Refining Unit Operations¶

This section describes a suite of shortcut unit operation models intended for conceptual design, mass/energy balance closure, and techno-economic screening of petroleum refineries. Each model is a steady-state, explicit calculation that captures the principal yield, conversion, and contaminant-partitioning behaviour of its reference process without requiring detailed kinetic parameter tuning. All models share a common contaminant vector (sulfur, nitrogen, mercaptan sulfur, Ni/V/Fe, CCR, asphaltenes, TAN) that is carried across the flowsheet through mass-weighted pseudocomponent partitioning.

Shortcut Crude Distillation Unit (CDU)¶

Overview¶

The Shortcut CDU models an atmospheric (optionally coupled with vacuum) crude distillation tower as a set of TBP-cut-point separations. The feed crude assay is characterised by a set of pseudocomponents generated from the bulk TBP curve following the method of Riazi and Daubert . No tray-by-tray calculation is performed; the tower is treated as an idealised sequence of sharp cuts with user-specified overlap expressed through a cut-point uncertainty band.

Stream Topology¶

Port	Direction	Description
Crude In	Inlet (material)	Preheated crude (post-desalter)
LPG / OffGas	Outlet (material)	Tower overhead light ends
Naphtha	Outlet (material)	Light + heavy naphtha combined cut
Kerosene	Outlet (material)	Kerosene / jet side-draw
Diesel	Outlet (material)	Light + heavy diesel (AGO) side-draw
AR / Bottoms	Outlet (material)	Atmospheric residue

Cut-Point Distribution¶

Each pseudocomponent \(c\) in the feed is assigned to the product cut whose TBP window contains its normal boiling point \(T_{b,c}\):

\[ w_{c \rightarrow k} = \tfrac{1}{2}\!\left[ \mathrm{erf}\!\left(\frac{T_{b,c} - T_{k-1}}{\sigma\sqrt{2}}\right) - \mathrm{erf}\!\left(\frac{T_{b,c} - T_{k}}{\sigma\sqrt{2}}\right) \right] \]

where \(T_{k-1}\) and \(T_{k}\) are the lower and upper TBP cut points of product \(k\) and \(\sigma\) (K) is the cut-point uncertainty. When \(\sigma \rightarrow 0\) [eq:cdu_cutweight] reduces to a sharp-cut assignment.

The product mass flow is obtained by summing the cut fractions over all pseudocomponents:

\[ \dot m_k = \sum_{c} w_{c \rightarrow k}\, \dot m_c \]

Contaminant Partitioning¶

Contaminants (S, N, Ni, V, Fe, CCR, asphaltenes) are distributed according to user-supplied boiling-point-dependent concentration curves or, alternatively, using the default correlations of Gary, Handwerk & Kaiser . Metals and CCR are concentrated in the residue according to

\[ x_k^{\mathrm{cont}} = f_{\mathrm{cont}}(T_{b,k}) \]

with a mass-balance closure step that renormalises the distribution so that the total contaminant mass across all products matches the feed.

Hydrodesulphurisation (HDS) Reactor¶

Overview¶

The Shortcut HDS model represents a fixed-bed hydrotreating reactor operating on middle-distillate feeds. Sulfur conversion follows an \(n\)-th order power-law kinetic expression with explicit hydrogen partial-pressure dependence , evaluated at liquid hourly space velocity (LHSV). Nitrogen removal is optionally computed using an independent Arrhenius law.

Stream Topology¶

Port	Direction	Description
Feed In	Inlet (material)	Hydrocarbon feed
H\(_2\) Make-up	Inlet (material)	Hydrogen feed
Product Out	Outlet (material)	Desulfurised hydrocarbon
Sour Gas Out	Outlet (material)	H\(_2\)S + NH\(_3\) + unreacted H\(_2\)

Kinetics¶

The pseudo-homogeneous sulfur conversion rate is

\[ r_{\mathrm{S}} = k(T)\, C_{\mathrm{S}}^{\,n}\, P_{\mathrm{H_2}}^{\,m} \]

with

\[ k(T) = k_0\, \exp\!\left(-\frac{E_a}{R T}\right) \]

where \(n\) is the reaction order in sulfur (typically 1.5–2), \(m\) is the hydrogen partial-pressure order, \(k_0\) is the pre-exponential factor, \(E_a\) is the apparent activation energy, \(T\) is the reactor temperature, and \(C_{\mathrm{S}}\) is the sulfur mass concentration in the liquid phase.

Integrating [eq:hds_rate] over the reactor residence time \(\tau = 1/\mathrm{LHSV}\) for \(n \ne 1\) gives the sulfur conversion:

\[ X_{\mathrm{S}} = 1 - \left[\,1 + (n-1)\, k\, P_{\mathrm{H_2}}^m \, C_{\mathrm{S,0}}^{\,n-1}\, \tau\,\right]^{-1/(n-1)} \]

Mercaptan sulfur is removed at a user-specified fraction \(X_{\mathrm{RSH}}\) and the removed sulfur is added to the sour-gas stream as H\(_2\)S.

Hydrogen Balance¶

Chemical hydrogen consumption is proportional to the mass of sulfur and (optionally) nitrogen removed:

\[ \dot n_{\mathrm{H_2}}^{\mathrm{chem}} = \alpha_{\mathrm{S}}\, \dot m_{\mathrm{S,rem}} + \alpha_{\mathrm{N}}\, \dot m_{\mathrm{N,rem}} \]

with stoichiometric coefficients \(\alpha_{\mathrm{S}}\) (mol H\(_2\) / kg S) and \(\alpha_{\mathrm{N}}\) (mol H\(_2\) / kg N).

Fluid Catalytic Cracking (FCC) Unit¶

Overview¶

The Shortcut FCC supports two yield methods: an empirical yield-slate adjusted by feed quality (CCR) and a simplified three-lump Weekman kinetic model . The unit models the riser/regenerator pair lumped as a single conversion volume operating at a specified temperature, pressure, and catalyst-to-oil ratio.

Stream Topology¶

Port	Direction	Description
Feed In	Inlet (material)	VGO / gas-oil feed
Dry Gas	Outlet (material)	C\(_2^-\) cracked gas
LPG	Outlet (material)	C\(_3\)–C\(_4\) olefin-rich cut
Gasoline	Outlet (material)	FCC gasoline (naphtha range)
LCO	Outlet (material)	Light cycle oil
Slurry	Outlet (material)	Heavy cycle oil / slurry
Flue Gas	Outlet (material)	Regenerator effluent (CO\(_2\), O\(_2\), N\(_2\))

Weekman Three-Lump Kinetics¶

The classical Weekman & Nace three-lump formulation tracks gas-oil (\(A\)), gasoline (\(B\)), and coke+gas (\(C\)):

\[ \begin{align} A \xrightarrow{k_1} B, \qquad A \xrightarrow{k_3} C, \qquad B \xrightarrow{k_2} C \end{align} \]

Assuming second-order gas-oil cracking and first-order gasoline overcracking, the lump balances along the riser are

\[ \begin{align} \frac{\mathrm d A}{\mathrm d\tau} &= -(k_1 + k_3)\, A^2 \\ \frac{\mathrm d B}{\mathrm d\tau} &= k_1\, A^2 - k_2\, B \end{align} \]

where \(\tau\) is the riser residence time. All rate constants are modulated by a catalyst deactivation function \(\phi(t_c)\) dependent on the catalyst-on-stream time \(t_c\) . The conversion \(X = 1 - A/A_0\) is obtained by explicit integration.

Yield-Slate Method¶

When the Slate method is selected, product yields \(Y_i^0\) at a reference CCR are tabulated and corrected for feed CCR according to

\[ Y_{\mathrm{coke}} = Y_{\mathrm{coke}}^0 + \beta_{\mathrm{CCR}} \left(\mathrm{CCR}_{\mathrm{feed}} - \mathrm{CCR}_{\mathrm{ref}}\right) \]

with all other yields rebalanced to close the mass balance.

Regenerator Heat Release¶

The coke combustion duty in the regenerator is

\[ \dot Q_{\mathrm{regen}} = \dot m_{\mathrm{coke}}\, \Delta H_{\mathrm{coke}}^{\mathrm{comb}} \]

with \(\Delta H_{\mathrm{coke}}^{\mathrm{comb}}\) a user-editable coke combustion heat (default 32 500 kJ kg\(^{-1}\)).

Hydrocracker (HCR)¶

Overview¶

The Shortcut HCR converts VGO / atmospheric residue into light ends, naphtha, kerosene, diesel, and unconverted oil (UCO) according to a target conversion and a yield slate that is adjusted as conversion moves away from the reference point. Hydrogen consumption is derived from conversion and heat of reaction.

Stream Topology¶

Port	Direction	Description
Feed In	Inlet (material)	VGO / DAO feed
H\(_2\) Make-up	Inlet (material)	Hydrogen
Light Ends	Outlet (material)	C\(_1\)–C\(_4\)
Naphtha	Outlet (material)	C\(_5\)–C\(_{11}\)
Kerosene	Outlet (material)	C\(_{11}\)–C\(_{14}\)
Diesel	Outlet (material)	C\(_{14}\)–C\(_{20}\)
UCO	Outlet (material)	Unconverted oil
Sour Gas	Outlet (material)	H\(_2\)S + NH\(_3\) + unreacted H\(_2\)

Yield Model¶

At the target conversion \(X\), the yield of each light cut is

\[ Y_i = Y_i^{\mathrm{base}} \left( 1 + \gamma_i \left( X - X_{\mathrm{base}} \right) \right) \]

where \(Y_i^{\mathrm{base}}\) is the tabulated yield at the reference conversion \(X_{\mathrm{base}}\) and \(\gamma_i\) is the conversion sensitivity. The UCO yield is obtained by mass-balance closure. Sulfur and nitrogen are removed at user-specified fractions \(X_{\mathrm{S}}\) and \(X_{\mathrm{N}}\) with the removed mass reported as H\(_2\)S and NH\(_3\).

Heat of Reaction¶

The net reactor exotherm is

\[ \dot Q_{\mathrm{rxn}} = - \dot m_{\mathrm{feed}}\, X\, \Delta H_{\mathrm{conv}} \]

with \(\Delta H_{\mathrm{conv}}\) the enthalpy released per kg of gas-oil converted.

Delayed Coker¶

Overview¶

The Shortcut Coker models a delayed-coking drum at a specified heater outlet temperature and drum pressure. Product yields are computed from empirical CCR-based correlations in the Gary, Handwerk & Kaiser style, with explicit partitioning of dry gas from total gas and LGO from total gas-oil.

Stream Topology¶

Port	Direction	Description
Feed In	Inlet (material)	Vacuum residue
Dry Gas	Outlet (material)	C\(_2^-\)
LPG	Outlet (material)	C\(_3\)–C\(_4\)
Naphtha	Outlet (material)	Coker naphtha
LGO	Outlet (material)	Light gas-oil
HGO	Outlet (material)	Heavy gas-oil
Coke Out	Outlet (material)	Petroleum coke (solid)

Yield Correlations¶

The coke yield is a linear function of feed Conradson carbon residue (CCR):

\[ Y_{\mathrm{coke}} = f_{\mathrm{coke}}\,\mathrm{CCR}_{\mathrm{feed}} \]

Total gas and total naphtha are expressed as

\[ \begin{align} Y_{\mathrm{gas}} &= a_{\mathrm{gas}} + b_{\mathrm{gas}}\,\mathrm{CCR}_{\mathrm{feed}} \\ Y_{\mathrm{nph}} &= a_{\mathrm{nph}} + b_{\mathrm{nph}}\,\mathrm{CCR}_{\mathrm{feed}} \end{align} \]

Total gas is split into dry gas and LPG through the fraction \(\phi_{\mathrm{dg}}\); total gas-oil (obtained by mass-balance closure) is split into LGO and HGO through \(\phi_{\mathrm{LGO}}\). Metals (Ni, V, Fe) are concentrated quantitatively in the coke. The fired heater duty is \(\dot Q = h_{\mathrm{heater}}\, \dot m_{\mathrm{feed}}\).

Catalytic Reformer¶

Overview¶

The Shortcut Reformer models a fixed-bed catalytic reformer operating on heavy naphtha feed. Product yields (H\(_2\), light ends, reformate) depend on the target RON severity; the RON-sensitivity coefficients allow the user to calibrate the model to a specific catalyst generation (semi-regenerative, cyclic, or CCR).

Stream Topology¶

Port	Direction	Description
Feed In	Inlet (material)	Hydrotreated heavy naphtha
H\(_2\) Rich	Outlet (material)	Net hydrogen gas
Light Ends	Outlet (material)	C\(_1\)–C\(_4\)
Reformate	Outlet (material)	High-octane aromatic product

Yield-Severity Correlation¶

Let \(R\) be the target reformate RON and \(R_{\mathrm{base}}\) the reference severity. Product yields are

\[ Y_i = Y_i^{\mathrm{base}} + s_i\,(R - R_{\mathrm{base}}),\qquad i \in \{\mathrm{H_2, LE, Reformate}\} \]

where \(s_i\) is the RON sensitivity (typically \(s_{\mathrm{H_2}}>0\), \(s_{\mathrm{LE}}>0\), \(s_{\mathrm{Reformate}}<0\)). The heater duty is \(\dot Q = h_{\mathrm{heater}}\, \dot m_{\mathrm{feed}}\).

Amine Treater¶

Overview¶

The Shortcut Amine Treater separates H\(_2\)S and CO\(_2\) from a sour gas using an aqueous alkanolamine solvent. Rather than resolving the vapour–liquid equilibrium of the reactive H\(_2\)S/CO\(_2\)/amine system, the unit uses user-specified removal fractions calibrated against rigorous column simulations or operating data.

Stream Topology¶

Port	Direction	Description
Sour Gas In	Inlet (material)	H\(_2\)S/CO\(_2\)-containing gas
Sweet Gas	Outlet (material)	Treated gas
Acid Gas	Outlet (material)	Stripper overhead (H\(_2\)S + CO\(_2\))

Component Removal¶

For each acid gas component \(i \in \{\mathrm{H_2S, CO_2}\}\),

\[ \begin{align} \dot n_{i}^{\mathrm{acid}} &= X_i^{\mathrm{rem}}\, \dot n_{i}^{\mathrm{feed}}\\ \dot n_{i}^{\mathrm{sweet}} &= (1 - X_i^{\mathrm{rem}})\, \dot n_{i}^{\mathrm{feed}} \end{align} \]

A small hydrocarbon slip \(\phi_{\mathrm{HC}}\) (mol fraction of sweet-gas hydrocarbons) is co-absorbed and reports to the acid-gas stream. The amine circulation rate is computed from

\[ \dot V_{\mathrm{amine}} = \lambda\, \dot n_{\mathrm{H_2S}}^{\mathrm{acid}} \]

where \(\lambda\) is the user-specified amine circulation (L of amine solution per mol of H\(_2\)S absorbed, typical 25–50 for MDEA).

Claus Sulfur Recovery Unit¶

Overview¶

The Shortcut Claus models a two- or three-stage modified Claus SRU as a lumped conversion reactor. The unit takes acid gas (H\(_2\)S + CO\(_2\), typically from the Amine regenerator) and converts H\(_2\)S to elemental sulfur at a specified recovery fraction that reflects the global equilibrium + kinetic limitations of the Claus reactors and condensers.

Stream Topology¶

Port	Direction	Description
Acid Gas In	Inlet (material)	H\(_2\)S-rich feed
Sulfur Out	Outlet (material)	Elemental sulfur (S\(_n\) surrogate)
Tail Gas Out	Outlet (material)	Unconverted H\(_2\)S + SO\(_2\) + inerts
Waste Heat	Outlet (energy)	Steam-generator duty

Reaction Stoichiometry¶

The overall Claus reaction is

\[ 3\,\mathrm{H_2S} + \tfrac{3}{2}\,\mathrm{O_2} \rightarrow \tfrac{3}{n}\,\mathrm{S}_n + 3\,\mathrm{H_2O} \]

The moles of H\(_2\)S reacted are

\[ \dot n_{\mathrm{H_2S}}^{\mathrm{rxn}} = X_{\mathrm{rec}}\, \dot n_{\mathrm{H_2S}}^{\mathrm{feed}} \]

where \(X_{\mathrm{rec}}\) is the sulfur-recovery fraction (typically 0.95–0.99). The sulfur produced is reported using a user-selected elemental sulfur surrogate (default S\(_1\)).

Waste Heat¶

The reaction releases a user-tuneable heat per mol of H\(_2\)S reacted \(\Delta H_{\mathrm{rxn}}\) (default 220 kJ/mol):

\[ \dot Q_{\mathrm{WHB}} = \dot n_{\mathrm{H_2S}}^{\mathrm{rxn}}\, \Delta H_{\mathrm{rxn}} \]

Product Blender¶

Overview¶

The Product Blender pools multiple material-stream inlets into a single outlet, aggregating mass, energy, composition, and contaminant load. It is most commonly used upstream of a storage tank or to form a combined off-site feed in TEA/LCA flow-sheet closures.

Stream Topology¶

The blender has a variable number of inlet ports (\(n_{\mathrm{inlets}} \ge 2\)) and a single outlet.

Mixing Equations¶

The outlet mass and enthalpy follow simple conservation:

\[ \begin{align} \dot m_{\mathrm{out}} &= \sum_{j=1}^{n_{\mathrm{inlets}}} \dot m_j \\ \dot H_{\mathrm{out}} &= \sum_{j=1}^{n_{\mathrm{inlets}}} \dot H_j \end{align} \]

The outlet pressure is either

\[ \begin{align} P_{\mathrm{out}} &= \min_{j}\, P_j \qquad\text{(mode \textit{Min}, conservative)} \\ P_{\mathrm{out}} &= \frac{\sum_j \dot m_j\, P_j}{\sum_j \dot m_j} \qquad\text{(mode \textit{Average}, mass-weighted)} \end{align} \]

The outlet temperature is the mass-flow-weighted average of the inlet temperatures. Contaminant concentrations of the pool are mass-weighted:

\[ x_{\mathrm{out}}^{\mathrm{cont}} = \frac{\displaystyle\sum_{j} \dot m_j\, x_j^{\mathrm{cont}}} {\displaystyle\sum_{j} \dot m_j} \]

Octane and RVP Blending¶

When a gasoline pool is blended, the blender can also estimate the octane numbers and the Reid vapor pressure (RVP) of the product from per-inlet values, following the correlations in Gary and Handwerk (chapter 12). Each inlet \(j\) carries a research octane number \(r_j\), a motor octane number \(m_j\), an RVP, and the olefins and aromatics contents \(O_j\), \(A_j\) (vol%). The values are read from the inlet stream extra properties (RON, MON, RVP in Pa, OlefinsVolPct, AromaticsVolPct) when present, and from the per-port configuration otherwise. An inlet without a value is left out of that blend, and the share of the pool it represents is reported as uncovered. When no MON is given, \(m_j = r_j\) is assumed. When no aromatics content is given and the inlet carries PNA data, the aromatics mass fraction of the feed is used.

The blend fractions \(f_j\) are the inlet volumetric flows at stream conditions (basis Volume, the default) or the inlet mass flows (basis Mass); the volume basis falls back to mass when a volumetric flow is unavailable. With bars denoting fraction-weighted averages, \(\bar r = \sum_j f_j r_j\), \(\bar m = \sum_j f_j m_j\), \(\overline{rs} = \sum_j f_j r_j s_j\) and so on, where \(s_j = r_j - m_j\) is the sensitivity, the octane methods are:

Linear: \(R = \bar r\), \(M = \bar m\).
Ethyl (Ethyl RT-70 interaction method):

\[ \begin{align} R &= \bar r + a_1 \left(\bar r\,\bar s - \overline{rs}\right) + a_2 \left(\bar O^2 - \overline{O^2}\right) + a_3 \left(\bar A^2 - \overline{A^2}\right) \ M &= \bar m + b_1 \left(\bar m\,\bar s - \overline{ms}\right) + b_2 \left(\bar O^2 - \overline{O^2}\right) + b_3 \left[\frac{\bar A^2 - \overline{A^2}}{100}\right]^2 \end{align} \]

with \(a_1 = 0.03324\), \(a_2 = 0.00085\), \(a_3 = 0\), \(b_1 = 0.04285\), \(b_2 = 0.00066\) and \(b_3 = -0.00632\).

User Coefficients: equation [eq:blender_ethyl] with the six coefficients entered by the user (they default to the published set).

The anti-knock index \((R + M)/2\), the sensitivity \(R - M\) and the linear averages are reported alongside the blended values.

The RVP methods are Linear, \(\mathrm{RVP} = \sum_j f_j\, \mathrm{RVP}_j\), and Chevron Index, which blends the Chevron blending index \(\mathrm{RVPBI} = \mathrm{RVP}^{1.25}\) linearly:

\[ \mathrm{RVP}_{\mathrm{blend}} = \left(\sum_j f_j\, \mathrm{RVP}_j^{1.25}\right)^{1/1.25} \]

The power law is scale invariant, so the pressure unit does not matter.

The blended RON, MON, olefins, aromatics and RVP are written to the outlet stream extra properties under the same keys, so a downstream blender picks them up automatically.

Isomerization Unit¶

Overview¶

The Shortcut Isomerization unit converts light straight-run naphtha (C\(_5\)–C\(_6\)) to a high-octane isomerate through \(n\)-paraffin to iso-paraffin rearrangement over a Pt/Al\(_2\)O\(_3\)-Cl or zeolitic catalyst. Yields are parameterised as a function of the octane uplift from feed RON to isomerate RON.

Stream Topology¶

Port	Direction	Description
Feed In	Inlet (material)	Light straight-run naphtha
Isomerate	Outlet (material)	High-octane iC\(_5\)/iC\(_6\) product
Light Ends	Outlet (material)	C\(_1\)–C\(_4\) by-products

Yield Model¶

Let \(R_{\mathrm{iso}}\) and \(R_{\mathrm{feed}}\) be the isomerate and feed RON. Yields are

\[ Y_i = Y_i^{\mathrm{base}} + s_i\,(R_{\mathrm{iso}} - R_{\mathrm{feed}}),\qquad i \in \{\mathrm{Isomerate, LightEnds}\} \]

The heater duty is \(\dot Q = h_{\mathrm{heater}}\, \dot m_{\mathrm{feed}}\). Sulfur and nitrogen in the feed partition to the products according to user-specified fractions (\(\mathrm{SFrac}_{\mathrm{iso}}\), \(\mathrm{SFrac}_{\mathrm{LE}}\), \(\mathrm{NFrac}_{\mathrm{iso}}\), \(\mathrm{NFrac}_{\mathrm{LE}}\)).

Alkylation Unit¶

Overview¶

The Shortcut Alkylation unit models the acid-catalysed alkylation of isobutane with light olefins (propylene and butenes) to produce a high-octane, low-sulfur alkylate. The model is a lumped olefin-conversion yield calculation at a specified reactor temperature, pressure, iC\(_4\)/olefin mol ratio, and olefin conversion.

Stream Topology¶

Port	Direction	Description
Olefin Feed	Inlet (material)	Propylene / butylene cut
iC\(_4\) Feed	Inlet (material)	Isobutane (fresh + recycle)
Alkylate	Outlet (material)	High-octane alkylate
Purge	Outlet (material)	nC\(_4\) + heavy ends purge
Cooler Duty	Outlet (energy)	Reactor heat-removal duty

Yield and Duty¶

Let \(\dot m_{\mathrm{ol}}\) be the olefin mass flow and \(X_{\mathrm{ol}}\) the olefin conversion. The alkylate mass is

\[ \dot m_{\mathrm{alk}} = y_{\mathrm{alk}}\, X_{\mathrm{ol}}\, \dot m_{\mathrm{ol}} \]

where \(y_{\mathrm{alk}}\) (typical 1.7–2.2) is the alkylate yield per kg of olefin reacted. Unreacted olefins plus any nC\(_4\) reporting to the purge form the purge stream. The reactor exotherm is

\[ \dot Q_{\mathrm{cool}} = -\, X_{\mathrm{ol}}\, \dot m_{\mathrm{ol}}\, \Delta H_{\mathrm{alk}} \]

with \(\Delta H_{\mathrm{alk}}\) the heat release per kg of olefin reacted (typical 1350–1500 kJ kg\(^{-1}\)).

Shared Contaminant Vector¶

All refining blocks above carry and update a common contaminant vector that travels with every material stream:

Property	Symbol	Unit
Total sulfur	\(x_{\mathrm{S}}\)	wt%
Total nitrogen	\(x_{\mathrm{N}}\)	wt%
Mercaptan sulfur	\(x_{\mathrm{RSH}}\)	wt%
Nickel	\(x_{\mathrm{Ni}}\)	mass ppm
Vanadium	\(x_{\mathrm{V}}\)	mass ppm
Iron	\(x_{\mathrm{Fe}}\)	mass ppm
Conradson carbon	\(x_{\mathrm{CCR}}\)	wt%
Asphaltenes (C\(_7\) insol.)	\(x_{\mathrm{asph}}\)	wt%
Total acid number	\(\mathrm{TAN}\)	mg KOH/g

In any unit operation, the outlet-stream contaminant vector is computed by mass-weighted partitioning using user-editable or default split fractions; removed contaminants (e.g. sulfur reporting as H\(_2\)S) are converted to the corresponding molecular surrogate and reported in the by-product stream.

PNA-Aware Yield Modulation¶

In addition to the contaminant vector, each pseudocomponent may carry a paraffin / naphthene / aromatic (PNA) composition triplet \((x_{\mathrm{P}}, x_{\mathrm{N}}, x_{\mathrm{A}})\) estimated at characterisation time from the TBP curve and bulk SG (see the Petroleum Characterization section) or overridden by the user. The feed-average PNA composition is aggregated as a mass-weighted mean over those feed pseudocomponents that have a PNA triplet set:

\[ \bar{x}_{i} \;=\; \frac{\displaystyle\sum_{c \in \mathcal{C}_{\mathrm{PF}}} \dot m_{c}\, x_{i,c}} {\displaystyle\sum_{c \in \mathcal{C}_{\mathrm{PF}}} \dot m_{c}}, \qquad i \in \{\mathrm{P}, \mathrm{N}, \mathrm{A}\} \]

where \(\mathcal{C}_{\mathrm{PF}}\) is the set of petroleum-fraction pseudocomponents for which at least one of \(x_{\mathrm{P}},\,x_{\mathrm{N}},\, x_{\mathrm{A}}\) is defined. When \(\mathcal{C}_{\mathrm{PF}} = \varnothing\) the PNA machinery is inactive and default (PNA-independent) yields are used. Triplets are renormalised to sum to unity before use. The feed PNA triplet is reported by every PNA-aware unit operation (output properties Feed xP, Feed xN, Feed xA).

Each PNA-aware unit applies a multiplicative modulation of the form

\[ y'_{j} \;=\; y_{j}\,\bigl(1 + \alpha_{j}\,(\bar{x}_{i} - \bar{x}^{\mathrm{ref}}_{i})\bigr), \qquad \bar{x}^{\mathrm{ref}}_{i} \;\text{user-configurable} \]

to a base yield or consumption \(y_{j}\) (with the modulated yield set renormalised when \(j\) indexes a complete product slate). The sensitivity coefficient \(\alpha_{j}\) and the reference PNA fraction \(\bar{x}^{\mathrm{ref}}_{i}\) are unit-specific configuration parameters. The PNA-dependent behaviour of each block is summarised below:

Unit	Effect of feed PNA
Reformer	H\(_2\) and light-ends yields scale with \((\bar{x}_{\mathrm{P}} - \bar{x}^{\mathrm{ref}}_{\mathrm{P}})\), so paraffin-rich naphthas dehydrocyclise to aromatics, releasing H\(_2\) and gas.
Isomerization	Aromatics are inert under C\(_5\)/C\(_6\) isomerisation; an aromatic fraction above \(\bar{x}^{\mathrm{ref}}_{\mathrm{A}}\) linearly penalises isomerate yield and increases the light-ends slip.
FCC	Target riser conversion is scaled by \((1 + \alpha_{X}(\bar{x}_{\mathrm{P}} - \bar{x}^{\mathrm{ref}}_{\mathrm{P}}))\) (paraffin-rich gas oils crack more readily); the coke yield is bumped by \((1 + \alpha_{\mathrm{coke}}(\bar{x}_{\mathrm{A}} - \bar{x}^{\mathrm{ref}}_{\mathrm{A}}))\) on top of the CCR-driven baseline.
HCR	Per-pass conversion is paraffin-accelerated through \((\bar{x}_{\mathrm{P}} - \bar{x}^{\mathrm{ref}}_{\mathrm{P}})\); chemical H\(_2\) consumption per kg feed is aromatic-accelerated through \((\bar{x}_{\mathrm{A}} - \bar{x}^{\mathrm{ref}}_{\mathrm{A}})\) to account for ring saturation.
HDS	An aromatic-saturation term adds \(\dot m_{\mathrm{feed}}\,\bar{x}_{\mathrm{A}}\,f_{\mathrm{sat}}\,(\mathrm{H}_2/\mathrm{kg~Ar})\) to the hydrogen demand computed from the sulfur/nitrogen HDS/HDN kinetics.
Coker	Coke yield is boosted by \((1 + \alpha_{\mathrm{coke}}(\bar{x}_{\mathrm{A}} - \bar{x}^{\mathrm{ref}}_{\mathrm{A}}))\) beyond the CCR anchor; gas and naphtha yields are boosted by \((1 + \alpha_{\mathrm{gas}}(\bar{x}_{\mathrm{P}} - \bar{x}^{\mathrm{ref}}_{\mathrm{P}}))\). Yields are renormalised so that \(\sum_j y_j + y_{\mathrm{coke}} = 1\).

PNA awareness can be disabled per unit through the UseFeedPNA flag; it is likewise inactive whenever the feed has no PNA data attached, making the correlation strictly additive with respect to the baseline yield slates.

Common Assumptions and Limitations¶

Steady state. All blocks are solved as steady-state conservation models; no dynamic inventory or start-up effects are captured.
Lumped conversion volume. Reactors are treated as a single lump; intra-reactor temperature, pressure, or composition profiles are not resolved. LHSV and residence time enter only through their integrated effect on the global conversion.
Pseudocomponent basis. Feed streams are characterised through the standard DWSIM pseudocomponent machinery; model accuracy therefore depends on the quality of the feed assay (TBP curve, bulk sulfur, nitrogen, CCR, metals).
No detailed VLE for reactive systems. Reactive systems (Amine, Claus) are modelled through user-calibrated removal / recovery fractions rather than rigorous reactive VLE. This is adequate for conceptual design but should be verified against rigorous column/reactor simulations before detailed engineering.
Contaminant split fractions are user data. The default partitioning coefficients are indicative only and must be calibrated against operating data for any given refinery.
Energy balance closure. Reactor exotherms / endotherms are returned as side-streams (energy ports or duty results). The user is responsible for placing the corresponding utility streams (fired heater, cooler, steam generator) on the flowsheet to close the overall heat balance.

Numerical Solution Procedure¶

All refining unit operations follow the same explicit calculation pattern:

Resolve the connected inlet material and energy streams; read temperature, pressure, composition, and contaminant vector.
Initialise the pseudocomponent list from the feed SelectedCompounds collection.
Evaluate the conversion / yield equations of the unit as a function of the configured operating conditions ([eq:hds_conversion], [eq:fcc_lumps], [eq:hcr_yield], [eq:reformer_yield], [eq:claus_hsrxn], etc.).
Partition the feed mass among the outlet streams; close the mass balance by adjusting the residue / UCO / purge stream.
Update the contaminant vector of each outlet by applying the user-specified split fractions and accounting for contaminants converted in-reactor (S \(\rightarrow\) H\(_2\)S, N \(\rightarrow\) NH\(_3\), S \(\rightarrow\) S\(_n\)).
Set outlet-stream temperature and pressure; when not independently specified, the outlet temperature is computed from a \((P, H)\) flash that reflects the reactor exotherm.
Write the energy-balance duties to any connected energy-stream outlet (heater, cooler, waste-heat boiler).
Collect per-unit key performance indicators (conversion, severity, recovery, CCR, RON) into the results object for display in the editor and for downstream TEA / LCA consumption.

No iteration is required for any of the refining shortcut models; all equations admit an explicit closed-form or one-pass evaluation given the inlet stream conditions and configuration parameters.

Typical Usage Workflow¶

Build the crude assay by adding pseudocomponents from a TBP curve (see Characterise Crude Assay in the main Tools menu) and set the bulk contaminant properties of the feed material stream.
Drop the Shortcut CDU on the flowsheet and connect the crude feed; verify that the product cuts close the feed mass balance and contaminant distribution.
Build the downstream refinery by chaining the remaining refining unit operations (HDS, FCC, HCR, Coker, Reformer, Isomerization, Alkylation) to the CDU cuts as appropriate.
Connect an Amine Treater to any sour-gas stream that requires H\(_2\)S removal, and route the acid gas to the Claus SRU.
Use Product Blenders to pool streams into gasoline, diesel, and fuel-oil pools prior to storage.
Review the Results tab of each unit for conversion, yield, severity, and contaminant-vector summaries, and feed the resulting pool qualities into the DWSIM TEA / LCA extensions for economic and environmental assessment.