P2 Solar Mini-Grid: Design Optimization

👋🏾 Hey there! We'd really appreciate it if you took the time to complete our 🗳️ SURVEY 🗳️ and provide us some feedback! Thank you 🙇🏽‍♂️

🛠️ Demo!:

Description:

The state-of-the-art in mini-grid design is poor. Most of it runs on rules of thumb, and the few software tools that exist are notoriously oversimplified and return useless results. This leads to over- and under-sized mini-grids that are unprofitable and/or don't provide reliable power to communities. Renewvia, like many companies, has begun developing small, Excel-based software tools to simulate mini-grid performance. A talented software developer with an interest in machine learning could build on the simulation work already done to test different convex optimization tools to create an automatic mini-grid design tool. Users would input the location of the mini-grid, as well as information about potential customers and financing. The model would, ideally, output a mini-grid design optimized for 20-year IRR, including a shopping list and financial model. We built a proof-of-concept last year that used a genetic algorithm to optimize design that produced promising results. Refining and expanding that tool could finally bring engineering maturity to the mini-grid industry.

Team Members and Roles:

  • Earth - Full-Stack Engineer - Python, SQL, Java, JS
  • Pranav - Front-End Engineer - Angular, React, JavaScript, Java, Python, C++, HTML, AWS
  • Kunle - Front-End Engineer - Python, Go, Java, Javascript, React

Project Goal:

Improve the design optimization of mini-grids by creating an automatic mini-grid design tool that outputs a mini-grid design optimized for 20-year IRR, including a shopping list and financial model.

Lighthouse Scores Desktop:

  • Performance: 100
  • Accessibility: 100
  • Best Practices: 100
  • SEO: 100

Lighthouse Scores Mobile:

  • Performance: 99
  • Accessibility: 100
  • Best Practices: 100
  • SEO: 100