Solution to a Data Science challenge aimed at optimising project planning in an industry
A description of the problem follows below; for a more detailed version, refer to challenge_statement.pdf
- CONTEXT
Imagine an Oil & Gas operator with a portfolio of projects to execute, faced with the decision on how to sequence__these in time. Each project has project attributes such as a production profile (correlating with how much revenuewill be generated over time), a maturity (indicating from when onwards a project is ready for execution), and the type of hydrocarbon that will be produced (Oil or Gas).
- BUSINESS QUESTION
How to optimally plan this sequence of projects, i.e. in what sequence should I execute which projects from my__total portfolio of available projects?
- GIVEN
Assume the following:
3.1 PORTFOLIO
A portfolio with projects to be planned is provided here as an Excel table
Every project has the following properties:
- its name
- whether it's an Oil or Gas project
- the earliest date the project can be executed ('earliest spud year')
- how long it takes to execute ('duration')
- its production profile (how much Oil/Gas is produced, i.e. how much revenue will this project generate)
All projects are assumed to have the same cost profile.
Note: The earliest spud year is not necessarily the year of execution – it's the earliest possible year of execution. (The actual year of execution is a variable to be optimized by you.)
3.2 OPTIMIZATION
To address the business question, please consider the following scenarios:
- Scenario 1: optimize the project sequence for maximum Oil production in 2021-2025.
- Scenario 2: optimize the project sequence for maximum Oil production in 2021-2025 with a desire that gas remains as long as possible around 1M m3/d from 2021 onwards.
- DELIVERABLE
Design one or more approaches that are capable of addressing the business question. Demonstrate the efficiency & robustness of your approach(es).
