OilAuthority.com

Boundary Conditions in Pressure Transient Analysis: No-Flow and Constant-Pressure Derivative Signatures, Distance-to-Boundary Estimation, and WCSB Pool Edge Identification

Boundary conditions in reservoir engineering and pressure transient analysis (PTA) are the mathematical constraints placed on the outer edges of the reservoir drainage volume that govern how pressure disturbances propagate through the reservoir and what late-time pressure behavior a producing well will exhibit during a buildup or drawdown test — the two primary physical boundary types are the no-flow boundary (Neumann condition: zero pressure gradient, zero flux, at the boundary) and the constant-pressure boundary (Dirichlet condition: fixed pressure at the boundary), each producing a distinctive and diagnostically unambiguous signature on the log-log pressure derivative plot that is the centerpiece of modern well test interpretation. The no-flow boundary — physically realized by a sealing fault, a low-permeability stratigraphic barrier, a pinch-out, or the outer limit of a finite drainage area — reflects the pressure transient back into the reservoir with the same sign as if a mirror-image production well existed at the same distance on the other side of the boundary (the method of images), and this superposition of the real well pressure transient and the image well pressure transient causes the pressure derivative on the log-log plot to increase sharply from its radial flow value (a horizontal stabilized derivative) and eventually double (slope 2× the radial flow value) at the time the transient reaches the boundary: if the reservoir is bounded on all sides by sealing faults, the derivative continues to rise as successive image wells are added by the method of images until it achieves a unit slope (straight line on log-log) indicating pseudosteady-state depletion — the signature of a fully closed tank. The constant-pressure boundary — physically realized by an active aquifer with sufficient transmissibility and storage to maintain pressure at the reservoir edge, a gas cap, a strongly supported waterflood pattern, or a fracture with high conductivity connecting to a large fluid volume — causes the pressure derivative to decrease from its radial flow value and eventually stabilize at or near zero as the boundary provides pressure support that prevents the well pressure from continuing to decline: on the pressure plot, BHFP stabilizes at a constant value during production once the constant-pressure source is providing full voidage replacement, and on the derivative plot, the stabilizing derivative approaches a downward slope toward zero as time progresses. The time at which the pressure transient first reaches a boundary (t_boundary) can be converted to a distance estimate using the diffusivity equation: d = 0.0328 × sqrt(k × t_boundary / (phi × mu × ct)) where d is in metres, k in mD, t_boundary in hours, and phi × mu × ct is the product of porosity, viscosity, and total compressibility — providing an independent distance-to-boundary estimate from well test data that can be compared against seismic fault interpretation or geological pool edge picks to validate or calibrate the structural model. In WCSB petroleum exploration, boundary detection from PTA is routinely used to identify the sealing-fault boundaries of Devonian carbonate reefs (where the reef mound is surrounded by tight argillaceous off-reef lime mudstone), the stratigraphic pinch-out edges of Viking sandstone channels, and the aquifer boundaries of Cardium sandstone pools where active water encroachment transitions from a no-flow boundary signature early in well life to a constant-pressure signature as aquifer influx strengthens.

Key Takeaways

  • No-flow boundary derivative signature: doubling and the method of images: When a pressure transient traveling radially from a producing well encounters a single linear sealing fault, the derivative doubles from its radial flow plateau value (from stabilized at m on the Horner plot slope to 2m, or from stabilized at 0.5 × kh/mu in PTA units to 1 × kh/mu). The doubling occurs because the image well across the fault contributes an equal pressure transient to the observation well, effectively halving the drainage area that the real well can access on one side. Multiple parallel or intersecting faults create a stepwise doubling pattern: two parallel faults create a channel (derivative rises toward infinite slope on a log-log plot as channeling develops), while a four-sided bounded fault compartment produces a unit-slope derivative (pseudosteady state) at late time. In WCSB Devonian reef tests, a single sealing no-flow boundary detected at 400-1,200 m distance from a tested well confirms the reef edge position inferred from seismic amplitude and reservoir geologist mapping, and the derivative doubling time pins the boundary distance to within ±50-100 m.
  • Constant-pressure boundary signature: derivative downturn and aquifer connectivity assessment: A constant-pressure boundary is identified on the pressure derivative plot by a derivative that decreases from its radial flow value toward zero at late time — the derivative turns downward instead of doubling. The physical interpretation is that the pressure sink created by production is being filled by fluid flux from a high-pressure source at the boundary (aquifer, gas cap, or injector), and the well pressure approaches a new steady state rather than declining continuously. In WCSB Cardium pool PTA, the transition from no-flow to constant-pressure boundary behavior at the same boundary distance (as interpreted from the time to derivative inflection) indicates the aquifer becoming "connected" as the pressure differential drives aquifer water into the reservoir over months to years — an observation that predicts the timing of water breakthrough and guides the waterflood management strategy for the pool. A constant-pressure boundary detected at 2 km from a producing well with an aquifer transmissibility-to-reservoir transmissibility ratio of 10:1 provides sufficient pressure support to maintain plateau production for 3-5 years before water breakthrough begins.
  • Distance to boundary calculation and comparison with seismic fault picks: The distance-to-boundary formula d = 0.0328 × sqrt(k × t_bd / (phi × mu × ct)) requires accurate permeability (from the radial flow analysis of the same test), porosity and compressibility (from core and fluid PVT), and the correct identification of the boundary detection time t_bd from the derivative plot. Uncertainty in t_bd identification introduces distance uncertainty of ±20-30% in typical WCSB well tests where derivative noise at late time may obscure the exact inflection point. For a Cardium well with k = 20 mD, phi = 0.20, mu = 2 cp, ct = 10^-5 /kPa, and t_bd = 8 hours: d = 0.0328 × sqrt(20 × 8 / (0.20 × 2 × 10^-5 × 10^6)) = 0.0328 × sqrt(160/4) = 0.0328 × 6.32 = 0.207 km = 207 m. Seismic fault pick at 185 m from the well — consistent within PTA uncertainty. The 22-metre discrepancy guides the updated structural map to shift the fault position 22 m closer to the well, improving the volumetric calculation for proven reserves in the bounded fault block.
  • Combined no-flow and constant-pressure boundaries: identifying composite reservoir systems: Many WCSB reservoir settings involve composite boundary conditions — a no-flow boundary on one or more sides (sealing fault, pinch-out) combined with a constant-pressure boundary on another side (active aquifer edge, injector support). The derivative plot for a composite system shows intermediate behavior: the derivative first doubles from the no-flow boundary, then turns down as the constant-pressure boundary provides support, potentially stabilizing at a value between the original radial flow value and twice that value depending on the geometry. This composite signature is characteristic of WCSB Devonian carbonate pools trapped against a fault (no-flow on the up-thrown side) with aquifer support from the basin-adjacent direction (constant-pressure on the down-dip side). Correctly identifying the composite boundary system from PTA allows the reservoir engineer to distinguish the "supported" drainage area (contributing to production without pressure decline) from the "unsupported" fault-bounded portion (where pressure declines as a closed tank) for accurate material balance calculation and infill well placement decisions.
  • Boundary effects in WCSB horizontal well tests and their interpretation complications: Horizontal wells in WCSB Montney and Duvernay formations have complex boundary geometries: the wellbore itself creates a linear no-flow boundary condition at the top and bottom of the reservoir (the sealing top and bottom of the pay zone), lateral boundaries are defined by fault or pinch-out positions, and the toe and heel of the horizontal lateral define further no-flow boundaries that create early-time linear flow signatures. The pressure derivative of a bounded horizontal well in a rectangular drainage box shows a succession of flow regimes: early time wellbore storage (unit slope), early linear flow (half-slope), infinite-acting pseudoradial flow (flat), then boundary-dominated linear flow (half-slope return) from the no-flow top and bottom seals, and finally pseudosteady state (unit slope) when all boundaries are reached. Correctly diagnosing each flow regime requires the reservoir thickness, lateral length, and permeability anisotropy to be simultaneously estimated from the derivative shape — a model-dependent interpretation process where multiple combinations of these parameters can fit the observed derivative, introducing ambiguity that is resolved only by integrating PTA with geomechanical data and 3D seismic attribute interpretation.

Boundary Detection in a WCSB Devonian Carbonate Reef Test

A WCSB Devonian Leduc reef exploration well in central Alberta runs a 72-hour DST with a 48-hour buildup on a 25 m perforated interval in the reef core. Log-log derivative analysis: radial flow identified at 2-8 hours, stabilized derivative = 4.2 kPa per log cycle. At 22 hours on the buildup, the derivative begins to double, reaching 8.4 kPa per log cycle at 35 hours — classic single no-flow boundary signature. Boundary distance calculation: k = 55 mD (from radial flow), phi = 0.08 (sonic/density log), mu = 0.5 cp (black oil PVT), ct = 1.2×10^-5/kPa; t_bd = 22 hours: d = 0.0328 × sqrt(55 × 22 / (0.08 × 0.5 × 1.2×10^-5 × 10^6)) = 0.0328 × sqrt(1210/0.48) = 0.0328 × 50.2 = 1.65 km. Seismic reflection off-reef carbonate shale contact interpretation placed the reef edge at 1.80 km from the well. The PTA boundary distance of 1.65 km is 150 m inside the seismic pick — consistent with the reef-to-off-reef acoustic impedance contrast being a diffuse transition rather than a sharp contact, as confirmed by the off-reef mud log (anhydrite-carbonate transition over 8 m in offset well). The 1.65 km boundary distance defines the minimum drainage radius of the reef for volumetric reserves calculation: pi × (1,650)^2 × 0.08 × 0.25 × 0.6 / 1.03 = 3.1 million m³ (31 million barrels OOIP) — confirming commercial field size for development well authorization.

Fast Facts

The method of images — the mathematical technique that explains why a no-flow boundary causes the pressure derivative to double — was adapted for petroleum reservoir engineering from the classical electromagnetic field theory method used to compute potentials near grounded conductors, applied to well testing by D. Bourdet and colleagues in a landmark 1983 SPE paper (SPE 12777, "A new set of type curves simplifies well test analysis") that introduced the pressure derivative plot as the primary interpretation tool for boundary detection. Before the derivative method, boundary detection from buildup tests required the analyst to visually identify "humping" on the Horner plot — a subjective and error-prone interpretation that the derivative plot replaced with an objective, quantitative doubling signature. The 1983 paper is among the most cited in SPE literature and directly enabled the widespread adoption of modern PTA software throughout the WCSB oil patch in the late 1980s and 1990s.

The bounded reservoir that results when no-flow boundaries on all sides create a closed tank — and the pseudosteady-state depletion, drainage area calculation, and material balance behavior that follow — are described under bounded reservoir, where the PSS pressure decline rate, P/z plot linearity for gas, and WCSB Cardium and Viking pool boundary identification are covered alongside recovery factor implications of bounded vs aquifer-supported reservoirs. The pressure buildup and drawdown testing methodology that produces the derivative plot used to identify boundary conditions — including the Horner analysis, log-log type curve matching, and flow regime identification sequence — is described under pressure transient analysis. The sealing fault that is the most common geological cause of no-flow boundary behavior in WCSB Devonian carbonate and Cretaceous sandstone reservoirs, and how seismic interpretation and fault seal capacity analysis compare with PTA boundary distance estimates, are covered under fault seal.

Browse the Oil and Gas Glossary