Release of Various Substances to the Environment by Regular Railway Operation

Michael Burkhardt, Luca Rossi, Markus Boller

Eawag: Swiss Federal Institute of Aquatic Science and Technology, Dübendorf, Switzerland

Abstract

There is increasing concern on environmental pollution by emissions of the transport sector. Protecting the natural resources requires sustainable measures also for railways. For the first time a study was conducted to quantify all substances released by railway operation and maintenance to the environment. We investigated the relevant sources (e.g. brakes, track-switches, wheels, contact lines, wooden sleepers, vegetation control) and the amounts emitted by normal operation within the whole network of the Swiss Federal Railways SBB . The substances identified are mainly heavy metals (e.g. iron, copper, zinc, manganese, and chromium), hydrocarbons (e.g. polycyclic aromatic hydrocarbons) and one herbicide (glyphosate). Substances are released by specific abrasion processes and leach out from construction material. Heavy metals reach the environment mainly as particles due to abrasion. The release of hydrocarbons occurs, for example, by exudation and leaching at wooden sleepers area-wide or at track-switches point wise. The mass balance and pattern of emissions reflected a spatial and temporal distribution within the railway network. However, the exposure to soil, surface water and groundwater and the fate of substances released to the environment is incomplete. On one hand, the results of our study give comprehensive information to public authorities as well as to other railway companies determining the sustainability and the environmental risk. On the other hand, the presented results are an outstanding basis for selected studies investigating the fate and behaviour of the substances in soil, drainage water and groundwater along railway tracks.

Introduction

Sustainable protection of soil, surface water and groundwater require an environmental impact assessment of pollutants emitted by the transport sector. Knowing sources and pathways of crucial substances, mass flow analysis can be established and subsequent measures for the reduction of undesired impact can be taken. Railway operation is associated with the release of inorganic and organic substances into the environment (Fig. 1) [1,2,5,6,7,9]. Currently, knowledge on emissions of regular railway operation and the fate and behavior of the substances in the environment are scarce comparatively to traffic. It has to be anticipated that substances entering railway ballast and soil may leach to groundwater or surface waters (Fig. 1).

Figure 1: Distribution of substances emitted by regular railway operation and leaching to groundwater.

Due to incomplete information, railway companies are not able to evaluate the efficiency of typical drainage systems for mobilized substances. Up to now, technical reduction measures are uncertain in terms of efficiency and design (Fig. 2). With the lack of essential data and expertise, it seems to be clear that an environmental impact assessment can presently not be fully established. Nevertheless, there are strong indications that environmental the exposure has to be evaluated for several relevant substances [7]. In the context of the EU Dangerous Substances Directive (76/464/EC) and EU Water Framework Directive (2000/60/EC), member states are expected to establish pollution reduction programmes including the control of emissions, discharge and measures. Several substances released by railway operation are mentioned in the directive as priority and specific pollutants (List I, II), e.g. copper, zinc, chromium, and polycyclic aromatic hydrocarbons (PAH). Due to the lack of information of sources and emissions, the pathways are unknown. Thus, railway companies are not able to estimate the environmental fate of these emissions or to assess if quality standards for water and soil protection are met (Fig. 2) [7,12]. In view of sustainable railway operation and maintenance as well as of measures that are economical and specific for different purposes, railway companies have the responsibility to assess the fate of substances released into the environment. Within a first step of an environmental impact assessment, a proper evaluation of emissions from all relevant sources may improve our knowledge. Later on, selected experimental investigations along tracks might replenish the first assessment leading to a reliable impact assessment and to reduction measures. In collaboration with the Swiss Federal Railways (SBB), we investigated sources and substances released in the entire railway network, equivalent to the first step of the environmental impact assessment (Fig. 2). The emitted amounts of substances and their expected environmental fate and behavior as well as the present knowledge on runoff and substance properties were described [4]. To our knowledge it is the first time that all substances emitted by railway operation have been identified and quantified for a whole railway network. The presented emission pattern is focusing on the most relevant sources and substances extracting information from [3] and [4].

Figure 2: Overview of the entire study evaluating the environmental risk of railway operation and maintenance.

Material and Method

Based on an intense review of several surveys, reports, and scientific literature (overall n=50), the relevance of the substance sources was described for the past, the present, and the future taking into account future developments in railway operation and maintenance. The complete list of references can be found in [3] and [4]. The composition of the emission was calculated for every source from the composition of construction materials and products used [e.g. 9,10,11]. The amounts released implicate the material consumption, recycling quota documented and emission factors estimated by SBB. Most of these data are not for public up to now. The release of substances was cumulated for all relevant sources in 2003. Additionally, the results were normalized to the entire railway network (7200 km) and expressed in gram per kilometre of railway track. If necessary, the average frequency of 127 trains per track and

day (reference year 2003) was used to estimate emissions [10]. A few substances linked to railways are judged as harmless in the environment (e.g. calcium, carbon, magnesium, phosphorus, sulfur; Tab. 1: harmless substances) and therefore not displayed in detail here. The influence of weather conditions on leaching rate and deposition was not considered. Accidents and technical disturbances are also not taken into account as we focused on regular operation and maintenance of railway tracks.

Results and Discussion

The results show that currently most the substances released to the environment are metals and hydrocarbons (Tab. 1, 2). The main sources of metals are friction processes at brakes, tracks, wheels and contact lines. The emission of hydrocarbons is related mainly to wooden sleepers, lubrication of track-switches and wheel flanges. Characteristic operation losses are diffuse area-wide as well as at point-sources. The results can be divided into a) operation emissions (depending mainly on the train frequency and type), and b) substance release almost independent from railway traffic. a) Operation Emissions The main source of operation emissions seems to be at brakes. The brakes used by SBB are made of gray iron (G-Brakes), composite (C-Brakes), and iron sinter (S-Brakes) [1,11] . The total consumption of friction brake pads reached 2003 about 2,390 t. Based on the returned amount of 20%, in total 1,912 t were emitted by braking. In detail, the G-Brakes reached 1,670 t (87%), C-Brakes 209 t (11%) and S-Brakes 33 t (2%) (Tab. 1) . Most of the abrasion of brakes is attributed to freight trains (67%) due to the predominant use of G-Brakes in freight wagons (75%). Iron is the outstanding metal, followed by manganese, copper and chromium (Tab. 1). The ingredients of binder in C-Brakes could not be clarified as well as arising secondary transformation products (e.g. PAH) [1].

Source Abrasion Substance Proportion Emission (Material) (t/a) (%) (t/a)

Gray Iron Brakes 1,670 Iron 93.3 1'558.1 Harmless Substances 5.8 97.1 Manganese 0.6 10.0 Chromium 0.15 2.5 Copper 0.1 1.7

Composite Brakes 209 Iron 44.9 94.0 Harmless Substances 32.0 67.0 Binder 23.1 48.0

Iron Sinter Brakes 33 Iron 67.9 22.4 Copper 20.0 6.6 Harmless Substances 11.3 3.8 Boron 0.05 0.02 Tin 0.09 0.03 Antimony 0.01 0.003 Lead 0.01 0.003 Molybdenum 0.01 0.003

Wheels 124 Iron 96.0 120 Harmless Substances 1.3 1.6 Manganese 1.2 1.5 Chromium 0.3 0.4 Copper 0.3 0.4 Nickel 0.3 0.4 Molybdenum 0.08 0.1 Vanadium 0.05 0.06

Rails 400-550 Iron 97.0 390-530 Harmless Substances 1.0 4-5.6 Chromium 1.0 4-5.5 Manganese 1.0 4-5.5

Table 1: Composition of abrasion emitted by three types of brakes, wheels, and rails used SBB in 2003. Harmless substances are e.g. calcium, carbon, magnesium, phosphorus, and sulfur [3].

The contact between wheels and rails generates 124 t of abrasion material at wheels and, depending on the recommended references, 400-550 t at rails (Tab. 1). Due to the significant increase in kilometric train performance, the current losses are expected to be still higher. The comparable composition of tyred wheels and monobloc wheels [11] allowed summarizing both abrasion losses. The composition of wheels and rails are >95% iron, followed by manganese and chromium (Tab. 1). Additionally, during abrasion of wheels copper, nickel, molybdenum, and vanadium occur. Contact lines consist to 99.8% of copper and 0.2% of silver [11]. The average train frequency 2003 lead to abrasion losses of 38 t/a copper and 80 kg/a silver emitted as particulates. The amount is similar to an emission of about 5.3 kg/a copper per kilometre track. On winding tracks and tracks with significant slope the wheel flanges of trains are lubricated by spraying special oils [9]. For this purpose the amount of lubricating oil and grease in 2003 reached 33 t/a and 6 t/a, respectively. About 11 t/a of the lubricating oil is a mixture of different inorganic synthetic oils with an unknown part of additives and solid particles, but without any unwanted heavy metals, PAH and halogen compounds. The other 22 t/a lubricating oil contained harmful perchlorethene and the presence of heavy metals was not declared. Most of the oils and grease used were applied for loss lubrication of e.g. engines, gearing, buffer, and bearing. The amount used reached about 197 t/a and 69 t/a for oil and grease, respectively. Neither data on the ingredients nor on the amount used for different purposes could be found. There was also no information available on emission factors. However, it is not expected that the total mass was released to the environment. b) Substance release almost independent from railway traffic Substances release from sources as corrosion-resistant objects (e.g. poles, bridges), track-switches, wooden sleepers and vegetation control, are almost independent from railway traffic. The investigation of corrosion-resistant steel objects consider a bout 144,000 poles, 650 bridges, and unknown number of platform roofs. The poles were hot zinc dipped for 20 years. The galvanizing consists to more than 99% of zinc and traces of cadmium [5]. The annual amount of zinc and cadmium dissolved at poles reached 20 t/a and 2 kg/a, respectively, resulting in 140 g/a zinc per pole. An unknown number of bridges and platform roofs are still coated with lead. For example, in the area of Zurich about 90% of both objects consisted of lead coated steel [5]. However, the losses of lead are not quantifiable. Within the railway network 40% out of 14,000 track-switches are completely lubricated [10]. Normally, all flexible parts were maintained between twice a week and every second week yielding to a lubricant requirement of 68 t/a. The quantity required for lubrication-free railway switches is nearly negligible. The introduced lubricants were composed of synthetic oils, additives, and inorganic thickener except heavy metals, PAH and halogens. It is assumed that 40% of the 55 t/a lubricants needed for saddles volatized, resulting in a release of 33 t/a to the railway ballast. At Swiss level, 41 t/a hydrocarbons used at track-switches emitted to the environment and 6.8 kg/a at each track-switch, respectively. The proportion of wooden sleepers in the SBB network reached up to 43% corresponding to about 5.1 million single sleepers [6]. Sleepers made from beech are pressure impregnated with 15 kg and oak sleepers with 6 kg creosote. The bituminous mixture creosote contains 200–10,000 different substances [2,6,8]. About 80-85% belongs to PAHs (20-40% belong to the 16 US EPA PAHs), 5-15% to heterocyclic hydrocarbons, 1-12% to phenolic compounds, and about 5% to a variety of heterocyclic compounds. Overall, a creosote stock in wooden sleepers of 65,000 t was calculated containing 13,100 t EPA-PAHs and 130 t phenolic compounds. Components of creosote may enter the environment by volatilization, exudation (Fig. 3), leaching or inside particles abraded [8,6]. Within the lifetime of a wooden sleeper (25 years), about 30% of creosote may be released leading to losses of 5 kg creosote at each wooden sleeper [6]. With respect to the uncertainty of 20-30% of the estimated emission rate the annual emission of creosote may reach in average 990 t/a containing about 800 t/a PAHs and 2.2 t/a phenolic compounds [13]. The 80 t/a EPA PAHs emitted annually are dominated by the volatile naphthalene, acenaphthene, fluorene und anthracene. The low volatile PAHs are exuded or released via particulate matter. In particular, creosote release is strongly non-linear decreasing with time [8]. At Swiss level, a rough estimation of 94,000 wooden sleepers used in 2003 lead to a release of 525 kg/a PAH within the first year and 230 kg/a in the second. The uncertainty of the calculated PAH emission is particularly high.

Figure 3: Exudation losses of creosote from impregnated wooden sleepers in hot weather [3]. Along railway tracks plants and weeds have to be eliminated for security and stability reasons. A common maintenance procedure of vegetation control is the application of herbicides. Just like other European railway companies SBB spread the non-selective herbicide glyphosate. This herb icide is only absorbed by the leaves of plants and it is not absorbed by roots. In 2003 about 3.9 t/a of the active ingredient was applied [10]. The application rate of 2.2-2.9 kg/ha is corresponding to an area of 1'345-1'770 ha, equivalent to 50% of the total railway network. In order to improve controlled dosing, the application is conducted manually with a backpack sprayer since 1999. During regular railway operation and maintenance, various substances are emitted from several sources. The ten most important metals are displayed in Tab. 2. In total, about 2,270 t/a metals, 1,360 t/a hydrocarbons and 3.9 t/a glyphosate reached the surrounding of the SBB railway network (7200 km tracks). Abrasion from rails is the largest source of chromium and contact-lines the largest source of copper, both occurring area-wide. The data set seems to be reliable. Nevertheless, the varying boundary conditions (e.g. exposition, products used and emission rates) cause a broad spatial and temporal variability. For instance, braking and acceleration (e.g. railway stations, signals) increase the average emission values estimated from flat tracks.

Substance Emission Emission per kilometric track

Source

(t/a) (g/km) Iron 2,176 302,000 Brakes >>> Rails > Wheels Copper 46.6 6,480 Contact Lines >> Brakes Zinc 19.8 2,750 Galvanization Manganese 15.5 2,170 GG-Brakes > Rails > Wheels Chromium 6.9 960 Rails > GG-Brakes Nickel 0.4 50 Wheels Vanadium 0.06 8.5 Wheels Lead 0.003 0.5 S-Brakes Antimony 0.003 0.5 S-Brakes Cadmium 0.002 0.3 Galvanization Binder 21 2,900 K-Brakes

Hydrocarbons 1'357 176,800 Wooden Sleepers >>> Loss Lubrication >> Track Switches > Wheel Flange

Glyphosate 3.9 540 Vegetation Control Table 2: Cumulated emission of the most important substances presented for the relevant sources on the entire railway network of SBB.

Conclusion and Outlook

The release of metals to the environment by regular railway operation correlates mainly with particulate matter from abrasion processes like braking, followed by losses from rails, wheels and contact lines (Tab. 3). On a quantitative basis, abrasion of brakes is the largest emission source. Since 2005, SBB started to substitute G-Brakes by C-Brakes in freight wagons. Therefore, a significant decline of the mass emitted by G-Brakes is expected simultaneous with

an increase of substances from C-Brakes. Due to the increasing kilometric performance the release by friction processes increase at the same time. The size distribution and the fate of the emitted particles are still unknown. Due to this uncertainty a risk assessment of the heavy metals including the transport behavior as well as the corrosion and dissolution of particles, respectively, is crucial. The losses of zinc and cadmium leached at poles depend on the exposition and the age of the object. Leaching in the luff seems to be significantly smaller than in the lee and aged hot galvanized poles are protected by a layer of dust particles. It is expected that the total amount of used hydrocarbons reached the environment by leaching, exudation, drip losses, and volatilization. Compared to metals, the mass balance of released hydrocarbons is more arguable. Nevertheless, the wooden sleepers seem to be the most important sources of hydrocarbons and of the toxic PAHs area-wide. Track-switches are well known point-sources of hydrocarbons resulting in polluted ballast. The periodical cleaning of track switches with hot water steam, sometimes combined with surfactants, may foster the leaching significantly [4]. The increasing use of concrete sleepers and lubrication-free switches may reduce the emission of these substances. The herbicide glyphosate is classified as a moderate toxic substance and seems to be less toxic than the herbicides atrazine and diuron used earlier [4] . Nevertheless, compared to soils, the fate and behavior in ballast and track profile with less organic matter and inhibited microbial activity is unclear.

Source of Abrasion Emission Proportion (t/a) (%)

Brakes 1,912 73 Rails 550 21 Wheels 124 5 Contact Lines 38 1 Total 2,624 100

Table 3: Sources of abrasion by SBB railway operation on a mass quantity basis calculated (2003). The main sink of all released substances seems to be the railroad embankment, and a smaller proportion deposit in soil nearby the track. For instance, hydrocarbons occurred in the embankment to the bottom of the track profile of 1 m after regular operation period, and metal contents in soil are increased mainly within 5-10 m distance to the tracks [4]. Although most of the hydrocarbons and PAHs, respectively, are potentially degradable, track profiles are highly polluted after lifetime. The available results of analyses from embankment material indicate a serious exposure and unknown leaching risk via drainage systems to surface water and groundwater [4,9] . Neither the pattern of the substances released nor their environmental impact on groundwater and surface waters are well known. Because it is for the SBB railway network still impossible to predict what will happen when a potentially toxic substance enters the natural environment (Fig. 2), it is essential to establish selected investigations under realistic conditions. To assess exposure, appropriate investigations should keep track of the distribution of the substance and should give information on the fate and behavior of pollutants. The information summarized in this paper and the report [3,4], respectively, is a reliable basis to select a worst-case site for further studies. Finally, the mass flow analysis of the emitted substances might be used to investigate sustainable measures reducing potential environmental impact (Fig. 2). It will be possible to build efficient solutions to protect the environment. These must be confirmed by field studies. Technical (e.g. new track profiles, drainage systems) and operational (e.g. use of lubricants completely degradable) measures have to be evaluated. Overall, the study may help to construct and operate railways in a more sustainable way. Even if the emissions are far away from motorized traffic emissions, railway companies may promote their environmental considerations thanks to our investigations.

Acknowledgements

We thank Fernande Gächter and Helmut Kuppelwieser, both from SBB, for assisting the study and all persons of SBB and engineering companies delivered essential information. The SBB, Swiss Federal Office for the Environment (FOEN), and Swiss Federal Office of Transport (FOT) is thanked for initiating and funding the project.

References

[1] P. Bregy. “Emissionen von Verbundstoff-Bremsklotzsohlen”, Swiss Federal Railways SBB AG, BahnUmwelt-Center, Bern, p. 42, (2004). [2] K. M. Brooks. “Polycyclic Aromatic Hydrocarbon Migration from Creosote-Treated Railway Ties into Ballast and Adjacent Wetlands”, U.S. Department of Agriculture, Forest Service, Madison/WI, p. 53, (2004) [3] M. Burkhardt, L. Rossi, M. Boller. “Stoffemissionen durch Bahnanlagen und Bahnbetrieb”, Der Eisenbahningenieur, 12, pp. 18-22, (2005). [4] M. Burkhardt, L. Rossi, N. Chèvre, M. Boller, L. Steidle, J. Abrecht, F. Gächter, S. Knabl und H. Kuppelwieser. “Gewässerschutz an Bahnanlagen - Emittierte Stoffe im Normalbetrieb der SBB sowie Grundlagen zu deren Umweltverhalten”, Eawag, Dübendorf, p. 120, (2005). [5] Carbotech. “Schwermetallbelastung der Böden in der Umgebung korrosionsgeschützter Stahlobjekte”, Kanton Zürich - Fachstelle Bodenschutz, Basel, p. 30, (1994). [6] M. Kohler. “Gehalte und Emissionen von polyzyklischen aromatischen Kohlenwasserstoffen in und aus teerölbehandelten Holzschwellen des schweizerischen Eisenbahnnetzes”, EMPA, Dübendorf, p. 117, (2000). [7] M. Osborne, M. Montague. “The potential for water pollution from railways”, Report CIRIA C643, CIRIA, London, UK, p. 20, (2005). [8] OSPARCOM. “PAH Emission Factors with Procedural Guidance for the Submission of PAH Emission Data”, Report DIFF 97/7/3-E(L), Norwegian Pollution Control Authority, Oslo, p. 19, (1996). [9] SBB. “Gleisabwässer und Gewässerschutz”, Swiss Federal Railways SBB AG, Bern, p. 48, (1993). [10] SBB. “SBB und Umwelt”, Swiss Federal Railways SBB AG, BahnUmwelt-Center, Bern, p. 15, (1999). [11] SBB. “Unterlagen zur Untersuchung über PM10 - Beilagen 1-4”, Swiss Federal Railways SBB AG, BahnUmwelt-Center, Bern, p. 10, (2001). [12] A. Zysset, C. Hugi, R. Pfammatter. “Wegleitung - Gewässerschutz bei der Entwässerung von Verkehrswegen”, Swiss Agency for the Environment, Forests and Landscape (SAEFL), Bern, p. 57, (2002).