Availability of an Overhead Contact Line System for the Electrification of Road Freight Transport

Abstract

The electrification of road freight transport on highways using an overhead contact line system is being tested in Germany. In this study, we investigated availability of an overhead contact line system based upon unique real-world data gathered in the ELISA field test. Based on our investigation, we anticipate a high system availability in future operations. We concluded that the newness of the system can justify the longest downtimes. To confirm our findings, we first categorized causes of the downtimes. Building upon these data, we analyzed these downtimes, specifically focusing on the longest occurrences as they posed significant barriers to future operation and market implementation. Subsequently, we illustrated how the system’s availability changed when individual causes of downtimes were reduced over time. Our research is paving the way to chart potential future operational scenarios. Our contribution aids decision-makers and all individuals who need to determine whether overhead contact line technology should be employed on highways in the future.

1. Introduction

Road freight transport is currently undergoing a major technological change, triggered by the goal to offset or obviate negative effects of greenhouse gas emissions. The potential technologies offer different advantages in terms of production, provision, use, and local emissions in order to make a significant contribution. In addition to the possibility of using fuel cells or electric vehicles powered purely by charging stations, dynamic charging options are being analyzed worldwide. Various dynamic charging options (conductive and inductive) are promising, but they are at different stages of development in terms of technological readiness; however, they are also considered to have a high potential for reducing greenhouse gas emissions.

The electrification of road freight transport is being tested in Germany with the implementation of an overhead contact line (OCL) system as a new large technical system (LTS) on highways. The aim of the ELISA project is to analyze the new technology in terms of its feasibility and potential for reducing greenhouse gas emissions by considering various stakeholders. ELISA stands for “electrified, innovative heavy traffic on highways” and represents the first and largest public OCL test track in Germany. The key advantage this technology offers is supplying trucks with energy while driving, which means that smaller batteries and lower charging rates can be realized. An OCL system allows suitably equipped trucks to consume electrical energy from overhead lines while driving and dynamically charge an on-board electricity storage system; this means sections without an OCL can also be driven electrically. In addition, the direct use of electricity is very efficient, and the electricity can be provided by adjacent decentralized energy parks.

The assurance of availability is a key factor in implementing an OCL system. Our research, based on real data from the ELISA project, shows such a system achieves an availability of more than 98%. After implementing an OCL system as a new LTS, the operator’s key tasks are maintaining its functionality and guaranteeing its availability. This is particularly the case if the LTS in question is to replace or extend a previously existing and competing LTS within a transformation process. Such transformation processes can be found, for example, in telecommunications (wireless technology or the Internet) or energy supply (decentralized energy supply and renewable energies). Electrification and digitalization, however, are also leading the transport sector to undergo a transformation process [1,2,3].

The implementation of an OCL system for road freight transport forces decision-makers to face major challenges, such as the high cost and the initial low demand from a lack of potential users. However, the decision to implement and expand such a system depends on functionality and availability, as these are mandatory requirements for users. Thus, the following research questions arise: (1) Which challenges exist for the implementation as an LTS?; (2) Which causes of downtimes limit availability?; (3) How high is the general availability of an OCL system?; (4) How can functionality and availability be increased?

In this paper, we answered these questions by analyzing the OCL system built in Germany, which has been in pilot operation on a public road (highway A5 in Hesse) since May 2019. For this purpose, we used data generated from the OCL system during a field test of its first three years in operation.

Our results show negative external influences from traffic operations, such as incidents on roads, damage to the OCL (e.g., caused by vehicle contact), and general operations (e.g., by fire and rescue services), which in particular impair functionality and availability. In contrast, disruptions to the OCL itself or external influences that cannot be prevented, such as weather conditions, rarely occur. The longest downtimes in the first three years, however, are related to unexpected problems that were identified and solved during the field test (system-related teething problems). Despite many and varied causes of downtimes, we observed high availability. For instance, the option of shutting down the OCL in only one direction of travel was subsequently implemented to reduce downtimes and also increase availability, especially during planned downtimes (e.g., for track repair work). As the OCL system initially operated in the introduction phase and only switched to regular operation in 2020, we also examined the advantages and disadvantages of these operating modes.

Based on our results, we estimated the functionality and availability of an OCL as one type of the electric road system (ERS) if it is to be implemented as a new LTS in the road network. We support decision-makers, operators, and users who need to balance whether to implement an OCL system as a new LTS in road networks. To do so, we presented primary approaches to evaluating OCL systems for electrification of trucks and compare with other technologies. As the OCL is also a critical infrastructure, knowledge of possible malfunctions is of particular importance to reduce downtimes and thus high negative impacts [4,5,6]. This paper is the first to analyze influences on the availability of an OCL system, thereby contributing to the future design of a sustainable and resilient transportation sector and to the preservation of human health and the human environment.

2. Literature Review

The course is currently being set to make the transport sector sustainable and resilient. One promising option for road freight transport is electrification based on an ERS, although the necessary infrastructure must first be built. Our literature review identifies relevant specifications of an LTS and classifies the ERS in this context. For analysis, availability, and functionality considerations, the current technology readiness level (TRL) of the OCL and the infrastructure lifecycle is to be considered, thereby drawing on challenges in the system expansion. Furthermore, we show the downtime causes using energy supply as an example in order to be able to describe the challenges of an ERS in this context.

2.1. The Importance of the Availability of Electric Road Systems as Large Technical Systems

The concept of large technical systems (LTS) was first highlighted in the 1980s by Hughes [7]. He [7] illustrated the importance of the complex electricity network not only by showing examples of such systems and their developments in the US, Britain, and Germany, but also by outlining their social meanings. In the meantime, the term LTS has been established in the literature and encompasses a complex system consisting of technical components and operators who manage the systems, which itself makes a high social contribution. Gökalp [8] (p. 58) describes this as “global in scope and global in structure”. In addition to the electricity network, telecommunications, the Internet, transportation and rail networks, and urban transportation systems, for example, are referred to as LTS [7,8,9,10,11,12].

An important characteristic of LTS is that once they have been established, they can hardly be replaced by other systems. Hughes [13] also recognized this characteristic and called it technological momentum. The current climate problem makes it necessary to transform established LTS to obtain a sustainable transportation system [14,15,16,17], even though LTS are, by nature, resistant to these changes [18,19]. In the transport sector, these changes are unprecedentedly evident in road freight transport. While in passenger transport the expansion of charging stations makes the electrification of vehicles possible [20,21,22], the high-energy consumption of heavy vehicles means this solution cannot (currently) be applied [23,24,25]. For this reason, the expansion of electric road systems (ERS) is being discussed worldwide, especially in Europe [26].

As the first OCL systems were tested on public roads in the US, Sweden, and Germany [27,28] show, the electrification of heavy vehicles not only contribute to reducing greenhouse gas emissions [29,30], but also do so quickly. In the meantime, several studies have aimed to identify suitable corridors in which such a system should be established [31,32,33,34]. With technical feasibility, acceptance and political will are now required to drive forward the expansion of ERS, e.g., OCL systems (compared with the Strategic Niche Management by Hoogma et al. [35]). This is where the technological momentum of the current transportation system is revealed. While Tongur and Engwall [19] describe it as a “chicken-and-egg” dilemma, it is much more a problem of political assertiveness based on a lack of evidence of successful transformation of the LTS.

It, therefore, remains to be seen whether the new ERS will be able to break down the existing technical structures and develop into a new LTS. An important factor for decision-makers in this regard is knowledge about the availability and functionality of the energy-providing infrastructure. We contribute to this gap in knowledge by evaluating the data available from the ELSIA field trial. Certainly, it is not only the electrification of transportation that contributes to the transformation process, but also there are several other factors that have an additional inhibiting or supporting effect. These factors also include digitization in combination with autonomous driving [36], the changing values of societies [37], and the scarcity of resources [38].

To evaluate the availability and functionality of the new infrastructure, the infrastructure lifecycle must also be considered. LTS will not just be built: these systems will become part of the society and indispensable for users [17]. Gökalp [8] divides the lifecycle of an LTS into a total of four phases: the initial phase, the accelerated development phase, the stabilization phase, and the decline phase. Bolton and Foxon [39] transferred these phases into an infrastructure lifecycle model, which includes the following sections:

• System building and establishment;
• System expansion and momentum;
• Stagnation and inertia;
• System renewal and transition.

The first section is characterized by the fact that starting from small niches an establishment can take place within competing systems and the standardizations have been already achieved. The field tests conducted so far and the expansion plans promoted in Germany illustrate the description in the infrastructure model [40]. Concerning the TRL, the existing competing ERS technologies currently are not mature enough to compete with the OCL [41]. On the other hand, electrification is leading to a decline in conventional systems and thus to a technological transition [42].

A significant property of the initial phase and thus in the arena of system building and establishment is the high susceptibility to errors. Usually, the failure rate in the initial phase is quite high. The susceptibility to error of different components is often represented in the so-called bathtub curve and failure rate functions [43]. After the initial phase, the failure rate usually decreases, and the technical system then starts running smoothly. Instead of early failures (e.g., teething problems), random failures affect availability [44]. Accordingly, not only are teething problems to be expected with an OCL, but the entire ERS technologies for the electrification of road freight transport, remain in their infancy. Ensuring availability is not only a prerequisite, but also a major challenge, as the problems and their effects can hardly be estimated in the initial phase. Currently, no studies exist on the availability of an OCL over time, beginning with the system building and establishment up to the system expansion. With our investigations, we close this knowledge gap.

2.2. Causes of Downtime

The downtime causes of OCL systems, in general, are varied and result, for example, from extreme weather events [45,46], technical faults [47], or human error. At present, OCL systems are used for power distribution and also for the electrification of rail vehicles, both regionally and in cities. In some cases, trolleybuses are also used instead of streetcars [48,49]. Implementation of OCL systems on highways is, therefore, initially just another application, but it should be noted that the requirements (speed of vehicles compared to trolley buses and track keeping compared to the rail guided) differ significantly. If we align the downtime causes of an ERS with the downtime causes of an LTS (e.g., power grid), the main causes are shown as follows [50,51,52,53,54]:

• Weather events, temperature, and force majeure (e.g., strong winds and extremely high temperature);
• Electrical equipment and network congestion and overload (e.g., technical failures and load shedding);
• Faulty network operation (e.g., planned shutdowns and inadequate maintenance);
• Human violence (e.g., sabotage and cyber-attacks);
• Human errors;
• Short circuit.

Behnert and Bruckner [50] have determined that 90% of all transnational blackouts since 1965 can be assigned to the first three categories; however, it is important to note that the power network has only a few direct points of conflict with the transportation network (except for aviation corridors).

It should also be noted that failures can be predictable or unpredictable [55], expected or unexpected [56], and can occur for internal or external causes [57]. One important difference from the existing systems is that the OCL system has to be built from scratch over a larger section. This would be the first-time scaling of the system, and the causes of downtime cannot be fully captured at this phase. If we consider the fact that a road traffic incident can occur under an OCL, which leads to an OCL system shutdown, it becomes clear that external influences in particular affect availability. Additionally, such incidents could damage the OCL. The effects of traffic on the OCL availability are correspondingly high, so the causes of downtimes differ when compared to power networks or overhead contact lines in rail traffic. So far, there has been no consideration of which causes of downtimes are relevant for an OCL in the future and how road incidents as external influencing factors limit availability. Additionally, autonomous vehicles will also have an impact on traffic accident frequency [58]. We fill this gap with our research on the OCL.

3. Methodology

In order to prepare a systematic analysis of availability, our first step was to develop a superordinate classification of causes of downtime. An important pillar of our research is our intensively, carefully waged OCL infrastructure diary. This OCL infrastructure diary includes the exact times during which the OCL system was switched on or off and thus the times of energy supply and the downtimes. The necessary data were collected in a data logger in the system’s substation and can be retrieved as a csv file. In the event of a downtime, we documented the cause and sorted it into the corresponding category. For this purpose, we obtained additional information on the cause of the downtime from the OCL system operator. We determined the downtime’s duration, depending on the OCL scheduled operating time. In the ELISA Project, we defined two operation periods, which we consider in this paper as follows:

• Introduction phase of the OCL (7 May 2019–31 December 2019);
• Regular operation of the OCL (1 January 2020–30 June 2022).

During the introduction phase, the downtimes on a single day had a maximum of 8 h, as the planned operating time was only between 8 a.m. and 4 p.m. five days per week. In the regular operation, the OCL system operated 24 hours 7 days and was, therefore, limited only by operational downtimes.

To analyze availability, we considered downtimes in terms of both the duration and the category. In terms of the duration, we calculated the average of the OCL availability A_OCL by comparing the observed operating time t_O and the expected operating time t_E according to Equation (1) [59]. The expected operating time depended on the operation period. As the system could be switched on and off separately in the direction of travel, availability was determined for both directions of travel:

$${\mathrm{A}}_{\mathrm{O}\mathrm{C}\mathrm{L}}={\displaystyle \frac{{\mathrm{t}}_{\mathrm{O}}}{{\mathrm{t}}_{\mathrm{E}}}},\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h} {\mathrm{t}}_{\mathrm{E}}=\left\{\begin{array}{c}8 \mathrm{h} \mathrm{o}\mathrm{n} \mathrm{o}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s} \mathrm{i}\mathrm{n} \mathrm{i}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n} \mathrm{p}\mathrm{h}\mathrm{a}\mathrm{s}\mathrm{e}\\ 24 \mathrm{h} \mathrm{o}\mathrm{n} \mathrm{o}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{d}\mathrm{a}\mathrm{y}\mathrm{s} \mathrm{i}\mathrm{n} \mathrm{r}\mathrm{e}\mathrm{g}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{r} \mathrm{o}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\end{array}\right.$$

(1)

In addition to the downtimes’ durations and associated availability of the OCL, we analyzed in the next step causes of downtimes. It is necessary to consider the downtimes’ causes to be able to derive trends in how future availability will develop. It should be noted here that the OCL system is operating for the first time, and therefore, a large number of initial difficulties (teething problems) will occur; these teething problems must be identified to make consistent statements.

Figure 1 shows the methodological approach within the current work. The results in the respective chapters and sections serve as interim findings for the subsequent processes.

4. Findings

Based on our literature review and the OCL infrastructure diary, we analyzed the downtimes’ causes and availability. As a key result of our findings, we defined suitable categorizations of downtimes’ causes. This categorization included six different types of downtimes’ causes and is suitable to support the evaluation of future OCL availability. Based on the characterized downtimes’ causes and their durations and frequencies of occurrence, we analyzed availability and provide decision-makers with recommendations.

4.1. Categorization of the Downtimes

Based on the possibilities listed for the characterization of downtimes’ causes, a differentiation between internal and external as well as predictable and unpredictable downtimes was suitable. External causes included downtimes not caused by the OCL system itself. These were, for example, road traffic incidents that required the power supply to be shut down and severe weather conditions that caused downtimes. Each downtime that occurs was assigned to one of these categories.

Table 1. Categorization of the downtimes’ differentiated sphere of influence and predictability.

Sphere of influence	Predictability	Category	Description
Internal	Unpredictable	Malfunction-related downtimes	Downtimes of the system in which the reason for the malfunction results from the system itself and its nature (no external influences). This includes, general faults in the OCL system.
Internal	Predictable	Maintenance-related downtimes	Downtimes of the system due to planned inspection and maintenance work on the OCL infrastructure. If additional (unscheduled) maintenance work is required after a malfunction-related or incident-related shutdown, e.g., repair or replacement of components, this is assigned not only to the maintenance-related, but also to the malfunction-related or incident-related downtimes.
External	Unpredictable	Weather-related downtimes	Downtimes of the system due to weather conditions, e.g., storm damage to the system or manual shutdowns due to increased operating temperature in the substation of the OCL infrastructure.
External	Unpredictable	Incident-related downtimes	Downtimes of the system due to mechanical effects on the OCL infrastructure (e.g., incidents with the OCL) and due to targeted (short-term) switch-offs of the OCL infrastructure, e.g., at the request of the emergency response authorities in the event of an incident on the OCL section. The necessary time for repair work after an incident is also included in the incident-related downtimes.
External	Unpredictable/ predictable	Operation and system-related downtimes	Downtimes of the system that result from an unavailable operating possibility, e.g., in 8/5 operation on bridge days, due to an unscheduled lack of operating personnel, malfunctions in the operations center, human errors, or malfunctions in the communication to the OCL infrastructure.
External	Predictable	Work site-related downtimes	Downtimes of the system that are necessary due to construction measures in the track section, e.g., repair work on the road surface and repair work on the existing bridge structures as well as the installation of traffic cameras at the edge of the track.

shows the formed categories.

The critical downtimes were those that occurred externally and unpredictable. In the case of power networks, downtimes related to weather and human error were included. For an OCL system, the incident-related downtimes must also be considered. In order to be able to minimize the impact for such downtimes, suitable incident programs must be developed in the future. Conversely, the internal and unpredictable causes of downtime cannot be prevented, but the probability can decrease with continuous maintenance of the OCL system. Teething problems, which can occur especially in the initial phases in the lifecycle of an LTS (see Section 2.1), also belonged to this category and must be viewed as lessons learned in the system expansion.

4.2. OCL Availability in the ELISA Field Trail

First, we analyzed availability using statistical characteristics. The monthly evaluation of availability indicates that in some months, the OCL system could barely or not be used in a certain direction of travel (see Figure 2): on average, the availability values of the OCL system were 79.2% in the southbound direction and 86.16% in the northbound direction. As the separate directional switching of the OCL infrastructure was not set up until mid-February 2020, values in the two directions of travel showed the same numbers.

The availability calculated separately for each year showed values between 69.2% (in the southbound direction in 2020) and 96.4% (in the northbound direction in 2022). The expected operating time was significantly lower in 2019, as operations were in the introduction phase (see Section 3). The 69.2% availability represented a clear minimum value in regular operation, while the remaining figures had values of at least 80%. Accordingly, the total downtimes (per year and direction of travel) varied widely, ranging from 154.5 h to 2703.6 h. The sum of the total downtimes was 8006.94 h. Of these, a total of 3043.99 h (38%) was found in the northbound direction, and a total of 4962.95 h (62%) was found in the southbound direction.

Table 2. Availability of the OCL system in the ELISA project over the period from May 2019 to June 2022.

Year	2019		2020		2021		2022
Direction of travel	South	North	South	North	South	North	South	North
Expected operating time [h]	1376.0	1376.0	8784.0	8784.0	8760.0	8760.0	4344.0	4344.0
Observed operating time [h]	1055.6	1055.6	6080.4	7849.5	7012.0	7125.4	4153.1	4189.5
Availability [%]	76.7	76.7	69.2	89.4	80.0	81.3	95.6	96.4
Downtimes in total [h]	320.4	320.5	2703.6	934.5	1748.0	1634.6	191.0	154.5

presents the parameters for the availability of the OCL system and the downtimes in total.

4.3. OCL Downtimes in the ELISA Field Trail

In the next step, we applied the categories to analyze the downtimes. We noted the majority of the downtimes’ causes were incident-related (46.5% of the total duration of the downtimes). Malfunction-related (26.7%) and operation and system-related (17.8%) downtimes’ causes were also frequent reasons for the shutdown of the OCL system. The weather-related (4.1%), work site-related (2.7%), and maintenance-related (2.3%) were less frequent downtime causes.

Table 3. Downtimes causes of the OCL system in the ELISA project over the period from May 2019 to June 2022.

Year	2019		2020		2021		2022
Direction of travel	South	North	South	North	South	North	South	North
Downtimes in total [h]	320.4	320.5	2703.6	934.5	1748.0	1634.6	191.0	154.5
Malfunction-related [h]	3.1	3.1	0.0	0.0	1119.8	1012.8	0.0	0.0
Maintenance-related [h]	16.3	16.3	63.8	63.8	0.0	0.0	16.7	6.2
Weather-related [h]	98.5	98.5	0.0	0.0	0.0	0.0	64.9	64.9
Incident-related [h]	14.2	14.2	2557.3	775.4	118.9	127.7	50.0	64.5
Operation and system-related [h]	140.4	140.4	82.3	82.3	484.2	484.3	4.7	2.4
Work site-related [h]	48.0	48.0	0.1	12.9	25.1	9.8	54.7	16.5

shows the absolute downtimes in relation to the direction of travel by year.

In addition to a large number of minor downtimes, a total of 5 major downtimes (each more than 150 h) occurred during the pilot operation and significantly impacted availability. These 5 downtimes represented a total of 6033.25 h (75.3% of the total duration of the downtimes). Of these, a total of 4057.94 h (81.8%) was found in the southbound direction, and a total of 1975.31 h (64.9%) was found in the northbound direction. A chronology of the five longest downtimes was shown as follows:

• The first longest downtime (3045.18 h) involved an incident caused by a vehicle whose improper storage of transported goods collided with the OCL, damaging it in the southbound direction on 24 January 2020 (incident-related downtime). However, due to the impending COVID-19 pandemic, a repair within a short period was not possible. In addition to the contact ban, the delivery of spare parts during the COVID-19 pandemic conditions added to the complexity. Ultimately, the OCL could not be put back into operation until May 2020 in the southbound direction (2476.90 h downtime in the southbound direction). As the OCL in the northern direction of travel did not show any damage, the possibility of separate switching of the direction of travel was examined and finally implemented. In mid-February, the OCL system was finally put into operation in the northbound direction (568.28 h downtime in the northbound direction).
• The second (1183.96 h) and the third (948.51 h) longest downtimes occurred for the same initial reason. Here, the rock salt, which came into the area of the OCL system during the highway road salt spreading operations, reacted with the OCL insulators, requiring some to be replaced (malfunction-related downtime of 591.98 h per direction of travel). Thus, the OL infrastructure had to be shut down from mid-March to early April (the second longest downtime). In December 2021, all insulators were then replaced with suitable insulators over a period of about three weeks (the third longest downtime). The insulators replacement in the southbound direction caused a downtime of 527.71 h and a downtime of 420.78 h in the northbound direction.
• The fourth longest downtime (747.98 h) occurred from the end of 2020 to the beginning of 2021. During this period, the OCL system was completely switched off for more than two weeks. The reason was the change of responsibilities and thus the handover of the system (operation and system-related downtime of 373.99 h per direction of travel).
• The fifth longest downtime (174.72 h) occurred in 2019. Here, a very strong storm event damaged the OCL (weather-related downtime of 87.36 h per direction of travel); however, this was the only weather-related damage to the OCL. In 2022, the system was shut down as a preventative measure in advance of a storm warning for a total of 129.88 h. No damage occurred in the system.

Our analyses show that the causes of the longest downtimes are not expected to occur in the future. The longest downtime could only increase to such an extent, because the repair was not possible due to a complex tense situation. It should be noted, however, that while only the northbound direction was in operation, an incident-related downtime of 126.96 h occurred because of emergency response operations. Our analyses show that if the system malfunctioned, it could be repaired within one day. On the other hand, the second and third longest downtimes were typical teething problems that could be identified and solved within the project.

Planned shutdowns also occurred along with the ELISA test track expansion in 2022. As only the southbound direction was extended, this direction of travel was largely affected (work site-related downtime of 37.91 h); however, as both directions of travel were initially switched off for the one-sided operation of the OCL system, the northbound direction of travel was also marginally affected (work site-related downtime of 0.03 h).

If these five longest downtimes and the downtimes related to the test track expansion were excluded from the consideration of availability, a value of over 95% was obtained for the period under consideration in regular operation (see

Table 4. Adjusted values for availability and downtimes causes of the OCL system in the ELISA project from May 2019 to June 2022.

Year	2019		2020		2021		2022
Direction of travel	South	North	South	North	South	North	South	North
Availability [%]	76.7	76.7	95.3	95.1	96.7	96.8	98.0	97.9
Downtimes in total [h]	320.4	320.5	416.4	433.3	285.0	278.6	88.1	89.5
Malfunction-related [h]	3.1	3.1	0.0	0.0	0.1	0.1	0.0	0.0
Maintenance-related [h]	16.3	16.3	63.8	63.8	0.0	0.0	16.7	6.2
Weather-related [h]	98.5	98.5	0.0	0.0	0.0	0.0	0.0	0.0
Incident-related [h]	14.2	14.2	270.2	274.2	118.9	127.7	50.0	64.5
Operation and system-related [h]	140.4	140.4	82.3	82.3	140.9	141.1	4.7	2.4
Work site-related [h]	48.0	48.0	0.1	12.9	25.1	9.8	16.8	16.5

). Overall, the downtime duration was reduced to 2231.8 h (1121.8 h in the northbound direction and 1110.0 h in the southbound direction). The distribution of causes showed clear differences between the introduction phase and the regular operation, as operation and system-related downtimes dominated in the introduction phase. In regular operations, incident-related downtimes represented the largest share of total downtimes (56.9%). After extensive work on the IT infrastructure between the control center and the substation, the proportion of operation and system-related downtimes in 2021 was relatively high; therefore, in the period under consideration, they comprised the second largest proportion (28.5%). Maintenance-related downtimes (9.5%) and work site-related downtimes (5.1%) were less frequent, while malfunction-related downtimes and weather-related downtimes were almost non-existent.

We can also show that the possibility of separate switching in the direction of travel of the OCL had a positive effect on availability. Since 2020, this separate circuit has been used specifically in order to carry out different work in the road area. Accordingly, the work site-related downtimes showed very different values for the respective direction of travel. It should be noted that the separate switching of the OCL is not currently used at the request of the emergency response authorities for safety reasons.

5. Discussion about Future Availability

The availability and the downtime causes differed significantly between the introduction phase and the regular operation. While we can quantify the average of the availability in the regular operation at over 95% by excluding the five longest downtimes and the downtimes due to test track expansion, the introduction phase still only shows an availability of 76.7%. In this case, manual control of the OCL system caused the low availability; the system, for example, not only was switched on as planned in the 8/5 operation, but also remained completely switched off on bridge days. Accordingly, the share of operation and system-related downtimes in the introduction phase was high. In contrast, in 8/5 operations, predictable events such as the OCL infrastructure maintenance and the work site-related downtimes can be planned and scheduled even outside the guaranteed time of operation.

Based on our research, we can take a look into future operation, namely when such an ERS has expanded and manifested itself as a technical momentum. In this case, not only will the failure rate of the OCL infrastructure be low, but the total duration of downtimes will be reduced significantly; external events will thus primarily limit availability. Road traffic incidents, which lead to an OCL system shutdown at the request of the emergency response authorities, will take the largest share here. The number of downtimes will then depend in particular on the highway accident rate and the length of an electrified stretch of track. Recent data from the German Federal Statistical Office show the crash rate on Germany highways before the COVID-19 pandemic was about 13.5 incidents per year and route kilometer [60]. This means a 5 km OCL section (based on both directions of travel) will have approximately 68 incidents per year (34 incidents per direction of travel), which may limit system availability. For safety reasons, the OCL system is not currently shut down separately for each direction of travel at the request of the emergency response authorities. There is still great potential here to reduce the downtimes caused by incidents. Furthermore, the electrified sections should not be so large that, in the event of a necessary downtime, only short sections are affected; however, the length of such an electrified section depends on many factors and cannot be specified in general terms at this point. In addition to incident-related downtimes, however, maintenance-related downtimes and work site-related downtimes still have to be considered particularly in future operation.

Based on our research of operating the OCL system in the ELISA field trial, we assumed an availability of more than 98% for future operation. However, such high availability requires the possibility of a separate shutdown of the respective direction of travel. Furthermore, it remains to be seen how weather conditions will change. If the probability of extreme weather events increases, this will be accompanied by negative effects on availability. On the other hand, the potential exists that expansion of autonomous driving will reduce the number of incidents on the highway [58]. This reduction, on the other hand, would have a positive effect on availability. The expansion of the system creates additional redundancies. Planned shutdowns can then be implemented in a network-optimized manner to reduce disruption to users. Current developments in the transport sector can be integrated positively into the OCL system design.

6. Conclusions

Our results provide an initial insight into the general availability of an OCL system on public highways. The availability achieved in the field test has been presented and described and provides an important reference point for decision-makers to evaluate the suitability of such a system. With reference to the study’s research questions, we can conclude the general availability depends on launch duration and system expansion. The operating mode also has a significant impact on availability. In the initial phase under consideration, particular teething problems limit availability, whereas after a system expansion, we have to consider maintenance-related downtimes, incident-related downtimes, and work site-related downtimes. With data from the pilot operation, we can derive an availability of over 95%, with 98% possibly achieved in future operation. The implementation of an ERS, specifically an OCL system, is therefore a suitable option for electrifying heavy goods vehicles and achieving climate targets.

In order to increase availability and functionality, we proposed the OCL infrastructure to be extended for the deployment of rescue forces in such a way that a safe separate shutdown of the affected direction of travel is also made possible. This would produce considerable savings in downtime. In conclusion, we found only a few malfunction-related downtimes during operation. In order to be able to achieve an increase in availability, future investigations must directly address external events.

Finally, it should be noted that an ERS is one, but not the only, way to achieve climate neutrality. For each technology, there are essential areas of application and opportunities for improvement [61,62]. Only through the interaction of all possible technologies will we make a decisive contribution, especially in the transformation process, which is why research must continue to cover all areas in order to make appropriate technology comparisons.