We’ve noted increased interest in an iron making process known as HiSmelt, which appears to have been recently resurrected in China.
Some shareholders have asked us for our thoughts on HiSmelt and how our own Matmor process compares.
Comparing technologies can be difficult due to a lack of reliable, sufficiently detailed, publicly available information, so this is by no means a complete assessment.
In general, technology commercialisation can be a long and expensive road, especially if you’re developing a new iron- or steel-making process.
Typically, it takes an average of 10 years and $100 million to bring a new process to market[i].
For context, HiSmelt has taken about 36 years and over $1 billion to reach the point where it is claimed to be functioning stably and economically.
We’ve seen some notable failures in the race to develop new processes here in Australia. The two most prominent are BHP’s Hot Briquetted Iron (HBI) plant[ii] at Port Hedland, and Rio Tinto’s HiSmelt plant at Kwinana[iii].
The driver for their development resulted from a requirement to establish a downstream mineral processing plant in return for mining approvals by the Western Australia government.
BHP had lost a reported $2.5 billion on its HBI plant when it finally closed in 2005. Aside from process problems[iv] and safety issues[v], reports[vi] identified various elements of the design and development process which had resulted in out of spec product.
And while the BHP HBI case study is interesting, this article is about HiSmelt.
Beside government policy, what drove HiSmelt development in particular? Apparently, Rio had a lot of high-phosphorus iron ore. Removing phosphorus via traditional methods was uneconomic at the time and HiSmelt promised to liberate $30 billion[vii] worth of this otherwise stranded resource.
HiSmelt development went through several scale steps, with a one-tonne per hour plant being built in the mid-1980’s. Rio (then CRA) then spent $105 million ($195M in 2017 dollars) developing an 11-tonne per hour plant at Kwinana in 1992. That plant was the basis for research and development for seven years until 1999, during which it encountered significant technical issues. As a 2006 article in Forbes[viii] mentions:
Trials at the second plant lasted for seven years until 1999, during which HiSmelt hit a wall. Its original design was for a horizontal furnace, but that design didn’t scale up well. The engineers substituted a water-cooled vertical design that performed as promised, delivering a continuous flow of pig iron at 2,600 degrees Fahrenheit (1426°C).
What the article doesn’t mention is our very own Keith Henley-Smith was the man who first advised Rio to change the orientation of the HiSmelt furnace.
This eventually resulted in a sufficiently stable operation and led Rio to attract three new investors into a joint venture to develop the process to an 800,000 tonne per annum plant at a cost of $1 billion; Nucor took 25%, Mitsubishi 10% and Shougang Steel 5%.
As the Forbes article highlights, the aim was to produce ‘pig iron’ at a 20% saving compared to traditional blast furnace production, but plant stability and economics didn’t pan out as hoped.
By 2008 Rio had written off most of the carrying value of the plant, most of the 130 employees were made redundant and the plant had been put into ‘care and maintenance’ mode[ix].
The decision to shut down completely was made in 2011, driven by the ongoing technical issues and poor economic outlook[x].
Rio then signed an agreement in 2011 with Indian steel maker JSPL[xi] to sell the HiSmelt plant and relocate it to Orissa.
After announcing the India deal (but making little subsequent progress) Rio signed a new deal with a Chinese consortium[xii].
And while Rio has been silent on how much of the $1 billion research and development investment they’ve been able to salvage, the move to China seems to have resulted in the resurrection of HiSmelt following further process development.
So, how does HiSmelt compare with Matmor?
A source quoted in a recent LinkedIn post, who claims to have had recent direct experience with the HiSmelt project in China[xiii] says HiSmelt is on par with blast furnace cost of production in China.
Originally, HiSmelt was projected to be 20% cheaper than blast furnace production in Australia.
This gap indicates the process has not met original performance parameters. Given HiSmelt uses lower cost thermal coal in place of metallurgical (coking) coal and iron ore fines in place of premium lump iron ore, this suggests HiSmelt’s capex and non-feedstock opex may be higher than for a blast furnace.
The source also gives an account of the challenges encountered in achieving steady-state operation.
As mentioned, publicly available data on the cost of HiSmelt is difficult to find. We only have this anecdotal reference that suggests it is on par with a blast furnace, which we assume is in the context of Chinese economic and labour market conditions.
As part of developing our India project for Coldry and Matmor we conducted a Techno-Economic Feasibility (TEF) Study with NLC and NMDC[xiv]. The study indicated a better IRR for a new Matmor facility compared to a new Blast Furnace, so it may also be the case compared to a new HiSmelt plant, if the above source is accurate and our R&D outcomes match projections at subsequent commercial scale.
An update to our TEF financial snapshot was provided to the market in December 2017[xv] to illustrate the advantage of Matmor in being decoupled from iron ore and coking coal price increases.
The chart, above left, shows the compelling IRR for the proposed 500,000 tonnes per annum Matmor plant compared to a new Blast Furnace of the same capacity. The prevailing price of coking coal and premium lump ore in early 2016 were at a low. The chart, above right, shows the improvement in the business case for Matmor as traditional blast furnace input costs returned toward historic averages, driving BF IRR into negative territory, despite corresponding firming iron prices. Since then, the gap between premium lump iron ore and fines has widened and coking coal pricing has increased, further improving the case for Matmor.
In short, we believe that following successful research, development and commercialisation outcomes Matmor will be a better choice than HiSmelt in markets with lignite for the following reasons:
- Lower operating temperature: Lower temperature typically means lower capex, higher efficiency and lower operating cost
- Less complex furnace and auxiliary equipment design: Lower complexity typically means lower capital cost and lower risk of faults, downtime & repairs
- Lignite: Matmor can use the cheapest coal, lignite, resulting in lower raw material costs
- Lower CO2 emissions: Matmor is expected to emit significantly less CO2 than HiSmelt, but it’s difficult to quantify with the limited available data.
We have no doubt HiSmelt will be considered for deployment in places like China where the right combination of raw materials are available, and where the economics are equal to or cheaper than alternatives. We believe the same rationale will apply to Matmor in locations with lignite and iron ore or iron oxide waste such as millscale and nickel tailings.
China has over 7 billion tonnes of proved lignite reserves[xvi] and past interest in Matmor, subject to demonstration at scale, makes it a highly attractive and prospective market.
The issue of process complexity should also be considered in any comparison. In the above LinkedIn article, the writer highlights the consequences for HiSmelt of a simple pump failure, resulting in significant downtime to return the furnace to operation. By contrast, a failure in any component of the Matmor furnace simply results in discharge of the material, repair of the faulty component and return to service.
Matmor is the only lignite-based iron making process we are aware of. HiSmelt can’t use lignite, and as such is not a direct threat to Matmor in providing a solution for lignite resource owners.
Following commercialisation, we expect Matmor will allow customers to decouple from the metallurgical and thermal coal markets, increasing resource diversity and security. HiSmelt is tied to the thermal coal market.
We anticipate Matmor’s unique features will result in such benefits as lower capital intensity and operating expense compared to HiSmelt, providing a convincing business case for broad adoption in markets with cost-effective access to appropriate lignite reserves.
[i] HATCH Associates Report 2009