Group2 ProjectDeliverables

World Energy Production & Consumption: A Visual Study

Overview

Proposal

Analysis Report

Poster

Application

Motivation

Energy growth drives the well-being and prosperity across the globe. Growing demand for energy has to be met in a safe and environmentally conscious manner. Rapidly changing energy dynamics determine the course of our economic development, geopolitics, technological breakthroughs, massive investments and trade flows. The main objective of this project is to study the dynamics of regional energy production and consumption of the different countries of the world to identify the dominant as well as weaker player in the world energy market Though reports from various organizations such as EIA, IEA, OPEC, BP are available, most of the visualizations are static and do not aid exploration, limiting the scope of further drilling down to the areas of particular interests. Hence, we've decided to take up rich data-set provided by BP & explore the information related to energy production and consumption across countries, which might otherwise be hidden in the existing static visualizations.

Objective

Data visualization is often an after-thought for many of the practitioners who are collecting and analyzing data. And yet, without clear and compelling communication, analysis will never drive insights and action. Visualization can itself be used effectively in the process of insights discovery. Static visualization and reports do not allow for direct interaction with data, enhanced assimilation of information, quick access to relevant insights and drill-down analysis. By using interactive, visual data analytics techniques we will be bringing the above-mentioned capabilities to the interface. Following are the objectives of our work: (a) A user-friendly and interactive visualization application interface for data exploration that supports both aggregated and drilled down views and analysis which can be used by personnel in energy departments, policy makers and general public alike. (b) Interactive visualization to understand which parts of the world today’s energy requirements are sourced from. (c) Regional energy consumption portfolio and trends for each energy type. (d) Understanding and identifying country clusters as they are set to adopt sustainable energy usage.

About The Data Source

BP plc, formerly British Petroleum, is a British multinational oil and gas company headquartered in London, England. It is the world's sixth-largest oil and gas company, the sixth-largest energy company by market capitalization and the company with the world's twelfth-largest revenue (turnover). It is a vertically integrated company operating in all areas of the oil and gas industry, including exploration and production, refining, distribution and marketing, petrochemicals, power generation and trading. It also has renewable energy interests in biofuels and wind power. For 66 years, the BP Statistical Review of World Energy has provided high-quality objective and globally consistent data on world energy markets. The review is one of the most widely respected and authoritative publications in the field of energy economics, used for reference by the media, academia, world governments and energy companies. A new edition is published every June. For the purpose of this project, we are using the BP Statistical Review of World Energy data. The dataset consists of energy consumption and production data from around 1965 for 2017 for more than 90 countries spanning across regions Africa, Europe, APAC, CIS, Middle-East , North America, S. & C. America. Energy resources are categorized as Primary Energy(Oil, Coal, Gas, Hydro, Nuclear); Other Renewable resources(Solar, Wind, Biofuels, Geothermal etc.). Below figure shows the different data we have for Oil in the dataset: A closer look at the tab Oil: Production - Tonnes (from 1965) reveals the below information regarding Oil production across different countries of the world: Similarly, we have different datasets each for the different forms of energy like Natural Gas, Coal, Nuclear Energy, Biofuels etc.

Critique of the Existing Visualizations

Visualizations available on the energy outlook reports consists of basic graph types such as bar, line and pie charts which are static in nature and do not facilitate any discovery apart from what they are made to deliver. A sample visualization is presented below: As shown above, the visualization does not allow user to dig deeper into the dataset. There is a lot of scope for improvement in the visualization methodology used and with the open source community contributing aggressively to the plethora of R packages, the possibilities are endless. Hence, to enhance usability of the the data, we have come up with interactive plots in order to gain deeper insights.

Dashboard Design

Global trend in Energy consumption for years 2006-2017 across energy types: Stacked Area Chart A Stacked area chart gives us a good understanding of the rate at which world energy consumption is changing with time along with the trends for individual components. Tools Used: Plotly Country-wise percentage production across various energy types: Sunburst Plot Usage: World energy profile consists of Non renewable energy sources such as coal, oil and gas as well as renewable sources like hydro energy, nuclear energy and other renewables like solar, wind, biogas etc. Sunburst Chart allows the user to interactively drill down and understand the percentage contribution of a region to World energy Production detailed for each energy type. Critique: Sunburst is a form of a radial chart which breaks out general categories into subsets to better understand the components that make up or contribute to the whole. This form a visualisation is similar to a treemap however it is easier to see multiple layers of data with sunburst, while the treemap is better for comparing categories within the same hierarchical layer. Disadvantages: Deeper slices exaggerate their size, and look visually larger. This type of visualization requires the quantitative comparison of angles, instead of lengths, which is difficult for the human eye. Tools used: SunburstR Figure above shows the Sunburst visualization of the World Energy production. Figure 1 in the visualization represents the country-wise percentage production across various energy types. The sunburst plot is interactive and can be used to dig deeper to understand the percentage production of the energy type to the country level. For example,as revealed in Figure 2, China produces 13.4% of the total energy of the world through Coal. Similarly, we can explore different energy type productions. Primary Energy consumption for 2017 across the World: Interactive World Map To understand who the major primary energy consumers of the world and their respective CO2 emissions are, we have implemented an interactive world map and a bar graph both in plotly, which are then integrated using shiny as shown in the figure below. World map is being made using leaflet and icon markers are used to indicate OECD and OPEC countries. Critique: This method was chosen to make the exploration interesting. Choropleth Map was not considered as countries are custom aggregated in the report and for some countries values are not listed. Tools Used: Plotly In the map, bars are sorted in descending order and when a user clicks on a bar, corresponding country is highlighted in the map and a tool-tip displays the Energy consumption in Mtoe, along with the CO2 emissions and whether the country belongs to OECD/OPAC. Alternatively, the user can click on any country of interest on the map to get the information on the primary energy consumption and CO2 emissions for the particular country. This is designed to make the exploration interesting and intuitive. In this case, choropleth map will not be an efficient visualization as countries are custom aggregated in the report and for some countries values are not listed. Country-wise Energy Consumption Portfolio and energy trends: Geofacets Geofacets are used to visualize data for each geographical entity, with the resulting set of visualizations being laid out in a grid that mimics the original geographic topology as closely as possible for the region. Each geofacet visualization for a selected region consists of a horizontal bar chart for energy consumption portfolio for all the countries in the region. Figure 1 in the visualization below shows Country-wise Energy Consumption Portfolio which allows viewers to compare energy consumption distribution across types within a country as well as across countries in the region. Another geofacet graph which has an additional option to select the energy type is used to display the trends in consumption of the energy type across countries of the region. Figure 2 in the visualization above shows evolution of Energy usage per country over time for each energy type. (time-series in geofacet) Critique: Given the custom aggregations of countries in the report as well as the unavailability of data for few countries , geofacets were the best choice however comparisons across countries form different regions is not direct. Choropleth and Tile maps limit analysis to one variable which is color encoded to aid analysis. Quantifying intensity of color is difficult and choropleths are known to favour large geographic entities over smaller ones. Disadvantages: A geofacet grid is only meaningful if the person already has an understanding of the underlying original geography. Also, this form of visualization takes more space and representing some of the geographical entities on grid layout can be challenging. Tools used:”geofacet” in R, Geo Grid Designer app for creating grids for each region. Production vs Consumption landscape for each energy type: Interactive Scatter Plot We have come up with an interactive scatter plot to understand how does each country stand in the production vs consumption landscape for each energy type. Total energy production and consumption for each country are calculated and percentage of production and percentage of consumption that the country attributes to each energy type is calculated. For each of the energy categories namely oil, gas, coal, other renewables(Hydro and Nuclear data is not considered as a pane as production data is not available for these energy types.) countries are plotted on a scatter plot of percentage production vs percentage consumption. The size of the bubble on the scatter plot indicates the absolute amount that the country produces and the color of the bubble indicates the absoule amount that the country consumes of that energy type, as shown in the visualization below. The plot is interactive and the selection of a country/bubble on one pane, highlights the respective positions of the country in other panes aiding insights discovery, as shown in the visualization below: Tools used: Plotly with Crosstalk for interactivity Relative positioning of countries in terms of usage per energy type: Ternary Plot A Terenary plot is used for understanding the relative positioning of countries in terms of usage of Renewable, Hydro & Nuclear energy and non renewable energy sources. Usage: This plot at a glance helps in identifying countries that source their energy needs from renewables compared to non renewables etc. Also the bar chart of the absolute measure of the usage updates with the country selected in ternary chart to aid exploration and discovery. Year slider lets user to visualize the positioning of the country each year. And a slide through years 2006-2017 helps in identifying if a country has taken a serious stand on adopting sustainable energy options. Critique: Given that energy data logically can be divided in to three categories i.e non-renewable, hydor and nuclear and other renewables, ternary plot aids in plotting large number of countries. Dominant characters are identified and clusters emerge enabling classifications/identification of trends Tools used: plotly

Some Useful Insights

The Geofacet plot reveals that Japan is reducing their nuclear power production drastically after 2011 nuclear power plant leak, Fukushima Daiichi due to tsunami and earth quake Japan’s core focus has shifted from nuclear to other renewables. Last 5 years growth is remarkable, it has replaced about 33% of its nuclear consumption through other renewables Taiwan which is at higher risk for tsunami and earthquakes is following similar the same trend and has made significant reduction in their nuclear consumption. In 2016, winning government has stated phasing out nuclear power generation in their agenda With the reduction in the prices of Oil, the consumption from 2013 to 2017 has gone up for Oil across all countries. This can be visualized using the time-series pattern in the Geofacet graph for Oil For Finland, Iceland & Sweden, it is observed that a major proportion of energy consumption is from other renewable sources of energy From the Sunburst plot we can see that China produces 13.4% of the total energy of the world using Coal; Saudi Arabia produces 4.31% of the total energy of the World using Oil China, US & India are the top 3 countries in the Primary Energy Consumption in 2017 From the Geofacet plots we can see that China has been trying to reduce it's Coal consumption after 2012. If we look at Africa, Algeria has been able to reduce it's Coal consumption significantly over the years. The ternary plot shows us that the major energy source for Norway in terms of consumption is Hydro energy, which is a good indication. However, if we look at India, the major energy consumption is for Coal and Oil which are non-renewable sources of energy. US primarily consumes Oil, followed by Gas. China, over the years has tried to increase the percent contribution from renewable, hydro and nuclear energy sources as opposed to non-renewable energy sources.

Conclusion/Future Work

The original dataset has several other metrics like prices of each energy source, reserves in each country, electricity generation, trade movements etc. These variables also must be included for analysing energy dynamics and the world energy outlook. Data driven visualization and analysis on energy dataset reveals patterns in production/ consumption of energy across countries with time. The relative contribution of different energy resources in a country’s consumptions helps in quantifying the effort that the country puts in moving towards the sustainable energy initiative. Having said that, the domain knowledge and experience of personnel in energy industry is essential to support or rule out the insights that visual analysis has revealed . This application is a starting point for interactive analysis of energy data using R visualization packages and shiny application platform. Further, detailed breakdown of renewable sources, energy prices data and trade movements data can also be used to enrich current data and derive additional insights. Adding electricity generation data and power distance would reveal the dynamics of today’s world energy landscape.

R Packages Used

We have used the following R packages to come up with our visualizations: dplyr: A Grammar of Data Manipulation. It is a fast, consistent tool for working with data frame like objects, both in memory and out of memory. tidyr:It's designed specifically for data tidying (not general reshaping or aggregating) and works well with 'dplyr' data pipelines reshape:Casts a molten data frame into the reshaped or aggregated form you want readr :The goal of 'readr' is to provide a fast and friendly way to read rectangular data (like 'csv', 'tsv', and 'fwf'). It is designed to flexibly parse many types of data found in the wild, while still cleanly failing when data unexpectedly changes ggplot:A system for 'declaratively' creating graphics. You provide the data, tell 'ggplot2' how to map variables to aesthetics, what graphical primitives to use, and it takes care of the details Plotly:Easily translate 'ggplot2' graphs to an interactive web-based version and/or create custom web-based visualizations directly from R SunburstR:Make interactive 'd3.js' sequence sunburst diagrams in R with the convenience and infrastructure of an 'htmlwidget'. Crosstalk:Provides building blocks for allowing HTML widgets to communicate with each other, with Shiny or without (i.e. static .html files) Geofacet:Provides geofaceting functionality for 'ggplot2'. Geofaceting arranges a sequence of plots of data for different geographical entities into a grid that preserves some of the geographical orientation rgdal:Bindings for the 'Geospatial' Data Abstraction Library leaflet: Library to create Interactive Web Maps with the JavaScript 'Leaflet' shiny: Web Application Framework for R shinythemes: Themes for use with Shiny. Includes several Bootstrap themes shinydashboard: Create dashboards with 'Shiny'. This package provides a theme on top of 'Shiny', making it easy to create attractive dashboards

References

[1] Shneiderman, B. (2005) “The eyes have it: A task by data type taxonomy for information visualization” IEEE Conference on Visual Languages (VL96), pp. 336-343 [2] About BP: https://en.wikipedia.org/wiki/BP [3] Data: http://www.bp.com/statisticalreview [4] http://ryanhafen.com/blog/geofacet [5] https://hafen.github.io/geofacet/ [6] https://github.com/timelyportfolio/sunburstR [7] R Packages Description: https://cran.r-project.org