The Infill Drilling application is a quicker way of utilising large backend oil well data through a friendly graphical user interface (GUI) compared to the existing command line interface (CLI) used server side. Not only is it quicker to use due to its interface being easier to operate and accessible on most client-side devices, it quicker computationally by breaking down large processor intensive models into smaller bite-sized datasets without losing fidelity.
This project has taken an industrial computational process that takes up to 24 hrs to run (assuming adequate queries) down to minutes for users to see the results. As drilling operations even for smaller operations have overheads in the $100,000’s of dollars, the risk of downtime waiting on data driven operations is significantly lower.
My Role:
User Research, Interaction, Visual design, Prototyping & Testing
Duration:
January'20 - April'20
The Infill drilling App
Infill drilling analytics helps the user to identify robust infill well targets under uncertainty. This application gathers relevant information from the ensemble and automatically identifies targets and calculates hydrocarbon volumes within these target regions.
1. Create study
A list of studies will appear, with general information regarding previously made studies. Here the user can see that status, date of creation, ensemble, creator, and last modified by of each listed study.
2. Add criteria
In this part, the user can choose the basis of the target search from five different criteria: remaining oil, oil recovery factor, depleted pressure, remaining gas, and gas recovery factor.
3. Risk Filters
In this part, the user can implement their risk-preference when searching for infill targets. For each selected criteria in the property filter page, there will be a corresponding risk histogram illustrating the occurrence (probability) of cells which values are inside the previously-defined range across all realizations.
4. Well Targets
In the analytics page, the user will be able to visually learn where the targets are based on the search criteria and compare the statistics of in place volume between these targets.
Understanding the Context of Infill
Stakeholder interviews
Identifying pain points of users
Brainstorming different ideas
Wireframing & Prototyping
User Testing & Evaluation
The users of the solution are teams of Geologists and Reservoir Engineers. They work in collaboration to assess data driven research from secondary sources, pilot studies such core samples taken before site development and on-site geological sampling to cross-validate predictions based on existing studies and data analysis. As more information about the well is gathered they continually process primary data and cross-reference existing data. Computational time is a major bottleneck in well drilling operations as failure to predict well conditions could lead to wasted time and associated overheads.
I collaborated with the researchers and domain experts in the team to understand our core users, their environment, and majors tasks to accomplish. We interviewed few participants and gathered data on their current workflow and major challenges. This data helped us in finding opportunites to solve the problems.
We chose four Reservoir Engineers to interview. Interviews conducted were semi- structured in nature. Most of the interview were done with our internal team member who are Reservoir Engineers.
1. What's the context?
2. What are your major tasks?
3. How do you report & analyze data?
4. What are the major problems faced?
5. How do collaborations happen?
6. What tools do you currently use?
Observations were conducted to get and understanding of the Infill drilling process from various Reservoir Engineers and Asset team managers to study the tools / mediums used in their current workflow.
After conducting interviews and observations I created user personas with information about their behaviors, technologies used, pain points & their needs. View Persona
Every piece of data you have to import one by one which take a lot of effort and time
Users want to see the results in a concise and visual format which is easy to understand, share with stakeholders.
Running simulation often requires submitting hundreds of jobs on a cluster.
Due to the complexity of calculations they often fail to complete & deliver meaningful analytics on time.
1. Time consuming calculations in Petrel
2. Organizing & complex work flow
3. Long reports and high human error
4. Tedious process to analyze the data
How might we empower Reservoir Engineers and Geologist to find new target to drill wells, so that they can quickly make appropriate recommendations for corrective actions?
As we moved on to designing the interface of the application, I wireframed all the UI screens and conducted an intial user testing session.
Version 1
ISSUES
Repetitive links and buttons on the page
we might have more details of ensemble and studies
Version 2
ISSUES
Long page with too much data
Users probably be confused to have so many action on same slide
Version 3
ISSUES
Side navigation leaves less space for study details
Details are important for users
Currently the app is used by few companies.The asset team using the software are so happy that currently a second version of the software is in development to expand the filters for more diverse well finding scenarios.
User testing lead to the addition of assessment of other local well sites to the application. Useful information that may not have been discovered without following an established design methodology.
Points which are not covered in this case study but could be worth discussing in person:
Thanks for making this far. Here's a gif for you.
Have any suggestion or interested in working together on a project?
Go ahead, I'd love to connect with you.
© Designed & Coded by Raheel Ahmed