Virtual Power Plant

China Energy Storage Network News: 1. Background

The virtual power plant is actually not a new thing. It has been proposed for more than 20 years. It emerged in European countries such as Germany, Britain, France, and the Netherlands in the early 21st century and has many mature demonstration projects. It mainly focuses on the reliable grid connection of distributed energy and builds a stable business model in the electricity market.

So why are we hearing more and more about it in China recently? Because our power system and policy market are changing, gradually meeting the environment for the birth of virtual power plants.

Those who pay attention to the power industry know that China is now in the transition period of the old and new power systems. The market players will change from single to diversified, and power transmission will change from "generation, transmission, distribution and use" to "source, grid, load and storage". Especially after the establishment of the dual carbon goals, the addition of distributed energy and energy storage has made the original power system more complicated. The virtual power plant is a relatively cost-effective solution.

For example: In the past, peak shaving and valley filling of the power system was basically achieved through thermal power plants. According to estimates, meeting 5% of the peak load may require an investment of 400 billion yuan in power plants and supporting power grids. However, through virtual power plants, investment in construction, operation, incentives and other aspects only requires 50-80 billion yuan.

2. Concept

A virtual power plant (VPP), as the name suggests, is a virtualized power plant. It is not a physical power plant, but it plays the role of a power plant.

What are the functions of a power plant? One is to generate electricity and participate in the energy market; the other is to participate in the ancillary service market (peak and frequency regulation, etc.) by regulating power.

Therefore, we can conclude that a virtual power plant is a kind of "black box" that can play the role of a power plant. It does not require a new power plant to be built, but can serve as a "positive power plant" to supply power to the system, or as a "negative power plant" to absorb the system's electricity.

So what is in the “black box” that can achieve such a magical effect?

The answer is actually very simple: distributed power sources, energy storage facilities, controllable loads and other resources.

But these things alone are not enough, virtual power plants also need a set of technologies and systems to aggregate them intelligently. Therefore, a virtual power plant is essentially a software system that aggregates existing distributed resources and participates in the electricity market through collaborative control, thereby replacing the construction of new real power plants.

Here we give a complete definition: A virtual power plant is a management system that uses information technology and software systems to aggregate and coordinate the optimization of multiple distributed resources such as distributed power sources, energy storage, controllable loads, electric vehicles, etc., so that it can participate in the electricity market and grid operation as a special power plant.

It sounds like virtual power plants and demand-side response are quite similar, and their essential connotations are also relatively consistent.

In fact, broadly speaking, the virtual power plant is an extended version of demand-side response. The demand-side response is mainly peak shaving, mainly targeting user loads; the virtual power plant takes into account both peak shaving and valley filling, and includes sources, networks, loads and storage.

3. Composition

Next, we will disassemble the composition of the virtual power plant in detail. Let's take a look at a picture first:

Based on the definitions in the second part, we should be able to see the meaning of the picture more clearly.

Since virtual power plants gather distributed energy (power generation), energy storage (charging/discharging), controllable loads (power consumption), etc., they can be classified according to their actual composition:

According to their respective characteristics, we can draw the following conclusions:

1) Power supply type: has the ability to sell energy and can participate in the energy market; and participate in the ancillary service market depending on the actual situation

2) Load type: has power regulation capability and can participate in the ancillary service market; lacks energy selling attributes

3) Energy storage type: It can participate in the ancillary service market and sell electricity by discharging during certain periods of time.

4) Hybrid: An all-round player.

In foreign cases, each country has its own characteristics. Japan and Germany use energy storage and distributed power sources as the main body of virtual power plants, while the United States mainly relies on controllable loads, which have accounted for more than 5% of peak loads.

Let’s expand it:

1) Controllable load

The key areas of controllable load resources mainly include industry, construction and transportation. Among them, industry is divided into continuous industry and discontinuous industry; construction includes public, commercial and residential, etc., and air conditioning load is the most important in the construction field; transportation includes shore power, public transportation and private electric vehicles.

There are large differences in both the quality and quantity of controllable load resources. In terms of quality, they can be judged from the dimensions of adjustment willingness, adjustment ability, and adjustment and aggregation cost performance. In general, non-continuous industry is the preferred high-quality resource with "three highs" of willingness, ability, and aggregation, followed by electric transportation and building air conditioning. In terms of quantity, the development of adjustment and aggregation technology and the reduction of costs are constantly increasing the amount of adjustable load resources.

2) Distributed power supply

Distributed power generation refers to smaller generator sets configured at the user site or near the power consumption site, including small gas turbines, small photovoltaic and small wind power, hydropower, biomass, fuel cells, etc. or a combination of these power generation methods.

From the perspective of virtual power plants, the definition of distributed power generation resources lies in the scheduling relationship. All power generation resources that are not in the existing public system or can be separated from the public system can be included in virtual power generation resources. In this sense, all self-contained power plants are potential resources for virtual power plants.

3) Energy storage

Energy storage is the most revolutionary element in the electric energy industry. The rapid improvement in the economic efficiency of energy storage technology has enabled electricity to break through the limitation of large-scale economic storage and has also changed the industry's control and optimization mechanism.

According to the difference in storage form, energy storage equipment can be roughly divided into four categories: one is mechanical energy storage, such as pumped storage, flywheel energy storage, etc.; second is chemical energy storage, such as lead-acid batteries, sodium-sulfur batteries, etc.; third is electromagnetic energy storage, such as supercapacitors, superconducting energy storage, etc.; fourth is phase change energy storage.

4. Business Model

After aggregating multiple types of resources, virtual power plants must eventually participate in the electricity market to gain revenue. They are generally divided into:

1) Electricity trading

Virtual power plants can act as electricity sales companies and trade directly with users, or purchase power generation rights from thermal power plants.

2) Ancillary services market

Virtual power plants participate in peak load service transactions by reducing output (increasing load) during load troughs or increasing output (shedding load) during load peaks, as well as similar service markets such as frequency regulation.

V. Operation and Case Studies

Case 1:

In 2020, the virtual power plant project in Huangpu District, Shanghai. The largest trial run so far, with more than 50 buildings participating and a load release of about 10,000 kilowatts. How is it achieved? During peak hours of electricity consumption, the system automatically adjusts multiple characteristic parameters such as temperature, air volume, and speed of the central air conditioners of relevant buildings in the virtual power plant area, with little impact on user experience.

There is some workload on the technical level, including control, metering, scheduling, trading, etc. The granularity of previous energy regulation was relatively large. Virtual power plants target the equipment level, and if automatic response is required, the complexity will be very high.

How do users participate in virtual power plant projects and realize benefits? Taking the Shanghai project as an example, load integrators bid on the system platform. There are three levels of architecture: platform, load integrator, and user.

Subsidy prices vary according to response time. If a user performs peak load shaving within 30 minutes, the subsidy will be 3 times the price (there is a base value), 2 times the price between 30 minutes and 2 hours, and lower for longer periods.

The source of subsidies currently mainly comes from the surplus part of the power purchase price difference in inter-provincial renewable energy spot transactions among provinces, so there are still some restrictions, and many provinces have not yet started spot trading.