The energy provider marketplace is rapidly changing. Ongoing deregulation has pushed many utilities to develop new commercial offerings that provide both enterprises and households with increased contractual flexibility. In most European and Asian countries and in the United States, clients can now potentially receive proposals from 10 or more utility providers. The ongoing energy transition also opens the market to new producers who generate their energy from very different sources. Renewable energies such as solar, wind, and hydro complement existing fossil fuel energy resources and now produce up to 25 percent of the total production during certain periods of the year. Renewables address a marketplace need for low carbon power generation and sustainable energy development.
These types of marketplace variables significantly add to the complexity of forecasting energy consumption behaviors across grids. As a result, the ability to forecast energy prices using traditional methods becomes less accurate, and grid operators have a more difficult time remaining profitable. To address this issue, grid operators can now turn to AI-driventechnologies to forecast energy prices.
Energy Consumption Pattern Changes Impact Pricing
Energy consumption patterns have also changed. Enterprises and households now have more control over how and when they consume energy. This new-found flexibility allows them to lower their cost of consumption. In some cases, they may even act as temporary energy producers when using their own solar panels. They are also shifting their energy consumption to periods of the day when energy prices and overall energy demand is lower.
Retailers and offices, for example, control their lighting, and channel the intensity of their heating or cooling to times of the day or night when prices are lower. Residential homes are also starting to use devices like smart thermostats to better control their heating and cooling and can schedule the recharging of their electric vehicle (EV) batteries at night when rates are low. For the energy producers who support these customers, the traditional top/down model of managing a grid is no longer a sustainable energy business model.
Market and Lifestyle Changes Alter Electricity Pricing Mechanisms
In this context of changing regulations, power sources, and consumer preferences, the way energy is priced has evolved. In fact, many regions of the world have developed their own specific demand/supply/pricing rules. Energy is now traded in public market forums like theEuropean Energy Exchange (EEX). In addition to being linked to primary energy prices (i.e., prices for Brent crude and gas), electricity contracts are also priced for periods representing future days, weeks, months and even into the next year–and these prices fluctuate on a daily basis.
Such prices may evolve by a factor of 1 to 2 during their trading periods. In fact, intraday prices regularly evolve by a factor of 1 to 5 on an average basis. Very high price peaks are observed during certain days–for example, when a large industrial asset unexpectedly requires maintenance in winter and is taken off-line. Negative prices are also becoming more and more of a market reality as renewable production sometimes results in an unexpectedly high volume (high supply) of available energy due to misinterpreted weather forecasts.
Digital Solutions Enable Forecasting and Increase Accuracy
As a result of these conditions, it has become much more difficult for both utilities and major energy consumers to predict both energy production and portfolio consumption. Therefore, they find themselves in a situation where they cannot access an optimally low energy price. The solution is to enable a level of forecasting that increases the accuracy of supply, demand and prices. Only then can electrical utilities and distributors maintain and/or increase their profitability.
This is wherenew artificial intelligence(AI) digital solutions play an integral role. We atPredictive Layer, a Schneider Electric Energy ManagementTechnology Partner, have developed such a system to help large enterprises, traditional utilities, and producers of renewable energy to better forecast energy consumption and pricing.
AI and Machine Learning Provide Higher Accuracy and Adaptability
Forecasting technologies, statistics, and models over the last 50 years were developed by small teams of expert energy forecasters. Their approach worked well in managing the previous generation of grid operations.
Now, accelerating levels of complexity across grids create limits on the effectiveness of these human forecasting teams. However, new generations of distributed IT and infrastructure tools allow very affordable and resilient high-performance computing that enables the creation of more accurate forecasting models. These models now supplement forecasting on a daily basis. In fact, technology evolution is such that it is no longer necessary to manually build new forecasting models. New AI and machine learning tools automatically evolve the forecasting models by incorporating new trending and network supply and demand behaviors on a daily basis. With the proper AI tools, forecasting teams can produce and manage forecasts with 10x more accuracy than before.
Oliver Cognet, CEO, Predictive Layer, is a business leader, corporate developer, and entrepreneur. He has 25+ years of experience in High Tech, Software Development, and Telecom markets focusing on enterprise and consumer markets. Mr. Cognet knows that technology and innovation can help businesses better understand the behavior and needs of consumers, thus helping businesses to grow. Predictive Layer brings the full power of machine learning artificial intelligence (AI) to businesses so they can simply and intuitively model and establish forecasts and predictive analytics of their key target indicators.