PE Electrical common mistakes
PE Electrical · Understand the traps and how to avoid them.
In NEC-style ampacity and feeder/OCPD sizing, the candidate adds the 125% continuous-load factor to the wrong quantity - e.g., applies 125% to noncontinuous loads, applies it to the conductor ampacity check but not the overcurrent device, or forgets to apply it separately to feeder and to the largest-motor contribution in a multi-load feeder.
The 125% rule is learned as a rote number rather than as a factor attached to a specific element (conductors, OCPD, neutral, feeder). Under time pressure the candidate applies it globally or forgets which loads are continuous.
What to do instead
Before calculating, draw a two-column scratch table: 'continuous loads' and 'noncontinuous loads'. Then build the load-side sum (125% continuous + 100% noncontinuous + any special motor/continuous-motor adders such as 125% of the largest motor) and a separate conductor check. Re-read the question to confirm whether it asks for the OCPD rating, the conductor ampacity before adjustment, or the ampacity after adjustment and derating - these are three different numbers.
For a load with harmonic content, the candidate computes power factor as cos θ from the fundamental displacement angle alone and reports it as the load power factor, giving an answer that is too high relative to the true (distortion-inclusive) power factor.
Textbook single-frequency phasors dominate practice, so 'PF = cos of the impedance angle' becomes automatic and the RMS-current penalty of harmonics is invisible.
What to do instead
Whenever a waveform, rectifier, VFD, or current-distortion data appears, switch to the definition PF = P/S = real power divided by the product of true RMS volts and true RMS amps. For the common case of negligible voltage distortion, compute PF = cos θ1 × (I1,rms / I,rms) and sanity check that the result is lower than the displacement factor, never higher.
The candidate substitutes a line-to-line voltage where a line-to-neutral value belongs (or vice versa) and/or drops the square-root-of-two factor when converting between RMS and peak - for example, using 480 V directly as the per-phase voltage in a wye-connected per-phase impedance calculation, or computing instantaneous peak stress as 480 V instead of 480×√2.
Phasor magnitudes in problems are nearly always RMS, but the formulas the candidate memorized for single-phase circuits are silently reused in three-phase circuits where the per-phase voltage is V_LL/√3. The peak/RMS distinction is also easy to skip because both are called 'volts'.
What to do instead
Adopt a written convention at the top of every solution: state 'all quantities RMS unless labeled peak', then explicitly write the conversion you are using (V_LN = V_LL/√3 for wye; V_peak = √2 × V_rms) before substituting any number. Finish three-phase problems by checking total power against √3 × V_LL × I_L × PF.
In AC nodal or mesh analysis the candidate drops a minus sign on a current source that opposes the assumed loop direction, or adds admittances/impedances as real scalars instead of complex quantities, ending with a magnitude that looks plausible but is wrong.
Reference directions are kept only in the head rather than on paper, and the imaginary part is treated as a small correction under time pressure. Complex arithmetic by hand also invites sign slips.
What to do instead
Sketch the circuit and mark every assumed current arrow and node polarity before writing equations; carry j through symbolically and only substitute at the end. Verify with an independent check: total complex power delivered by sources must equal total I²R plus the reactive balance, and the phase angle of the result should be physically reasonable (e.g., inductive current lagging).
The candidate confuses poles with pole pairs, or mis-signs slip - computing synchronous speed as 60f/P instead of 120f/P, reporting a slip greater than 1 for normal motoring, or giving a negative slip for a motor operating as a generator.
The 120 factor already embeds the pole-pair conversion, so the candidate who reasons 'pairs' and then also divides by two double-counts. Slip sign conventions are rarely exercised because most practice problems are motoring only.
What to do instead
Write n_s = 120f/P with P = number of poles, then s = (n_s − n)/n_s and immediately classify the operating mode from the sign: 0 < s < 1 motoring, s < 0 generating, s > 1 braking. Sanity check the rotor speed is below synchronous speed for motoring and above it for generating, and check rotor frequency f_r = s·f_s is small in normal operation.
On three-phase transformer banks the candidate applies the square-root-of-three factor on the wrong side of the ratio (treating a delta-wye bank as if its line-to-line ratio equalled the turns ratio), or ignores the 30-degree phase shift, and on per-unit problems re-refers impedance through the transformer when the base has already been chosen consistently.
Single-phase transformer ratio drills are over-learned; the delta and wye terminal relationships are the part that actually changes, and per-unit's main advantage (no impedance referral needed across a bank when bases are chosen per winding) is not trusted.
What to do instead
For every bank, write the winding voltage ratio first (per-phase turns ratio), then derive the line-to-line ratio from it explicitly (delta-wye gives √3 × turns ratio). Note the phase shift for a delta-wye bank. For per-unit work, set voltage bases in the ratio of the transformer's winding voltages and then state 'no referral needed' before computing; verify by checking that the ohmic value on each side transforms by the square of the turns ratio.
The candidate uses the conductor radius in place of the geometric mean radius when computing inductance, forgets to combine the subconductor GMR with the bundle spacing for bundled conductors, and mixes per-conductor quantities with per-phase quantities.
The radius is what appears in capacitance formulas and in resistance-adjacent intuition, so it gets carried into inductance by habit; bundle GMR is a rarely used formula that is misremembered as a simple average.
What to do instead
Before computing inductance, write the correct quantity by name: GMR (not radius) for inductance, and for a bundle of n subconductors use GMR_bundle = (GMR × product of spacings)^(1/n) - e.g., for two subconductors at spacing d, √(GMR × d). Confirm the final answer is per phase (per phase = per conductor for a single conductor per phase) and check magnitude against a typical rule of thumb (roughly 0.4–0.6 mH/km for overhead lines).
The candidate computes voltage drop or regulation by subtracting scalar voltage magnitudes, or plugs in the wrong regulation definition - using the sending voltage as the denominator, or comparing no-load receiving voltage to full-load sending voltage instead of to full-load receiving voltage.
The phasor nature of the sending/receiving relationship is lost when the problem is set up as arithmetic, and 'regulation' is memorized as a phrase rather than a ratio with a specific denominator.
What to do instead
Use the phasor relation V_S = V_R + I(R + jX) (or the per-unit equivalent), then take magnitudes only at the end. For percent regulation write it out fully: 100 × (|V_R,no-load| − |V_R,full-load|)/|V_R,full-load|, holding the sending voltage constant. Check the sign - regulation is positive for a normal lagging load - and confirm voltage drop rises as load power factor becomes more lagging.
For unbalanced faults the candidate plugs the single-line-to-ground formula into a line-to-line or double-line-to-ground case (or forgets the factor of 3 on the ground current), or connects the sequence networks in the wrong configuration for the fault type.
All the fault formulas look alike and are memorized as bare expressions rather than as consequences of how the sequence networks are tied together at the fault point. The 3× factor on I0 for ground current is easily lost.
What to do instead
Memorize the sequence interconnection rather than the final formula: series for SLG, parallel for LL, parallel with the zero-sequence branch for LLG, and positive-sequence-only for a balanced three-phase fault. Reconstruct the current from the interconnection - for SLG, I_a1 = V_f/(Z1+Z2+Z0) and the ground current is I_f = 3I_a1; for LL, I_a1 = V_f/(Z1+Z2) and the line current magnitude is √3 × |I_a1|. As a check, the SLG fault current should be the largest of the four for a solidly grounded system with low Z0.
When combining a motor's subtransient reactance with a generator or utility source reactance on a common base, the candidate adds the per-unit values directly without first converting each to the chosen MVA and voltage base - or performs the base change with the voltage ratio inverted.
Per-unit values look dimensionless and therefore 'portable', so the impedance-base scaling step is skipped. The voltage-ratio term is squared and inverted easily when done from memory.
What to do instead
Tabulate every machine's nameplate MVA, kV, and reactance, choose the system base, and apply X_new = X_old × (S_base,new/S_rated) × (V_rated/V_base,new)² line by line before any series or parallel combination. Keep voltages consistent with the transformer ratios so no further referral is needed, and check the result against the physical ohmic value on one side of the system.