Chemical-Process Fundamentals — Engineering Reference

1. At a glance

Chemical-process engineering is the discipline that designs, builds, operates, and safeguards the industrial systems that transform raw materials into products through physical and chemical change. Where mechanical engineering shapes solids and electrical engineering moves charge, chemical engineering moves streams of matter — gases, liquids, slurries, solids in suspension, multiphase mixtures — and rearranges their composition by reaction, phase change, mixing, and separation. A working chemical engineer holds a single mental model that simultaneously contains:

• Mass and energy balances for any control volume, in any frame, at any steady-state or transient condition.
• Thermodynamic equilibrium — VLE, LLE, SLE, reaction equilibrium — which sets what is feasible.
• Reaction kinetics + reactor design — Arrhenius, rate laws, batch / CSTR / PFR / packed-bed / fluidized / membrane — which sets how fast and how selectively.
• Unit operations — distillation, absorption, extraction, adsorption, membrane, crystallization, drying, filtration — which physically partition the streams.
• Heat exchange + pinch analysis — which couples the energy balance to economic plant integration.
• Process control + instrumentation — DCS, PLC, P&IDs, SIS — which keeps the plant inside its operating envelope.
• Process safety — HAZOP, LOPA, SIL, relief sizing, PSM — which is the non-negotiable layer that turns a working plant into a safe working plant.

The field is the foundation for refining, petrochemicals, specialty chemicals, pharmaceuticals, food, water treatment, mining/hydrometallurgy, pulp and paper, fertilizers, environmental remediation, hydrogen, carbon capture. The same toolkit sizes a 500 000 bbl/d refinery FCC unit and a 5 kg/batch pharma reactor — only the numbers and the regulators change.

Place in the engineering stack: thermodynamics + transport phenomena → unit operations + reaction engineering → process synthesis + integration → control + safety → operations. This note assumes thermodynamics, heat-transfer, and fluid-mechanics as background.

2. Why it matters

Chemical-process industries (CPI) account for roughly 5 % of global GDP, more than US $5 trillion in annual revenue, and ~40 % of total industrial energy use (IEA Tracking Industry 2024, World Bank value-added series). The world’s nitrogen fertilizer (Haber–Bosch, ~180 Mt NH₃/yr) feeds roughly half of all humans; the world’s polymers (~400 Mt/yr) clothe, shelter, and package nearly everything else. Every liter of motor fuel, every active pharmaceutical ingredient, every gram of food preservative, every microgram of pesticide has passed through a chemical process.

The discipline therefore carries an unusually heavy economic + safety burden. Good process design produces on-spec product, cheaply, safely, predictably, for decades. Bad process design produces:

• Bhopal 1984 — Union Carbide MIC release, 30+ tonnes of methyl isocyanate vented from a runaway reaction triggered by water ingress to a storage tank; estimated 3 000 immediate deaths and 15 000+ delayed; the deadliest industrial accident ever.
• Flixborough 1974 — Nypro caprolactam plant, cyclohexane vapor cloud explosion from a bypass-pipe failure; 28 deaths, plant destroyed.
• Piper Alpha 1988 — North Sea platform, condensate leak following a permit-to-work failure; 167 deaths.
• Texas City 2005 — BP refinery ISOM unit, hydrocarbon geyser at startup; 15 deaths, 180 injured.
• West, TX 2013 — ammonium nitrate fertilizer storage fire and detonation; 15 deaths.
• Buncefield 2005, Jaipur 2009, La Mède 1992, Skikda 2004 — tank-farm and refinery fires/UVCEs.

OSHA Process Safety Management (29 CFR 1910.119, 1992) and the EPA Risk Management Program (40 CFR 68) were direct regulatory responses to these losses. Every chemical engineer working with covered processes — flammable inventories above 10 000 lb (4 540 kg), or any of 137 highly hazardous chemicals above their threshold quantities — must demonstrate the 14 PSM elements (Section 9p). The corresponding international regime is the EU Seveso III Directive (2012/18/EU), named for the 1976 Italian dioxin release.

The two pay-offs of mastering chemical-process fundamentals: (1) ability to size and economically optimize a unit (mass + energy balance, equilibrium, kinetics); (2) ability to identify and mitigate the credible failure modes before commissioning (HAZOP, LOPA, relief sizing).

3. First principles — material + energy balances

Every chemical-process analysis begins with a control volume and the conservation statements applied to it.

General balance equation

For any conserved quantity Q (total mass, component mass, energy, momentum, atoms of a chemical element) and any control volume:

Input  −  Output  +  Generation  −  Consumption  =  Accumulation

For total mass, generation and consumption are zero (mass is conserved). For components in a reactor, generation and consumption come from stoichiometry. For atoms (C, H, O, N, S balances), generation and consumption are zero even with reaction — useful for combustion checks.

Steady-state simplification

At steady state, Accumulation = 0, so:

Input  +  Generation  =  Output  +  Consumption

For a non-reacting unit (mixer, splitter, distillation column, heat exchanger) at steady state: in = out per component.

Component vs total balance, and the bookkeeping trick

Given n components and one stream, you can write n component balances and 1 total balance — but only n are independent (the total is the sum of components). The engineering convention is to pick n − 1 component balances + 1 total balance, or all n component balances; never mix.

Element balances

For combustion or any redox system, atom counts on each side must match. Burning 1 mol C₃H₈ in 5 mol O₂:

C₃H₈ + 5 O₂ → 3 CO₂ + 4 H₂O

C: 3 = 3, H: 8 = 8, O: 10 = 10. Elements balance regardless of reaction extent — the test that catches half the homework arithmetic errors.

Energy balance for an open (control-volume) system

The steady-flow energy equation (SFEE) from thermodynamics is the chemical-process workhorse:

Q̇ − Ẇ_s  =  Σ_out ṁ (h + V²/2 + gz)  −  Σ_in ṁ (h + V²/2 + gz)

For most CPI units (single-phase pipe, distillation column, reactor) the KE and PE terms are negligible vs the enthalpy term. The simplified form:

Q̇ − Ẇ_s  =  Σ_out ṁ h  −  Σ_in ṁ h

Reactor enthalpy balance includes heat of reaction (sensible + latent + reaction terms) — usually treated by reference to a 25 °C standard state and tabulated ΔH°_rxn.

Degrees-of-freedom (DOF) analysis

DOF = (number of unknowns) − (number of independent equations). The system is well-posed when DOF = 0; under-specified when DOF > 0 (need more design choices or measurements); over-specified when DOF < 0 (inconsistency).

A streamlined DOF count for a unit:

• Unknowns = stream flows + compositions + temperatures + pressures + extents of reaction
• Equations = mass balances (one per component, or n − 1 component + 1 total) + energy balance + equilibrium relations + given specifications + reactor design equations

The DOF check is the first thing a chemical engineer does before writing any equation. A flowsheet with 50 streams may have hundreds of variables; without DOF you do not know whether the simulator has a unique solution.

Recycle, bypass, purge

Feature	Why used	Calculation method
Recycle	Return unreacted feed to reactor to raise overall conversion (e.g. ammonia synthesis with single-pass X ≈ 25 %, overall X ≈ 96 %)	Tear-stream method: guess recycle composition, iterate until consistent
Bypass	Route part of feed around a unit to blend at target spec (e.g. natural-gas dehydration overdrying then re-wetting to spec)	Algebraic lever-arm on the splitter
Purge	Bleed off recycle to prevent inert buildup (Ar in NH₃ loop, light ends in alkylation)	Component balance on the inert: purge × inert mole fraction = inert in fresh feed

Conversion, selectivity, yield

For a reaction A → products (limiting reagent A):

• Conversion X_A = (n_{A,in} − n_{A,out}) / n_{A,in}
• Selectivity S_{wanted/total} = mol of wanted product formed / mol of A reacted
• Yield Y = X · S = mol of wanted product / mol of A in feed

A high-conversion process can have low selectivity (lots of byproducts) and is no better than a lower-conversion + high-selectivity scheme once recycle is factored in. The economic trade-off between conversion and selectivity is one of the two or three most important process-design decisions.

4. Thermodynamic equilibrium

Process design lives or dies on phase + reaction equilibrium. The methods inherit from thermodynamics but specialize for multicomponent mixtures.

Phase equilibria — VLE, LLE, SLE, gas–solid

System	Governs	Tool
VLE (vapor–liquid)	Distillation, flash, absorption, stripping	Raoult’s law (ideal) → γ-φ or φ-φ for non-ideal
LLE (liquid–liquid)	Solvent extraction, decanters	NRTL, UNIQUAC
SLE (solid–liquid)	Crystallization, leaching	Solubility data, modified Raoult’s
Gas–solid (adsorption)	PSA, TSA, drying	Langmuir, Freundlich, Toth isotherms

Raoult’s law (ideal mixture VLE)

For an ideal liquid mixture in equilibrium with its ideal-gas vapor:

y_i · P  =  x_i · P_i^sat(T)

y_i, x_i are vapor and liquid mole fractions; P_i^sat(T) is the pure-component saturation pressure (Antoine, Wagner, DIPPR fits). Ideal Raoult applies to very similar molecules — adjacent n-alkanes, isomers, isotopes — and almost nothing else.

Activity coefficients (non-ideal liquid)

Real liquid mixtures use an activity coefficient γ_i:

y_i · P · φ_i^V  =  x_i · γ_i · P_i^sat · φ_i^sat        (γ-φ form)

For low P, φ_i^V ≈ φ_i^sat ≈ 1, collapsing to modified Raoult: y_i · P = x_i · γ_i · P_i^sat. The γ_i is computed from a Gibbs-energy model:

Model	Best for	Notes
Wilson (1964)	Polar + non-polar, no LLE prediction	Two binary parameters per pair; widely used in distillation
NRTL (Renon–Prausnitz 1968)	Polar + LLE	Three binary parameters; standard for extraction
UNIQUAC (Abrams–Prausnitz 1975)	Most non-electrolyte systems	Two binary parameters; basis for UNIFAC
UNIFAC (Fredenslund 1975)	Predictive (no measured data needed)	Group-contribution; first-pass estimate when data lacking
eNRTL (Chen)	Aqueous electrolyte systems (CO₂ capture, sour-water stripping)	Electrolyte-NRTL

Henry’s law (dilute gas–liquid)

For a gas dissolved in liquid at low concentration:

P_A  =  H_A(T) · x_A

H_A is the Henry’s constant [Pa or Pa/mol-fraction]. Used in absorber design, dissolved-gas calculations, environmental fate models.

K-value and relative volatility

The K-value of component i: K_i = y_i / x_i. For ideal Raoult: K_i = P_i^sat / P.

The relative volatility of light key over heavy key in a binary or pseudo-binary distillation:

α_LK,HK  =  K_LK / K_HK

α > 1.05 is the rule of thumb for distillation feasibility; below that, switch to extractive, azeotropic, or pressure-swing distillation, or to a different separation entirely.

Common K-value / EOS choices

Model	Best for	Where used
Ideal Raoult	Adjacent alkanes, low P	First-pass C5+ separations
Peng–Robinson (PR, 1976)	Hydrocarbons + light gases at high P	Refining, gas processing — industry default
Soave–Redlich–Kwong (SRK, 1972)	Similar to PR; slightly older	Refining alternative
PR–BM (Boston–Mathias)	PR with improved α(T) for polar	Modified hydrocarbons
NRTL-RK / UNIQUAC-RK	Polar liquid + non-ideal vapor	Specialty chemicals, pharma
eNRTL + Helmholtz	Aqueous electrolytes	CO₂ capture (MEA), sour gas (DEA, MDEA)
GERG-2008	Custody-transfer natural gas	Pipeline metering
SAFT, PC-SAFT	Polymers, associating fluids	Modern research, growing industrial use

Reaction equilibrium

For a reaction at equilibrium:

ΔG°_rxn = − R T ln K_eq
K_eq    = Π (a_i)^ν_i

ν_i is the stoichiometric coefficient (positive for products, negative for reactants), a_i the activity (≈ partial pressure in atm for gases, ≈ x·γ for liquids, ≈ 1 for pure solids/liquids in standard state).

Qualitative shifts follow Le Chatelier: increasing T favors the endothermic side; increasing P favors the side with fewer moles of gas; removing a product (reactive distillation, membrane reactor) pulls equilibrium forward.

5p. Reaction kinetics + reactor design

Rate law

For a reaction A + B → P:

r  =  k(T) · f(c_A, c_B, ...)

with the Arrhenius temperature dependence:

k(T)  =  A · exp(−E_a / (R T))

A is the pre-exponential factor, E_a the activation energy (typically 40–250 kJ/mol for chemical reactions; lower for diffusion-limited; higher for combustion + cracking).

An order-of-magnitude memory aid: rate roughly doubles per 10 K for E_a ≈ 50 kJ/mol near room temperature (the “Q10” rule borrowed from biochemistry). For E_a = 100 kJ/mol the factor is closer to ×4 per 10 K.

Reactor models

Model	Mixing	Time vs space	Equation (per V or per F_0)	Use
Batch	Perfectly mixed, closed	Time-dependent	dC_A/dt = r_A (V const)	Pharma, fine chem, specialty
CSTR (continuous stirred-tank)	Perfectly mixed, open	Steady-state, time-independent	V = F_{A0} X / (−r_A) (at exit conditions)	Large-volume liquid reactions, polymerization
PFR (plug-flow)	No axial mixing, complete radial	Steady-state, position-dependent	V / F_{A0} = ∫₀^X dX/(−r_A)	Gas-phase, high-temperature, refining
Semi-batch	Mixed, feed during reaction	Time-dependent	Time-varying mass + species balance	Exothermic with controlled feed (Grignard, nitration)
Packed bed (heterogeneous)	Plug-flow over solid catalyst	Steady-state	W / F_{A0} = ∫ dX/(−r’_A), η effectiveness factor	Catalytic — refining FCC, reforming, oxidation
Fluidized bed	Vigorously mixed solids in gas	Steady-state	Two-phase / Kunii–Levenspiel models	FCC, gasifiers, polyethylene
Membrane reactor	Plug-flow + selective removal	Steady-state	PFR + permeation term	Equilibrium-limited (H₂ from CH₄ reforming)
Bubble column / slurry	Gas dispersed in liquid + solid	Hold-up + mass-transfer limited	Two-film + reaction	Hydrogenation, oxidation, F-T synthesis
Trickle bed	Gas + liquid downflow over catalyst	Plug-flow with wetting	Holdup + transport corrections	Hydrotreating, hydrocracking

Core design equations

Batch reactor (constant volume, isothermal):

−dC_A/dt = (−r_A)              with C_A(0) = C_{A0}
t        = ∫_{C_A}^{C_{A0}} dC_A / (−r_A)

For first-order: t = (1/k) · ln(C_{A0}/C_A) = (1/k) · ln(1/(1−X)). For second-order in A only: t = (1/k) · (X / (C_{A0}(1−X))).

CSTR (steady-state, isothermal, single reaction):

V / v_0  =  τ  =  (C_{A0} · X) / (−r_A)      (−r_A evaluated at exit composition)

PFR (steady-state, isothermal, single reaction, constant volumetric flow):

V / F_{A0}  =  ∫₀^X dX / (−r_A)
τ            =  V / v_0   (constant v only)

Selectivity–yield optimization

Reactor design knobs:

• Temperature: balances rate vs equilibrium vs selectivity (often a selectivity-temperature trade-off in series-parallel networks).
• Residence time / conversion: longer τ → higher X but lower selectivity in series reactions A → P → byproduct.
• Reactant ratio / dilution: excess of one reactant pushes selectivity for parallel reactions; inert diluent moderates exotherms.
• Recycle: unreacted feed back to reactor; raises overall X without driving single-pass X to the selectivity-killing region.
• Reactor type: CSTR favors slower-rate selective routes (well-mixed, single low concentration); PFR favors fast reactions and series networks where intermediates are the target (concentration profile lets you stop midway).

6p. Unit operations — separations

After reaction, the next biggest cost in any chemical plant is separation — pulling product cleanly from byproducts and unreacted feed. The choice depends on driving force (volatility, solubility, size, charge, density) and the target purity.

Method	Driving force	Best for	Typical equipment
Distillation	Volatility (α > 1.05)	Bulk liquid mixtures, high throughput	Tray / packed columns
Absorption	Solubility in selective solvent	Removing minor vapor from gas (acid-gas, sour gas)	Packed / tray absorber
Stripping	Reverse of absorption	Removing dissolved gas from liquid (steam, air, N₂)	Packed / tray column
Liquid–liquid extraction	Partition coefficient	When distillation infeasible (azeotrope, T-sensitive)	Mixer-settler, RDC, Karr
Adsorption (PSA / TSA)	Selective adsorption on solid	Drying, H₂ purification, air separation	Fixed-bed cycles
Membrane	Selective permeation	Bulk gas (H₂, CO₂), desalination, juice concentration	Hollow-fiber, spiral-wound, flat sheet
Crystallization	Solubility, supersaturation	Pharma, sugar, salts, urea	Cooling / evaporative / anti-solvent
Drying	Vapor-pressure deficit	Solids dewatering	Tray, rotary, fluidized, spray, freeze
Filtration / centrifugation / sedimentation	Particle size, density	Solid–liquid	Plate-and-frame, rotary vacuum, decanter, cyclone

Distillation — the workhorse

McCabe–Thiele method (1925) gives a graphical solution for binary distillation, equivalent to step-counting on an x–y diagram between the equilibrium curve and the operating lines.

For multistage column design the shortcut method (FUG: Fenske–Underwood–Gilliland) is universal:

• Fenske equation — minimum number of stages at total reflux:

N_min + 1  =  ln[(x_LK / x_HK)_D · (x_HK / x_LK)_B] / ln(α_avg)

• Underwood equations — minimum reflux ratio R_min:

Σ_i  α_i · z_i / (α_i − θ)  =  1 − q          (feed equation — solve for θ ∈ (α_HK, α_LK))
Σ_i  α_i · x_{i,D} / (α_i − θ)  =  R_min + 1   (top equation — gives R_min)

• Gilliland correlation (1940) — N at finite R, given N_min and R_min:

(N − N_min) / (N + 1)  ≈  f((R − R_min) / (R + 1))   [chart or Eduljee correlation]

Rigorous column simulation in Aspen Plus, HYSYS, ProMax, PRO/II, gPROMS uses Naphtali–Sandholm (or Inside-Out) tray-by-tray methods with full MESH equations (Material, Equilibrium, Summation, Heat) per stage.

Tray vs packed columns:

• Trays — sieve, valve (Glitsch V1, Sulzer VG), bubble-cap. Robust, broad turndown, well-defined N_actual.
• Random packing — Pall ring, Intalox saddle, Raschig ring. Cheap, simple, moderate HETP.
• Structured packing — Sulzer Mellapak, Koch FlexiPac, Raschig Super-Pak. Low HETP (~0.3–0.5 m), low ΔP, used in vacuum and high-purity service.

Absorption / stripping — the Kremser equation

For dilute, isothermal, countercurrent gas–liquid contact with N theoretical stages and absorption factor A = L / (m·G):

(y_{N+1} − y_1) / (y_{N+1} − y_1^*)  =  (A^{N+1} − A) / (A^{N+1} − 1)        [Kremser 1930]

y_1^* = m · x_0 is the gas concentration in equilibrium with the inlet solvent. For A > 1 the absorber is feasible; A ≈ 1.4 is the economic optimum for amine acid-gas removal (MDEA/MEA).

Liquid–liquid extraction

Ternary phase diagrams (right-triangle or equilateral). The Hunter–Nash graphical method (1939) handles single-stage and countercurrent multistage extraction. Equipment: mixer-settler banks (slow, large hold-up — pharma standard), rotating-disc contactor (RDC), Karr reciprocating-plate column, pulsed sieve-plate column, centrifugal contactors (Podbielniak, Robatel).

Adsorption — PSA and TSA

• Pressure-swing adsorption (PSA) — typically 4–12 beds of zeolite or carbon molecular sieve cycling between high P (adsorb) and low P (regenerate). Standard for H₂ purification (steam-methane reforming), small-scale O₂ (medical), and air separation.
• Thermal-swing adsorption (TSA) — slower cycle, regenerated by hot purge gas. Standard for natural-gas dehydration (molecular sieve 3A or 4A), removing trace water before cryogenic processing.

Membrane separations

Type	Pore / mechanism	Use
MF (microfiltration)	0.1–10 µm	Bacteria, cells
UF (ultrafiltration)	1–100 nm	Proteins, macromolecules
NF (nanofiltration)	0.5–2 nm	Divalent salts
RO (reverse osmosis)	< 1 nm, solution-diffusion	Desalination, ultrapure water
Gas separation	Solubility-diffusion	H₂/CO₂ (Air Products PRISM), N₂/O₂
Pervaporation	Selective evaporation	Dehydration of organics (ethanol/water past azeotrope)
Electrodialysis	Ion exchange membranes + DC field	Brackish water, dairy whey

Crystallization, drying, mechanical separations

Crystallization modes: cooling (sucrose), evaporative (NaCl), anti-solvent (pharma), reactive (precipitation). Population-balance modeling on MSMPR (mixed-suspension, mixed-product-removal) — the “CSTR of crystallization.”

Drying uses constant-rate + falling-rate periods, mapped on drying curves. Equipment selection: tray dryer (small batch), rotary (granular, high-throughput), fluidized-bed (heat-sensitive), spray (liquid feed → fine powder, e.g. milk powder), freeze (heat-sensitive pharma/food).

Mechanical separation for solid–liquid: pressure filtration (plate-and-frame, leaf, candle), vacuum filtration (rotary drum, horizontal belt), centrifugation (decanter, disc-stack, basket), sedimentation (thickener, clarifier), hydrocyclones.

7p. Heat exchange in chemical process

Heat exchangers are everywhere in CPI — typical refinery has 200+ shell-and-tubes per train. Detailed methods are in heat-transfer; here are the process-specific pieces.

Rating methods: LMTD with correction factor F (for cross-flow and multipass) or ε-NTU. Refinery practice uses HTRI Xchanger Suite for rigorous shell-and-tube; Aspen EDR for design.

Equipment types (per TEMA 10th ed.):

• Shell-and-tube — workhorse; AES, BEM, NEN, AKT designations encode head + shell + rear-end types.
• Plate-and-frame (Alfa Laval, GEA) — high U, compact, gasketed (process-fouling risk).
• Plate-fin / brazed-aluminum — cryogenic LNG, air separation.
• Air-cooled (fin-fan) — process condensers, water-scarce sites.
• Spiral, double-pipe, scraped-surface — niche services.

Pinch analysis (Linnhoff, 1979)

The systematic method for heat-exchanger network (HEN) synthesis + utility targeting.

1. List all hot and cold streams with their (T_supply, T_target, CP = ṁ c_p).
2. Choose a minimum approach temperature ΔT_min — typically 10–20 °C in refineries, 5–10 °C in chemicals, 2–5 °C in cryogenics (LNG plate-fin).
3. Construct composite curves (hot + cold) on T–H axes; shift the cold curve to enforce ΔT_min.
4. The pinch point is the closest approach; above pinch needs hot utility, below needs cold utility.
5. Minimum utility targets Q_H,min and Q_C,min are read directly; HEN design follows.

Tools: Aspen Energy Analyzer, KBC SuperTarget. Properly done pinch can cut utility cost 10–40 % on a brownfield revamp.

8p. Control + instrumentation

Process control links the steady-state design to operating reality. Detailed control theory is in classical-control and mpc-control; here are the chemical-specific pieces.

P&ID conventions (ISA-5.1)

A piping and instrumentation diagram shows every line, vessel, valve, instrument, and control loop. Symbols:

• Circles = instruments (locally mounted = single line; DCS = horizontal line through circle; logic-implemented = diamond inside circle).
• First letter = measured variable (F flow, L level, T temperature, P pressure, A analyzer, S speed, J power).
• Second letter = function (I indicator, R recorder, C controller, T transmitter, V valve, Y computing, A alarm).
• Example: FIC-101 = flow indicating controller, tag 101.

Common loop types

Loop	Manipulated	Notes
Flow control (FC)	Control valve	Fast (~1–10 s); often inner loop in cascade
Level control (LC)	Outlet flow	Integrating process; tune for moderate response
Temperature control (TC)	Steam, fuel gas, coolant flow	Slow (minutes); often cascade
Pressure control (PC)	Vent, makeup, compressor recycle	Fast for gas, slow for liquid
Composition control (AC)	Reflux ratio, reactor T, ratio	Slow + noisy (analyzer dead time)
Cascade	Inner FC + outer TC/LC	Standard for reactor T control (TC sets cooling-water FC SP)
Ratio	Maintains ratio of two flows	Combustion air/fuel, blending
Feedforward + feedback	Disturbance + error correction	Distillation feed-disturbance compensation
Override / selector	High/low signal selector	Anti-surge for compressors, safety overrides

DCS and PLC platforms

Platform	Vendor	Strength
Honeywell Experion PKS	Honeywell	Refining, large continuous
Emerson DeltaV	Emerson	Chemicals, pharma, batch
Yokogawa CENTUM VP	Yokogawa	Oil & gas, LNG
Siemens PCS 7	Siemens	Chemicals, integrated with PLC
ABB Ability 800xA	ABB	Pulp + paper, mining, power
Rockwell PlantPAx	Rockwell	Mid-size hybrid plants on Logix PLC base
Schneider Foxboro Evo	Schneider	Refining, legacy I/A migration
PLC — ControlLogix (Rockwell), S7-1500 (Siemens), M580 (Schneider), MELSEC (Mitsubishi)	Various	Discrete + skid + safety logic

Advanced process control (APC)

A model-predictive control (MPC) layer above the DCS regulatory control. Standard products: AspenTech DMC3, Honeywell Profit Controller, Shell SMOC, Yokogawa SMOC. MPC handles multivariable, constrained, dead-time-heavy units (FCC, crude tower, ethylene cracker) and typically returns 1–5 % yield/throughput improvement on a refinery unit — millions of dollars per year per implementation.

9p. Process safety — the non-negotiable

OSHA PSM (29 CFR 1910.119) — the 14 elements

#	Element	What it requires
1	Employee participation	Workers consulted on PSM program
2	Process safety information (PSI)	Chemistry, technology, equipment data
3	Process hazard analysis (PHA)	HAZOP / What-if / FTA every 5 years
4	Operating procedures	Written, current, accessible
5	Training	Initial + refresh + documentation
6	Contractors	Selection, training, audit
7	Pre-startup safety review (PSSR)	Before commissioning new/modified
8	Mechanical integrity (MI)	Inspection, testing, QA of critical equipment
9	Hot work permit	Written permits for ignition sources
10	Management of change (MOC)	Any change reviewed + approved + documented
11	Incident investigation	Within 48 h, root-cause analysis
12	Emergency planning + response	ERP, evacuation, mutual aid
13	Compliance audits	Every 3 years
14	Trade secrets	Information access regardless of confidentiality

EPA RMP (40 CFR 68) overlays Program-3 requirements for the largest covered facilities.

HAZOP (hazard and operability study)

Structured deviation analysis using guide words (No, More, Less, As well as, Part of, Reverse, Other than) applied to parameters (Flow, Pressure, Temperature, Level, Composition, Time). For each node:

1. Identify deviation (e.g. “No flow in line 23”).
2. Identify causes (pump trip, valve closed, blockage).
3. Identify consequences (vessel under-cools, overpressure, runaway).
4. Identify existing safeguards (LL alarm, interlock, PSV).
5. Recommend further action if needed (add HH alarm, SIL-rated SIF, mechanical change).

A typical unit HAZOP is 2–5 days of 6–10 people; the cost is justified by avoided incidents.

LOPA (layer of protection analysis)

Quantifies the gap between initiating-event frequency and tolerable consequence frequency.

Mitigated freq  =  Initiating freq  ×  Π PFD_IPL

Each independent protection layer (IPL) — alarm + operator action, BPCS interlock, SIS interlock, mechanical relief, dike, plant ERP — contributes a PFD (probability of failure on demand). The SIS must close the residual risk gap to meet the target SIL per IEC 61511.

SIL — Safety Integrity Level (IEC 61511)

SIL	Demand-mode PFD	RRF	Example
1	10⁻² – 10⁻¹	10 – 100	Reactor high-T trip on cooling water loss
2	10⁻³ – 10⁻²	100 – 1 000	Vessel high-P trip with redundant sensors
3	10⁻⁴ – 10⁻³	1 000 – 10 000	Compressor surge prevention, 2oo3 voted
4	10⁻⁵ – 10⁻⁴	10 000 – 100 000	Rare in CPI; typical of nuclear/aerospace

Implementation: SIL-2/3 logic solvers — HIMA HIMax, Triconex Tricon/Trident, Siemens S7-1500F, Rockwell GuardLogix; SIL-rated transmitters and valves with certified SFF (safe-failure fraction) and HFT (hardware fault tolerance).

Relief sizing (API 521 / 520)

Every pressurized vessel needs relief protection sized for the worst credible scenario:

• Fire (external pool fire — heat input via API 521 / 14E equations).
• Blocked outlet (downstream isolation valve closed).
• Loss of cooling / reflux.
• Heat-exchanger tube rupture (high-P shell into low-P tube or vice versa).
• Runaway reaction (vent capacity may need DIERS two-phase methodology).
• Thermal expansion (block-in liquid heated by sun).
• Utility failure (electricity, instrument air, cooling water).

API 520 Part I gives PSV (pressure-safety-valve) orifice sizing equations for gas, vapor, liquid, two-phase, and steam. API 521 lists credible scenarios and gives flare-system + vent-stack sizing. The sized capacity goes into a rupture-disc + PSV + flare stack (KO drum → flare header → flare tip).

Inherently safer design (Trevor Kletz)

Four principles, ordered by preference:

1. Minimize — smaller inventory of hazardous material (intensified reactor, just-in-time delivery).
2. Substitute — replace with less hazardous chemistry (water-based solvent for benzene).
3. Moderate — lower T, lower P, dilute (liquid storage at atmospheric instead of pressurized).
4. Simplify — eliminate failure modes (no bypasses, no manual valves on critical lines).

Other code obligations

Standard	Scope
ASME B31.3	Process piping design + fabrication
ASME BPVC Section VIII	Pressure-vessel design
API 510	Pressure-vessel inspection
API 570	Piping inspection
API 653	Atmospheric tank inspection
NFPA 30	Flammable + combustible liquids
NFPA 68 / 69	Deflagration venting / explosion prevention
IEC 61511	Functional safety of process SIS
IEC 62443	OT/ICS cybersecurity
API 14C	Offshore safety analysis
CSB recommendations	US Chemical Safety Board incident findings + non-binding lessons

10p. Worked examples

Example A — Mass balance with recycle (ammonia synthesis)

Problem. Haber–Bosch ammonia synthesis: N₂ + 3 H₂ → 2 NH₃ at 150 bar, 450 °C. Fresh feed is 1 kmol/s N₂ + 3 kmol/s H₂ (stoichiometric). Single-pass conversion of N₂ across the converter is 25 %. Argon inert enters as 0.5 mol% of the N₂ feed (5 mol/s). After the converter, NH₃ is condensed out completely; unreacted N₂ + H₂ + Ar are recycled. A purge is bled off to control Ar buildup. Find: recycle rate, purge rate, overall conversion.

Setup. Let R = recycle molar flow [kmol/s], P_purge = purge flow [kmol/s], with mole fractions of N₂, H₂, Ar in the recycle (= reactor feed minus fresh feed). Let X_op = overall conversion (NH₃ out / N₂ fresh in).

Argon balance (Ar in = Ar out, since Ar is inert):

Ar in fresh feed  =  Ar in purge
0.005 kmol/s      =  P_purge · y_Ar

If we let the recycle Ar concentration reach 10 mol% (typical operating choice), then y_Ar = 0.10 in the loop, so P_purge · 0.10 = 0.005, giving P_purge = 0.050 kmol/s.

Overall N₂ balance (steady state, NH₃ produced uses N₂ stoichiometrically):

N₂ fresh in  =  N₂ in purge  +  N₂ converted to NH₃
1.0          =  0.050 · y_{N₂,purge}  +  (1/2) · NH₃_out

With y_{N₂,purge} ≈ 0.225 (typical, N₂ + H₂ in 1:3 in the loop, balance Ar + small NH₃ slip):

1.0 = 0.050 × 0.225 + 0.5 × NH₃_out
0.9888 = 0.5 × NH₃_out
NH₃_out ≈ 1.978 kmol/s

Overall conversion = NH₃_out × (1/2) / N₂_fresh = 0.989 → 98.9 %, consistent with industry experience (~95–98 % overall vs ~25 % per pass).

Recycle ratio. With per-pass X = 0.25 on N₂ and reactor inlet N₂ flow = N₂_fresh + N₂_recycle = 1 + R · y_{N₂}, balance gives R · y_{N₂} ≈ (1 − 0.25)/0.25 × 1.0 = 3.0 kmol/s of recycled N₂; recycle total ≈ 3.0 / 0.225 ≈ 13.3 kmol/s (about 13× the fresh N₂). The Aspen Plus or HYSYS flowsheet would close this by tear-stream iteration with Wegstein acceleration.

Example B — McCabe–Thiele for benzene/toluene distillation

Problem. Feed F = 100 kmol/h, x_F = 0.40 benzene (mole fraction), saturated liquid (q = 1). Distillate D = 50 kmol/h at x_D = 0.95, bottoms B = 50 kmol/h at x_B = 0.05 (overall benzene balance check: 100·0.40 = 40 = 50·0.95 + 50·0.05 = 47.5 + 2.5 = 50 ✗ … let’s correct).

Re-balance. F · x_F = D · x_D + B · x_B → 40 = D · 0.95 + (100 − D) · 0.05 → 40 = 0.05·100 + 0.90·D → D = 35/0.90 = 38.9 kmol/h, B = 61.1 kmol/h. (The original F splits don’t balance with x_F = 0.40; this is the corrected version.)

Relative volatility. At column-average T ≈ 95 °C, α_BT ≈ 2.4 (Antoine + Raoult).

Step 1 — Fenske (N_min, total reflux).

N_min + 1  =  ln[(x_LK/x_HK)_D · (x_HK/x_LK)_B] / ln α
            =  ln[(0.95/0.05) · (0.95/0.05)] / ln 2.4
            =  ln[19 · 19] / 0.875
            =  ln(361) / 0.875
            =  5.889 / 0.875
            =  6.73
N_min       =  5.73  →  ~6 ideal stages at total reflux

Step 2 — Underwood (R_min). With q = 1 (sat liquid), the feed equation:

Σ α_i z_i / (α_i − θ) = 1 − q = 0
α_B z_B / (α_B − θ) + α_T z_T / (α_T − θ) = 0
2.4 · 0.40 / (2.4 − θ) + 1.0 · 0.60 / (1.0 − θ) = 0
0.96 (1 − θ) + 0.60 (2.4 − θ) = 0
0.96 − 0.96 θ + 1.44 − 0.60 θ = 0
2.40 − 1.56 θ = 0
θ = 1.538

Top equation:

R_min + 1 = Σ α_i x_{i,D} / (α_i − θ)
          = 2.4 · 0.95 / (2.4 − 1.538) + 1.0 · 0.05 / (1.0 − 1.538)
          = 2.28 / 0.862  +  0.05 / (−0.538)
          = 2.645 − 0.093
          = 2.552
R_min = 1.55

Step 3 — Operating R and Gilliland. Pick R = 1.5 · R_min = 2.32. Gilliland correlation (Eduljee fit):

(R − R_min) / (R + 1)  =  (2.32 − 1.55) / 3.32  =  0.232
(N − N_min) / (N + 1)  ≈  0.75 (1 − x^0.566) = 0.75 (1 − 0.232^0.566) ≈ 0.418

Solve: N − N_min = 0.418 (N + 1) → N − 5.73 = 0.418 N + 0.418 → N (1 − 0.418) = 6.148 → N ≈ 10.6 ideal stages.

Step 4 — Real trays. At overall tray efficiency η_O ≈ 0.65 (typical for B/T system on sieve trays): N_actual ≈ 10.6 / 0.65 ≈ 17 real trays + reboiler. Feed-tray location from Kirkbride correlation; final answer normally rounded up + design margin added.

In practice this column would be re-rated rigorously in Aspen Plus RadFrac or HYSYS Distillation column block; the shortcut method here is just for sizing the first-pass equipment.

Example C — PFR sizing for a second-order liquid reaction

Problem. Irreversible 2nd-order liquid reaction A + B → C, rate −r_A = k · C_A · C_B, k = 0.05 L/(mol·s) at 80 °C. Feed: C_{A0} = 2 mol/L, C_{B0} = 4 mol/L (so θ_B = C_{B0}/C_{A0} = 2). Volumetric flow v_0 = 10 L/s. Target conversion X_A = 0.90. Find PFR volume.

Stoichiometric relations (constant ρ, constant v_0):

C_A = C_{A0} (1 − X)
C_B = C_{A0} (θ_B − X)
−r_A = k · C_A · C_B  =  k · C_{A0}² · (1 − X) · (θ_B − X)

PFR design equation:

τ  =  C_{A0} · ∫₀^X dX / (−r_A)
   =  C_{A0} · ∫₀^X dX / [k · C_{A0}² · (1 − X) · (θ_B − X)]
   =  (1 / (k · C_{A0})) · ∫₀^X dX / [(1 − X) · (θ_B − X)]

The integral (partial fractions, θ_B ≠ 1):

∫ dX / [(1 − X) · (θ_B − X)]  =  (1 / (θ_B − 1)) · ln[(θ_B − X) / (θ_B · (1 − X))]

For X = 0.90, θ_B = 2:

∫₀^0.9 = (1/1) · ln[(2 − 0.9) / (2 · (1 − 0.9))]
       = ln[1.1 / 0.2]
       = ln(5.5)
       = 1.705

Then:

τ = (1 / (0.05 · 2)) · 1.705 = 10 · 1.705 = 17.05 s
V = τ · v_0 = 17.05 · 10 = 170 L

Check by alternative reactor. CSTR at same X (evaluated at exit composition):

τ_CSTR = C_{A0} · X / (−r_A)|_exit
       = 2 · 0.9 / [0.05 · 2 · 0.1 · 2.2]
       = 1.8 / 0.022
       = 81.8 s
V_CSTR = 818 L  ≈  4.8× the PFR

For positive-order kinetics, PFR is always smaller than CSTR for the same conversion — the classic chemical-process-fundamentals trade-off. A bench-scale PFR (170 L) is feasible; a CSTR option would push to 818 L, more capital + more residence-time selectivity loss in any series reaction.

11p. Edge cases / gotchas

• Equilibrium-limited reactions — methanation, ammonia synthesis, water-gas shift, methanol synthesis. Push by lowering T (if exothermic) until kinetics drop too far, then use multistage with inter-stage cooling; excess reactant dilutes products; product removal (reactive distillation, membrane reactor) drives equilibrium forward. Reactive distillation in methyl-acetate manufacture (Eastman) replaced 11 unit operations with one column.
• Catalyst deactivation — coking (FCC, reforming), sintering (Ni at high T), poisoning (Pt by S, Cu by Cl, Pd by Hg), fouling. Design includes spare reactor + scheduled regen; catalyst-life economics are the hidden cost of the whole reactor block.
• Runaway exotherms — Seveso 1976 (TCDD release from a TCP plant), Bhopal 1984, T2 Laboratories 2007. Whenever ΔH_rxn × concentration exceeds the heat-removal capacity of the cooling system, T accelerates and the reaction tip-toes onto the exponential Arrhenius slope. Design rule: thermal mass + jacket area + emergency relief sized assuming worst credible scenario (loss of cooling, runaway, two-phase venting per DIERS).
• Non-ideal VLE — azeotropes — ethanol–water at x = 0.956 mol (95.6 wt%) at 1 atm. Cannot be separated by ordinary distillation past the azeotrope. Solutions: pressure-swing distillation (azeotrope composition shifts with P), entrainer azeotropic distillation (benzene historically, now cyclohexane), extractive distillation (ethylene glycol), pervaporation (zeolite A membrane → 99.5 % ethanol, used at commercial fuel-ethanol plants).
• Solids handling — bridging, ratholing, segregation — Jenike shear-cell test before designing any hopper. Mass-flow vs funnel-flow geometry; eccentric outlets cause segregation by size or density. Adding vibration, air pads, mechanical activators when needed.
• Heat-exchanger fouling — every CPI exchanger fouls. Design includes a fouling resistance R_f (TEMA Table A.1 typical values: 0.0001–0.001 m²·K/W) added to U; choose cleanable geometry (U-tube, plate-frame) for high-fouling service; schedule chemical or mechanical cleaning. CIP (clean-in-place) cycle is mandatory in food/pharma.
• Trace impurities can kill catalysts at ppb levels — Pt reformer poisoned by 1 ppb S; methanol synthesis Cu/ZnO/Al₂O₃ poisoned by Cl. Guard beds (sulfided activated C, ZnO bed, chloride trap) handle low-level removal before the main reactor.
• Material compatibility for tough service — anhydrous HF needs PFA-lined steel or Hastelloy C-276; high-T high-P hydrogen needs Cr-Mo steels per the Nelson curves (API 941) to avoid hydrogen attack; chloride stress-corrosion cracking limits 304/316 SS in marine vapor service; wet H₂S sour service triggers NACE MR0175. See materials-steel.
• Scale-up traps — pilot scaled-up to commercial often fails on mixing time (impeller power per volume), heat-transfer area-to-volume ratio (drops with scale, exotherms become harder), residence-time distribution (lab CSTR may behave like commercial PFR), and side reactions that didn’t appear at lab T but do at the slightly higher commercial T. The pilot phase is not optional.
• Cyber-physical attacks — Stuxnet (2010, Iranian centrifuge controllers), Triconex Triton/Trisis (2017, Saudi petrochemical SIS) showed that ICS networks are real attack surfaces. IEC 62443 zoning + conduit segmentation + air-gapped SIS where credible.
• Process intensification — microreactors, spinning-disc reactors, oscillatory-baffled reactors, supercritical CO₂ extraction. Promising but mostly niche; specialty chemicals + pharma are the easiest fits.
• Hidden inventory in piping + thermal mass — relief-case analysis must account for fluid trapped in long pipe runs and thermal mass of vessels. Block-in scenarios on long subsea hydrocarbon lines have caused several incidents from solar heating alone.

12p. Tools (ecosystem)

Steady-state process simulation:

• Aspen HYSYS — gas processing, refining, hydrocarbons; PR / SRK / NRTL backbone; industry default in O&G.
• Aspen Plus — chemicals, polymers, electrolytes; richer property package (NRTL, eNRTL, UNIQUAC, PC-SAFT); industry default in chemicals + pharma.
• ProMax (Bryan Research & Engineering) — gas processing, amine sweetening, sulfur recovery; strong on amines + glycols.
• PRO/II + DynSim (AVEVA / SimSci) — refining, petrochemicals; legacy strong base.
• UniSim Design (Honeywell) — HYSYS-like, integrated with Honeywell DCS.
• Symmetry (Schlumberger) — upstream + midstream gas.
• gPROMS (Siemens PSE) — equation-oriented, custom unit operations, pharma + advanced modeling.
• DWSIM — open-source (.NET, cross-platform), good for teaching + small-scale design.
• CHEMCAD (Chemstations) — mid-tier commercial; broad CPI use.

Dynamic simulation: Aspen HYSYS Dynamics, Aspen Plus Dynamics, UniSim Dynamic, Petro-SIM Dynamic, gPROMS, ProTreat.

Reaction engineering: Cantera (open-source, gas-phase kinetics + transport), Aspen RPlug/RBatch/RCSTR, Athena Visual Studio, COMSOL Chemical Reaction Engineering module.

Heat-exchanger design: HTRI Xchanger Suite (industry standard, shell-and-tube + plate-fin + air-cooled), Aspen Exchanger Design and Rating (EDR), Aspen B-JAC (legacy shell-and-tube), HTFS+ for tube-side.

Pinch + energy integration: Aspen Energy Analyzer, KBC SuperTarget, ProSimPlus Energy.

CFD for reactors and mixers: ANSYS Fluent, Siemens Star-CCM+, COMSOL CFD, OpenFOAM (open-source).

Properties / thermodynamics databases: NIST REFPROP 10.0, DIPPR Project 801, CAPE-OPEN compliant Aspen Properties, ChemSep + ChemBrain, Yaws compilations.

Process safety:

• PHA-Pro (Sphera) — HAZOP/LOPA/What-if facilitation tool, industry standard.
• PHAST + Safeti (DNV) — consequence modeling (releases, fires, explosions, dispersion).
• ALOHA (EPA/NOAA, free) — emergency dispersion modeling.
• DEGADIS / SLAB — dense-gas dispersion (heavier-than-air, e.g. chlorine, LNG vapor cloud).
• CANARY by Quest (Worley) — vapor cloud dispersion + radiation.
• HEM (HSE methodology) — toxic + flammable consequence for UK/Seveso.

P&ID + 3D: AVEVA E3D / PDMS / Diagrams, Bentley OpenPlant, AutoCAD Plant 3D, Intergraph SmartPlant 3D / P&ID.

DCS: Honeywell Experion PKS, Emerson DeltaV, Yokogawa CENTUM VP, Siemens PCS 7, ABB 800xA, Rockwell PlantPAx, Schneider Foxboro Evo.

PLC: Rockwell ControlLogix + GuardLogix, Siemens S7-1500 + S7-1500F, Schneider Modicon M580, Mitsubishi MELSEC iQ-R, Omron NX/NJ.

Safety logic solvers: HIMA HIMax / HIMatrix, Triconex Tricon CX / Trident, Siemens S7-400F/H, ABB AC800M HI, Rockwell GuardLogix.

13. Cross-references

• thermodynamics — VLE, reaction equilibrium, energy balance, exergy.
• heat-transfer — exchanger sizing, fired-heater, condenser/reboiler design, LMTD/ε-NTU.
• fluid-mechanics — pipe sizing, two-phase flow, pump NPSH, compressible flow, Crane TP-410.
• pumps-turbomachinery — process pumps (ANSI B73.1, API 610), compressors, expanders.
• materials-steel — Nelson curves for high-T H₂, NACE MR0175 sour service, stainless selection.
• classical-control — PID tuning, cascade, feedforward, anti-windup.
• digital-control — DCS/PLC implementation, discretization.
• mpc-control — APC layer (DMC3, Profit, SMOC) on top of DCS.
• microcontrollers — embedded controls in skid systems + smart instruments.
• scientific — Aspen, HYSYS, gPROMS, DWSIM input formats; CAPE-OPEN; Modelica thermo libraries (planned).
• industrial-automation — IEC 61131-3 Structured Text for PLC + DCS function blocks (planned).
• chemical-process-fundamentals — domain applications (planned).

14. Citations

1. Felder, R. M.; Rousseau, R. W.; Bullard, L. G. Elementary Principles of Chemical Processes, 4th ed. Wiley, 2015. ISBN 978-1118431221. Canonical undergraduate mass + energy balance textbook.
2. Himmelblau, D. M.; Riggs, J. B. Basic Principles and Calculations in Chemical Engineering, 8th ed. Prentice Hall, 2012. ISBN 978-0132346603.
3. Smith, J. M.; Van Ness, H. C.; Abbott, M. M.; Swihart, M. T. Introduction to Chemical Engineering Thermodynamics, 9th ed. McGraw-Hill, 2022. ISBN 978-1260597684.
4. Sandler, S. I. Chemical, Biochemical, and Engineering Thermodynamics, 5th ed. Wiley, 2017. ISBN 978-0470504796. Strongest treatment of solution thermodynamics + phase equilibria.
5. Fogler, H. S. Elements of Chemical Reaction Engineering, 6th ed. Pearson, 2020. ISBN 978-0135486221. The canonical reactor-design textbook; ISE / global standard.
6. Levenspiel, O. Chemical Reaction Engineering, 3rd ed. Wiley, 1999. ISBN 978-0471254249. Classical reactor-engineering text; PFR/CSTR comparisons.
7. Geankoplis, C. J.; Hersel, A. A.; Lepek, D. H. Transport Processes and Separation Process Principles, 5th ed. Pearson, 2018. ISBN 978-0134181028.
8. McCabe, W. L.; Smith, J. C.; Harriott, P. Unit Operations of Chemical Engineering, 7th ed. McGraw-Hill, 2005. ISBN 978-0072848236. Legacy reference for unit ops.
9. Seader, J. D.; Henley, E. J.; Roper, D. K. Separation Process Principles, 4th ed. Wiley, 2016. ISBN 978-1119139621. Modern standard separations text.
10. Wankat, P. C. Separation Process Engineering, 4th ed. Prentice Hall, 2017. ISBN 978-0133443653.
11. Linnhoff, B. et al. A User Guide on Process Integration for the Efficient Use of Energy, IChemE 1982; revised editions through Linnhoff March / KBC. The pinch-analysis canonical reference.
12. Crowl, D. A.; Louvar, J. F. Chemical Process Safety: Fundamentals with Applications, 4th ed. Prentice Hall, 2019. ISBN 978-0134857770. The canonical process-safety textbook.
13. CCPS (Center for Chemical Process Safety). Guidelines for Hazard Evaluation Procedures, 3rd ed. Wiley-AIChE, 2008. ISBN 978-0471978152. HAZOP / LOPA / What-if methodology reference.
14. CCPS. Guidelines for Initiating Events and Independent Protection Layers in Layer of Protection Analysis, Wiley-AIChE, 2014. ISBN 978-1118777930.
15. API Standard 520 Part I (sizing) + Part II (installation), 10th ed., American Petroleum Institute, 2022. PSV sizing methodology.
16. API Standard 521 Pressure-relieving and Depressuring Systems, 7th ed., American Petroleum Institute, 2020. Relief-scenario inventory + flare-system design.
17. API RP 14C Analysis, Design, Installation, and Testing of Safety Systems for Offshore Production Facilities, 8th ed., 2017.
18. API RP 14E Design and Installation of Offshore Production Platform Piping Systems, 6th ed., 2013.
19. ASME B31.3-2024 Process Piping. American Society of Mechanical Engineers.
20. OSHA 29 CFR 1910.119 Process Safety Management of Highly Hazardous Chemicals, 1992 (current). https://www.osha.gov/laws-regs/regulations/standardnumber/1910/1910.119
21. IEC 61511 Functional safety — Safety instrumented systems for the process industry sector, 2nd ed., 2016, with Amendment 1 (2017). Three parts: framework, requirements, guidance.
22. IEC 61508 Functional safety of electrical/electronic/programmable electronic safety-related systems, 2nd ed., 2010. The generic FS standard underlying IEC 61511.
23. NFPA 30-2024 Flammable and Combustible Liquids Code.
24. NFPA 68-2023 Standard on Explosion Protection by Deflagration Venting.
25. NFPA 77-2019 Recommended Practice on Static Electricity.
26. ISA-5.1-2009 Instrumentation Symbols and Identification. International Society of Automation. P&ID symbol standard.
27. ISA-101.01-2015 Human Machine Interfaces for Process Automation Systems.
28. Kletz, T. What Went Wrong? Case Histories of Process Plant Disasters and How They Could Have Been Avoided, 6th ed. Butterworth-Heinemann, 2019. ISBN 978-0128105399. Inherent-safety canon.
29. Aspen Technology. Aspen Plus User Guide V14 + Aspen HYSYS User Guide V14. Bedford, MA, 2024. Vendor reference for industry-standard process simulators.
30. DIPPR Project 801. Evaluated Process Design Data, Design Institute for Physical Properties, AIChE. Annual updates. Property database underlying most CPI simulators.