5 applications of artificial intelligence in the energy industry

The energy industry generates a lot of data. In order to turn these data into driving forces to increase productivity and cut costs, major energy industry companies – oil and gas giants, and renewable energy companies – have turned their attention to artificial intelligence.

Since 2012, news of putting artificial intelligence and the energy industry together for reporting has begun to increase. This paper briefly describes the five application directions of artificial intelligence in the energy industry, and the corresponding cases.

Storage

According to a recent Greentech Media report, US energy storage reached a new milestone in the fourth quarter of 2017: between 2013 and 2017, cumulative reserves exceeded 1,000 MWh. The report also predicts that this number will double this year. With the increase in storage capacity and the emergence of new technologies, artificial intelligence is increasing the efficiency of this market.

Case 1: Stem

Stem, Calif., has developed a program codenamed Athena that uses artificial intelligence to map energy usage and allows customers to track energy price fluctuations to use stored energy more efficiently.

Stem has raised more than $37 million from a number of investors, including the US Department of Energy, GE Ventures and Singapore's sovereign wealth fund Temasek Holdings.

The Autonomous Grid

Today, there are many sources of energy for the grid. In addition to conventional power generation, there is wind and solar energy, which complicates the process of operating the grid system. The artificial intelligence is used to analyze large-scale data sets. This multi-source collection process is more stable and efficient.

Case 1: US Department of Energy

In September 2017, the US Department of Energy awarded a research award to SLAC researchers at Stanford University to reward them for using the artificial intelligence technology to improve grid stability. By using past data to program power fluctuations and grid weaknesses, the new “smart grid” will automatically respond quickly and accurately to major events.

Case 2: Siemens

Smart grids can also better manage different types of energy at the same time. Siemens has released a software package to operate the network, the so-called "active network management" (ANM). The principle of ANM is to improve efficiency by tracking how the grid interacts with different energy loads to adjust its adjustable components. Although this was manually adjusted before, when new energy producers (such as solar power plants) start working, or when new energy consumers start to access the grid, ANM will adjust the grid accordingly. Therefore, ANM also laid the foundation for electric vehicles to use the smart grid for charging.

Case 3: UK National Grid

In March 2017, DeepMind, an artificial intelligence company acquired by Google, announced jointly with the UK National Grid that they plan to add DeepMind's artificial intelligence technology to the UK's power system. The project will handle massive amounts of information such as weather forecasts and Internet searches to develop predictive models that are experiencing a surge in demand.

Case 4: Grid Edge

A British company called Grid Edge (which provides cloud-based power management software services) claims that they use artificial intelligence technology to predict and optimize energy allocations, returning control to power users. Specifically, the Grid Edge operates a VPN that connects and analyzes the energy consumption data of the user's building. Using this information, the Grid Edge communicates with the connected grid and develops a scheduling strategy. The purpose of these strategies is to save energy and avoid overloading.

Failure Management

In November 2017, a coal-fired power plant in northern India exploded, killing 32 people because of a gas pipeline blockage that caused the boiler to explode. This is a type of failure that often occurs in the energy industry. The cause of the accident is that there is no regular inspection of the equipment, and there are no strict regulations in many parts of the world, so equipment failure is very common. Using artificial intelligence to observe equipment and detect failures before an accident can save time and money and even save lives. Currently, many startups are trying to provide this service to the energy industry.

Case 1: SparkCognition

In December 2017, the US Department of Energy awarded SparkCognition an award to use artificial intelligence to increase the amount of electricity generated by coal-fired power plants. The company combines analytics, sensors and data generated in operations to predict when critical infrastructure will collapse.

Case 2: AES Corporation

In September 2017, US energy giant AES Power announced plans to enter artificial intelligence as a means of improving the company's vigilance, efficiency and protection of its assets, primarily for their solar power plants and grid systems.

Upstream Exploration

Case 1: BP Ventures (British Petroleum Ventures) Beyond Limits

BP Ventures invested in an artificial intelligence company called Beyond Limits. The company has been involved in exploration experiments conducted in outer space. Investing in Beyond Limits, BP Ventures said it plans to use Beyond Limits' oil and gas exploration technology to find new oil reserves.

Case 2: Chevron

Oil giant Chevron is using artificial intelligence to find new wells across California and old wells with added value.

Energy Consumption

By monitoring the energy consumption behavior of individuals and businesses, artificial intelligence can provide solutions to optimize energy consumption processes.

Case 1: Alphabet's Nest

Nest, a subsidiary of Alphabet, has developed an intelligent thermostat that can reduce energy consumption by automatically adapting to user behavior. Once Nest is installed in the user's home, it will begin to learn the living habits of the occupants and adjust the temperature accordingly. According to Nest, the company's technology has saved its users 10% to 12% of heating costs.

Case 2: Nnergix

Nnergix, Spain, uses machine learning technology to predict the impact of atmospheric and weather conditions on renewable energy capacity, such as estimating the amount of electricity generated per hour by a photovoltaic power plant.

Case 3: Google Sunroof

Google released a tool called Sunroof to calculate the impact of solar energy on American families. The project uses several factors to calculate how much money can be saved using solar energy, including weather data, electricity bills, 3D modeling, and shadow calculations.