INTRO: Faced with replacing half the bodies of its 2224 car fleet over five years, BHP Iron Ore worked with Lynx Engineering to develop an innovative design which not only boosts fuel efficiency but also cuts substantially the whole life cost of buying and maintaining rolling stock
BYLINE: Mike Moynan, Alex Cowin, Andrew Stevens and Kris Kilian*
MOUNT NEWMAN MINING, now known as BHP Iron Ore, was the third iron ore mining company to commence operations in the Pilbara region of Western Australia. Mining began back in 1969 with a designed production and rail capacity of 5 million tonnes/year, but was quickly expanded1. Three decades later it stands at 65 million tonnes/year. Relentless pressure for operating economies has seen BHPIO’s cost per tonne hauled halved since 1994 (RG 6.99 p377).
Today, BHPIO’s fleet of gondola wagons totals 2224. With the exception of 248 recently purchased cars, manufactured from 3CR12 stainless steel, this fleet has carbon steel bodies. Most were introduced between 1969 and 1979.
With 79% of the fleet over 21 years old, many wagon bodies are reaching the end of their physical and economic lives. BHPIO has therefore started to introduce into its fleet a new generation car that is lighter, more fuel efficient, more aerodynamic, and has a lower whole life cost.
Fleet maintenance philosophy
The original 1969 cars, 90 of which remain in service, were purchased second-hand from the Oroville Dam project. They have Corten carbon steel bodies. The majority of the fleet was built in Australia by Commonwealth Engineering from Austen 50 carbon steel, with 6 to 8mm plate on the walls and 9 to 10mm on the floor.
Average age of the BHPIO fleet is 24 years, compared to the average age of the North American gondola car fleet which ranges between 13 to 19 years2. Despite this, BHPIO’s performance record compares very favourably with those of other operators around the world.
The most widely used parameter for assessing wagon productivity is net tonne-km per wagon-year. On this scale, BHPIO scores 11·2 million compared with 10·5 for CVRD’s Carajas line in Brazil, 7·9 for Hamersley Iron, and 7·4 for Spoornet’s OREX iron ore haul3.
BHPIO’s maintenance philosophy is based on monitoring the main component parts of an ore car rather than the complete car. Data is referenced to the ore car body but essentially tracked for major components such as wheels, bearings, axles and brakes. This ensures that the differing maintenance cycles associated with each major component are associated with a full history for each item, should any inherent manufacturing or maintenance defects associated with a batch of components or a particular servicing technique occur.
It has also allowed BHPIO to develop unique R&D strategies for each individual car component with the objective of maximising component life. As a result, nearly every component exceeds the original equipment manufacturers’ life expectancy by a significant margin.
The assembly of any one set of components into an operational ore car is not a permanent arrangement; rather, it is regularly re-arranged during the routine maintenance process. Individual components or entire assemblies may be replaced with new or reconditioned components to minimise turnround time for servicing the car. This has allowed BHPIO to view an ore car as having an infinite rather than a finite life.
To support this philosophy, BHPIO has developed its unique Ore Car Component Tracking system4. OCCT tracks components throughout their lives. Every repair is recorded, along with every position they have occupied on an identified car.
This information allows optimal maintenance strategies to be formulated. It allows accurate prediction of when components near the end of their useful lives. Trends can be established, and OCCT facilitates a swift response as soon as things start to go wrong.
The component with one of the longest life cycles is the car body. It also represents the single largest cost at 30% to 40% of the whole wagon. BHPIO has been very successful in maximising the life of its car bodies. This in part explains why the average age is higher than comparable fleets, in spite of operating at the world’s highest axleloads.
This has been achieved through a combination of tried and tested preventative maintenance strategies, and development of a world class, innovative body maintenance management expert system called Car Care5 (box right).
Next generation car body
The need to replace 50% of bodies in the BHPIO fleet over the next five years provided a tremendous opportunity to improve on the current gondola design - which basically has not changed for the last 50 years - and to take advantage of the latest materials for construction.
The intention was to choose a cost-effective design that offered significant operational cost reductions and an increase in carrying capacity. This new design was to replace the old body without the need to modify existing operational parameters, such as loading and unloading facilities. It would fit existing bogies, and be fully compatible with all existing car components. For future proofing, it would be designed to exploit a 40 tonne axleload limit.
Lynx Engineering, building on the knowledge gained through developing Car Care, designed a body that meets all these requirements.
Mild steel car bodies are heavy because the designer must allow for metal loss due to corrosion and abrasion, so this is not the preferred material today. Aluminium is corrosion resistant, but more expensive than steels. Its impact energy resistance is low, and it is notoriously difficult to weld to steel.
Stainless steel - a group of corrosion resistant steels containing at least 12% chromium - has been promoted for coal wagons since the mid-1970s. There are two main types of stainless steel that can be used for body manufacture: austenitic and ferritic.
Austenitic steels are non-magnetic. In addition to chromium, typically at the 18% level, they contain nickel which increases their corrosion resistance. AISI304 (18% chromium, 10% nickel) austenitic stainless steel has all the required chemical and engineering properties but is very expensive. However, BHPIO has trialled AISI301 quarter-hard stainless steel for floors in a number of replated car bodies.
The ferritic stainless steel 3CR12 was developed by Columbus Stainless to bridge the performance and cost gaps between carbon steels and the high alloy stainless steels. Although 3CR12 is cheaper than conventional austenitic stainless steels because it does not depend on expensive elements like nickel and molybdenum, it still has 250 times the corrosion resistance of unpainted mild steel. In atmospheric corrosion testing, 3CR12 has a given corrosion rate of only 0·001 to 0·002mm/year in a marine environment.
BHPIO has been unable to measure material loss in 132 wagons replated with 3CR12 five years ago. The mechanical properties of 3CR12, though not as good as the higher cost austenitic steels, are superior to both aluminium and carbon steels6. Spoornet found no reduction in wall thickness or cracking during a five year test of 3CR12 coal cars7.
So 3CR12 was chosen by BHPIO for its new car bodies. The improved mechanical properties allowed BHPIO to reduce car tare weight by 3·5 tonnes, using 5mm plate for both walls and floors, whilst at the same time increasing the load carrying ability to permit future exploitation of 40 tonne axleloads.
Fabrication properties of 3CR12 are similar to carbon steel in terms of weldability and ease of forming, and it is easier to weld than the austenitic steels8. It develops a polished surface in service which in turn allows for a faster, cleaner discharge with minimal hang-up, and hence carry back.
Queensland Rail found the initial cost for their 3CR12 coal cars to be less than aluminium (14% higher) and 301 austenitic steel (6% higher), but Austen carbon steel was 13% cheaper9.
Similarly, Spoornet found that initial costs for 3CR12 were higher than Corten, but that the total 30 year life cycle cost of a 3CR12 car, including maintenance and initial costs, was two-thirds that of a Corten car (Table I).6
To overcome the higher first cost of a 3CR12 body, the car designer was asked to minimise manufacturing costs. As a result, the new body will be around 8% cheaper to manufacture than the existing design, using the same material.
The ruthless quest to both reduce cost and greenhouse gas emissions has resulted in fuel consumption currently standing at 0·83 litre per tonne of ore delivered to Port Hedland, a reduction of 30% since 1991.
In spite of this, in one year an ore car will ’consume’ twice as much money in fuel usage as in maintenance. The current gondola design is a ’brick on wheels’, and a prime target for aerodynamic improvement even at the modest speeds achieved by BHPIO trains.
The new aerodynamic design has reduced the drag coefficient by 19% at 75 km/h, confirmed on both full scale cars and on wind tunnel scale models. Welding the car walls on the outside of the body framing not only improves aerodynamics, but also allows an additional 10 tonnes (8%) of iron ore to be carried without any modifications to the existing loading facilities.
Total fuel saving predicted for the new cars is 18·3%. This is made up of 9·5% from better aerodynamics, 7·4% due to the additional payload, and 1·4% due to lower tare weight.
Testing the new cars
BHPIO took delivery of three prototype car bodies in December 1998. They are manufactured from black 3CR12 steel, meet applicable AAR and ARA criteria, and include internal conduits for ECP braking using either radio or wires.
Mounted on existing bogies, the new cars have been extensively tested in the workshop, and under all the operational conditions they will experience. They were instrumented with transducers to assess their dynamic responses, to confirm finite element analysis results, and to gather enough information to confirm their expected operational life.
Wind tunnel tests on scale models were performed to supplement full-scale aerodynamic tests. Thorough analyses were carried out to check every important aspect of the new car, from its structural integrity under various types of loading to its dynamic behaviour on track. Among the parameters checked were: