Introduction
My first interest was visualizing a comparison of different renewable energy sources currently used in New York State. I started with data on the Solar Photovoltaic Incentive Program. Prior to 2012, incentives were provided by several smaller programs organized by the New York State Energy Research and Development Authority (NYSERDA), Long Island Power Authority (LIPA), PSEG Long Island, and the New York Power Authority (NYPA). In 2012, Governor Cuomo launched NY-Sun to provide monetary incentives to residents, commercial operations, and nonprofits to install Photovoltaic panels to provide solar power in lieu of traditional energy sources.
After combing through the data, I found several areas that could be potentially visualized. I chose four areas which would answer the following questions:
- How does Photovoltaic (PV) production change by location in NYS?
- Which sectors (residential, nonprofit, etc.) benefit the most from these incentives?
- Which energy companies are responsible for PV installation through these incentives?
- How does the total cost of PV projects compare to the amount of incentives provided?
Inspiration
This visualization shows estimated solar energy production in the United Kingdom in 2014 broken down by season, month and hour of the day. It uses a dataset very similar to my own, but focuses just on just energy generated. At first glance it is visually pleasing and thematically appropriate, but does not provide the level of detail I wanted in my visualization. Nor does it offer any means of comparison except for production over one year’s time.
A second visualization compares solar energy generated by countries of the world and breaks down these into regions. While this visualization uses several different methods to make this comparison (line graphs, a map, and a treemap) it reuses the same data for each comparison. The second page does address growth of the solar industry, but again only measures growth in terms of energy generated. Ideally I would like more detailed and nuanced comparisons; I’d like to address issues other than simply energy production, such as cost.
The final visualization I considered is also the most robust. Authored by the International Renewable Energy Agency (IRENA), it includes comparisons of energy production as well as cost, investments, employment in the renewable industry, and even patents related to renewable energy. The visualization begins with a World map showing the most basic level of information, then allows the user to dive deeper based on their interest. It also includes other types of renewable energy, not just solar. This is perhaps the best example I found that inspires the design and detail I want in my visualization.
Methods
Data
Searching for public information on sustainable energy led me to the public data site for New York State (data.nyc.gov). Under the Energy & Environmental data subset I found the dataset called: Statewide 200kW or Less Residential/Non-Residential Solar Photovoltaic Incentive Program: Beginning 2000. After reviewing the data in Open Refine I found that the data was already normalized with very few NULL values. While filtering the data in Open Refine I also found the four areas I wanted to visualize in Tableau.
Tableau
A bad bar graph
After uploading the data to Tableau, I started plugging different variables into the row & column fields to compare different styled visualizations. I wanted to choose four visualizations that would best address the questions above and be visually understandable. For each question, I tried different styles of charts & graphs to find the one that worked best at representing the data at hand. For example, the comparison of inverter quantity & energy production by sector was originally a bar graph, then a treegraph, before settling on a bubble chart. Because the residential sector was so much greater than the other sectors, the visualization as a bar graph did not work as intended; the bars for three of the sectors was so small that it appeared to be non-existent. A bubble chart can better display the smaller values in comparison to one extremely larger value. The large, yellow Residential bubble also alludes to the Sun, which helps subtly illustrate the connection between the Sun and solar power.
Bubble chart as the Sun
Tableau also allows fine level selection of color for visualizations. I chose the gradient from black to yellow as the best way to show the range from low to high of annual energy production in KWh. Yellow is often associated with electricity and black often associated darkness or the lack of electricity (e.g. a blackout). To represent inverters, I chose cobalt blue which contrasts nicely with the electric yellow. For the line graph showing project cost versus incentives, I used green to represent cost and teal for incentives. Green is the color associated with money in the US and the teal color was suggested by Tableau in the same color palette as the money green.
Results & Discussion
Now with four individual charts, they could be arranged together on a dashboard with filters and legends added to help illustrate the data’s story to the viewer. Estimated solar energy production was mapped over time by zip code; the default view shows the running total for each zip code, but the user has the choice to view by each year as well. This also filters the other elements of the dashboard by year. Users can also filter the dashboard by clicking on the Sector and Utility Company visualizations to see data only from one sector or one particular energy company.
After completing the dashboard, I saw several trends. The first is that residential solar power dominates the incentive programs in both inverter quantity and energy production. Geographically, Long Island and the Hudson Valley produce the most energy from these incentive programs, which is especially highlighted due to color selection. There was also a great rise in project costs and a slight rise in incentives paid from 2012 onward. After researching possible explanations I found that in 2014 Governor Cuomo announced $400 million in incentives over four years to stimulate the growth of the solar power industry. This might explain the large jump so I annotated this event on the line chart to give viewers background on the solar industry in NYS at the time.
Future Direction
The goal of this visualization was to show the change in solar energy production in New York State through incentive programs. The future of many data visualizations rely on more data to be published. The data I found was limited to incentive recipients and 200kW or less. It also reports the estimated energy production not the actual contribution. If explored further I would add other sustainable energy sources such as wind or hydroelectric power. Perhaps including oil & gas production would add more depth to the picture of New York’s state of energy.