Module prices are on the rise, yet there remains a lot of manufacturing overcapacity available. AECEA itself predicted mid-year that 390 GW of capacity would be reached by the end of 2021, in “stark contrast” to the 160 GW to 240 GW of anticipated global demand, as you put it. How do you see the situation today?
Frank Haugwitz, founder, AECEA: Conditions have changed. Module prices are at a high. Right now, you pay CNY 2.10 ($0.33) to CNY 2.20 per watt, so that has prompted developers or investors to either postpone, shift demand into next year, or just kick off the project with a small phase this year and the larger bulk of the project to be executed next year. At the same time, given this high-price environment we are in right now, some of these manufacturers have started to reconsider their expansion plans. Because at some point, they ask the question, “does it really make sense to keep on building up more production capacities if we cannot pass all these high-price modules to the end users?”
Because the demand has shifted to next year, I don’t see this overcapacity situation as severe as three, four months ago. It has softened. Of course, oversupply conditions will remain in place. I mean, it has always been an oversupply situation, anyway.
Let’s zero in on polysilicon and the rise from CNY 56/kg in April-May 2020 to CNY 270/kg currently. Where do you see these costs next year, as the very exclusive club of just 10 polysilicon producers in China is widened to include as many as 16 or so producers, as you noted in your July 8 briefing?
Right now, the price of polysilicon is about CNY 276/kg. It could be that by the end of next year, provided all projects will execute as planned, China could be home to around about 1 million plus tons of nominal annual polysilicon production capacities. The actual output, given the ramp-up of new production capacity, will not reach this level until 2023. By the end of next year maybe the actual output could be between 700,000 and 750,000 tons, which is indeed an increase. But nevertheless, this will pretty much become available in the latter half of next year.
We may witness a steady, soft price decline. So possibly by the second half of next year or early second half next year, falling to around about CNY 170 per kilogram. That would already be quite a drop by about CNY 100, but compared to CNY 56 in 2020, it’s more than double.
Whether that will translate into a significant reduction of module prices remains to be seen, because of China’s dual carbon policy, power restrictions, and how the pricing of PV components and other materials develops.
You mention power constraints, which are both at the provincial and regional levels. This has left many PV manufacturers without power to carry on with their manufacturing. How do you see this other form of curtailment impacting our industry in 2022?
These kinds of power restrictions are expected to be in place for the foreseeable future. As of now, let’s say the broad consensus is until the end of this year or after the winter period. So, it is basically from April onwards. From the second quarter next year onwards, we are back to normal.
If I put myself into the shoes of the Chinese government and I ask, what do I want? I would like to have a lasting change of my energy mix, so I’m not turning off the switch and then turning it on again. Otherwise, we would have an annual exercise like this. In the long run, Beijing does not want to have these kinds of situations recurring every year. This can’t be. So that’s why the significant and lasting change ahead will be the introduction of so-called time-of-use electricity tariffs.
Distributed solar PV had a rather lackluster performance last year, with just 5.4 GW installed in the C&I segment and 10.1 GW in the residential segment. How do you see distributed solar in China being influenced by time-of-use tariffs?
This year the distributed generation segment, consisting of residential PV and C&I, is outperforming. Time-of-use tariffs means that at a given time throughout the day you pay a certain tariff, including C&I incentives, which used to be subject to a fixed tariff – now no more. In almost all provinces, C&I tariffs are no longer set in stone. They too are subject to this time-of-use scheme and are subject to market demand and supply and other drivers. It is one more driver, why C&I solar will become much more popular: It helps me as a manufacturer to reduce my own factory carbon footprint.
At some point you drill down to the company level in each separate province. Thus far we’ve talked about provincial-wide decarbonization targets. But now we drill deeper and deeper. At some point we will be on a factory level. At that level, the factory owner will be asked how to conserve energy, how to save energy, how to reduce carbon footprint, etcetera.
So, you have this kind of policy switch with this introduction of time-of-use electricity tariffs, that will bring a lasting change in both the consumption and generation of electricity.
As a commercial or industrial user, if you have your own rooftop system, maybe combined with storage, you can schedule your production according to a schedule you determine and not be resigned to the following: “All right. Now I have power from midnight until eight o’clock in the morning.” Often you will be informed at very short notice. I learned that sometimes you have only an hour or two lead time.
We’ve seen similar drivers elsewhere in the world for solar+storage. How much can the tariff structures make an impact on solar+storage’s competitiveness in China?
It is cheaper with this kind of time-of-use electric tariff and the so-called “floating tariff.” Basically, now you have a fixed tariff and a certain percentage of the tariff can float according to demand and supply. It can go up to 20% and sometimes these time-of-use differences between early morning and peak hours are up to CNY 1 and more. That is a lot.
Solar PV is very competitive – that will drive some of the demand for distributed generation. But as well, it brings in the longer term a change of the overall, let’s say, consumption habits.
Looking more closely at China’s end market, this summer AECEA highlighted the new NDRC Circular Economy Development Plan for 2021-25. This plan prioritized distributed solar PV in combination with electrical energy storage as the second most important key project. We’ve been talking about the C&I segment, but doesn’t it appear that residential has really impressed with its expansion?
In the first nine months of this year residential PV installations grew by about 120% year on year. In the first three quarters of this year, already about 11.6 GW of residential PV systems were installed. So, 120% plus, and the C&I segment is still lagging behind. It’s around about 5 GW so far this year, so almost equal to the entire C&I installation in 2020.
Distributed overall, we’ll see that there was massive demand because in August as the NEA basically approved a new nationwide distributed generation PV pilot program. Accordingly, counties shall make sure 50% of government buildings accommodate rooftop PV systems, 40% of public buildings, 30% of commercial buildings, and 20% of residential buildings in semi-rural areas. As of today, out of a total of 2,851 counties, over 670 of them have been approved by the NEA. And if you assume about 200 MW to 250 MW per county, that makes around 130 GW to 170 GW of additional demand between now and the end of 2023.
This is outstanding growth and looming future expansion. But it was the utility scale that formerly led the way in China. What is going on there?
If you talk about large-scale utility ground-mounted systems in, let’s say eastern and central provinces, there is a significant land constraint. The province of Shandong recently released a notice saying no new ground mounted systems shall be registered and approved until further notice. They will look into each and every ground-mounted project installed since June 2018 to investigate the status quo and verify if everything is in compliance and the positive and negative impacts.
But what is on top of all this is that they want to make sure no more agricultural land would be covered by solar PV plants. So you have a conflict with land needed for securing food supply. Is it more important than generating green electricity? Shandong is the No. 1 province in terms of installations. They have installed about 26 GW of PV, around 10% of China’s cumulative total, which is not a small number.
Yunnan also just released a notice regarding the protection of grasslands and forests. Inner Mongolia early this year released a plan to save 52% of the Inner Mongolian territory. It’s a no-go area for future PV system development in order to protect the grasslands in this case. So you have two issues: environmental protection and making sure in the long run you can guarantee food supply. That’s why distributed generation can only increase in terms of importance in the future.
For this year, I expect residential installations to grow to about 16 GW, from 10 GW last year. For the C&I segment I expect around 6-7 GW, from 5.4 GW last year.
Definitely for the remaining years of the 14th Five-Year Plan the share of the distributed segment in the overall PV build will increase.