PE Control Systems common mistakes
PE Control Systems · Understand the traps and how to avoid them.
Converts a differential-pressure flow transmitter's 4–20 mA signal linearly into flow (e.g., calls 12 mA '50% flow') instead of taking the square root of percent DP - or takes the square root twice when the problem states the transmitter or DCS already extracts it.
The 4–20 mA scaling rule and the square-root rule are learned as separate facts, and exam stems often say only 'DP transmitter' without restating the signal path, so candidates default to linear scaling out of habit.
What to do instead
Write the whole chain on the scratch sheet before calculating: DP in inH2O → % of DP span → sqrt → % of max flow → flow units. Check the stem for 'square-rooted output' or a flow-calibrated range; if present, do not root again. Sanity anchor: on a linear (unrooted) DP transmitter with a 0–100% DP span, 12 mA is 25% of maximum flow, not 50%.
Ranges a closed-tank level DP transmitter as simply 0 to the liquid height above the tap, ignoring the reference leg (wet leg/dry leg) head, so zero elevation or zero suppression is never applied; also uses SG = 1 for a process liquid whose specific gravity is given as something else.
Open-tank intuition carries over, and the elevation/suppression adjustment is a calibration detail rarely used outside the field, so the reference-leg head is mentally invisible on a text-only problem.
What to do instead
Draw the two impulse lines and compute the pressure at each tap for the 0% and 100% level cases; the transmitter span is (h_max × SG) of liquid head, while the lower range value shifts by the wet-leg head (SG_ref × h_ref). Confirm the required LRV is negative for a wet leg on the high side, and always substitute the stated SG rather than water's.
Picks controller action from a memorized rule of thumb ('level controllers are reverse acting') without tracing the sign through the valve and process, producing a loop whose output drives the PV further from setpoint.
Controller action is usually learned as a table to memorize rather than as a sign calculation, and the exam adds an air-to-open or air-to-close valve that flips the answer.
What to do instead
Do an explicit sign trace: (1) does increasing controller output open or close the valve (this reverses sign for air-to-close); (2) does the resulting flow increase or decrease the measured variable. If increasing output increases the PV, the controller must be reverse acting; if it decreases the PV, direct acting. Confirm by imagining a small positive PV error and checking that the resulting output moves the PV back toward setpoint.
Reads gain margin at the 0 dB gain-crossover frequency and phase margin at the −180° phase-crossover frequency, or reports a phase margin in decibels.
The two crossover frequencies look interchangeable on a Bode plot and both margins are 'distance from instability', so the definitions get swapped under time pressure.
What to do instead
Fix the anchors: gain margin is the factor by which gain can rise before instability, measured at the phase-crossover frequency where phase = −180° (GM in dB = −20·log|G|); phase margin is the degrees of phase above −180° measured at the gain-crossover frequency where |G| = 0 dB. State the frequency you used next to each answer - if the two frequencies are the same number, one of them is wrong.
Sizes a gas, steam, or flashing-liquid control valve with the incompressible-liquid Cv equation, substitutes gauge pressure where absolute pressure belongs, ignores choked flow when the downstream pressure falls below roughly half the upstream absolute pressure, and mixes Cv with Kv.
The liquid equation is the only one most candidates have used, gauge pressure is what gauges read, and Cv/Kv differ by only about 15%, so a wrong choice still looks plausible in the answer options.
What to do instead
Select the equation by service: liquid (with allowance for flashing when P2 drops below vapor pressure), gas/steam (absolute pressures, expansion factor, choked-flow limit at ΔP ≈ 0.5·P1 absolute for many gases), and always convert to absolute before dividing. If both Cv and Kv appear in the options, remember Kv ≈ 0.865·Cv and check which unit system the flow and pressure are stated in.
Treats 'fail-closed' as a synonym for 'air-to-close', and assumes the valve moves to a safe position on loss of electrical signal without checking whether the actuator is spring-return and how the solenoid and positioner are arranged.
The words 'fail' and 'close' seem to go together, and candidates answer from the tag description rather than from the physical force balance of spring versus diaphragm pressure.
What to do instead
Reason from forces: air-to-open with a spring-to-close actuator means loss of air drives the valve closed (fail-closed); air-to-close with spring-to-open means loss of air drives it open. Then add the signal path: a solenoid that exhausts on de-energization produces the air-failure position, whereas a lost 4–20 mA signal may only drive the positioner to its own configured limit. Base the answer on the stated actuator and accessory arrangement, not on the word 'fail'.
In a cascade scheme, tunes the secondary (slave) loop slower than the primary, or writes the primary controller's output as the valve signal instead of as the secondary controller's setpoint.
Cascade is drawn as two ordinary loops stacked together, and if the candidate has only tuned single loops there is no reason to think the time scales must be separated.
What to do instead
Enforce the structure: primary output = secondary setpoint; secondary output = valve. The inner loop must be at least three to five times faster than the outer loop (typically it is tuned first and with the outer controller in manual), otherwise the two loops interact and the outer loop sees a moving target. If a problem gives two time constants closer than that ratio, the cascade is the wrong architecture.
Computes a feedforward controller gain with the wrong sign or with the disturbance and manipulated-variable units mismatched, so the disturbance is amplified rather than cancelled.
Feedforward is often memorized as 'Gff = −Gd/Gp' as a pure formula, and the sign and unit check never gets done because the derivation is not re-created.
What to do instead
Derive it in one line at steady state: for zero error under a disturbance step, Kff = −(Kd/Kp), where Kd is disturbance-to-PV gain and Kp is manipulated-variable-to-PV gain, each in consistent engineering units. Then test directionally: if the disturbance raises the PV and the manipulated variable also raises the PV, the feedforward must decrease the manipulated variable - if your sign says otherwise, flip it.
Credits the basic process control system loop as an independent protection layer in a LOPA or SIL determination, or claims a 1oo2 sensor arrangement reaches a given SIL without accounting for common-cause failure.
Both layers reduce the same hazard frequency, so the BPCS 'feels' like protection, and redundancy is assumed to square the failure rate even though identical devices share calibration, impulse lines, and utilities.
What to do instead
Apply the IPL tests explicitly: specificity, independence, dependability, and auditability. A BPCS control loop normally fails independence and is not credited (unless the standard being applied allows a specifically designed high-integrity BPCS, and even then only with justification). For redundant architectures, use PFDavg formulas with a beta factor for common cause rather than λ^2 alone, and state the proof-test interval you assumed.
Approves an intrinsically safe loop by matching only the barrier voltage and current to the field device, ignoring the device's Ci and Li and the cable's capacitance and inductance, or substitutes explosion-proof enclosure reasoning for an IS entity evaluation.
Voltage and current are the visible headline numbers on a datasheet, while the reactive entity parameters sit lower on the page and are easy to omit; 'IS' and 'explosion-proof' are both recalled as 'safe in a hazardous area' without the distinct logic of each.
What to do instead
Check all four entities with cable included: Voc ≤ Vmax, Isc ≤ Imax, Ci + Ccable ≤ Ca, and Li + Lcable ≤ La. Explosion-proof containment is a different protection concept (no energy limitation) and cannot be substituted into the entity comparison. If the loop length is given, compute cable L and C from the per-unit-length values before comparing.