THE BOILER used on John Heald’s Darjeeling and Himalayan Railway B-class locomotive is, in 7.25inch gauge, large, to say the least. Which begs the question: what material is best for such boilers? Copper is not economic, stainless steel has safety problems and the usual material, mild steel, has its drawbacks.

These days you look to Australia for boiler design innovation and approval.

It all started when Les Irwin from the Penwood Railway near Sydney, Australia contacted several experts for advice on an alternate to mild steel without the attendant problems of ordinary stainless steels.  This group included Dr. Ian Ward, metallurgist, Kurt Staub, welding engineer and Robert Brunton, BSc in Metallurgy and University of New South Wales lecturer.

All recommended use of Duplex SAF 2205 steel alloy developed in Sweden by Sandvik Engineering Group as an alloy with a low expansion rate that would resist cyclic stress cracking at the welds, resist corrosion, and be suitable for pressure vessels.

The microstructures of 2205 consist of a mixture of austenite and ferrite phases and duplex steels display properties characteristic of both. Duplex alloy steels have actually been around since the 1930s, but the 2205 alloy is a nitrogen enhanced duplex steel that became available in the 1980s. The nitrogen significantly improves the corrosion resistance properties of the alloy, which also exhibits a yield strength that is more than double that of conventional austenitic stainless steels especially when welded.

Building the B-Class boiler

As discussions went on, John Heald needed a boiler for the B-Class (see article here) and was the ideal testbed for the first duplex boiler. As scaled from prototype, the boiler would exceed the maximum volume allowable for hobby boilers under New Zealand code (50L), so it was decided  to make it shorter, with a false extension at the front to fill the gap. 

After fine tuning the drawings for the boiler to assure compliance with the preliminary 2205 code, the drawings were given to Sandvik NZ where a laser-cut kit-set of parts was produced.   All tube holes including inspection and tapping holes were cut in the various plates by Sandvik to an accuracy of  0.1mm. 

Dennis Collins, with 25 years of experience welding Duplex 2205 steel, rolled and cold formed the various plates and tack welded the parts together.  Dave Giles and Paul Newton inspected all the assembly fits and checked and  recorded  progress on a regular basis. 

John Heald spent several evenings machining the 2205 round bar making the stays for around the firebox and the crown sheet.   Dennis then welded the test boiler components in sequence and the boiler was completed.

Hydrostatic and stress testing

An initial hydrostatic test to 150 PSI found no leaks or weeps, and it was agreed that the 2205 test boiler group conduct a series of cyclic tests on the boiler from cold to 100 PSI pressure (another limit in New Zealand for hobby boilers) and back to cold again via cooldown and then blowdown at 30 PSI.  This was to be repeated 48 times to simulate about 5 years of use and the welds then tested for cracks.  

During testing, it was found that the boiler went from from cold to pop-off at 100 PSI in just 23 minutes thanks to the lighter construction.  An equivalent carbon steel boiler would take about 40 minutes to reach pressure. 

After the conclusion of the tests, the boiler was thoroughly cleaned and successfully crack tested. Special attention was paid to the area around the fire tube plates, and the throat area and foundation ring.

The length expansion of the boiler from ambient temperature to 165 C (329F)  working temperature was only 4.5 mm over the 1 meter length of the boiler, the same amount as a carbon steel boiler of equivalent size.

No water treatment was used at all, and after 48 tests, no appreciable sludge was collected in the blowdown catch containers. As the local water was full of all sorts of minerals, it was expected that calcium could be a problem, but no buildup was detected on internal inspection of the boiler after the cyclic tests.

How does the cost compare?

According to current sources, the costs of producing this boiler are expected to be comparable or a little more than a carbon steel vessel of similar size.  The major benefit is the expected longevity of the SAF 2205 material as compared to carbon steel, plus cost and time savings realised from greatly reduced boiler maintenance and water treatment tasks.  

Long term performance

The boiler, as of early March 2012, has been under steam for 452 hours (including cyclic testing). that is about 10 years worth of steam ups at 4 hours per session once per month.

A recent internal inspection showed the boiler to be in ‘as new’ condition internally after all those hours.


The conclusions reached by the boiler test group are that a SAF 2205 Duplex boiler constructed to the AMBSC part 4 code for miniature boilers will be more than satisfactory provided the overall design and welding procedure is strictly adhered to.  The life expectancy of the Duplex material SAF 2205 alloy steel should be similar to a copper boiler, and is suitable for boilers ranging from 25 – 50 litres (6 to 13 gallons).

For more details of the boiler and of the new Australian code for Duplex boilers go to the website http://sonadaworks.com/  Information for this article and our article on the loco is extracted from that website, with permission.