Production data is usually analyzed by plotting the normalized rate (q/Δp) Vs. Material Balance Time (MBT) on a log-log scale. A late-time unit-slope is usually interpreted as boundary-dominated flow (BDF) and can be used to calculate the reservoir volume. However, a decreasing flow rate not only causes MBT to be larger than Maximum Producing Time (MPT), in certain cases it can also create an artifact of false unit-slope giving the wrong impression of achieving BDF. There are claims in the literature (Agnia et al., 2012) that all data beyond MPT should be ignored.
This paper presents a detailed analysis of the issue, and investigates the conditions under which data beyond the MPT should be ignored. Three different reservoir models have been used with associated pertinent flow regimes to cover and demonstrate the subject. Specifically, it addresses questions such as: What causes MBT to be larger than MPT?; Under what conditions can that data be meaningful? and Should MBT data that extend past the MPT must always be ignored?
The question whether MBT is valid beyond MPT and under what conditions, has been investigated in great detail in this paper. The observations are that:
If flow is transient, then any data beyond MPT should be ignored (except for a constant flowing pressure).
If flow is boundary-dominated, data beyond MPT is valid and it accentuates the correct interpretation.
To differentiate between transient and boundary-dominated flow, use of external information is often necessary.
Bad data will often manifest themselves beyond MPT and should probably be ignored.
It has been shown that even when flow is transient, a unit-slope may be observed on the q/Δp Vs. MBT plot. This is a false trend and will give the wrong HCIIP. When the flow is Transient, the volume of investigation is still increasing with time, and any volume corresponds to a "minimum." Observations during this study indicate that the "calculated volume" based on the apparent (false) unit-slope data is approximately one-half of the "correct reservoir volume." However, when flow is boundary-dominated, the observed unit-slope on the q/Δp Vs. MBT plot is a true trend and will give the correct HCIIP.