Harness Oil and Gas Big Data with Analytics: Optimize Exploration and Production with Data Driven Models

Chapter 7Production Forecasting

I have seen the future and it is very much like the present, only longer.

Kehlog Albran, The Profit

The art of predicting production and estimating the ultimate recovery in oil and gas reservoirs has stimulated much debate among upstream engineers across recent decades. The literature in the early years of the twentieth century immersed itself in the study of percentage decline curves or empirical rate-time curves that found credence in expression of production rates across successive units of time, framed as percentages of production over the first unit of time. W. W. Cutler¹ opined that a more robust methodology defining a straight-line relationship was achievable when using log-log paper, the implication being that decline curves that reflected such characteristics were of a hyperbolic geometric type as opposed to an exponential.

We use the decline curve equations to estimate future asset production:²

1. Hyperbolic decline:
   q = q (1 + D bt)^−1/b
2. Exponential decline:
   q = qi exp^(−Dt)

In the equations, b and D are empirical constants that are determined based on historical production data. When b = 1, it is a harmonic model, and when b = 0, it yields an exponential decline model.

There are a number of assumptions and restrictions applicable to conventional decline curve analysis (DCA) using these equations. Theoretically, DCA is applicable to a stabilized flow in wells producing at constant flowing bottom-hole pressure (BHP). Thus data from the ...