There are several methods to forecast production and estimate oil and gas reserves in unconventional shale reservoirs and although they all require different inputs, all reservoir engineers can agree that the more production history available, the more reliable will be the forecasting no matter the selected methodology.
Decline Curve Analysis is probably the most common (and easiest!) method to estimate ultimate recovery and has been used by reservoir engineers throughout the years. Although little data is required to use this method in comparison with other forecasting processes such as Rate Transient Analysis and Numerical Simulation where more knowledge of the reservoir and fluid is needed, the question that arises is whether it is correct or not to use DCA during the early life of a well.
The goal of this work is to quantify the uncertainty associated with the estimation of the ultimate recovery with Decline Curve Analysis in order to be able to predict how much production history of a well is needed for the production forecast to be considered accurate and reliable. In other words, we must find, and stablish certain cut-off parameters related with the declinatory process that would allow the reservoir engineer to determine how soon can DCA be implemented in an unconventional shale reservoir to estimate reserves in a reliable way.
When an engineer faces the need to perform a reliable production forecast in an unconventional reservoir, there are various tools inherited from classical reservoir engineering that can be implemented among which are the simple and traditional Decline Curve Analysis (DCA in all its variants), Analytical Models, very useful but only valid under certain conditions, and a step further we have Numerical Models which are much more robust but additional effort and data are required to obtain a reliable production forecast (Figure 1).
Decline Curve Analysis (DCA) is one of the oldest methodologies used by production and reservoir engineers to predict future production performance, calculate reserves, estimate ultimate recovery, and generate an economic valuation of oil and gas assets. Moreover, this simple method requires a minimal amount of data to predict performance minimizing errors. This contribution uses the public production database of Vaca Muerta horizontal wells to explore results of different DCA methods with sensitivities that allow us to better understand the minimal production history needed to successfully estimate the ultimate recovery of a well. In order to be able to decline hundreds of wells with different sensitivities in a reasonable amount of time, an in-house tool was developed in a joint project that included reservoir