Original studies
Petroleum engineering studies and experiments
Explore one controlled question at a time: what changes, which inputs stay fixed and how to reproduce the result. These original synthetic studies include exact data, methods and interpretation. They are teaching experiments with stated assumptions.
Drilling hydraulics and rheology
Same flow rate different hole sizes a synthetic hydraulics study
This controlled study sends the same 600 gpm through several synthetic hole sizes around one 5 in string. Only annular area changes, so the velocity difference follows directly from continuity.
Reproducible study + data →Why a good rheology fit can fail outside the measured range
A small residual over measured shear rates does not guarantee a good prediction outside that interval. Extrapolation tests assumptions that the fitted observations may not constrain.
Reproducible study + data →Pressure depth and well control fundamentals
The error from using measured depth for hydrostatic pressure
Using measured depth in a hydrostatic calculation overstates the fluid head whenever MD exceeds the true vertical height. The relative error is MD/TVD − 1 for the same uniform fluid.
Reproducible study + data →One surface pressure two different equivalent mud weights
The same 500 psi surface term represents 1.9231 ppg at 5,000 ft and 0.9615 ppg at 10,000 ft. Equivalent density is inseparable from its evaluation depth.
Reproducible study + data →Cementing volumes and well geometry
How much a small washout changes a synthetic cement estimate
This synthetic study keeps casing and interval length fixed and enlarges only the hole. A small diameter change can produce a much larger percentage change in annular slurry demand.
Reproducible study + data →Why rounding at every step changes a cement worksheet
Rounding each interval's sack demand upward can overstate a shared batch. Sum unrounded compatible slurry volumes, calculate theoretical sacks and then round once for that batch.
Reproducible study + data →Reservoir flow and well productivity
The productivity gain from doubling permeability in a synthetic case
Doubling permeability doubles steady radial rate in this controlled case. That proportional gain follows from the equation when pressure boundaries, fluid properties, geometry and skin are unchanged.
Reproducible study + data →Two equally good permeability fits with different noisy data
Different noisy core datasets can produce similarly convincing fits but different predictions. Residual size alone does not capture sampling uncertainty or geological representativeness.
Reproducible study + data →Volumetrics material balance and displacement
The same rock volume with three different fluid volume bases
One reservoir gas amount can be described as reservoir cubic feet, reservoir barrels or standard cubic feet. Numerical differences between those volumes are not differences in gas inventory.
Reproducible study + data →How input rounding propagates into a volumetric estimate
Input rounding propagates through the product of net rock volume, porosity and gas fraction and through division by Bg. The effect depends on which inputs were rounded and in which direction.
Reproducible study + data →Decline curves and forecasting
Same first year different long term decline tails
Curves matched at the initial rate and the first-year endpoint can still have different long-term tails. Matching two observations does not identify a unique decline model.
Reproducible study + data →Why monthly rate samples do not exactly equal analytical cumulative
Monthly endpoint samples do not exactly reproduce an analytical cumulative integral. Summing rate samples requires a stated quadrature rule and time intervals.
Reproducible study + data →Petrophysics saturation and rock properties
Two rocks with the same porosity and different Archie response
Two rocks with identical porosity can have different electrical formation factors and calculated saturations because Archie parameters reflect different pore structures and calibration.
Reproducible study + data →Brooks Corey versus van Genuchten on a shared synthetic case
Brooks–Corey and van Genuchten can describe different capillary-curve shapes on a shared effective-saturation basis. Matching one point does not make their parameters equivalent.
Reproducible study + data →Black oil and gas PVT
The API coefficient switch in a synthetic solution gas study
A piecewise PVT correlation can change coefficients at an API threshold. A visible change near the branch is a property of that implemented formula, not evidence of a physical discontinuity in crude oil.
Reproducible study + data →Why a small pressure datum mistake can distort a gas calculation
A pressure-datum error has a larger relative effect on a low-pressure gas calculation. At fixed Z and T, substituting gauge pressure for absolute pressure creates a relative density error of Patm/Pabs.
Reproducible study + data →Production inflow and surface facilities
Three synthetic IPR curves from one test point
One rate–pressure test point can be matched by several inflow shapes. Agreement at that point does not establish equal absolute-open-flow predictions or equal rates elsewhere.
Reproducible study + data →Why the same standard gas rate has different pipe velocity
The same standard gas rate can occupy different actual volumes at different pressure and temperature. Pipe velocity must use that local volume, not the unchanged standard-rate number.
Reproducible study + data →Geomechanics and hydraulic fracturing
Doubling exposure time does not double Carter leakoff
For fixed area and zero spurt, doubling exposure time multiplies Carter leakoff volume by √2 rather than two. A nonzero spurt term makes the total-volume ratio smaller than √2.
Reproducible study + data →How a stress sign convention changes an apparent failure result
Changing stress sign convention is harmless only when stresses, pressure terms and the failure equation are transformed together. Flipping one input alone can create a false failure indication.
Reproducible study + data →Compositional thermodynamics and phase behavior
Same feed two equations of state a reproducible comparison
This reproducible PR–SRK study uses one six-component feed and identical numerical state inputs. It reports model differences and balance diagnostics without declaring either EOS more accurate for an unmeasured fluid.
Reproducible study + data →What a heavier C7 plus assumption changes in a synthetic flash
Changing a C7+ average molecular weight changes the assigned heavy-fraction model and can alter its flash result. The direction and size are conditional on the characterization, EOS and other fixed assumptions.
Reproducible study + data →Multiphase flow thermal behavior and flow assurance
The same wellbore with three synthetic thermal histories
This synthetic study compares the same thermal system under different shut-in durations. The different restart states come from stored formation heat, not a change in current material properties.
Reproducible study + data →Equilibrium wax onset is not a wax deposition forecast
Equilibrium wax appearance identifies when a solid phase becomes thermodynamically favorable in a selected model. Deposition additionally depends on transport, wall conditions, kinetics and removal.
Reproducible study + data →