Quartz sand price rises as polysilicon and modules fall

Two weeks ago we reported that the price of solar crucibles, used in the Czochralski process to produce solar polysilicon ingots, had risen from $1,700 per unit in November, to $2,200 in February, to $2,800 in early March. Back in 2021, the price was just $600. Prices are not reported as frequently and consistently as for other components, but the shortage is only getting worse and April prices will surely be higher.

Two weeks ago quartz sand, of which up to 100 kilograms are used per crucible, rose 50% in a week to $11,000 per ton, then to $12,000 a week later. The highest price reported as 9th March, for imported sand of the highest purity, (only the inner face of the crucible needs to reach the minimum purity requirement of 4N5, 99.995%) was over $20,000 per ton, courtesy of Norway’s The Quartz Company and Belgium’s Sibelco.

Currently there is a “subtle bottleneck” in the solar manufacturing chain in the wafer segment, which is part of the reason for the gently declining polysilicon price. PV Infolink expects wafer production capacity to increase disproportionately to the other segments in the course of Q2, and according to Shanghai Securities, this will result not so much in a renewed tightening of the polysilicon supply, as in a further shortage and price rise for crucibles. In which case – crucibles may soon pass 4% of total module cost, or even more, with the shortage reaching its peak in Q3, after which new production capacity begins to be activated.

This contrasts with the prices of polysilicon and solar modules, with polysilicon falling 5% and 2.5% in the past two weeks, while modules are at $264 per kW as of mid-March per EnergyTrend, down 3.2% from mid-February.

Like polysilicon, the high purity quartz price has been rising since mid-2021, but while the polysilicon price has gone into a (surprisingly slow so far) reverse in 2023, quartz sand and crucibles are still rising.

It takes between 18 and 24 months to expand high-purity quartz sand production, the same time period incidentally as it takes to build a new polysilicon factory. And in a further close resemblance to the polysilicon shortage, Western suppliers are not especially willing to expand their production capacity just to meet a shortage of a few years, and would rather continue a shift to serving electronics-grade customers.

On a monthly basis, production of high-purity quartz sand is estimated at up to 7,000 tons, which is enough for up to 38 GW of wafers – and that output level for wafers was already closely approached in November, when Chinese authorities reported 36 GW manufactured. This rose to only 36.6 GW in December and fell very slightly to 36.4 GW in February. In this context, wafer manufacturers will be able to seize profits much as the polysilicon makers did last year. At the same time the wafer segment’s largest players such as LONGi and TCL Zhonghuan will be able to use their access to large stocks of sand and crucibles, including higher-quality and more durable crucibles, to outcompete smaller wafer manufacturers, resulting in consolidation.

The latest Silicon Industry Branch figures show the polysilicon price at $32 per kilogram, down from $35 per kilogram two weeks ago. Having doubled in the course of 2022, Chinese polysilicon production output is barely growing at all from the 100,000 tons posted in December – with 316,000 expected to be manufactured in Q1, brought to 340,000-tons in total with Western imports. Hundreds of thousands of tons of new production capacity will be brought online later this year, but skewed into Q4, with total output in 2023 expected to reach only 1.5 million tons. This is enough for around 550 GW, so wider than the crucible bottleneck.

We estimate 24,000 tons of demand for high-purity quartz sand per 100 GW of wafer production, so around 120,000 tons sand demand this year, maximum, while a Chinese agency has predicted supply of only 80,000. Another source expects 25,000 tons bought from the West this year combined with 70,000 tons of Chinese production, which it says is enough for 423 GW of wafer production – still less than the latent global demand for installed photovoltaics of 500 GW in 2024. In 2024, it’s expected that the crucible shortage will be about as severe as in 2023.

The two Western quartz sand companies have production capacity expansions underway of 16,000 tons. The supply shortfall for 2023 may be as much as 40,000 tons, but for the specifically highest-quality quartz sand used in the crucibles’ inner faces, the shortfall may be only 13,000 tons.

The big Chinese player Pacific Quartz, which is also the third high-quality producer (though not quite as high-quality) alongside the two European companies, can soon reach 60,000 tons of production capacity. At least three other Chinese companies are planning to join this technically demanding field. Pacific Quartz posted a 274% net profit increase for 2022.

In the manufacturing of crucibles from quartz sand, expansion plans are more generous, lacking the requirement for high-quality mining deposits. Large-scale expansion plans were set in motion in 2022 and production capacity can double or triple over the next three years.

Due to the cost and bottleneck pressure, the quality of crucibles and wafers may have to be compromised, or at least unnecessarily strict procedures and standards maybe be relaxed. This means pulling ingots more rapidly, replacing crucibles less frequently, and more intensively doping the crucibles with lower-purity sand. This would cut against the trend towards higher-efficiency, higher-purity photovoltaics, specifically N-type cells which require as much as 33% more frequent replacement of crucibles to avoid impurities, compared to P-type cells. Even so, crucible usage per GW can be reduced by as much as 20% through these tactics.

Quartz crucibles slowly deform when used in the Czochralski process, which operates at close to the 1,100 degree Celsius deformation threshold of the crucible. A crucible uniformly made of the highest-purity quartz can last for 15 days, but for those which use lower-purity quartz on the other and middle sections, replacement must be made more frequently. Crucibles are very much a constantly expended consumable, part of the solar industry’s marginal cost of production like any other raw material. They are not a one-off speedbump to production capacity expansions. The constant year-on-year growth of the solar industry, now running at 50% annually, must be matched by the supply of quartz sand and crucibles.

One point we raised in last fortnight’s article was that quartz crucible production can definitely be expanded to match the solar industry’s demand for wafers – the question is whether this can be achieved with the existing industrial approach based around mining from a few high-purity deposits around the world. If not, then other sources such as recycling crucibles and makes synthetic quartz would be adopted, which are much more expensive. As we published that article, one quartz crucible company in China, Triumph Technology, stated in response to an investor question that it is already building a 5,000-ton synthetic quartz production line, but further expansion beyond that scale will depend on market conditions. Synthetic quartz costs $10,000 per ton in one recent Chinese pilot line, more than three times the current high-purity quartz price. But this is for a one-thousand times-higher purity and has not been fully scaled up.