I share a recent interview question asked by Baringa (this is a consulting company focused strongly on energy, utilities, and the energy transition). From a student forum / database (for internship). The question was: 'Why do some renewable power plants continue generating electricity even when electricity prices turn negative? and when do prices become negative ? do you know any example? '

A new course, "Deterministic Optimisation in Python", is now live in the Classroom (see screenshot attached). It is course number 114. This is a beginner-level course (no need to even know Python) with no prerequisites. Optimisation is used when we want to minimize or maximize something. In most cases we want to minimize some cost. Here we have a smart building and we want to minimize its daily operational cost. The simplest form of optimisation is "deterministic' , which means that all the data are known, and they are given to the Optimisation model. There is no 'uncertainty'. When an optimisation model is not 'stochastic' then by default it is 'deterministic'. So we typically do not use the word 'deterministic' because in 90% of the cases, it is implied. In reality there is always some uncertainty. For example, we may not know the electricity demand every hour. So, the smart building has residents who consume electricity. We do not know how much this electricity will be tomorrow. Instead of modelling this uncertainty using some advanced method (in which case, we would speak about 'stochastic optimisation') , we use 24 values for the demand, and 24 values for the renewable output, and this makes the model 'deterministic'. We use Gaussian (Normal) distribution and from this distribution we randomly select ('sample') 24 values for the electricity demand, and 24 values for the renewable output. In most of the energy projects we use deterministic optimisation with known data i.e. some official source will give us the data. In the event that they won't give us the data, we can make assumptions e.g. like in this course we make the 'Gaussian assumption' i.e. we assume the demand follows the Normal distribution. We will walk through the entire process step-by-step: from setting up the Python environment and generating synthetic data using numpy, to formulating the mathematical model in Pyomo and analyzing the results with pandas and matplotlib. This course provides aa very good practical experience, which is popular in real-world projects.

A new report on energy trends has been published and can be found by clicking on ‘Classroom’ and navigating to Section 6.2 (see the attached screenshot). You can use this report and the visualisations it includes, in your own projects, work, or studies, without limits. This report explains the progress for the UK’s hydrogen rollout. The report includes diagrams and flowcharts that provide context, and also a list of relevant sources that were used to complete this report. These sources are from the Financial Times, Wall Street Journal, the Economist and Investors Chronicle (all sources are available inside the report). Your subscription in this Skool community gives you access to paywalled energy-economics articles from these publications (Financial Times etc) indirectly through these reports. I have also included some explanations and additional text that explains some details. The text is written in beginner-friendly, easy-to-understand language. Reading these reports can help with interviews, meetings, presentations, networking, and public speaking. Strongly recommended.

The attached plot shows the level of electricity in Africa. How best can it increase? using smart grids? micrograms?

Since linear and logistic regression are supervised models and are frequently used for exploratory analysis in data mining, would it be accurate to say that the distinction between data mining and machine learning is primarily methodological (discovery vs prediction) rather than algorithmic? Please feel free to share any experiences choosing to use data mining for Energy or any other industry.