WIND ENERGY PRODUCTION PROJECT – presentation –
Nowadays, we are working to solve a short term prediction problem to solve the biggest wind farm production problem – Energy –
RENEWABLE ENERGIES PROJECT: Wind Farm Predictions
- The main problem of wind farms is the uncertainty of the availability of wind. It means that we cannot know what will be the production of electricity by the park.
- At 10 am, all electricity producers must submit to OMIE (Spanish energy market operator) its production of electricity for every hour of the day to create the aggregated supply and demand curves of electricity. Therefore, wind farms need to know in advance the electrical output of the day. This information is obtained by different forecasts.
- If the wind farms do not meet the contracted power supply by marketers, energy will be purchase from other non-renewable sources and renewable energy producers will be sanctioned.
- Our goal is to give a better estimation of daily energy production to reduce uncertainty better than previously used. This forecasting increases competitiveness in the wind sector and improves efficiency of services.
- The mission of our project is to increase the use of wind energy through R&D to have less dependence on other energies. Therefore, we focus on self-energy dependence of the Spanish electric system (currently, 81% of primary energy consumed in Spain is imported).
- The product that we offer is designed to wind energy producers. It is a booming market, because wind power is being the main source of electricity generation in Spain (covering 20.9% of demand). Furthermore, in Spain, around 85.5 million euros a year are invested in R&D and we estimate an increasing market demand.
- Our product is a “Software as a Service” (SAAS) enabling more accurately daily predictions of energy production. The system offers improved forecasts of the European system of predictions (ECMWF). The access to Software may be from any platform with Internet access and it offers a continuous service improvement and customer service.
- At first, a consultancy is performed to study the needs of predicting the wind farm. Following this study, the park will get a user in our website to get their predictions via online. An initial payment will be made by the consultancy process and, after that, they get a monthly / yearly subscription service.
1. AN IDEA WAS BORN
The idea was born based on the need to optimize the scheduling of daily energy to maintain the quality of supply energy by the OMIE (Spanish energy market operator) by reducing uncertainty about the production of electricity that will produce by a wind farm.
The initiator of the project was Borja V. SORLÍ SANZ [Franco-Spanish Engineer and Business IMBS], who worked on two research centers (including EDF research center) where advanced statistics and knowledge in Data Science were applied to get energy predictions based on wind and solar farms installations.
2. ENTREPRENEUR’S TEAM
Currently, we are a team of three people.
To begin, Melody ROMERO DIAZ [Student Degree in Industrial Technology – ETSII (UPV) -] joined the project to get EDPR University Challenge finalist award, where our project was not as developed as it is today.
On the occasion of the announcement of the II GDES Entrepreneurs Award candidature for Energy Sector, Néstor RUANO FOLCH [Degree in Industrial Technology Engineering – ETSII (UPV) -] was joined the team.
We complement each other very well since the three engineering speak the same engineering jargon. Competency: Borja V. is dedicated to the business vision, innovation and sales; Néstor working on research, innovation and product; and Melody has focused on programming, marketing and design.
We probably include a fourth person with design and usability expertise to get better end user experience.
3. DESCRIPTION OF THE BUSINESS IDEA
Since we believe that wind energy is the future of energy in Spain and Europe. Therefore, we aim to supply a wind farm with a prediction of production according to market needs. Increasing their competitiveness and improving efficiency by reducing uncertainty regarding the amount of power generated.
To achieve this objective, we intend to implement a “Software as a Service” (SAAS) that allows the user to access their personalized system and updated forecasts at all times. This system improves the prediction of the European system of predictions (ECMWF). Access to this Software may be done from any Internet connection, in addition to a maintenance service and customer care.
Basically, our business idea is to offer a SAAS to predict the daily production of energy generated in one geographical location. For this first phase, a consultation process will be developed to meet the real needs of the customer (eg in France, the prediction of energy production should be done the eve at 17h, while in Spanish market should be at 10 am; therefore, there is a variety of possibilities) and study the characteristics of the park to facilitate an online solution. This first step is not scalable but the fact of offering a SAAS in a web platform is. It means that, after first consultancy process, we can offer our product for a large number of customers without having to increase our resources to provide this service. This online service would be acquired with a monthly or annual fee subscription to the service.
In laboratory, it is shown that advanced statistical method requires far less time to give a solution. By using the same resources, the physical method (using fluid dynamics simulations) takes about 3 hours to give a similar to those given by our method in just 30 minutes. Therefore, we can speak of a minor computing needs.
For adaptation, it would be a desired step to other sectors because with our knowledge and resources we can solve with an incredible accuracy any prediction problem if we have the desired data.
Our solution has two main axes of innovation:
– On the one hand, the implementation of a commonly used technique in Artificial Intelligence (AI) to forecast Wind Energy production. Many large research centers around the world are studying the possibility of applying techniques such as Artificial Neural Networks (ANN) to predictions of this kind.
– On the other hand, applying a specific SAAS (Software As A Service) to the wind energy sector and to distribute it through Internet connection without external specific hardware or software. In addition, the product provides inter-user activity with which self-manage and adapt their needs may be possible.
After we conducted a market study, we decided to center our efforts on wind energy sector because it is currently booming. If we statewide the study, we found that wind power is the main source of electricity generation in Spain, covering the 20.9% of the electricity demand (2013). Being Spain, the fourth country in the world in installed wind power, after China, the US and Germany. With an installed capacity of 22.970,58MW (June 30, 2014).
If we add to these facts, the need to know in advance the electrical output, make this sector, the ideal market for our product. Then, the product that we offer is a Software as a Service (SAAS) enabling more accurately daily predictions of energy production. It will allow better regulation of wind farms and accurate curves will be produced for daily OMIE market. Curves must be submitted each day at 10am to create aggregated supply-demand electricity curves and get a price of electricity for each hour of the day.
Our product is intended for any wind energy producer in the European Union, but our initial market will focus statewide (Spain).
We believe that apply our acquired R&D methods is the key to achieving the European objectives of energy from renewable sources by 2020. In addition, it will help to get energy independence (81% of primary energy consumed in Spain is imported and comes from fossil fuels). Currently, in Spain, about 85.5 million euros per year are going to R&D. For this reason, and for all the above data, we estimate a fairly high demand of our product.
As barriers to entry in this business, we have found that some wind farms have existing contracts with other companies or their own software predictions. Still, being a rapidly developing market with great potential, we expect wind energy producers are open to innovation because it is making a big investment in the sector in these new forecasting techniques.
The strongest competitors in the industry are:
– Meteogroup. For almost 30 years working in the industry. It offers a broad consulting: to advice from where to position the park, construction and maintenance. It offers a 48h predictions view using a web service (recently installed) with a system of alerts via email, SMS and telephone and customer service 24h. It has about 1,300 stations worldwide and operating in any region of the globe.
– Gamesa. It is the biggest supplier of wind generators. It is dedicated to the construction and maintenance of wind farms, making it necessary for predictions. It also has a web application for predictions. Furthermore, it gives short-term predictions of generated power, predicts potential deterioration and malfunctions in the wind turbine components. Also provides an alert system as in the previous case and tablet applications. It works in Greece, Spain, India, China, Poland, Mexico and Sweden.
– Meteológica. It is a company with over 10 years of experience in the sector. It is dedicated to the service of weather forecasts for different sectors of energies. As for the predictions of wind generation provides a web interface with graphical displays and recent observations, contrasts quality viewing and modifying the availability of the park. His predictions come until 360h with hourly updates. Operates globally.
– Meteosim. It is a company with over 10 years in the sector. It is dedicated to the service of weather forecasts for different sectors of energies. It offers a consultancy service with short-term predictions and it serves companies like Acciona, Iberdrola, gasNaturalfenosa, Repsol and Mapfre, among others.
– National Renewable Energy Centre (CENER). It is a specialized in applied research and development and promotion of renewable energy technology center. It provides services and conducts research in 6 areas: Wind, Solar Thermal and Photovoltaic Solar, Biomass, Energy in Buildings and Grid Integration of Renewable Energies. They offer quality predictions in the Spanish daily and intraday markets based on an ensemble of models predicting short and long term with which service is provided to some of the largest electric utilities in the country.
7. MARKETING STRATEGIES
In order to provide greater visibility and transparency, we decided to develop a website for offering information about our “venture” based in training and delivering our services.
Therefore, it will pave the way for a general public who may be interested with related products and even opt for another study or product applied to solve a particular problem for other kind of company or sector.
However, we have focused our time and efforts on improving our SAAS to solve the problem of wind production in the short term, by which we choose in this convocation of the Second Prize GDES Entrepreneurs in the Energy Sector in collaboration with IDEAS.
Then, on one hand, we (1) training to solve some basic problems in data-science completely free and available on our website. And, secondly, we will have our (2) SAAS applied to wind energy producers.
Regarding our Energy Services product for wind parks, the pricing policy corresponds on two steps. A first phase in which we will realize an (i) study concerning the problem of the client (with a price of 15 000€ per client) to achieve their goals and provide the best possible service through our online SAAS. This step should not last more than a month of work.
After consulting this study would be given the exact solution to their problems via Internet connection to the company as can be seen prototype available on our website (ionds.com/saas). So, we would have (ii) a subscription fee (€ 500 / month or 5000 € / year) to maintain the service and the customer with the customer.
We begin a campaign of direct marketing for companies that might be interested in the Valencia community and then expand to other parts of Spain and Europe.
If expectations will meet, we could begin to study other markets forecasts given the experience. And therefore, we would begin to use other more convenient channels of communication with potential customers.
8. INTELLECTUAL PROPERTY, TRADEMARKS
As for intellectual property issues, we started working our brand image “ION data services” on our website for a general audience. However, we will focus, in the first instance, in wind energy producers.
As discussed with a specialist lawyer in tax issues for Startups, maybe we could get the tax benefits of a company with R&D charge when the company was constituted (by studying advanced methods of forecasting).
9. CURRENT LEVEL OF DEVELOPMENT
Previously, two pilot tests were performed with satisfactory results. The first project was subject to studies in ENSAM (ParisTech) by Borja V. over a region of Corsica (France). The second was held on the occasion of the EDPR Renewables Award on a coastal region of California (USA), where we get the Finalists Award for this project in a lower level of development.
Currently, we have worked to transform the idea into a viable and accessible to end users product. To do this, we developed a website (ionds.com) in which we have dedicated a special tab for this project (“Energy Services” section).
In a later stage, to improve our product and be able to serve a greater number of users; you can find other services based on Data Science as our project could easily be extended to other application areas easily.
10. ECONOMIC AND FINANCIAL ASPECTS
Since the software does not require many technical, operational and computational techniques employed are quite fast, does not require a prohibitively expensive computer equipment, which could be budgeted around 500 €. It should also be funds to get a direct contact with customers and be able to realize a marketing campaign.
By using open source software, such as R software for the implementation of our product, we do not require a large initial investment. The only important investment that should be done would be to purchase data from ECMWF (European Centre for Medium The-Range Weather Forecasts). However, as stipulated our marketing system, we work in a previous consulting study of the park. So, no big initial investment will be required for obtaining funds to get data. Furthermore, in many cases, the wind park has already the required data for us and they will offer it to deal with their problem.
11. PROJECT IMPACT
As explained above, power electricity plants must give their predictions of generated power for the next 24 hours, every hour, to OMIE (daily market). For predictive factors, there is an intra-day market that can avoid penalties marketer generator due to errors in predictions. It happens if generated energy does not match previously estimated energy production, so they should acquire by non-renewable sources or they will be punished.
Our project aims to improve predictions of wind farms as we have showed in previous projects. So, we would avoid impose sanctions for companies and we avoid to buy external energy due to had realized an estimate power generation below the real power generated. We optimize the scheduling of daily energy market from wind farms by improving the quality of predictions and give better supply to OMIE.
Regarding social benefits, we can talk about an improvement in wind energy is a step toward energy independence abroad. Because, as discussed above, 81% of primary energy consumed in Spain is imported and comes from fossil fuels.
We may also increase the benefits for the environment, since the more precise forecasts of the power generated, the producers could buy less non-renewable energy. And, we increase the benefits to the sector by improving its efficiency. It could increase renewable energy entrepreneurship by getting the lowest risks and, therefore, by increasing confidence in the sector. It would also increase the confidence of traders regarding renewable energy generators.