31 Jul 2017

Balancing European wind power

In a paper published in Nature Climate Change, we used a combination of a meteorological understanding of weather regimes and our Renewables.ninja wind and PV generation time series to show that by smart spatial deployment of future wind farms, based on our knowledge of weather regimes, it is possible to keep the European wind fleet’s multi-day variability to levels the power system is already dealing with today. Figure 1. Read more...
6 Sep 2016

Renewables.ninja: simulating PV and wind power plants

Together with Iain Staffell, I published two papers, available today, using reanalysis and satellite data to simulate PV and wind power plants across Europe (references at the bottom of this post). Alongside these two publications, we built a large database of measured power output data to validate the simulations and perform bias corrections. This now allows us to accurately model hypothetical European wind and PV power output over several decades, investigating issues like the impact of increasing PV deployment on net power demand: Read more...
18 Dec 2014 #python

Calliope, an open-source energy systems modelling framework

Energy system models are used by analysts in academia, government and industry. They allow researchers to build internally coherent simulations and scenarios of the extraction, conversion, transportation and use of energy, either today or in the future, at scales ranging from cities to entire countries or continents. These models are particularly important because they can help with understanding and planning the transition from a fossil-fuel dominated energy system to one primarily consisting of clean and renewable energy. Read more...
29 Aug 2014 #python

IPython notebook extensions to ease day-to-day work

Update 26 October 2015: The ipython-extensions repository has been updated with instructions for Jupyter/IPython 4.0, but for now, the CodeMirror customizations do not seem to work. Update 4 March 2015: As of IPython 3.0, it seems that code in custom.js (such as the examples below) should be attached to the app_initialized.NotebookApp event (the ipython-extensions repository is updated accordingly): $([IPython.events]).on("app_initialized.NotebookApp", function () { // your code here }); Since I use the IPython notebook for a lot of explorative data analysis, I wanted some extra functionality to make day-to-day notebook work just a little bit more pleasant. Read more...
22 Jun 2014

Can concentrating solar power (CSP) deliver reliable electricity?

We just published a paper in Nature Climate Change where we examine whether concentrating solar power (CSP) can provide the reliability we’re used to from conventional (fossil or nuclear) power plants. In other words, under what conditions can CSP plants provide baseload power, i.e. an essentially stable output of power over time similar to nuclear plants, and dispatchable power, i.e. power output that can be quickly called upon if needed to balance demand? Read more...
21 Apr 2014 #python

IPython extension to toggle dark code cells

Update 25 August 2014: As described in this newer post, this is now part of the ipython-extensions repository. Sometimes it’s more pleasant to read light text on a dark background, particularly when dealing with code. However, depending on ambient lighting, it can be necessary to switch back to dark text on a light background on short notice. To make this possible, I created a simple extension for the IPython notebook (version 2. Read more...
13 Apr 2014 #python

Mapping the world's nuclear power plants

This post demonstrates the use of Python to map the world's nuclear power plants. First, we procure data on plants currently operational or under construction. Then, we make some back of the envelope calculations on how many plants would power the world's current electricity demand. For this latter part, we use GeoPandas, a very interesting development for Python-based geospatial analysis, building on modules such as Fiona and Shapely to add geospatial functionality on top of one of the workhorses of scientific Python, pandas. Read more...
7 Apr 2013 #python

Doing a minor version update for homebrew-installed Python

I install Python with homebrew (on OS X) and manage my packages with virtualenv. I was interested in whether updating from Python 2.7.3 to 2.7.4 would cause my virtualenvs to break. It did, but fixing them turned out to be very simple: First, of course, $ brew upgrade python; brew cleanup python, which installed Python 2.7.4 and then removed 2.7.3. According to this post, re-running virtualenv suffices to update the paths inside a virtualenv (the installed packages remain, and you shouldn’t need to re-install or recompile anything after a minor version update): Read more...
19 Mar 2013

The discovery of sustainability

There are several events in the twentieth century commonly thought of as key milestones in developing a shared environmental consciousness. The first views of our planet from space (in particular the famous “blue marble” photo taken by Apollo 17 astronauts in 1972) and the publication of Rachel Carson’s Silent Spring a decade earlier are such milestones. So is the idea of sustainable development, as formulated by the Brundtland Commission: “Sustainable development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs. Read more...
23 Feb 2013

Linking DOIs on Scopus Preview

I use Scopus RSS feeds via my institutional Scopus access. Unfortunately, if I save links from a feed on the go I get redirected to a ‘Scopus Preview’ page when I later open the link in my browser. This preview offers no direct way to get the full paper, and no outbound links. Seizing the opportunity to experiment with making a jQuery-based user script, I threw together some simple code to add a DOI resolver link to the preview page: Read more...