BookmarkSubscribeRSS Feed
arati_mejdal

Staff

Joined:

May 21, 2014

Visual analytics and design of experiments for the petroleum industry

I am excited that JMP is going to the Society of Petroleum Engineers ATCE conference for the fourth time in five years. For the past 87 years, ATCE has been the leading technical conference in the Exploration and Production area of the oil and gas industry. JMP is so flexible that it can be used in every industry –- and I have multiple examples of using JMP for production in the oil industry. I'm looking forward to showing what JMP can do. So if you will be at the SPE ATCE conference Oct. 30-Nov. 2 in Denver this year, please stop by our booth and take a look at all of the applications of JMP in this field.

For example, design of experiments in JMP has traditionally been used in oil production because it is powerful and flexible for creating good designs for simulators. Some use the custom designer to create fewer runs, and some use the space-filling designs, which are specifically designed for use with simulators. JMP has multiple powerful and visual modeling platforms to analyze the results.

Also, JMP visual analytics can be used to increase production rates and reduce costs. Analysis of historical data, both geological and process-related, can provide valuable insights for improving production and reducing environmental effects. JMP reliability analyses will allow reliability engineers to forecast potential failure and determine how to target failures. The following graph compares the life cycle of the top-performing pump type (Pump B) to the lesser-performing pump types.

JMP Pro is a more advanced version of the software that brings the power of predictive analytics to production and reservoir engineers. JMP Pro has additional modeling techniques to create predictive models in adverse modeling situations. The Prediction Profiler found in both JMP and JMP Pro allows engineers to dynamically visualize the effects of each predictor on the response. They can then analyze various scenarios by instantly changing factor values and seeing the impact on the response. The graphic Monte Carlo simulator allows them to use simulation to see the effect while including the variability and the distribution information. The graph below shows a sorted list of predictors that JMP Pro determined affects the recovery factor in a specific well type.

The graph below shows the Prediction Profiler with the graphic simulator based on a Neural Net analysis on the six most important predictors of the same data.

1 Comment
New Contributor

Dear Dr. Mejdal,

 

I am Mina R. Mousa, a petroleum engineering PhD candidate at Texas A&M University. I am sending you this email to ask if you can kindly help me with a presentation I prepared about the role of design of experiments (DoE) in research design and analysis in the petroleum industry. I am planning to give it in a paper contest for PhD students organized by the Society of Petroleum Engineers. I am also planning to show it to the professors in our department. The reason I am contacting you is that I use JMP in all my work. I also found that you visited some conferences organized by the Society of Petroleum Engineers before. So you have an idea about what is going on in our industry.

 

In the petroleum industry, we have a problem applying statistics in general and DoE. This is more obvious in the lab experiments and field trials. For example, no one of the 20 accredited petroleum engineering programs focuses on teaching DoE. In our department at Texas A&M, we have 27 labs. None of them uses DoE. They mainly use one-factor-at-a-time and trial and error. Consequently, such studies are inefficient and could lead to inaccurate conclusions. I have prepared a review paper (not published yet) talking about the problem, current challenges, and how to help mitigate it.

 

Thank you so much for your consideration, and I am looking forward to hearing back from you. 

 

Sincerely,

Mina

E-mail: mina.shaker@tamu.edumina.r.mousa@gmail.com

Personal profile: Link