Registered Public Utility Engineer - Power common mistakes
Registered Public Utility Engineer - Power · Understand the traps and how to avoid them.
Treating a coincident peak demand forecast as if it were the sum of individual customer class peaks, then sizing generation or transmission capacity to that inflated total.
Candidates carry over intuition from simple addition and forget that class peaks occur at different hours, so diversity and coincidence factors must be applied; the exam deliberately gives class-level peaks to test whether you net them.
What to do instead
Always compute the system coincident peak by applying coincidence factors to each class peak, then add a planning reserve margin to the coincident total, not to the arithmetic sum. State the coincidence assumption explicitly in your work.
Computing reserve margin or loss-of-load expectation using installed capacity instead of dependable or accredited capacity, and ignoring forced-outage rates in the adequacy calculation.
Installed nameplate capacity is the most visible number, and candidates conflate it with what the fleet can actually deliver during peak; derates, ambient conditions, and forced outages are easy to skip under time pressure.
What to do instead
Use net dependable capacity (or capacity accreditation values for variable resources) as the numerator, apply forced-outage and derate assumptions, and keep the reserve-margin formula consistent: (dependable capacity − peak load)/peak load.
Performing per-unit fault or load-flow calculations without a consistent base change, mixing MVA bases between the transformer, generator, and line, producing a wrong impedance or current.
Candidates memorize the per-unit formulas but forget that every impedance must be referred to one common base; the base-change step is the most commonly skipped line of work.
What to do instead
Write the chosen S_base and V_base at the top of the page, convert every given impedance with Z_new = Z_old × (S_new/S_old) × (V_old/V_new)², and carry the base through to the final current conversion I_base = S_base/(√3·V_base).
Selecting a protective relay pickup or time-dial setting from a single fault current without checking coordination against upstream and downstream devices, then declaring the scheme correct.
Candidates solve the arithmetic and stop, treating coordination as an afterthought; exam questions often embed a CT ratio or a minimum fault current that breaks the chosen setting.
What to do instead
Build a coordination table: for each device list pickup, time-dial, and the fault currents at its reach points, verify the primary/backup time margin (typically 0.2–0.4 s) at the maximum and minimum fault levels, and check CT saturation and ratio before finalizing.
Computing a revenue requirement or rate design by allocating costs with a single factor (e.g., all demand-related costs to energy) instead of matching cost causation to the appropriate billing determinant.
Candidates default to the simplest allocation and overlook that production, transmission, and distribution costs have different drivers (energy, demand, customer count), which the exam tests through class cost-of-service problems.
What to do instead
Classify each cost as customer-, demand-, or energy-related first, allocate using the matching determinant (minimum-system or average-and-excess demand for demand costs, kWh for energy), then design rates that recover each class's allocated revenue requirement.
Assuming a tariff or rider applies uniformly to all customers and computing a bill without checking eligibility, demand ratchets, or time-of-use periods.
Candidates read the rate schedule but skip the applicability section and the ratchet clause, which frequently changes the billed demand in the exam scenario.
What to do instead
Before computing, extract the applicability clause, the billing demand definition (including any ratchet and power-factor adjustment), and the TOU period boundaries; then compute the bill line by line and label each charge.
Evaluating power quality against a single nominal limit without distinguishing steady-state voltage variation, harmonic distortion, and transient events, or without checking the measurement interval and percentile basis.
Candidates treat all power-quality limits as one threshold and forget that standards specify different evaluation windows (e.g., 10-minute or 3-second aggregation) and percentile compliance rather than instantaneous maxima.
What to do instead
Identify the phenomenon first (voltage regulation, unbalance, harmonics, flicker, sag/swell), then apply the correct limit, aggregation interval, and compliance percentile; state the measurement basis in your answer.
Sizing an interconnection or performing a grid-impact study using only the proposed generator's nameplate output and ignoring the point of interconnection short-circuit strength, reverse power flow, and the applicable screening threshold.
Candidates focus on the generator and skip the host-system conditions; the exam supplies feeder impedance and existing fault level precisely to test whether you check the screen before the detailed study.
What to do instead
First apply the applicable fast-track or screening criteria (capacity relative to feeder peak, aggregate penetration on the circuit), then compute the fault-level change and voltage rise at the point of interconnection; if the screen fails, specify the detailed study scope.
Answering a DER integration question by assuming the inverter behaves as a constant power source, ignoring volt-VAR, volt-watt, and ride-through settings that change the answer.
Candidates learn the classical synchronous-machine model and apply it unchanged to inverter-based resources, missing that smart-inverter functions alter voltage support and fault response.
What to do instead
State the inverter control mode (constant power factor, volt-VAR, volt-watt) and the ride-through category before solving; if the mode is unspecified, note the assumption and show how the result changes with it.
Recommending a maintenance or replacement decision from a single condition reading or age, without weighing failure probability, consequence, and the asset's criticality in the reliability model.
Candidates equate age or one test result with end of life and skip the risk-based framework the exam expects; they also ignore that condition monitoring changes failure probability, not just cost.
What to do instead
Frame the decision as risk = probability of failure × consequence, use condition and health-index data to update probability, and compare run-to-failure, refurbishment, and replacement on lifecycle cost and reliability impact before recommending.
Handling an ethics scenario by citing only the employer's instruction or only the client's interest, without identifying the public-safety duty and the required disclosure path.
Candidates treat ethics questions as opinion and answer with loyalty or confidentiality alone, missing that the engineer's paramount obligation is public safety and that there is a defined escalation sequence.
What to do instead
Name the specific obligation (public safety, competence, truthful reporting), identify the conflict, then state the escalation path: raise it internally with documentation, escalate to the responsible authority if unresolved, and disclose only what is necessary under the applicable rules.