Analysis of Failures of Automatic Level Crossing Devices ^†

Abstract

The safety of automatic level crossing devices (ALDs) is of utmost importance due to the need to ensure safe passage through railway crossings—accidents must not be allowed. In this paper, an analysis of failures of ALDs in the railway infrastructure in Bulgaria for the period 2020–2022 is performed. First, a statistical evaluation of failures and their duration are calculated. Then, simple reliability indicators have been defined (the time for failure-free operation for different periods, the statistical assessment of the intensity of the failures flow and MTBF) in two ways—considering all failures and their duration or only those due to technical reasons. Finally, the coefficients of availability and unavailability are determined (also in both ways). A comparison was made with the availability coefficients of the indoor devices. The results obtained can be used in determining guidelines for the reconstruction and renovation of signaling systems and railway lines, as well as in identifying conflict points where multi-level crossings should be implemented. It should be noted that performing only a statistical analysis is not enough, it is necessary to determine the reliability indicators and, above all, the availability coefficient.

1. Introduction

A level crossing is a junction where a railway line crosses a road at the same level [1,2], i.e., the railway line does not cross the road via an overpass or tunnel. Automatic level crossing devices (ALDs) are used to signal and protect road vehicles when a train is about to pass through the level crossing. They also provide information to the locomotive driver about the activated ALD, as well as to signal about a working ALD in the absence of a train in the area. Therefore, ALDs are a system for protecting railway crossings between stations and for providing information to the locomotive driver about safe crossings. They signal to road vehicles when a train is about to pass through the crossing: if the crossing device is in working order and there is no railway vehicle in the crossing notification zone, the signal indicates—with a white flashing light, and with a red flashing light—if there is a railway vehicle in this zone.

The safety of ALD operation is of the utmost importance, as safe passage through railway crossings must be ensured and accidents must not be allowed. In this paper, a statistical evaluation of data on the failures of ALDs is performed, and then reliability indicators are determined. The results obtained can be used in determining guidelines for the reconstruction and renovation of railway lines and signaling systems.

2. State of the Problem and Research Methodology

ALDs are a part of the inter-station devices of signaling systems. The “Signalling and Telecommunications” Departments (STDs) at the National Railway Infrastructure Company (NRIC) are responsible for the operation, repair and maintenance of signaling systems at Bulgarian Railways. The three units in Sofia, Plovdiv and Gorna Oryahovitsa are accountable for the activities in the respective areas. They collect data on failures of all devices of signaling system—their number and duration. The causes for the appearance are also indicated—as a result of technical or other reasons (assaults, theft, vandalism, incidents, activities of other organizations, natural disasters, etc.). Every year, NRIC issues an annual consolidated report of activity based on the collected data. The failures and their duration are systematized by device type:

• indoor station devices—these are interlocking stations (ILSs), and the following types operate in Bulgarian railways: route-computer interlocking stations (RCISs), route-relay interlocking stations (RRISs), electromechanical interlocking stations (EMISs), relay systems for key dependences (RSKDs) and temporary traffic lights control panels [3];
• outdoor station devices (OSDs)—point machines, traffic lights, rail current circuits and axle counters (RCS_ACs);
• inter-station devices (IsDs)–ALDs, relay semi-automatic interlocking (PAB type), automatic interlocking (AB type), automatic locomotive signaling and points on open road [4];
• telecommunications (TCs)—automatic telephone switches (ATC type) and external telecommunications lines (TCLs).

This research analyzes failures of automatic level crossing devices across the country over a three-year period (2020–2022) and compares them with failures of other equipment.

The study begins with a statistical evaluation of device failures using the information given in the annual consolidated reports of the activity. The following calculations are performed [5]:

• total number n of device failures for the monitored period:

$$n={\sum }_{i=1}^N{r}_i\left(\Delta t\right),$$

(1)

where N—total number of types of devices monitored in the interval Δt;

r_i(Δt)—number of failures of a particular device type in the monitored interval Δt;

Δt = 365 × 24 = 8760 h (1 year) or Δt = 3 × 365 × 24 = 26,280 h (3 years).

• total duration t of failures of devices for a period:

$$t={\sum }_{i=1}^N{t}_i\left(\Delta t\right), h$$

(2)

where t_i(Δt)—duration of failures of a particular type of signaling devices in the monitored interval Δt.

• share of failures n_f by types of devices over the monitored period:

$${n_f}_i={\displaystyle \frac{r_i\left(\Delta t\right)}{n}}·100, \%, i = 1\dots N$$

(3)

where n_{f i}—percentage ratio of failures of the ith type of signaling devices.

• share of duration of failures t_f by types of devices over the monitored period:

$${t_f}_i={\displaystyle \frac{t_i\left(\Delta t\right)}{n}}·100, \%, i = 1\dots N$$

(4)

where t_{f i}—percentage ratio of duration of failures of the ith type of signaling devices.

Using these expressions, the share of failures and their duration for all device types can be calculated and compared. Furthermore, the devices with the highest number of failures and the longest duration can be identified.

The study continues with the determination of reliability indicators of the ALDs using the following equations [6,7,8,9]:

• time for failure-free operation τ(Δt) of ALDs:

$$\tau \left(\Delta t\right)=\Delta t-t_{ALD}\left(\Delta t\right)$$

(5)

where t_ALD (Δt) is the duration of failures of ALDs for the monitored period Δt;

• statistical assessment of the intensity of the failure flow:

$$\stackrel{̑}{\omega }(\Delta t)={\displaystyle \frac{r_{ALD} \left(\Delta t\right)}{N_{ALD}·\tau \left(\Delta t\right)}}$$

(6)

where r_ALD(Δt) is the number of failures of ALDs for the period Δt;

N_ALD = 524—number of ALDs in Bulgarian railways (it did not change during the three-year study period).

The following ratio can be assumed to represent the average number of failures per one ALD during the observed period:

$$r_{ALDav}(\Delta t)={\displaystyle \frac{r_{ALD}\left(\Delta t\right)}{N_{ALD}}}$$

(7)

• Mean Time Between Failures MTBF:

$$\mathit{MTBF}={\displaystyle \frac{1}{\stackrel{̑}{\omega }\left(\Delta t\right)}}$$

(8)

• stationary coefficient of availability:

$$A=1-\stackrel{̑}{\omega }\left(\Delta t\right)·t\left(\Delta t\right)$$

(9)

• stationary coefficient of unavailability:

$$U=\stackrel{̑}{\omega }\left(\Delta t\right)·t\left(\Delta t\right)$$

(10)

It should be noted that the ALDs are recoverable objects, and MTBF is the basic parameter for fault tolerance. The coefficients $A$ and $U$ are complex indicators—they characterize both the failure-free operation and repairability of the devices and give a more accurate evaluation of their reliability.

3. Research Results

3.1. Statistical Evaluation of Failures and Their Duration

Information on ALD failures across the country during the monitored period (2020–2022) is systematized in

Table 1. Failures of ALDs and their duration.

Year	Failures Due to Technical Reasons		Failures Due to Other Reasons
	Number	Duration	Number	Duration
2020	134	633.96 h	422	1200.74 h
2021	164	688.25 h	309	832.34 h
2022	273	996.29 h	444	930.50 h
Total (2020–2022)	571	2318.50 h	1175	2963.58 h

, with the number and duration of ALD failures due to various causes grouped separately [10,11]. Data are based on [12] and are processed according to Equations (1) and (2).

Then the share of failures n_f and their duration t_f were calculated using Equations (3) and (4), not only for ALDs but also for all types of signaling devices. The correlation of failures on different equipment types are shown graphically in Figure 1, Figure 2, Figure 3 and Figure 4 (by year and for the three-year period, respectively). As can be seen, the percentages of failures and their duration for ALDs are some of the highest—about 20% for each period, reaching 25% for 2022. A higher percentage is only reached for failures of rail current circuits and axle counters.

In general, failures on external devices (inter-station IsDs and outdoor station OSDs) are higher than those on internal devices (interlocking stations ILSs and automatic telephone exchanges ATC type). They are under the influence of climatic conditions (temperature, rain, snow, etc.), and the risk of theft, vandalism and other encroachments is higher.

In particular, regarding ALDs, the risk of transport accidents is the greatest, e.g., breaking a barrier beam when a vehicle attempts to pass when prohibited due to an oncoming train. In such a case, the duration of the failures can hardly be predicted, as it is usually also related to a police investigation. That is why the values of failures for other reasons and their duration in Table 1 are much higher compared to those for technical reasons. In addition, it is necessary to move the repair crew to the accident site, which also requires time. However, the duration of failures is relatively lower than that for track current circuits and axle counters, as no time is needed to accurately locate the fault. It is also valid for the external telecommunications lines—the percentage duration of failures in them is much greater than the percentage of failures.

3.2. Analysis of Reliability of ALDs

First, simple reliability indicators have been defined—the time for failure-free operation τ(Δt) for different periods Δt using Equation (5), the statistical assessment of the intensity of the failures flow ${\widehat{\omega }}_i\left(\Delta t\right)$ using Equation (6) and MTBF using Equation (8). All indicators are calculated in two ways—considering all failures and their duration or only those due to technical reasons. The results obtained for the main simple indicator MTBF are shown graphically in Figure 5. It is obvious that the mean time between failures due to technical reasons significantly exceeds the indicator when all failures are considered. This once again confirms the conclusion that the downtime due to non-technical reasons can be difficult to determine.

Second, the coefficients of availability and unavailability have been determined using Equations (9) and (10) again in two ways. The results obtained for the availability coefficients in both ways (considering all failures and their duration or only those due to technical reasons) are shown graphically in Figure 6. Again, it is obvious that the coefficients regarding technical failures significantly exceed the indicator when all failures are taken into account and are greater than 0.9. For 2020 and 2021 the coefficient is in the range 0.97–0.98, while for 2022 it is lower—0.933 (the number of failures in this year had increased, as well as their duration). For the three-year period it is close to 0.9, which is also a relatively high value.

This value was compared with the availability coefficients of the indoor devices, and the results are displayed graphically in Figure 7. It can be seen that the indicator of some of interlocking stations is even lower than that of ALDs.

Similarly, indicators can be calculated when considering only failures due to other reasons (non-technical) and their duration. In this case, lower values of the availability coefficients will be obtained.

4. Discussion

An analysis can also be carried out by sections (lines and specific ALDs), calculating the availability coefficients in both ways. It can then be determined for which devices the availability coefficient calculated in relation to failures due to technical reasons is lower and measures can be taken to modernize or replace the ALD type [10,11,13].

On the other hand, for devices with a lower availability coefficient calculated in relation to failures due to other reasons, it is necessary to consider a change in the infrastructure—construction of an overpass, tunnel, etc.

Thus, the results obtained can be used in determining guidelines for the reconstruction and renovation of signaling systems and railway lines, as well as in identifying conflict points where multi-level crossings should be implemented.

5. Conclusions

Automatic level crossing devices are the most critical facilities in railway infrastructure due to the intersection of the railway with the road and the danger of accidents due to the inattention of road vehicle drivers. ALDs must ensure safe passage, and their reliable and secure operation is of utmost importance.

This paper analyses failures of ALDs in the railway infrastructure in Bulgaria for the period 2020–2022 and determines reliability indicators.

It should be noted that performing only a statistical analysis is not enough; it is necessary to determine the reliability indicators and, above all, the availability coefficient.