Sita Dawanse:

GitHub_link

Trends in Renewable Energy Adoption

This dataset was obtained from the Kaggle dataset webpage Global Data on Sustainable Energy This dataset contains sustainable energy indicators and other useful factors across all countries from 2000 to 2020, including vital aspects like electricity access, renewable energy, carbon emissions, energy intensity, Financial flows, and economic growth. With the help of this data we expect to gain profound insights into global energy consumption patterns over time.Therefore the goal fo this project is to investigate the pattern of energy consumption and their consequences on climate by the emission greenhouse gas, global temperture rise.

We would like to examine electricity produced from different resources by different countries and how the production and consumption has changed over time, access of electricity to the percentage of population over time. This data provides a score for each country individually every year, we can use this data to compare the trends from different countries and dive deeper into how the production, consumption and accessibility are affected by financial flows, and economic growth. We can also answer questions such as, what issues are most prevalent in the world for the clean and renewal energy? Furthermore, we can examine the data alongside other country-wide factor datasets such as gdp_growth, gdp_per_capita, Land Area, Latitude, Longitude.

The transition towards a more sustainable energy mix is a critical challenge for countries worldwide. The generation of electricity plays a crucial role in this transition, as different sources of electricity generation have varying environmental impacts, particularly in terms of carbon dioxide (CO2) emissions. Understanding the relationships between electricity generation from different sources (fossil fuels, nuclear, and renewables) and CO2 emissions is essential for informing energy policies and strategies aimed at mitigating climate change.

The Questions that I'm addressing here is

  1. What is the relative contribution of fossil fuels, nuclear, and renewables to a country's total electricity generation, and how has this mix changed over time?
  2. What is the relationship between a country's renewable energy capacity and its share of renewable energy consumption, and how does this relate to CO2 emissions?

ETL (Extraction, Transform, and Load)

We load the dataset Global Data on Sustainable Energy to Pandas DataFrame. This dataset shows the year, country, each country’s access to electricity, access clean cooking fuel, financial flow, electricity from different fuels and all others variables. We will tidy the data by taking the the specific columns required for the analysis and visualization.

Retrieving the data corresponding to countries located in North America

The above plot shows a strongly skewed distribution to the right, with the majority of countries having relatively low primary energy consumption per capita, as represented by the tall bars on the left side of the plot. The top bar reflects a significant number of countries, with primary energy consumption per capita ranging from 0 to 25,000 kWh.As energy consumption per capita grows, the height of the bars decreases, showing that fewer countries are in the higher consumption range.There is a large tail on the right side of the plot, indicating that a small number of countries have extremely high primary energy consumption per capita, ranging from approximately 75,000 kWh/person to 150,000 kWh.

A bar plot with distribution is a suitable visualization choice for displaying the primary energy consumption per capita data as it shows the frequency or count of observations (countries/regions) for different ranges of primary energy consumption per capita values. This allows us to see the overall distribution shape and identify any clusters or outliers in the data.

Question: Is there a correlation between energy consumption per capita and its economic growth (GDP growth)?

I'm investigating relationships between variables so scatter plot will be suitable for it.

This graph shows the relationship between a country's GDP growth rate and primary energy usage per capita. The x-axis shows the GDP growth rate (in percentage), which ranges from negative (economic contraction)to positive(economic growth). The y-axis represents primary energy usage per person (kWh/person). Each data point on the scatter plot represents a specific country and is colored differently to make it easier to identify. The distribution of these data demonstrates that countries with greater GDP growth rates tend to have higher energy consumption per capita, and vice versa. The overall trend indicates a positive association, however,certain countries diverge from this pattern. Some countries have relatively high energy consumption despite poor or negative GDP growth, while others have relatively low.

This plot shows the trend of access to electricity as a percentage of the total population over time (from 2000 to 2020) for various countries in the Caribbean and Central American regions. Most major countries, including Canada, United state have relatively high and consistent access to power, hovering around 100% during the time. Haiti has by far the lowest access, with only about 40% expected by 2020, indicating severe electrical access deficits when compared to its neighbors. Jamaica Panama, the Dominican Republic also showed steady improvement, reaching near-universal access around 95-100% by 2020 after being in the 70%-90% range in 2000.

This visualization efficiently uses multiple line charts to display time-series data for "Access to electricity (% of population)" across various countries. Line charts are an excellent choice for illustrating trends and patterns in variables across time. The addition of multiple graphs allows for a comparative analysis of power access levels and trends within a given region.Given the importance of electricity access as a development indicator, this visualization provides useful insights into the current state and evolution of electrification in the specified region.

This box plot likely shows the distribution of access to clean cooking fuels across various countries. The horizontal axis (X-axis) represents countries and the vertical axis (Y-axis) represents a score indicating access to clean cooking fuels (higher values represent greater access). The box itself contains the middle half of the data (interquartile range). The line in the middle of the box represents the median (the 50th percentile, which divides the data in half). The whiskers extend from the box to the lowest and highest data points within 1.5 times the interquartile range. Any data points beyond the whiskers are considered outliers and are plotted individually.

This visualization offers a regional perspective on energy supply patterns. It reveals a common reliance on fossil fuels among the depicted countries. However, the adoption of nuclear and renewable alternatives varies considerably across these nations.

The countries' total energy supply levels differ greatly, with the United States having the greatest at over 4,000 TWh, while tiny nations like Saint Lucia and Saint Kitts and Nevis have far lower levels, around 1 TWh. Most countries continue to rely heavily on fossil fuels (blue bars), however, some, such as Canada and Costa Rica, have improved in adopting renewable energy sources (green bars). Nuclear power (orange bars) plays an important role in a few countries, such as the United States and Canada, but is missing or insignificant in many others.

This graph shows the electricity generated from nuclear power sources in North America by three countries: Canada, Mexico, and the United States, over the period from 2000 to 2020.The y-axis displays the amount of power generated by nuclear sources, measured in TWh (terawatt-hours), and the x-axis indicates the years.Throughout the provided time period, the United States has consistently produced the most nuclear-generated electricity of the three countries. The trend line for the United States (green) shows very consistent and high levels of nuclear energy generation, ranging around 800 TWh. Canada's nuclear electricity generation (blue line) has been consistent throughout the years at roughly 90-100 TWh, which is much lower than the US but greater than Mexico (orange line) has generated less nuclear electricity Throughout the period, its nuclear electricity generation has been less than 20 TWh.

The correlation matrix shows the relationships between various variables related to electricity generation sources, CO2 emissions, financial flows to developing countries, and economic indicators like GDP growth and GDP per capita.

Analysis of the correlation matrix reveals a strong positive association (0.98-0.99) between fossil fuel electricity generation and CO2 emissions, indicating that countries hat rely heavily on fossil fuels have higher emissions. Nuclear power exhibits a moderate positive correlation (0.81-0.97), suggesting lower emissions than fossil fuels, but not completely emission-free. Renewables display the least positive correlation (0.74-0.76) with CO2 emissions, indicating their potential as cleaner energy sources. Notably, the correlation matrix reveals a strong positive correlation (near 1) between all electricity generation sources. This suggests an interconnected energy system where countries utilize a diverse mix of generation methods to meet their energy demands. Overall, this matrix emphasizes the considerable impact of fossil fuel-based energy generation on CO2 emissions and the potential benefits of shifting to cleaner sources such as renewables and nuclear power to reduce emissions while meeting energy demands.

The x-axis indicates renewable energy capacity per capita, and the y-axis on the left displays renewable energy share (in percentage). On the right, the y-axis shows CO2 emissions (kt). Each data point on the scatter plot represents an individual country. Countries with higher renewable energy capacity per capita have a greater proportion of renewable energy in their overall energy mix. In general, there is an inverse link between renewable energy share and CO2 emissions, implying that nations that use more renewable energy have lower CO2 emissions. However, there is significant variation among countries, with some outliers having relatively high renewable energy capacity or share but still exhibiting high CO2 emissions, and vice versa. The data suggests a connection between renewable energy use and CO2 emissions; however, a significant spread is observed across countries. Some outliers might possess high renewable capacity but lack efficient infrastructure for integration into the energy grid, leading to continued reliance on fossil fuels. Conversely, other outliers might have limited renewable resources but have adopted stricter emission control measures for existing power plants.

The magnitude and sign of the coefficients indicate the relationship between each features variables and the target variables(in this case CO₂ emission). The positive coefficient means as the feature variable increases, the target variable tends to increase. The negative coefficient indicates that as the feature variables increases the target variable should also increase. Therefore, here we can see that as the production of Electricity from nuclear source increases the emission of CO₂ decreases and the CO₂ emission from the renewal sources is around 10 times less that from fossiles fules.

R² measures the proportion of the variance in the dependent variable that is predictable from the independent variables.R² ranges from 0 to 1, higher R² values here indicate a better fit of the model to the data. RMSE is similar to MSE but is in the same units as the target variable. It's more interpretable compared to MSE.

In conclusion::

A lower RMSE obtained using a Random Forest model compared to other models suggest that the Random Forest model performs better in terms of predicting the target variable.

Working with Asian countries Data set

The graph above shows a line graph of "Access to electricity" percentages for various countries across Asia from 2000 to 2020. It provides an overview of the progress made by different nations in providing access to electricity over the past two decades. Countries like Japan, Singapore, Qatar have full excess to electricity, but some countriel like Iqran, Malaysia, China have gradually improved overtime and reached the ultimate goal, for 100% excess to electricity.

This above graph shows how much each country have the contribution of different energy sources to its total electricity production over time. It also that most of the developed and developing nations are mostly relying on fossil fules. Also the countries with 100 % electricity excess are heavily relying on fossil fuels. Some of the countries like Kyrgyzstan, Nepal, Bhutan Afghanistan are relying renewal fules for the electricity generation

The correlation matrix shows the relationships between various variables related to electricity generation sources, CO2 emissions, financial flows to developing countries, and economic indicators like GDP growth and GDP per capita.

This model coefficients indicate that for the Asian countries, elctricity generation from Nuclear and Renewal sources have inverse effect on CO₂ emission. But the electricity generation from fossil fules have positive effect on CO₂ emission.

European Countries

From the graph above we can conclude that most of the countries in Europe have full excess to electricity, except some countries where we can see negligibly small fluctuations indicating countries in Europe have good excess from very long time.

Several countrie's, like as Norway, Iceland, Sweden, and Switzerland, have a significant share of electricity generation from renewables, likely due to their hydropower resources and emphasis on sustainable energy. Countries like France, Belgium, and Ukraine have a substantial nuclear component in their electricity mix, indicating their historical investments in nuclear power infrastructure. Fossil fuels remain a dominant source of electricity generation for many European nations, though some countries like the United Kingdom and Germany have made efforts to transition towards cleaner sources over the years.

From the over all analysis of the given data and the calculations done using different machine learning models. And finding the different metric values like MSE, R-Square value, RMSE a conclusion can be drawn for this dataset as follows.

  1. Countries highly relying on fossil fules are generating high co2 emission as compared to nuclear and renewal sources. This is expected as fossil fuel combustion is a major contributor to greenhouse gas emissions
  2. Countries relying on renewable and nuclear electricity generation exhibits a moderately positive to highly positive co2 emission, this could be due to several factors:

But over all we can conclude from the calculation that being able to switch to nuclear nad renewal will play a crucial role in environmental impact particularly in terms of Co2 emissions.