Intelligent storage - New frontiers in power provisioning

The history of storage is as long as the history of study and exploitation of electricity itself. People have been enjoying the benefits of portable power since the first dry-cell flashlights of the 1890s.
But it’s only now that storage technologies are beginning to offer a solution to the wider industrial-scale challenges of contemporary power production and usage, and renewables are at the core of this revolution.

Storage and sustainability are inseparable

Decarbonization imperatives rest on green power – and the continuity of supply of green power rests on storage. Intermittent and variable supply are, inevitably, inherent characteristics of the renewable power to which we must turn: solar energy can only be gathered in daylight and turbines can only turn when there is wind.
The implications are clear. Just as we use reservoirs to hold water until we need it, so we need the means to store surplus electricity until it is required.

But the need to capture and store power is not just about the technology of production and distribution in a model that puts renewables at the core. It is also about wider social expectations and demands, and indeed about emerging regulatory frameworks.
The end of our reliance on fossil fuels is not quite imminent, but it is on the horizon. Now is the time to prepare and implement strategies for sustainable storage, as there can be no long-term reliance on renewables without it.

Two distinct storage models

In preparing for a utility landscape in which storage takes a central role, it’s important to recognize the basic differences displayed in two storage models. These two models differ in more than composition and function. From the business perspective, they offer different service potential, and this in turn means that they will attract different investors, owners, and operating partnerships.

• Type One – the storage factory. These will be massive, stationary, industrial facilities. These storage implementations will typically use lithium-ion battery technologies and will initially offer power output in the region of 15-20 MW. They are most likely to be owned and operated by utilities and will be used primarily as a means to store renewable surplus which can then be deployed to balance grid demand.
• Type Two – the virtual power plant. This aggregated power model will primarily use and service the batteries of electric vehicles to establish virtual power storage capacity. This model is not only heavily reliant on massive data gathering and tracking. It also demands a highly distributed and agile infrastructure, allowing vehicles to both feed into and draw out from the grid. Although utilities will feature prominently, we can also expect automotive and real-estate companies to take an active role here along with local and regional government.

Sustainability and asset optimization

Although very different in form and purpose, these two models do have something critical in common.
Batteries themselves have a considerable environmental footprint. They are costly too. It is essential that these assets are optimized for both maximum useful life and minimum environmental impact from manufacture right through to decommissioning.

Much of the activity needed to protect and optimize these assets will be data-driven and knowledge-based. Whether located in an industrial power plant or in a family vehicle, every battery has an optimum cycle for charge and discharge, and with advances in the Industrial Internet of Things, every battery can be set-up to self-manage.
Asset optimization is about more than knowing how and when to charge and discharge for maximum longevity. It’s also about maximizing the financial value, and this means knowing when to sell surplus power and how to achieve the best price.

This activity too is data-driven, with accurate and granular forecasting being essential both the meteorological data needed to forecast renewable production and, on the demand-forecasting needed to ensure sustainable business return.

Intelligent storage and IT

In this new landscape, operation must become increasingly data-driven. Massive data flow and processing will cover, among other things: renewable production, storage management, grid management, demand forecast and billing for consumption and pay-out for prosumer production.
Combine all this with the shift to the increasingly agile operation of bi-directional low voltage grids, and it becomes clear that execution can only be achieved through digital. As these storage models gain momentum, they will rely on sophisticated automation, machine learning and AI, and indeed on massive demand for networking and computer processing.

Underpinning all these digital technologies is the need to establish an effective data platform, allowing data types from across the grid to be combined and analyzed for actionable intelligence. When, for example, vehicle-to-grid applications are deployed, data types from across the automotive, regional government and utilities sectors will be involved.
As a long-term partner in digital transformation for utilities, Atos is already deeply involved in these challenges. Working with the French national weather bureau, for example, we are already deploying AI and quantum computing in some of the most sophisticated meteorological forecasting, with our Greenforecast services. We are active too in the financial and market modelling activities which will underpin the effective commercial execution of new storage models.

Topics

• • Digital Transformation
  • Data and artificial intelligence
  • IoT and edge
  • Industry insights