Software Technologies : NNLAP

NNLAP: Generate Synthetic Log, Core, and Production Data
PST’s Neural Network Log Analysis Program, called NNLAP, applies advanced neural network technology to generate synthetic, validated wireline logs, core properties, and production data based on commonly available data.

NNLAP

NNLAP works with this data to learn the mathematical relationships between commonly available predictors such as SP, gamma ray, resistivity, and porosity data and the less commonly available targets, including density, acoustic, shear wave, permeability, NMR, and production data.

NNLAP provides a simple, experienced-based method of training the neural network in an interactive depth plot display. Geoscientists use NNLAP and their knowledge of log and rock relationships to select training points. Once these points have been selected, NNLAP provides immediate feedback on the results.

NNLAP Screen Capture The synthetic data that NNLAP generates can be confirmed by testing results against available target data from other wells. The application also provides cross-plot and statistical analyses, as well as automatic back-prediction of predictor data for validating results.

NNLAP allows you to create accurate synthetic logs, such as compressional and shear wave velocity data, for all the wells within a study area. These data can be tie to seismic data and used to determine rock properties in fracture stimulation design.

Other useful features of this application include:

Classification mode, used for rock type predictions
Cumulative mode, for evaluating target data expressed as a percentage
A Data Mining feature allows an end user to quickly evaluate basic mathematical relationships among all available data before beginning a project.

Companies using NNLAP edit logs up to 5 times faster than other available solutions, resulting in significant reductions of their development logging costs. In addition, the synthetic wireline logs, core properties, and production data generated by NNLAP enables the design of more cost-effective completions.

"I had only two wells with expensive core data. With Rt-Mod, SP-Mod and NNLAP I was able to generate synthetic permeability data over the entire well interval for all of the wells in my study area."

Petrophysicist, Major Domestic Oil Company