Algorithms to make energy smarter

Digital technology means that the energy we use in the future will be smarter, more distributed and harder to predict. One big change expected is the ability to operate and manage energy assets as flexibly as possible. That is the focus of AutoGrid, a start-up Total recently invested in. Its stake in the fledging business not only lets it embrace the use of digital tech in the energy industry now, it will help the company produce better energy in the future. Read on for a fuller analysis.

The internet may need energy — it now eats up an estimated 2 percent of the power generated worldwide1, but energy needs the internet, too. Pressured by global warming, ever-scarcer fossil fuel resources and various types of pollution and emissions, the future face of energy is being drawn now. It incorporates such key factors as saving energy, energy efficiency — smarter, more frugal consumption, if you will — and more energy from renewables. All while bearing in mind that the International Energy Agency's 450 Scenario2 still assumes a majority fossil fuel energy mix in 2050.

Big data pivotal to the energy challenge

So 100 percent renewable energy, at least at the global level, isn't coming any time soon. But the expectation is that the management of the world's fast-changing energy supply will be optimized by improving its day-to-day functioning in a way only digital technology can support. The best-known example is smart grids, which will balance energy supply and demand more efficiently, when and as needed, using the sources best suited to the purpose.

Can flexible management solutions help deliver clean, affordable and reliable energy worldwide? Watch this video to find out more.

But smart grids aren't the only application of digital to energy. It will take unprecedented technological innovation to create the more distributed, harder-to-predict global energy industry of the future. Even in so-called "hard" sectors such as extraction and production, digital technology can play a significant role, if only to control inputs and outputs.

Increased efficiency will also require embracing big data. That means acquiring and processing huge volumes1 of data, produced in real time through communication among often very disparate objects, and using them to operate and manage grids more efficiently.

• For more on the subject: Pangea, High Performance Computing for 3D Oilfield Modeling

Making energy smarter

That is right up AutoGrid's alley. Total Energy Ventures, the Group's corporate venture capital arm, invested in the California start-up in the summer of 2016. Established in 2011, it is one of the most advanced in the world at managing the flexibility offered by assets in the home and by distributed and centralized power grids. "In this still-new and mostly North American market, the company's software solutions have been selling briskly to major energy customers," says Christophe Tonion, Senior Investment Manager at TEV. "Discussions with its executives moved quickly and we won them over by talking about the planned changes in Total's business model. This gave AutoGrid a preview of the opportunities for it in the responsible energy field."

The start-up's solutions are based on three areas of technical knowledge and expertise.

First, it collects and processes mountains of data from highly disparate objects such as production, storage and consumption units. The objects are a grab bag, ranging from solar panels and wind turbines to pumps, individual thermostats and batteries for different kinds of electric cars. This technology applies to all types of energy, whether based on fossil fuels or renewables.

Second, once the big data are collected in a continuous stream and linked via a database, AutoGrid develops algorithms to process it. Their ultimate function is to predict resource behavior and failures and anticipate needs, consumption and availability.

Next comes the phase of optimizing energy assets and their management. In other words, factoring in the specific technical and structural features of each connected object, to control the objects as tightly as possible and make the most of their flexibility.

Lastly, AutoGrid customers can opt to manage their own asset, depending on what they do and the kind of installed technology base they have.

"AutoGrid is a credible, promising start-up operating in ecosystems that are innovative for Total," says Christophe Tonion. "Since our acquisition of an interest in it this summer, the company has won several contracts, beating out big energy names, and grown the volume it manages from 0.5 gigawatts to almost 2 gigawatts. That's pretty impressive when you consider that the market leader is at six."

A $26 billion-a-year market

And the market's potential is enormous. According to Amit Narayan, AutoGrid's CEO: "There's a market worth $26 billion a year worldwide for balancing supply and demand on complex and dynamic power grids. The flexible management approach is completely new and allows us to take full advantage of the growing number of connected electrical equipment. It's clearly an industry that will continue to grow."

But the different markets are nothing alike. Operators range from public monopolies to a slew of private companies. National regulations may or may not allow an energy asset such as a privately owned photovoltaic panel to be monetized. And the practice of financially rewarding demand response (see below) doesn't exist everywhere. Consequently, there's no such thing as a "typical" market. The situation varies by market or region or even within each region. A market such as Hawaii, for example, prefers own consumption — having a household equipped with a production unit use most of the energy it produces itself — to grid feed-in. That's tied to the island geography of the state, which would find it harder to export surplus production. Conversely, California allows producers to feed as much energy as they like into the grid, and sets daily rates.

This state of affairs means that AutoGrid needs a specific solution for each market or customer. The start-up has to identify needs, familiarize itself with the customer's technical, legal and business environment and develop one or more dedicated applications first. It also has to come up with the most universal solution it can, to avoid having to develop a new version every time and racking up additional costs.

• For more on the subject: Total Connects with "Microgrid" Start-Up to Improve Energy Access in East Africa

A strategic partner for Total

Very active in gas and renewables businesses, Total is now getting deeper into the power value chain. From production, the group is expanding into marketing, distributed generation and storage. Its ultimate goal is to make the most of the way its activities mesh in order to offer customers integrated, optimized solutions based on digital technology.

"We need a software layer like the one AutoGrid offers to link our different assets," says Christophe Tonion. "Even though this market still depends on varied regulations and economic factors such as the cost of storing power, tomorrow's power grids won't be able to function effectively without efficient feed-in and distribution management. Our investment in AutoGrid will enable us to learn a technology and to do so early enough in a market that is still new."

Total Energy Ventures' investments recently earned it a Corporate Investor of the Year award in San Francisco at the 2017 Cleantech Forum 100. The event brought together the top 100 most innovative and promising cleantech start-ups, the ones that are the best positioned to solve upcoming technological challenges in the near term. Total Energy Ventures' portfolio of equity holdings includes a fair share of them.

• For more on the subject: Total Energy Ventures Invests in the Future

Balancing supply and demand through intelligent demand response

When power supply and demand are out of kilter, as during peak hours in cold snaps or heat waves, grid managers may ask business or individual end users to voluntarily cut back on their consumption. This is called demand response.

Demand response reduces, or eliminates, the need to resort to gas-, fuel oil- or coal-fired power plants, which can be started up quickly but are bigger polluters and carbon emitters.

The usual way to manage demand response is to ask users to draw less from the grid. Customers may be asked to shut down production equipment, for example, or even to start up a generator on site. Individuals may be incentivized through specific rates "discouraging" them from using power when the grid is stressed.

The other solution, made possible by technologies such as the ones proposed by AutoGrid, balances consumption through smart management. For example, household heating could be shut off for 10 minutes without significantly interfering with user comfort. In industry, a pump, air-conditioning system or battery could be controlled — shut off and restarted — in the same way without compromising site efficiency or safety.

By multiplying the number of connected "things" such as smart meters in private homes, economies of scale would ensure balance and demand response in a way virtually unfelt by users — resulting in lower power bills and a smaller environmental tab.

1 Greenpeace even cited a figure of 7 percent in its Click Clean 2017 report. Its estimate was confirmed by Brad Smith, Microsoft's President and Chief Legal Officer, who says that "datacenters will rank by the middle of the next decade among the large users of electrical power on the planet."

2 The 450 Scenario sets out an energy pathway consistent with the goal of limiting the global increase in temperature to 2°C by limiting concentration of greenhouse gases in the atmosphere to around 450 parts per million of CO2 .

3 Digital data created worldwide is expected to total 40 zettabytes (1021 bytes) in 2020, compared to 2.8 zettabytes in 2012.

This page was paid for by Total. The editorial staff of CNBC had no role in the creation of this page.