Coal price forecasts: Why we wouldn’t be building a coal power plant anywhere

Forecasts and basis

ITK has made minor adjustments to its coal price forecasts. We will use these to update our NEM electricity price forecasts, which are available to ITK subscribers.

We have lowered our long-term exchange rate to AUD 1 =USD 0.72 from 0.75 previously. This pretty much offsets a small reduction in the US$/t price for coal assumed.

Our model essentially assumes China and India do not dramatically change course prior to 2030. We think that climate change will probably catch up with these countries by, say, 2025 but don’t have enough confidence in that to put it into our official forecast. That said, I wouldn’t be building a coal fired power station anywhere.

Figure 1: New v old forecasts NSW thermal coal. Source: ITK

Our forecast is built up as follows:

Figure 2: NSW thermal coal price forecasts. Source:ITK

High coal prices mean high electricity prices and provide headroom for new renewables investment and lower electricity consumption. However they also encourage new coal mine investment. Low coal prices work in the opposite direction.

ITK’s view is that thermal coal prices will be driven by Asian power demand; above all else by what China and India do. European and USA coal demand will fall structurally.

China can increase domestic production, India can’t easily do that. China can keep on making short-term investment decisions in coal mining and coal generation and China will do what it has to, to keep its economy ticking and leader in power. Still, we expect coal investment to be a real grudge investment as the economics won’t be good.

In any event, as we also noted, Japan spec (6300 kcal/kg GAR) coal prices have dived over the past few months. This is partly due to soft demand, firstly in Europe and just last month in China, and partly due to increased supply. In other words: high prices have, as ever, been the cure to high prices, and the market is working.

The price in A$ terms has declined $17/t in June 2019 alone, most of which occurred in the space of just three days, putting some life into the old market adage “up by escalator down by the elevator”.

NSW coal price forecasts

NSW coal generators pay a mix of long-term contracted prices – most of which expire over the next few years – short-term contracts, and some spot purchases.

Volumes also have to fit in with train-line availability, which can at times be constrained.

Over the past couple of years there is clear evidence of strain in NSW.

Firstly, Sunset Power, owner and operator of the 8TWh per year Vales Point B generator ended up buying out the Chain Valley coal mine because that mine had gone broke. The mine or variations of it is very old and at the price it was selling to Vales Point couldn’t find the cash to keep up the opex. Owning the mine exposes Sunset to the longer-term closure and remediation costs it doesn’t have with the Vales Point B station itself, as that goes back to the government.

Secondly, Mt Piper – another 8TWh station – has the Springvale 2 mine as its sole source of supply. That mine has been controversial, to say the least, from the beginning and, were it not for specific NSW government legislation, would have had to close while its water system was remediated. It’s becoming obvious that one reason Banpu didn’t want to spend the money is that the mine itself has issues and these are restricting Mt Piper output, at least in the short term.

As with all large business, the lack of long-term confidence feeds through into difficult short- and even medium-term decisions.

For instance, Eraring is due to close in 2032. Its owner Origin is unlikely to underwrite development of a new coal mine, via a long term take or pay contract when the power station has only a 13-year life. Allowing for at least three years of environmental approvals and another couple of years to get operational, such a “mine” wouldn’t be operating before, say, 2024.

So this means Eraring will tend to write short-term contracts with existing mines. The same goes even moreso with Vales Point. Its owners don’t have the balance sheet to guarantee a long-term take or pay contract.

Of course the federal government can always change this. The federal government could just hand some money to Vales Point owners or to CLP, or to Origin, and say ‘go for it.’ It could hand some money to its best buddy AGL and say keep Liddell open, or keep Bayswater open. The fact that this would be stupid from any rational point of view is unlikely to be a barrier.

Stupidity, pig headedness and myopia are the three legs of federal energy policy. Although intervention can’t be ruled out, ITK’s guiding principle is that global warming’s overarching imperative will drive policy over time and so we constantly bet against stupidity. Our bets don’t always pay off but that’s the way we play it.

The 15 per cent discount in the new coal contracts reflects the fact that NSW tax payers put up with low quality coal. Not good enough for Japan or even China but plenty good enough for the residents of the Hunter Valley and Central Coast.

By comparison, here are consensus forecasts for Japan spec export thermal coal.

Figure 3: Consensus coal price forecasts. Source: KPMG

Right now the spot price is trading around the low end of consensus.

Figure 4: NSW thermal coal, domestic v exports. Source: NSW Govt

Here at ITK we put together a small table that tries to allocate thermal coal by mine to individual NSW generators. We wouldn’t claim any great accuracy for the table – it’s just a “water colour” sketch.

Figure 5: Mines and generators. Source: ITKe

The coal market is global but our focus is Asia

As with a number of global commodities, seaborne trade is only a small fraction of the total coal supply market. For instance China imports about 280 mt of coal of all sorts compared to domestic production of around 3.5 bn tons.

Still thanks to the work of the “Coal Tracker” team (previously coalswarm) the global industry trend is summarized as follows:

Global new coal generation additions is now running at say 20GW a year, down from 40GW a year. There are many reasons for this, one of which is certainly cheap gas in the USA. Still, the familiar picture arises of China and India continuing to drive up global carbon emissions. Roughly speaking, a MWh requires, say, as a global average for new plant, about .38t coal.

Capacity utilisation for coal plant in China is about 50 per cent and in India currently about 70 per cent, but falling. Lets say 55 per cent overall for the new additions. This gives around 38-40 mt of new thermal coal required, a relatively modest amount.

Of course, subsidies still exist for coal generation. Large subsidies still exist. The following table shows just the international donation program, ie China. Its perhaps no surprise to see so much Chinese support to Pakistan.

Figure 6 Source: G20 Coal subsidies. NRDC, IISD, OCI

Although the analysis is partial, at best, there are four countries to really think about: China, Japan, India and South Korea

China: we have discussed before. Imported coal is government-limited now at about 280 mt. Therefore, the only way to expand coal generation is more domestic Chinese coal mining. That represents a reversal of previous policy, the coal is unlikely to be as high quality as the imports from Australia and in the end will represent more wasted investment. Increasingly, rather than looking forward, China policy is starting to look very 20^thcentury.

Japan: is a divided country. Japan secretly and not-so secretly is one of the world’s ring leaders in financing coal investment in developing Asian Power.

South Korea: Today South Korea appears to be moving away from coal. Most relevant to this discussion, coal imports fell 9 per cent in the first four months of this year and the government plans for renewables to go from 6 per cent to 20 per cent by 2030 even while reducing the 25 per cent share of nuclear. Coal supplies 42 per cent of electricity in South Korea. Plans to refurbish old plants are on hold but there are still 7 new plants under construction.

Finance

Once again the environmental movement is well organised and Banktrack.org using largely Bloomberg data now tracks the annual lending by individual banks to fossil fuel companies. Overall JP Morgan’s support for the oil industry makes it the number 1 financier of fossil fuels but our focus here is on coal.

Overall global coal mining and coal generation business is flat in the 2016-2018 period running at what I see as a relatively modest US$45 bn per year.

Figure 7: Bank lending to the global coal and coal generation industry. Source: Bank

If we turn to the individual banks, and here we look only at the top ten – accounting for about 75 per cent of the lending – we find that Chinese banks in the first four spots are responsible for about 59 per cent of global lending to coal mining and coal generation.

Pleasingly, no Australian bank is to be found anywhere in these lists.

Figure 8: Banks lending to coal and coal generation industry. Source: Banktrack

Thermal coal opex and capex

Coal opex is high compared to capex. And a large proportion of opex is often fixed: rail and port take-or-pay contracts can run for 10 years, and fixed price mining contracts can run for two to five years. This makes coal miners’ response to prices “sticky” over the short term, as the relatively low variable opex component provides an incentive to maintain output as prices fall.

But once fixed-price logistics and mining contracts expire, operators become far more sensitive to prices, and output can respond rapidly to changes in price. In the event of sustained lower prices over several years, the operations at the top end of the industry cost curve will enter care and maintenance and, eventually, shut down.

As a reference NERA [National Energy Resources Australia] produced in Dec 2016 a “Coal Industry Competitiveness Assessment”.  Their cost analysis in a global context gets you to around A$ 78/t but we think royalties would have to be added.

We pushed up their 2015 numbers for inflation but also bumped up the development capex (expressed as $/t of lifetime production) to allow for higher and ever increasing environmental pressure.

As we have previously written we can also look at Banpu (Centennial’s) costs and the publicly listed Whitehaven Coal which runs a selection of mines including Maules Creek in the Gunnedah basin.

Those numbers are roughly based on our (quick) examination of recent accounts:

Banpu’s portfolio is dominated by underground operations, while Whitehaven operates a mix of surface and underground mines. Royalties in NSW are 8.2 per cent of “value” of open cut production and 7.5 per cent for underground. ITK’s view is that a 10 per cent return on equity based on working these numbers through gets to about A$100/t and this supports our long term coal price.