Is virtual power plant practical?
2024-08-14
한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina
Chao News Client reporter Hu Jingyi correspondent Chen Lisha
The virtual power plant, which does not require a factory building and does not emit smoke, has turned from a concept into reality.
Recently, Zhejiang organized the first virtual power plant this summer to participate in peak shifting of the power grid, which also marked the official launch of the province's virtual power plant management platform.
"Virtual" means that the entity cannot be seen. This special power plant is composed of thousands of charging piles, air conditioners, new energy storage, self-provided power plants and other resources. They are scattered in thousands of households and gathered together to participate in grid dispatching with the support of digital technology.
Second, incomplete power generation. For example, air conditioners, data centers, and landscape lighting are all pure electricity users. Charging piles and new energy storage can discharge electricity, but it is all electricity stored in electric vehicles and batteries in advance. Only distributed power sources and self-provided power plants can actually generate electricity.
Zhejiang Province Virtual Power Plant Management Platform. Image provided by State Grid Zhejiang Electric Power
Before the peak summer season this year, Zhejiang built the first virtual power plant management platform covering the entire province, and plans to strive to achieve a resource aggregation scale of 4 million kilowatts this year. By the end of the 14th Five-Year Plan, the resource aggregation scale will strive to reach more than 5 million kilowatts.
Electricity consumption should be staggered
Why build this kind of power plant?
In recent years, high temperatures and industrial development have led to a tight power supply situation in Zhejiang in summer. In July this year, the province's highest power load increased by 11.9% year-on-year, and the highest residential air conditioning load reached 21.66 million kilowatts, which is equivalent to a maximum of 18.05 million household air conditioners turned on at the same time in the province.
In order to meet electricity demand, Zhejiang generates electricity at peak capacity within the province and strives to purchase electricity from outside the province.
However, such peak power consumption moments are rare. For example, in 2023, the maximum power load in Zhejiang Province exceeded 100 million kilowatts, and the peak time that reached or exceeded 95% of the maximum load was only 33 hours. It is obviously uneconomical to expand power plants and power grids based on peak loads.
Zheneng Lanxi Power Plant. Image source: Visual China
Instead of blindly increasing electricity consumption, it is better to save electricity scientifically. If users can be informed during peak hours and encouraged to reduce electricity consumption or use electricity alternately at specific times, peak shifting can be achieved. Since 2018, Zhejiang Energy and Power Department has begun to work on load management.
At present, electricity users, new energy storage and other entities can participate in demand response on their own, that is, accept the grid's demand invitation and reduce electricity consumption at peak times on their own. Due to the price difference between peak and valley electricity, participating entities can also reduce electricity costs.
As user types become more and more diverse, including flexible charging new energy vehicles, various air conditioners in public places, small-scale energy storage for corporate use, and distributed photovoltaics installed on rooftops, the provincial dispatching center finds it difficult to issue instructions and make precise adjustments one by one because they are small and scattered. This undoubtedly increases the burden on power grid dispatching.
Ensure supply and reduce costs
Changing your ideas and bringing them together will unleash tremendous energy.
At present, the provincial virtual power plant management platform has aggregated scattered load resources such as more than 25,300 air conditioners, 4,803 charging piles, 818 new energy storage, 81 data centers, 581 landscape lighting, 87 self-provided power plants and 250,200 distributed power sources in the province. They belong to nearly 20 virtual power plant operators.
In this peak shifting, a total of 17 virtual power plants in the province participated, with a maximum response load of 1.323 million kilowatts and an average response load of 1.003 million kilowatts. That is, at peak times, virtual power plants scientifically dispatched subordinate resources to reduce power consumption, increase power generation, and reverse discharge according to instructions.
Zhefeng Energy Storage Power Station. Image source: Green Power Ganpo WeChat Official Account
On the evening of August 5, the Quzhou Xin'an Energy Virtual Power Plant integrated fragmented power resources such as electric vehicles, energy storage, and distributed photovoltaics to shift peak demand.
The Zhefeng Energy Storage Power Station in Quzhou is one of them. The power station shifted its 14,000-kilowatt peak load to the valley period from 23:00 at night to 8:00 the next morning, helping to reduce the pressure on the power grid by joining the virtual power plant. At the same time, based on the price difference between peak and valley electricity prices, this adjustment can reduce its electricity costs by about 11,900 yuan.
Energy conservation is the fifth energy source, and virtual power plants participate in ensuring supply during the peak summer and winter seasons in this way.
In addition, it can use its flexibility advantage to deal with new energy sources with unstable properties and promote the consumption of new energy electricity. Therefore, it also has the green attributes of reducing dependence on fossil fuels and releasing land resources.
AI as Commander
The key problem is how to awaken these small and scattered load resources? Behind the "1.0 version" of Zhejiang Virtual Power Plant is a national key research project team from Tsinghua University.
After signing contracts with various resources, virtual power plant operators will install monitoring equipment on electricity users as needed, and realize package management by aggregating power data. These data are then connected to the provincial virtual power plant management platform and connected to the province's new power load management system, thereby participating in demand response and other services.
Different resources have their own characteristics in terms of regulation capabilities, costs, etc. For example, distributed photovoltaics generate a large amount of electricity during the day, self-contained power plants mostly cooperate with the start and stop of corporate production lines, and air conditioning is closely related to ambient temperature.
The coordination of various resources can not only form scale, but also achieve internal complementarity, achieving the effect of 1+1>2.
State Grid Hangzhou Xiaoshan District Power Supply Company put into operation V2G bidirectional charging and discharging equipment in Qianjiang Century City Park, which can be connected to a virtual power plant. Photo by correspondent Yao Jinglin
Among them, algorithms, 5G communications, the Internet of Things and other technologies are indispensable. The Tsinghua University team has formed a topic on AI virtual power plant technology, allowing AI to achieve accurate calculations in terms of decentralized resource aggregation, regulation, and time-space combination. At the same time, blockchain technology is used to achieve decentralized links between resources, making virtual power plants safer and more reliable.
These theoretical research results were first applied to the Zhejiang Virtual Power Plant Management Platform and will be continuously iterated into Version 2.0 and Version 3.0 during the demonstration operation process.
In short, the virtual power plant is like a contractor. It not only solves the problem of the power grid being "out of control", but also solves the problem of users being "not large enough and not eligible to participate" or "not understanding the rules and not knowing how to participate", saving time and effort for both parties.
Market to stimulate
In fact, virtual power plants have only been around for a short time. The concept first appeared overseas in 1997, but was not introduced into the domestic energy internet system until 2016.
In Zhejiang, in recent years, many places such as Hangzhou, Ningbo, Jiaxing, and Jinhua have begun to explore regional virtual power plants and established corresponding subsidy incentive mechanisms.
In order to match the responsiveness of a "real power plant", virtual power plants need to continue to expand their scale, which requires the participation of more entities.
The attraction comes from the electricity market mechanism. In addition to participating in demand response, virtual power plants can also participate in the electricity spot market and ancillary service market, trade their power generation and regulation capabilities, and obtain market-based benefits.
Hangzhou Chun'an Economic Development Zone, cherry photovoltaic greenhouse agricultural photovoltaic complementary project. Image source: Visual China
Since last year, the peak-shaving and load-regulating auxiliary service trading market has been opened to third-party independent entities known as "virtual power plant version 0.5".
As of now, a total of 89 third-party independent entities such as virtual power plants and load aggregators have completed platform registration, 74 entities have participated in transactions, involving 29,360 secondary users, and the maximum single-day response load is 431,200 kilowatts.
The difference between "Version 0.5" and "Version 1.0" is that there was no perfect virtual power plant management platform at that time, and the market also had certain start-up thresholds, with a small scale and low incentive intensity. In the future, it will be necessary to continue to improve the mechanism and enhance the continuity of the market in order to connect to the power market.
Looking around the world, Germany's virtual power plants focus on managing distributed energy, aggregating participation in the electricity market and providing power ancillary services. The United States' virtual power plants gain profits by responding to demand in real time and flexibly. Japan's virtual power plants are mainly built on user-side energy storage, and the degree of commercialization is relatively high.
The development of virtual power plants will generally go through the stages of invitation-based, market-based, and autonomous dispatching. At present, the domestic market is mainly based on invitation-based or market-based models. There is still a long way to go to enter the mature market and achieve autonomous dispatching.
"Please indicate the source when reprinting"
