Synopsis

The NREL Market Mapper is an application that visualizes the current state of the US solar photovoltaic market. It includes installation data for both state and county level along with charts showing the number of installs, total capacity and average cost over a ten year range.

Code Example

One of the main focuses for this application was performance. Instead of directly hitting NREL's API for every call, we decided to create our own internal API. The application uses rake tasks along with a crontab to hit the NREL API nightly to update the application's database with the most recent solar installation data.

Here is an example rake task that hits the NREL API and updates the application database:

desc "Updates the NREL datebase if there is new solar data"
task update: :environment do
  us = State.find_by(abbr: "US")

  service = NrelService.new
  data = service.summaries(mindate: YEARS["2016"][:min], maxdate: YEARS["2016"][:max])
  summary_2016 = us.summaries.find_by(year: "2016")
  summary_2016.update(avg_cost: data[:result][:avg_cost_pw],
                      capacity: data[:result][:total_capacity],
                      total_installs: data[:result][:total_installs])

  puts "Updated 2016 NREL data for US"
end

The NrelService.new creates a PORO that allows us to easily make API calls to NREL's endpoints within the application. Two of the endpoints that we can hit are the index endpoint and summaries endpoint. In the case below, we are hitting the summaries endpoint, passing in a hash for the minimum date and maximum date which represent the start and end of 2016.

The code:

service.summaries(mindate: YEARS["2016"][:min], maxdate: YEARS["2016"][:max])

Evaluates to the endpoint:

https://developer.nrel.gov/api/solar/open_pv/installs/summaries?api_key=YOUR_API_KEY&mindate=1451606400&maxdate=1483228799

Our internal API returns the summary data for the US as a whole and for each state. Our API returns only the specific data needed to create the map and charts, cutting down the amount of data that is sent to the front-end.

Here is the summaries endpoint for accessing the US data:

nrel-market-mapper.herokuapp.com/api/v1/summaries

And here is an example of hitting the summaries endpoint for Colorado:

nrel-market-mapper.herokuapp.com/api/v1/summaries/find?state=CO

Motivation

NREL previously created their Market Mapper application in 2009 using the now defunct FlexMapper. This caused NREL to pull the Market Mapper from the Open PV visualization gallery because the application was not able to be maintained using the original technologies. A complete overhaul was needed using modern technologies.

The motivation was to recreate the Market Mapper with a focus on performance and mobile-first design since the majority of users now access websites through their mobile devices. This meant implementing responsive design to account for the variety of device sizes.

Example Usage

Mobile usage:

Desktop usage:

Contributors

Kimiko Kano - github Julian Feliciano - github