferred variants of the invention and form an integral part of this specification.

Brief description of the figures

Further purposes and advantages of the present invention will become clear from the following detailed description of an embodiment thereof (and variations thereof) and the accompanying drawings, provided purely for explanatory and non-limiting purposes, in which:

Figure 1 shows a graph representing pressure measurements taken at the fuel rail;
Figure 2 shows a monthly processing of the measurements from the graph in Figure 1;
Figure 3 shows a succession of time diagrams representing indicators calculated according to the method of the present invention.

The same reference numbers and letters in the figures identify the same elements, components, or functions.
It should also be noted that the terms "first," "second," "third," "upper," "lower," and the like may be used here to distinguish various elements. These terms do not imply a spatial, sequential, or hierarchical order for the modified elements unless specifically indicated or inferred from the text.
The elements and features illustrated in the various preferred embodiments, including the drawings, may be combined with each other without departing from the scope of protection of this application as described below.

Detailed description of preferred embodiments

According to a first aspect of the present invention,
Figure 1 shows a time graph representing the acquisition of pressure measurements taken at the rail of the fuel system of an internal combustion engine, using methane as fuel.
The pressure measurements are grouped by interval, and each interval is associated with a criticality indicator:

1st alert level: rail pressure ≤ 3.5 bar
2nd alert level: 3.5 bar < rail pressure ≤ 5.5 bar
3rd alert level: rail pressure between 5.5 bar < rail pressure ≤ 6 bar
Optimal target rail pressure > 6 bar.

Obviously, 3.5 bar corresponds to a first threshold Th1, 5.5 bar to a second threshold Th2 > Th1, and 6 represents a third threshold Th3 > Th2. Three thresholds identify four intervals.
Therefore, the second threshold of 5.5 bar is the one that separates a critical condition from an acceptable or normal operating condition.
This second threshold is the most relevant for these purposes, while the others are optional.
A warning level, according to the prior art, would be converted into the illumination of a warning light or the activation of a written message on a vehicle display or a remote display intended to inform the driver and/or a remote fleet manager about the operating conditions of the vehicle's fuel system.
According to the present invention, this information is not directly sent to the driver or manager but is processed as follows. Precisely because a direct comparison with the different pressure thresholds would lead to the distribution of "hasty" and unstable information. According to the invention, a first IR indicator is derived as $I_{R} = [1 - \frac{Engine Hour s_{critica l_{range}}}{Total Engine Hours}]$
Specifically, the ratio between the "Engine Hour Critical Range" time, corresponding to operation in a critical condition-i.e., with a rail pressure less than or equal to 5.5 bar-and the entire observation period, called "Total Engine Hours," is subtracted from the unit.
This indicator provides insight into the behavior of a vehicle/fleet to help prevent/intervene in diagnoses related to improper use.
This first IR indicator therefore indicates the percentage of time spent in the "normal" range and is classified according to the following rules:

0 to 0.5: very low rail pressure, critical vehicle condition
0.50 to 0.70: low rail pressure, warning condition
0.70 to 0.90: OK, but requires monitoring
> 0.90: good condition.

Again, the thresholds 0.5; 0.7; 0.9 identifies four intervals (ITh1, ITh2, ITh3, and ITh4) that allow us to assign meaning to the key performance indicator (KPI).
As we will see later, the value of the key KPI is associated with a message to be transmitted to the driver and/or fleet manager.
The division could be made into three intervals or a number greater than four.
It is clear that the first IR indicator represents a fraction of a predetermined observation interval (Total Engine Hours), in which rail pressure measurements are acquired, indicative of normal fuel system operation.
Indeed, the ratio $\frac{Engine Hour s_{critica l_{range}}}{Total Engine Hours}$ represents a time fraction of the observation interval in critical operating conditions, and the complement of this quantity therefore indicates the fraction of "Total Engine Hours" in which the fuel system is in "normal" operating conditions.
Over a long period of time, i.e., on the order of months, it is possible to obtain a relative frequency distribution. By averaging the frequency distribution over a monthly basis, for example, it is possible to obtain an indicator of the operating condition of the fuel system of a vehicle or a fleet of vehicles.
Preferably, this KPI is monitored, for example, from week to week or from month to month, to assess the health of the vehicle or a fleet of vehicles and, as we will see later, also the health trend.
Figure 2 shows an example of a KPI obtained based on the first indicator described above.
It can be seen that in the month of May, the optimal engine operating conditions were on the order of 63.6%, while critical conditions were on the order of 36.4%.
According to the invention, a second IFi indicator is calculated, representing the time derivative of the first IR indicator. This involves calculating the incremental ratio between two consecutive values of the IR indicator, over two successive observation intervals: $I_{F} = \frac{Δ y}{Δ t} = \frac{(I_{Ri} - I_{Ri - 1})}{Δ t}$
Generally, the vehicle data network allows pressure values to be acquired at a frequency of 1 Hz.
Therefore, if IR is calculated every 24 hours of engine operation, this means that 86,400 samples are acquired as the basis for calculating formula (1).
This second indicator, IF, indicates the tendency for improvement or deterioration of the first indicator. In absolute terms, it is characterized by the following:

from 0 to 0.5: insignificant change
from 0.50 to 0.70: small change
from 0.70 to 0.90: medium change
> 0.90: significant change

Evidently, when the derivative is positive, an improvement in the IR indicator is expected; conversely, when the first derivative is negative, a deterioration in the IR indicator is expected.
Therefore, IF can be calculated as the variation in IR over two different time intervals of 24 hours each of engine operation, i.e., while it is consuming methane.
An improvement can be due to various factors, including temperature, refueling, or proper fuel system maintenance. For example, refueling can naturally lead to an increase in rail pressure, bringing the current IR indicator to a better value than a previous value on the same indicator.
Similarly, a tightened fitting could stop a leak, resulting in an increase in rail pressure.
This second indicator obviously impacts the information content of the first indicator.

For example, if the first indicator represents a very critical condition of the fuel system, but at the same time, the second indicator shows a significant positive change, then it is reasonable to mitigate the level of criticality indicated by the first indicator. For example, a first table is shown based only on the first indicator, and a second table is shown based on both the first and second indicators:

Table 1: Based on IR only

0 < I_R ≤ 0,5	Out1: Very critical situation requiring immediate intervention (RED SEVERITY)
0,5 < I_R ≤ 0,7	Out2: Critical situation requiring short-to medium-term intervention planning (RED SEVERITY)
0,7 < I_R ≤ 0,9	Out3: Non-optimal situation requiring monitoring (YELLOW SEVERITY)
0,9 < I_R	Out4: Optimal situation (NO SEVERITY)

Table 2: based on IR and IF

IR	IF	KPI
0<=I_R<= 0,5	I_F < +0, 5	Out1: Very critical situation requiring immediate intervention (RED SEVERITY)
0<=I_R<= 0,5	I_F > +0,5	Out1: Critical situation showing insufficient improvement (RED SEVERITY)
0 , 5<I_R<=0 , 7	\|I_F\| ≤ 0,5	Out2: Suboptimal, stationary situation requiring planning of a short-term but not immediate intervention (YELLOW SEVERITY)
0 , 5<I_R<=0 , 7	<= - I_F 0, 5	Out3: Non-optimal situation rapidly worsening (RED SEVERITY)
0 , 5<I_R<=0 , 7	I_F > 0,5	Out3: The situation is not optimal and is improving. No intervention is planned. (NO SEVERITY)
I_R > 0, 7	I_F < - 0,5	The situation is optimal but rapidly deteriorating. Requires medium-term intervention planning. (YELLOW SEVERITY)
I_R > 0, 7	I_F > - 0,5	Optimal situation, approximately stable or improving (NO SEVERITY)

As can be seen from the second table, a good, yet non-optimal, condition such as 0.5 < I_R ≤ 0.7 can lead to three different types of outputs: Out1, Out2, and Out3. Out1, or red light (RED SEVERITY), requires immediate intervention; Out2, or yellow light (YELLOW SEVERITY), requires monitoring of the situation and, in any case, medium-term intervention planning; while Out3 corresponds to NO SEVERITY, i.e., an optimal condition that requires no intervention other than cyclical routine maintenance.
According to a preferred variant of the invention, the second derivative of the first indicator over time is also calculated, resulting in a third I_Si indicator: $I_{S} = \frac{d I_{F}}{dt} = {(\frac{d I_{R}}{dt})}^{2}$
It corresponds to the first derivative of the second ISi indicator over time; however, this calculation is an indicator of the increase in improvement calculated with the second indicator.
The third indicator indicates the increase in improvement/deterioration, and its interpretation is given considering its absolute value:

0 to 0.5 saturated: insignificant change
0.50 - 0.70: low increase
0.70 - 0.90: medium increase
> 0.90: significant increase.

When the increase is significant, then the third indicator represents a positive modifier of the indication obtained with the other IR and IF indicators, influencing the key KPI.

According to a preferred variant of the invention, the monitoring system's output is based on all three indicators: IR, IF, and Is. Overall, the severity of intervention in the field is reported based on the combination of the three indicators. Note that in some combinations of indicators the Is field may be irrelevant and indicated with ANY:

Table 3: based on the combination of I_R, I_F, I_S indicators

I_R	I_F	I_S	KPI
0<=I_R<= 0,5	I_F < +0,5	ANY	Out1: Critical situation showing very little improvement or, if I_F is negative, further worsening is expected. Immediate intervention is required. (RED SEVERITY)
0<=I_R<= 0,5	I_F > +0,5	Is<= 0.7	Out1: A critical situation showing improvement, but the rate of improvement is insufficient to be considered definitive. Immediate intervention is required. (RED SEVERITY)
0<=I_r<= 0,5	I_F > +0,5	Is > 0.7	Out2: A critical situation showing improvement at a rate sufficient to be considered potentially curative.
			Requires medium-term intervention planning and ongoing monitoring. (YELLOW SEVERITY)
0 , 5<I_R<=0 , 7	I_F <= 0,5	Is<= 0.7	Out1: Suboptimal situation, worsening or stable. The speed of improvement is not sufficient to be considered curative. Immediate intervention is required. (RED SEVERITY)
0 , 5<I_R<=0 , 7	I_F <= 0,5	Is > 0.7	Out2: Suboptimal situation improving. The rate of improvement is sufficient to consider the overall situation improving. Requires medium-term intervention planning. (YELLOW SEVERITY)
0 , 5<I_R<=0 , 7	I_F > 0,5	ANY	Out3: The situation is not optimal, but there has been a substantial improvement. No extraordinary interventions are planned. (NO SEVERITY)
I_R > 0,7	I_F > - 0,5	ANY	Out3: Optimal situation and negligible improvement or worsening (NO SEVERITY)
I_R > 0, 7	I_F < - 0,5	ANY	Out2: The situation is optimal but significantly worsening. Requires medium-term intervention planning. (YELLOW SEVERITY)

In fact, the key indicator, based on the examples shown in the tables, is a sort of linear combination of the two or three basic indicators I_R, I_F, and I_S.
In other words, the contribution of the second (and preferably also the third) indicator to the first indicator is weighted with a value lower than or equal to the first IR indicator.
While the first indicator indicates a current condition, the second indicator indicates a trend and therefore influences a positive or negative forecast of the fuel system's operating conditions.
Given what has been described above, the forecast can be improved, as it does not analyze a single mechanical device subject to wear, but a complex system whose efficiency depends on various factors, including, but not limited to, maintenance.
Therefore, thanks to the present invention, the analysis of the first derivative and, more preferably, the second derivative, allows for the temporary effects of fuel refueling to be remedied.
A processing unit, appropriately configured to calculate the first, second, and preferably also the third indicator, and to generate a message or, more simply, turn on a warning light based on the key KPI calculated from the indicators.
The yellow light, indicated in the table as "YELLOW SEVERITY," indicates the need to schedule, at least in the medium term, extraordinary maintenance on the fuel system, while the red light, indicated in the table as "RED SEVERITY," indicates the need for immediate repairs in the workshop.
It is clear that, compared to the prior art, the red light does not simply indicate a critical but temporary condition, but rather a stable or even worsening critical situation.
This aspect makes the diagnostic system reliable, significantly reducing false positives.
Figure 3 shows four cascade graphs:

At the top, the acquired rail pressure measurements,
The calculation of the first I_R indicator,
The calculation of the first I_F indicator,
The calculation of the third I_S indicator.

It is seen that the first derivative tends asymptotically to unity, while the second derivative tends asymptotically to zero.
By extending the observation over longer periods, such as a monthly basis, it is possible to filter out the effects of replenishment.
The present invention can be advantageously implemented using a computer program that includes coding means for implementing one or more steps of the method when this program is executed on a computer. Therefore, it is intended that the scope of protection extends to said computer program and also to computer-readable media comprising a recorded message, said computer-readable media comprising program coding means for implementing one or more steps of the method when said program is executed on a computer.
Implementation variants of the non-limiting example described are possible, without departing from the scope of protection of the present invention, which includes all embodiments equivalent to the claims for a person skilled in the art.
From the above description, a person skilled in the art is able to implement the object of the invention without introducing further construction details.

Claims

Computer-controlled diagnostic method for a methane fuel system of an internal combustion engine, in which the fuel system comprises a rail, the method comprising the following steps in sequence:
- acquisition of a succession of pressure measurements at the fuel system rail over a predetermined observation time interval (Total Engine Hours) while the engine is operating,

- comparison of each measurement with a pressure threshold (Th2) that separates a normal condition from a critical operating condition (P < 5.5 bar) of the fuel system,

- calculation of a first indicator (I_R) representative of a fraction $(1 - \frac{Engine Hour s_{critica l_{range}}}{Total Engine Hours})$ of said predetermined observation interval (Total Engine Hours), indicative of "normal" fuel system operation, wherein said indicator is calculated based on all comparisons relating to each measurement in the series of measurements,

- calculation of a second indicator (I_F) as a gradient of the first indicator over two distinct predetermined observation time intervals,

- calculation of a key performance indicator (KPI) as a linear combination of the first and second indicators, wherein said key performance indicator indicates a current condition and a forecast of a change from the current fuel system condition, and

- generation and transmission of a message corresponding to said key performance indicator.
A method according to claim 1, wherein a forecast of a normal condition, imminent failure, or current failure, indicated by the first indicator, is modified by the second indicator, the second indicator having a lower weight than the first indicator.
A method according to claim 2, wherein when the second indicator is positive, it represents an improving modifier of the first indicator, and when the second indicator is negative, it represents a negative modifier of the first indicator.
Method according to any of the preceding claims wherein said first IR indicator is given by the following formula: $1 - \frac{Engine Hour s_{critica l_{range}}}{Total Engine Hours}$ Wherein "Total Engine Hours" represents said predetermined observation interval and Engine Hours_{criticalrange} represents a portion of the predetermined time interval in which said comparisons indicate a critical engine operating condition.
Diagnostic system for a natural gas fuel system of an internal combustion engine comprising a pressure sensor operatively associated with a rail of the fuel system, a human/machine interface device, and a processing unit operatively connected to the pressure sensor and the human/machine interface device, wherein the processing unit is configured to perform all the steps of any one of claims 1-3.
System according to claim 5, wherein the processing unit is located in a remote processing center and is configured to acquire said pressure measurements from a plurality of natural gas-fueled vehicles and to calculate said key performance indicator (KPI) and generate said message by said human/machine interface device in relation to each vehicle of said plurality of vehicles.
A computer program comprising instructions for causing the processing unit of claim 5 to implement the method of claim 1.
A computer-readable medium storing the program of claim 7.
A land vehicle comprising an internal combustion engine fueled by methane gas through a fuel system comprising:
- a rail and a pressure sensor associated with the rail,

- a human/machine interface device, and- a diagnostic system according to claim 5.