AleaSoft, November 23, 2020. Interview of the magazine Energética XXI with Antonio Delgado Rigal, PhD in Artificial Intelligence, founder and CEO of AleaSoft.
Antonio Delgado reviews in this interview the trajectory of more than 20 years of AleaSoft, a company that has been offering its prices forecasting in the energy markets for two decades to the main players in the sector with the artificial intelligence as the main solution to carry out its studies and reports.
AleaSoft turned 21 years in the market on October 8. What was the evolution of the company in its more than two decades of existence?
From the first years we worked for the main electricity companies in the country: Endesa, Iberdrola, Unión Fenosa and Viesgo. The electricity market started in 1998 and the companies needed to make good forecasting of prices and national electricity demand.
In a second stage we began to work for companies related to the combined cycle gas turbines and the wind energy such as Gas Natural, Bahía Bizkaia Electricidad, Shell, Electrabel, Gamesa and Acciona. The massive installation of wind turbines brought with it the need to forecast the wind energy generation both at the park level and at the national level. At that stage we started working for REE making demand forecasts. In addition, we started making forecasts for large companies in different countries of Europe, mainly Italy.
In a third stage, since 2017, we began to work for companies related to the development and operation of photovoltaic parks in Spain and the rest of the continent. In this third stage, our work for companies in the financial sector related to the investments in renewable energy production was intensified. With the fight against the climate change and the plans to reduce the electricity production with fossil fuels, a long‑term strategy to generate all energy with solar energy production and wind energy production begins. These long‑term investments require the 30‑year prices curves forecasting that is the basis of the financial projects, necessary to quantify the profitability of the investments in the new renewable energy technologies. The European banks and investment funds interested in the wind and photovoltaic energy sector become our clients at this stage.
One of the keys to success was the team that we brought together. Highly qualified staff, experts in mathematical and statistical subjects with extensive experience in the energy sector.
How was the AleaSoft’s use of artificial intelligence to shape its prices forecasting models in the electricity markets changing?
The artificial intelligence is in fashion, it is a concept that sells, which is linked to the future but which already became reality in general due to the computing power of the current computers.
At AleaSoft we created a methodology and a forecasting model type that we call the Alea Model, which combines artificial intelligence with statistical methods mainly related to time series such as the SARIMA models.
We have used this new type of model since the beginning, 21 years ago, with great success and coherence in the forecasting related to the energy sector. In this time, we have used it to forecast in the main European energy markets, making prices, demands and renewable energy production forecasting in all horizons from the short to the long term. It should be noted that in order to make quality electricity prices forecasting, it is essential to previously forecast the variables related to the supply and demand.
What differentiates the long‑term prices curves forecasting that the company makes from others that we can find in the market?
Above all, the long‑term prices forecasting must be robust, coherent and based on scientific methods, characteristics that our models fulfil.
We make 30‑year prices forecasting with hourly breakdown, which allows having a maximum level of detail to calculate the price that the renewable energy facilities that are being planned to build will receive each hour. The long‑term hourly forecasting at this time is essential to make decisions with more precision.
Another differentiating feature of our methodology is obtaining confidence bands with a probabilistic metric, that is, we can generate long‑term prices forecasting with the associated probabilities, for example, to P15, P50 and P85. This probabilistic metric is essential for the renewable energy portfolio valuation and the risk quantification.
Which type of customers comes to AleaSoft to seek this type of forecasting of the electricity markets?
As we mentioned before, in the historical evolution of AleaSoft in these 21 years, we first worked for the large electricity generating and retailing companies of Spain and, later, of the rest of Europe. Another type of company we work with is the Transmission System Operators (TSO) who control the electricity production and consumption in all time horizons.
The large consumers also need to have mid and long‑term prices forecasting to make purchases in the futures markets and reduce the risk of buying in the spot market. In addition, in order to plan the contracting of a Power Purchase Agreement (PPA) and guarantee the purchase of renewable energy at a predetermined price in the coming years, a forecasting that gives a clear vision of the electricity market prices in the future is also necessary.
All companies linked to the financing and development of renewable energy facilities need forecasting. In this field we work for developers of renewable energy facilities, banks, consultancies and investment funds.
During the pandemic, AleaSoft is organising numerous webinars on the energy sector. Which were the topics that were addressed?
We are organising a webinar every four to five weeks to give a vision of the future related to the energy markets. The topics are mainly related to the future prospects of the energy sector, the future vision of the renewable energies and the issues related to the financing of these renewable energy projects. We also address issues of interest to the large consumers or retailers and the strategies to buy or sell energy in the most efficient way and with less risk.
Specifically, and in relation to the financing of renewable energy projects in Spain, is the sector optimistic regarding the development of new projects?
In the European Union there is a firm commitment to the decarbonisation and that translates into multi‑million investments in the renewable energy sector. Spain has a great wind and solar resource. In a few years we will go from being net energy importers to being exporters if the necessary investments are made in wind and photovoltaic energy generation facilities. All the energy that currently comes from fossil fuels that are imported into Europe will be generated within the continent and Spain will have an advantage especially in the photovoltaic energy generation.
In these more complicated months due to the pandemic, the investor appetite did not wane, on the contrary, more and more foreign entities are paying more interest to take positions in Spain with a long‑term vision of the future.
The Government recently confirmed the call for new renewable energy auctions, the first of them before the end of the year. How will this call affect the sector in general and the electricity market in 2021 in particular?
Regarding the issue of auctions, it was created as a controversy: “auctions yes”, or “auctions no”. Like everything in general, the extremes are not good and there is a happy medium.
The Spanish electricity market began to function in 1998 and has been a hour‑by‑hour benchmark for setting the electricity price not only in the spot but also in the future markets and the very long‑term contracts. It is a marginalist market that integrates all the markets of the European Union, that is, a single integrated market and with the same rules for all participants. The market signal is essential for the future investments.
The auctions are useful if they are done carefully, in certain geographical areas, little by little and to bring benefit to those who participate and to the consumers. If they are done in this way, it will not affect the market signal. If the auctions are poorly implemented, they can backfire. It is necessary to remember that the market self‑regulates and over‑regulation can discourage the investors. We exposed in different forums that, although we do not have auctions, the NECP objectives for 2030 in terms of photovoltaic energy production will be met.
Green hydrogen or the development of the storage are some of the current trends in the energy sector. Which immediate future do you forecast for both of them?
When we think of the future, in a transport and industry without polluting emissions, we must think of the green hydrogen as a fundamental energy vector. Decarbonising these two sectors is essential to achieve the neutrality of the greenhouse gas emissions. According to Eurostat, in 2018 the European industry and transport consumed 52% of the final energy. In the case of the transport, the oil products and the gas accounted for 90% of the consumption and in the industry, for 47%.
The production of green hydrogen will allow the current electricity demand ceiling to be multiplied by 400% or 500% in 20 years when we do not have fuels with polluting emissions.Final energy consumption in the European Union.
Source: AleaSoft Energy Forecasting