Backend GIS Database cleaning and modeling + wind dispersion algorithm - Environmental monitoring.

Summary: Deliverable - Backend database cleaning, modeling Skills required: R, GIS knowledge, Python, Some Data Engineering Automate the backend process of cleaning and aggregating different data sets. Once data is cleaned and in one database there are five key steps: 1) Convert Legal Land Descriptions of well sites to GPS co-ordinates One potential option: [login to view URL] 2) Deduct wind speed and direction from vehicle speed (methane sensor is mounted on vehicle) 3) Run wind dispersion algorithm (This can be a manual step for this project) [login to view URL] 4) Output results of wind dispersion algorithm onto a map visually represent wind dispersion and concentration levels. (does not need to be heavily designed or visually impressive at this stage) A RFP with more details of the project to be shared with selected candidates. Problem: Currently, it takes a person 4-5 days to manually go through all the steps above. Although the end goal for this is full automation a few manual steps such as to initiate the wind dispersion algorithm is fine for this first phase. Goal is to get the time down to a few hours. Introduction : This request for proposal is an invitation to a select group of companies to submit proposals for a project to create a minimum viable product (MVP) that can transform raw methane emissions data from natural gas producers into actionable insights. This project will be established as “MethaneChain - Minimum Viable Product”. Problem: In this use case emission data is being acquired by a sensor on a truck that drives around collecting GPS, Atmospheric, wind and CH4 emission data in parts per billion. Currently the process of moving this data around is done manually or with very basic “R” scripts and it can take one person several days to scrub, prep and analyze the data. The current users come from environmental engineering or university research projects and do not have the computer science expertise to speed up this process. This causes this valuable emission data to go unused until it is often too late to be useful. Goal: The goal of the MVP is to reduce the time of this laborious data process to less than an hour. However the entire process does not need to be automated at this stage. A few manual steps are allowed as necessary to reduce costs. Data workflow. Intent of RFP: The intent is to obtain information leading to the selection of one (1) or more vendors to participate in this project. Through this process, MethaneChain also seeks to develop operational relationships with development vendors who can support our vision. Appendix A - Workflow A. Import Raw Data (Back-End) Import wellhead / facility location and emission data, and associated site IDs Data from publically available datasets Clean and aggregate public data Import raw emissions sensor and associated GIS data Clean data and sort into standardized fields Import environmental (wind, temperature and pressure) data Data sources include Environment Canada and local airport METAR measurements, among other potential sources Clean environmental data and sort into standardized field Import current price per volume of natural gas and other regulatory thresholds (i.e. allowable baseline emissions) Combine raw data into a new database Data should be geolocated to individual sites or regions depending on data type B. Calculate Parameters (Back-End) Run wind dispersion algorithm using site locations as centre points to estimate emissions downwind of sensors Estimate emission rate (i.e. volume per day and volume per year) and leak cost (i.e. dollars/year) for sites using dispersion results and current spot price of natural gas Calculate a rough estimate of the carbon tax liability or potential credit (as compared to baseline) for site. C. Visualizations and Reporting (User Interface) Clean and format calculate data for export and visualization Load data into visualizations