Putting carbon intensity on the map with electricity consumption data

Copenhagen-based Electricity Maps is an evolving platform for carbon emissions monitoring and accounting in near real-time.

With carbon emissions being of increasing concern as the key short-term 2030 targets come closer, monitoring and managing these emissions is becoming increasingly complex for companies and other parties to meet legislative and other requirements.

Among the plethora of solutions emerging, the Danish start-up Electricity Maps presents a novel approach with visualisation of the carbon intensity of the current electricity consumption among other information.

“Electricity Maps is a data platform that in real time figures out the origin of electricity and its associated carbon emissions,” explains founder and CEO, Olivier Corradi.

Have you seen?
‘Balancing decarbonisation with affordability remains a critical challenge’
Why Esbjerg is Denmark’s energy transition metropolis

It comprises essentially two major components – a free app displaying the current generation status of countries and an API for companies with real-time and historical data, typically for carbon accounting purposes.

With these, and a third predictive component, the company has been advancing rapidly, with Corradi saying that the most important development has been the expansion of coverage, particularly of countries in Asia.

A second aspect is improvements to the forecasting engine to enable more accurate carbon footprint forecasting up to two days ahead, which requires taking into account not only how the electricity is generated in an individual country but also how it is imported from neighbouring regions.

“These are big engineering feats requiring a lot of data power under the hood and are being leveraged by our customers in order to reduce their carbon footprints,” Corradi says.

As an example he cites the charging of an electric vehicle, saying that it doesn't necessarily make sense for a driver to come home at night, plug in their vehicle and for it to start charging immediately.

“It would be so much smarter if it could wait until the time when the electricity is the cleanest,” he comments, noting that in Denmark in extreme cases the proportion of clean energy in the mix can change from less than 10% to over 70% in just two hours.

Aligning carbon and price signals

As an extension of this work, Corradi says the company also has started to work on price signals with the ideal to align the carbon and price signals to ensure that electricity is used when it is both cleanest and cheapest.

However, several factors are limiting its implementation, he says.

One is on data availability, with the lack of open, accessible high-quality information on electricity systems in all countries. The second is regulatory, with most companies carbon accounting only on an annual basis with few incentives to optimise on shorter timescales and few incentives for generators to deliver electricity at its lowest cost.

“For us it is about how you build an IT system that is able to capture and harmonise all this information and make it available to end consumers. It’s a challenge we are faced with and hope to manage in the next couple of years.”

And perhaps coincidentally Electricity Maps’ location in Denmark provides an ideal testbed for its technology, Corradi says.

Not only is the country a leader in terms of digitalisation of the energy sector, but also it has uniqueness in having large wind potential along with strong interconnectors to import wind energy from neighbouring countries.

“It’s one of the first places we are seeing all the intermittency issues of renewables and showcases why the country is in a unique position to tackle this dual problem of flexibility with a digitalisation angle.”

Data and AI

With data at the core of Electricity Maps’ offerings, much effort has been put into building a data quality checking system to clean the data, including the inclusion of AI and machine learning to replace data that is found to be problematic.

“Because the whole electricity grid is completely interconnected, if you have a mistake at one place it propagates to other places in the system when you make your analysis, so it’s absolutely crucial to first detect and then fix these mistakes,” Corradi says.

“The issues we are dealing with range from data that is missing for a particular hour to a data provider telling us that the wind production in a given place is only half of what it actually is. And at the other extreme is an outage, which means we are left without any raw data for multiple days but still need to be able to utilise the signals that have been available.”

Looking ahead to the potential of generative AI, Corradi feels there is still a lot of work that needs to be done on it and given the underlying energy costs of AI, that the focus shouldn’t be on the “biggest model” but on the performance of the model per watt.

“What is happening in this space is extremely interesting, but right now the algorithms are trained to produce the most likely sequences of contents and not necessarily content that is truthful or logically coherent. As a technologist, I still see a very long way before the job is completely replaced and I wouldn’t let an AI write code and put it on a production system because it simply doesn't give me the guarantees that I'm looking for.”

With that said, Corradi states that the motivation for his work is that electricity and energy are the cornerstone of fixing climate change.

“As our energy system is 80% fossil fuel-based, there's no way we're going to fix climate change if we're not electrifying our energy system with clean electricity and as I'm actually working on this is the thing that makes me wake up every morning.”

Based on an interview by Florence Coullet, Content Director for Enlit Europe.