NPI (log-log)

Return to the data diagnostics flowchart

These questions apply to the normalized pressure integral (NPI) – log-log:

Does the normalized pressure continuously increase?

  • The log-log plot of normalized pressure (∆pp/q) versus material balance time (Gp/q) should have an increasing trend.
  • A step change in rate causes a sudden deviation from the inclining trend, but if the data is consistent, it reverts to the original trend.

Reasons for inconsistencies:

  • A decreasing trend indicates non-reservoir effects, such as near wellbore cleanup, improving skin, wellbore liquid unloading (if pressures have been calculated from wellhead data).

Does the NPI plot have a single continuous stem?

  • Changes in rate or pressure are accounted for in the NPI plot, and should not cause it to develop multiple stems.
  • In the case of multiple stems, the stem that represents the correct reservoir model should be used for analysis.

Reasons for inconsistencies:

  • The reservoir model has changed (for example, work-over, recompletion, refrac, etc.).
  • The wellhead pressures measurement location may have changed (for example, tubing, casing, before / after valve), or there may have been a change of flow path or tubulars.

Are outliers being misinterpreted as boundary-dominated flow?

  • Outliers often appear on a unit slope on the NPI plot. This can lead to an incorrect diagnosis of reservoir depletion, or boundary-dominated flow (BDF).

Reasons for inconsistencies:

  • Outliers have been included in the analyses. Apply the median filter to remove outliers.