Curve Matching

Key Takeaways

• The Bourdet pressure derivative transformed pressure transient analysis by making different flow regimes visually distinct on a log-log plot — before the pressure derivative became standard in the 1980s, interpreters struggled to identify the beginning and end of different flow periods (wellbore storage, infinite-acting radial flow, boundary effects) from the pressure buildup curve alone; the pressure derivative (the rate of change of pressure with respect to the log of time) creates a distinctive signature for each flow regime: a unit-slope line on the derivative plot indicates wellbore storage, a flat (horizontal) derivative indicates infinite-acting radial flow (the portion from which permeability is derived), and a rising derivative at late time indicates a closed boundary; matching the field derivative to the appropriate Bourdet type curve allows the interpreter to read off the radial flow permeability-thickness product and the skin factor directly from the match point, making pressure transient analysis faster and more reliable than the earlier straight-line methods that required the interpreter to identify which portion of the curve corresponded to radial flow — a judgment that was often ambiguous without the diagnostic power of the derivative.
• Decline curve analysis using Arps curves is the most widely applied curve matching technique in the oil and gas industry because it requires nothing more than the production history of a well — no well test, no special data, just the monthly oil, gas, or liquid rates over time; Arps (1945) showed that most production decline curves fit one of three mathematical forms: exponential (constant fractional decline, b=0), hyperbolic (variable fractional decline, 0 < b < 1), and harmonic (b=1); matching the observed production trend to the appropriate Arps curve allows the engineer to extrapolate the production forecast to the economic limit and integrate the area under the curve to calculate estimated ultimate recovery (EUR); the widespread adoption of unconventional shale production created significant tension with the Arps framework because shale wells often decline hyperbolically at early time (b greater than 1) before transitioning to boundary-dominated flow at late time, requiring modified hyperbolic or stretched exponential models to avoid grossly over-predicting EUR.
• Rate-transient analysis (RTA) for unconventional wells applies curve matching in the time domain of flowing conditions rather than shut-in pressure recovery, allowing engineers to extract reservoir and fracture properties from the production data without ever shutting the well in for a formal well test — a major practical advantage in unconventional operations where shut-in for a pressure buildup test can cause formation damage from fluid redistribution or simply represents lost production revenue that operators are unwilling to sacrifice; RTA type curves account for the complex geometry of a hydraulically fractured horizontal well in a tight formation (multiple fracture wings, finite-conductivity fractures, stimulated reservoir volume, and reservoir matrix characteristics), and matching the field production data to these type curves provides estimates of fracture half-length, fracture conductivity, matrix permeability, and drainage area that guide infill well spacing decisions and reserve booking; the quality of an RTA match depends critically on the precision of the flowing wellhead pressure data — an RTA performed on well rates without accurate pressure data is essentially a decline curve analysis dressed up with more complicated mathematics.
• Non-uniqueness is the fundamental limitation of curve matching that every petroleum engineer must acknowledge — many different combinations of reservoir parameters can produce indistinguishable type curves over the range of data typically available, meaning that multiple geological models can "match" the same pressure or production data equally well; a well test that ends before the derivative reaches the infinite-acting radial flow plateau leaves the permeability and skin factor both undetermined, because the data could be matched by a high-permeability, high-skin well or a low-permeability, low-skin well with equal mathematical validity; the practical response to non-uniqueness is to constrain the curve matching interpretation with all available independent data — core permeability, analog well performance, geological mapping, and structural position — to select the physically most plausible solution from the family of mathematically valid matches; an interpreter who presents a single matched solution without discussing the range of possible matches and the constraints that rule out alternatives is not communicating the uncertainty that exists in the analysis.
• Straight-line analysis methods are complementary to type curve matching and provide an independent check on the derived parameters — Horner plots (for pressure buildup), Miller-Dyes-Hutchinson (MDH) plots, and Cartesian plot analysis all transform the data into a coordinate system where the radial flow period appears as a straight line with a slope proportional to permeability-thickness; when the permeability derived from straight-line analysis agrees with the permeability from type curve matching on the same dataset, confidence in the result is high; when they disagree, the discrepancy signals either that the straight-line analysis is drawn through the wrong portion of the data, the type curve match is not unique, or both — and the discrepancy must be resolved by examining which interpretation is more consistent with geological and petrophysical constraints; modern commercial software presents both analyses simultaneously and highlights inconsistencies automatically, but the engineer must still understand what the inconsistency means and how to resolve it.