PE Petroleum common mistakes
PE Petroleum · Understand the traps and how to avoid them.
Substitutes numbers into the Darcy radial inflow equation without matching the constant to the unit set - e.g., uses 0.00708 (field units, r in ft, h in ft, B in RB/STB) while entering drainage radius in acres or flow rate in m3/d, and accepts an answer that is off by orders of magnitude.
The leading constants (0.00708, 0.001127, 141.2) are memorized as bare numbers rather than as unit-conversion packages, and the distractors are built to match the wrong constant, so the wrong answer looks plausible.
What to do instead
Before computing, write a one-line unit ledger under the equation listing the unit of every variable and the units the constant absorbs. Finish with a magnitude check: PI should come out in the range of roughly 0.1–50 STB/d/psi for a typical well, and q = PI x drawdown must reproduce the given rate.
Mixes surface and reservoir volumes in material balance and flow calculations - plugs STB/d production directly into a rate equation that requires RB/d, or computes OOIP without Bo, and ignores Bg/Rs entirely for gas and gas-cap problems.
Problem statements quote production and reserves in STB and OOIP in STB, while the governing equations are written at reservoir conditions; the conversion step feels like a formality and gets skipped under time pressure.
What to do instead
Tag every number with a subscript (STB, RB, scf, psia) the moment it is read, and convert at one declared reference condition before substitution. As a rule, any term inside a material balance that represents fluid leaving the reservoir must carry its formation volume factor.
Treats the gas as ideal - computes Bg, density, or rate with pV = nRT at 4000–8000 psia, or misuses the Z-factor correlation by confusing pseudo-reduced temperature with reservoir temperature or by using the wrong gas gravity basis.
Ideal gas is fast, and at low pressure the error is small enough that practice problems reward it; at reservoir pressures Z is typically 0.7–1.1 and the error is large but not obvious.
What to do instead
Always take the four-step route: Tpc and Ppc from Sutton (or the gas gravity table), then Tpr and Ppr, then Z from the chart/standing correlation, then Bg = 0.0283 zT/P (ft3/scf, T in deg R, P in psia). Sanity-check that Z at high pressure is below or near 1 for most natural gases.
Writes the material balance with an inconsistent sign convention for water influx and expansion terms, or omits rock and connate-water compressibility in a closed undersaturated system, so the drive index or OOIP comes out negative or several times too large.
The equation is reproduced from memory in whichever form was last seen, and the expansion terms (cf, cw, Swi) are small enough that candidates assume they cannot matter.
What to do instead
Adopt one convention (withdrawals positive, on the left) and write the equation out in full each time, including the (cw Swi + cf)/(1 - Swi) term. If the computed OOIP changes sign or the drive index does not sum to 1.0, the convention is wrong - stop and re-derive rather than forcing the arithmetic.
Reports the mud weight from the drilling program as the pressure the formation sees while circulating, or checks kick tolerance and lost-circulation limits against hydrostatic pressure only, omitting annular friction.
Mud weight is the number printed on the program and the one recorded on the daily report, so it feels like the operative pressure; equivalent circulating density requires an extra step that candidates drop when the question does not literally say 'while circulating'.
What to do instead
For every pressure-bound check, first ask whether the well is static or circulating. If circulating, compute ECD = MW + dP_ann/(0.052 x TVD). Check the fracture gradient against ECD at the shoe and the pore pressure against the static (or swabbed) equivalent mud weight, and report both numbers if the question is ambiguous.
Designs the casing string for a single load case - usually burst from a gas kick - and never checks collapse under an evacuated or partially evacuated annulus, or tension under buoyant weight plus bending, and then compares the design factor on the wrong side of the inequality.
Textbook problems often present one load case, so the habit forms that casing design equals kick burst; collapse and tension checks are also more tedious to set up.
What to do instead
Build a three-row table (burst / collapse / tension) before touching a table of pipe grades: state the governing load scenario, the design factor, and the check as capacity/load >= DF. Do the burst case with a gas gradient to surface, the collapse case with the annulus evacuated (or the specified backup condition), and the tension case with buoyant pipe weight in air minus buoyancy plus bending if a deviated section is involved.
Computes cement slurry volume from openhole volume alone, ignoring excess factor, washout, shoe track, and the plug-displacement volume; or confuses slurry yield (ft3/sack) with slurry density (lbm/gal) and mixes sacks with pounds.
The geometric volume is the obvious part of the calculation and looks complete; the additive terms feel like small corrections, and yield versus density is a vocabulary slip that is easy to make under time pressure.
What to do instead
Lay out the volume in named components - openhole (with an explicit 1.2–1.5 excess), casing-to-hole annulus, shoe track, plus a stated displacement - and sum them last. Derive yield from absolute volumes (API class cement plus water plus additives) and confirm units: sacks x ft3/sack = ft3, and density in lbm/gal x 7.48 = lbm/ft3.
Uses a straight-line productivity index to compute inflow below the bubble point (or applies Vogel above it), and then confuses inflow performance with tubing/vertical-lift performance when locating the operating point.
PI is the first IPR model taught and gives a clean linear answer; the operating point question superficially looks like a rate calculation rather than an intersection of two curves.
What to do instead
Screen the problem first: if the flowing bottomhole pressure is below Pb, use Vogel with qmax from the test point (and subtract the undersaturated portion when Pi > Pb); if above Pb, use PI. For an operating-point question, state that you are intersecting IPR with the vertical lift curve and read the rate at the intersection, not at either curve alone.
Applies a Souders-Brown K value to a separator without checking whether the vessel is vertical or horizontal and whether a mist extractor is fitted, and drops the conversion between gas rate in MMscf/d, lbm/ft3 density, and the constant in the capacity equation; also sizes gas capacity but never checks liquid retention time.
One K factor gets memorized and reused; the vessel orientation and internals are usually buried in a single sentence of the stem, and liquid capacity is forgotten because the question appears to be about gas.
What to do instead
Write the K value with its service label next to it (vertical, no mist extractor ~0.03–0.06; horizontal ~0.4–0.5 ft/s depending on configuration), carry units explicitly through the gas capacity equation, then compute retention time = liquid volume / liquid rate and check it against the specification (typically 1–3 min for oil, longer with water). Report the controlling (larger) diameter.
Computes pump or compressor power with a rate in the wrong unit - bbl/day substituted into HP = Q dP/1714 (which takes gpm) - or omits pump/compressor efficiency and the number of compression stages, and uses isentropic head for a compressor without the (k-1)/k exponent or without Z.
The hydraulic horsepower formula is short and memorable, so the 42 gal/bbl conversion and the 24 h/day conversion into gpm are the first things dropped; efficiency and staging are treated as refinements rather than part of the definition.
What to do instead
Convert flow to gpm (bbl/d x 42/1440) before using 1714, and divide by the stated efficiency. For compression, compute the isentropic head with (k-1)/k, average Z, and inlet temperature, and add stages until the per-stage ratio stays within the mechanical limit (typically about 3–4:1). Finish with a brake-horsepower figure and an explicit efficiency assumption.