Engineer Chemical Industry common mistakes
Engineer Chemical Industry · Understand the traps and how to avoid them.
Carrying the wrong reference state through an energy balance: tabulating heats of formation and sensible heats relative to different base states (elements at 298 K versus a process stream at 25 °C), then adding them without a consistent datum, producing a plausible but wrong enthalpy change.
Candidates learn formation enthalpies and heat-capacity integrals as separate recipes and rarely force every term onto one declared reference. The arithmetic looks correct, so the inconsistency is invisible during a timed exam.
What to do instead
Before any energy balance, write the datum explicitly (elements in their standard states at 298 K, or the inlet stream at T0) and convert every term to that datum on paper. Do one worked check: sum the sensible heats along a path from datum to each stream temperature. If the datum line is not written down, treat the answer as unverified.
Treating equilibrium conversion as the achievable conversion in a reactor problem, ignoring kinetics, residence time and catalyst deactivation, and then sizing the reactor from that number.
Equilibrium is the first quantity taught and the easiest to compute, so it anchors the answer. Candidates conflate the thermodynamic ceiling with the design target.
What to do instead
Separate the two questions explicitly: state the equilibrium conversion, then state the rate-limited conversion at the specified space velocity or residence time. If the problem gives kinetics, use them; if it gives only equilibrium data, say in the answer that the design conversion cannot exceed equilibrium and is likely well below it.
Selecting a separation sequence without comparing relative volatilities or checking azeotropes, then discovering an infeasible distillation (near-unity relative volatility or a minimum-boiling azeotrope) only after committing to a flowsheet.
Candidates default to ordinary distillation because it is the most practised unit operation, and skip the quick screening step that would flag infeasibility.
What to do instead
Run a fixed screening routine before any sequence: list component boiling points, compute relative volatilities at the operating pressure, and check for azeotropes in the data given. If relative volatility is close to unity or an azeotrope exists, name the alternative (extractive or pressure-swing distillation, liquid-liquid extraction, adsorption) and justify it.
In pinch analysis, placing the minimum approach temperature on the wrong side of the composite curves, or adding utility before maximising process-to-process heat recovery, which yields a wrong minimum utility target and an over-designed network.
Pinch problems demand careful bookkeeping of shifted temperatures (hot streams shifted down by half the approach), and candidates mix hot- and cold-side shifts or skip the shifting step entirely.
What to do instead
Adopt a strict procedure: shift hot stream temperatures down by delta-T-min/2 and cold streams up by delta-T-min/2, build the problem table, and read the utility targets off the cascade before drawing any network. Report the pinch temperature and the two utility duties as a checkable triplet.
Answering a control-loop question by tuning for the fastest response, or by prescribing PID action without checking that the controller direction (direct versus reverse) matches the process gain sign, so the loop would drive the valve the wrong way.
Candidates focus on tuning rules and controller form and treat loop direction as an afterthought, even though a sign error makes any tuning irrelevant.
What to do instead
For every loop question, first state the sign of the process gain and the fail-safe valve action, then derive the required controller action. Only after that discuss tuning, and justify the choice (aggressive for tight setpoint tracking, detuned for robustness against dead time) rather than defaulting to the fastest.
Performing a hazard review by listing hazards in prose instead of applying a structured method (HAZOP guide words, LOPA layers, or a fault tree with quantified cut sets), so the answer names risks but never shows the initiating cause, consequence and safeguard chain.
Under time pressure candidates write descriptive safety answers and skip the method, which is exactly what a structured-method question is testing.
What to do instead
Use the named method mechanically: for HAZOP, state node, parameter, guide word, cause, consequence and existing safeguard for each line; for LOPA, list each independent protection layer with its probability of failure on demand and compute the mitigated frequency. Show the table structure even when the answer is short.
Giving a generic environmental answer ("reduce emissions", "apply best practice") instead of a quantified mass balance, emission factor calculation, or comparison against a permit or BAT limit.
Sustainability topics feel qualitative, so candidates drop the numbers, but the exam marks the calculation and the regulatory comparison.
What to do instead
Convert every environmental question into a balance: inlets minus outlets equals accumulation plus emission, with units stated. Compare the result to the relevant limit (permit condition, BAT-associated emission level, or discharge standard) and state the margin. A qualitative sentence without a number earns little.
Choosing a material by corrosion resistance alone, ignoring mechanical requirements and fabrication consequences (temperature limits, weldability, susceptibility to chloride stress-corrosion cracking or sensitisation), or diagnosing a corrosion failure without naming the mechanism.
Candidates memorise alloy-versus-environment tables and treat material selection as a lookup, forgetting that the selected grade must also survive the mechanical and fabrication duty.
What to do instead
For each selection, state the corrosion mechanism at risk (uniform, pitting, crevice, stress-corrosion cracking, erosion-corrosion), the controlling environment variable (chloride, temperature, pH, velocity), and the mechanical or fabrication constraint. If a mechanism is named, state the mitigation, not just the alloy.
Mixing real and nominal cash flows or discounting at the wrong point in time: applying a discount factor to already-discounted values, using a nominal rate on real cash flows, or omitting working capital and salvage from a net present value.
Project economics problems bundle several conventions, and candidates apply the discount formula mechanically without tracking which year each cash flow lands in.
What to do instead
Draw a time line from year 0 to end of life, place capital, working capital, operating cash flows, tax and salvage on it, then discount each to year 0 with an explicitly stated rate and state whether it is real or nominal. Cross-check the result against a simple payback or levelised figure.
Answering an ethics or regulatory question with a general statement of good intentions rather than citing the specific obligation, the applicable code or standard, and the escalation route when a client or employer pressures for a non-compliant decision.
Candidates treat ethics as common sense and compliance as background knowledge, so they omit the standard, the duty holder and the documented action.
What to do instead
Structure the answer in four parts: the technical fact that creates the conflict, the specific obligation or standard breached (for example a pressure-system code, a permit condition, or a professional code of conduct), the duty to document and escalate, and the consequence of silence. Name the standard and the escalation path rather than describing good character.