Produced by | Big Whale Project Group
Author | Under the rainforest
Header image | Robot Story
In the face of global climate change and environmental protection pressure, with the proposal of China's double carbon 3060 target, both energy production enterprises and energy consumption enterprises are considering how to adapt to the upcoming "new power system with new energy as the main body".
Digitalization is a just-needed technological infrastructure for power systems, so upgrading them is also seen as an important way to accelerate energy change. At the same time, digital intelligent technology is also an important supporting element to achieve the goal of double carbon, a quantitative study by the world-renowned Boston Consulting Group shows that AI applications are expected to reduce carbon emissions by 2.6-5.3 billion tons by 2030, which can account for 5%-10% of the total emission reduction.
This article is a series of special reports on the "2022 Whale List • Industry Digitalization" (click here to review the details of the list), Tiger Sniff interviewed a number of expert judges, presenting the stories and trends of key industries, in this issue, we chatted with Huang Feng, the mentor of The Whale, about energy digital innovation in the context of double carbon.
Huang Feng is the general manager of Baidu Intelligent Cloud Industrial Energy Product R&D, and is currently responsible for the product and R&D management of Baidu's intelligent cloud industry and energy industry, and has a deep understanding of the application scenarios of AI, big data and other technologies in the industrial and energy industries.
The following is a transcript of Tiger Sniff's dialogue, edited and compiled:
Tiger Sniff Survey: Under the background of double carbon, what energy transition scenarios can digitalization provide convenience?
Huang Feng: It took Western countries 200 years to achieve carbon peaking, while China will reach the target in 2030, from peak to neutralization, the United States from 2007 to the target 2050 interval of 43 years, the European Union interval of 71 years, and China only 30 years. At present, our carbon emissions have exceeded 10 billion tons, accounting for nearly 30% of the world's total emissions, so the time is tight and the pressure is very high.
Domestic electricity and industrial energy account for a high proportion, in the next few decades, we have to gradually change from "thermal power generation as the main body" to "new power system with new energy as the main body" to support the double carbon standard, it can be said that the core of all problems lies in solving the energy problem.
The series of artificial intelligence, Internet of Things, big data and other technologies contained in energy digitalization are also constantly bringing breakthroughs to the industry.
Typically, for example, on the energy production side, the randomness and intermittent characteristics of new energy have long been difficult problems in grid access, grid connection and consumption. Through digitalization, it can monitor the fluctuation of power supply, accurately predict the power generation of clean energy, regulate energy storage coordination, reduce the emission of backup generators, and promote the creation of zero-carbon power systems.
On the energy transmission and distribution and consumption side, digitalization can open up equipment, data and computing power across time and space and multiple dimensions, provide large-scale, long-distance transmission services, monitor the supply and consumption of energy in the system in real time, carry out comprehensive energy efficiency analysis and coordinated control, improve energy efficiency, and optimize the regional energy structure, so that distributed energy resources can be better consumed locally.
Tiger Sniff Research: With the maturity of 5G and the addition of new technologies such as the Internet of Things, big data and AI artificial intelligence, what innovative applications can digitalization upgrade?
Huang Feng: The upgrading effect that these new technologies can play is mainly reflected in prediction, optimization, and scheduling.
Through the prediction ability of AI, the prediction accuracy of new energy power generation output can be improved and the uncontrollability can be reduced. For example, by using a prediction model established by deep learning algorithms to predict the power of wind turbines, the wind power output prediction error can be less than 15%, which can effectively promote its large-scale consumption and solve the problem of uncontrollable production of renewable energy enterprises.
Wind power output prediction based on deep learning algorithm
Through the optimization capabilities of AI, the process of fossil energy can be adjusted to achieve as much output as possible with as little carbon emissions and raw material input as possible. Taking thermal power generation as an example, one of the key links is to improve the energy conversion efficiency in the coal-fired power generation process. Through AI modeling + mechanism model, the optimal operating parameter curve of each system operating with the output of the boiler and steam turbine of the thermal power plant can be calculated to ensure the optimal energy efficiency of the whole range.
Optimization of operating parameters of coal-fired power plants
At the same time, through the scheduling ability of AI, it can also reduce transmission and distribution losses, improve efficiency and network stability. Like some large food cold chain enterprises, the logistics network is all over the country, with thousands of departure lines every day, and the pressure of freight vehicle scheduling is large. The digital base can open up the key nodes of order, transportation, billing, and scheduling, comb the business data of warehouses, customers, carriers, and business scenarios such as cold chain and normal temperature distribution. On this basis, we will further use AI artificial intelligence to develop a vehicle-cargo matching platform, combine orders and capacity types, output the optimal vehicle scheduling scheme, improve distribution efficiency, reduce vehicle empty driving rate, reduce logistics costs and carbon emissions.
Tiger Sniff Survey: To optimize the efficiency of intermediate links, do digital service providers need to be very familiar with the specific operation of the industrial chain?
Huang Feng: Therefore, it is inseparable from the close cooperation and collaboration between enterprises and digital service providers. Share the actual transformation project of a water affairs group, the industrial chain of water affairs has raw water, water production, water supply, drainage, sewage, water saving and other links, with wide geographical distribution, scattered business ownership departments, and although the demand for energy consumption optimization is not as good as energy, manufacturing, construction, and transportation, the unit output value has high energy consumption and belongs to energy-intensive industries.
Although the traditional water affairs group uses a frequency conversion pump, it does not predict the demand for regional water consumption, there is an error in the water balance model, the data is inaccurate, and it is difficult to regulate the frequency of the pump in real time on demand with a constant frequency, resulting in large electricity consumption and waste.
Baidu Intelligent Cloud has established a water affairs brain for it, which can precipitate water data, rely on the REGIONAL water consumption prediction AI model, calculate the pump efficiency according to the parameters such as water demand forecast, pump flow, import and export pressure, motor voltage and current, analyze the actual operating conditions of the pump group, give recommended values for control parameters such as shutdown and frequency, and carry out accurate intelligent pressure regulation control.
In addition, the sewage treatment link can also use AI technology to accurately dose and aerate to reduce electricity consumption and drug consumption. It has been estimated that the complete set can reduce the unit energy consumption of water supply by 8%, and reduce the indirect carbon emissions caused by the use of electricity in the water supply link.
Intelligent pressure regulation technology route of a water affairs group
Tiger Sniff Survey: Compared with manual and camera, what are the advantages of new technologies in safety production monitoring?
Huang Feng: Explosion-proof inspection robots are now very popular, that is, in scenarios such as oil and gas storage and transportation, refining and production, instead of inspectors to identify the unsafe state and dangerous behavior of the environment. Basically, an oil and gas pipeline corridor explosion-proof intelligent inspection program can increase the frequency of inspection by 1.5 times, identify 28 types of abnormal risks, identify more than 99% of the accuracy rate, and AI monitoring and identification can give each link a more timely security guarantee than manual.
Tiger Sniff Survey: How receptive are business owners towards digital transformation? What are the relevant suggestions from the perspective of service providers?
Huang Feng: In the context of economic slowdown, enterprises are facing greater economic pressure to survive, if digitalization can not create real value for enterprises, it is difficult for enterprise decision-makers to have the initiative to invest enough in digitalization.
However, from another point of view, this investment is more like "short pain", if it can improve the overall efficiency and competitiveness of enterprises, and even solve the "long pain" problem, then it is worth it, but this requires decision-makers to have strong courage and foresight.
At the same time, the solution of digital service providers also needs to be able to truly face the market demand and solve the core pain points of customers, rather than just "going to the cloud for the sake of going to the cloud", and retreat after building the underlying infrastructure.
Digital upgrading is a long-term and continuous topic, the focus at each stage is different, in the final analysis, the need for digital service providers and enterprises in the win-win situation to continuously achieve iterative upgrading, in order to truly benefit the development of the industry.
"Big Whale List" is an industry list initiated by Tiger Sniff, which aims to continue to look for growing enterprises with vision and continuous innovation with a forward-looking vision, and empower "Big Whale" enterprises from the aspects of communication, customer acquisition, capital and cognition. Tiger Sniff invites industry experts, top institutional investors and head companies to work together to discover the hidden whale.