Refining, Midstream & Energy Transition

From crude to products and from well to market. Learn crude assays and the atmospheric and vacuum distillation units, conversion (FCC, hydrocracking, coking), treating (hydrotreating/desulphurisation, reforming), refinery configuration and complexity, product specifications (BS-VI), gas processing, LNG, pipeline hydraulics and storage, petrochemicals, emissions and the energy transition (biofuels, green hydrogen, CCS, crude-to-chemicals) — with Python for refinery yields and crude pipeline pressure drop.

Crude oil is useless until it is separated and converted into fuels and chemicals. A refinery first distils crude into boiling-range cuts, then converts heavy fractions into lighter valuable products and treats them to meet specifications.

Midstream moves oil and gas via pipelines, tankers and LNG trains. India, with refineries like Jamnagar (the world's largest refining complex), is a major refining hub.

An analogy

A refinery is like a sophisticated kitchen that takes one raw ingredient (crude) and splits it by boiling point, then cracks the heavy leftovers into lighter, more useful dishes (petrol, diesel, jet fuel, plastics feedstock).

Distillation cuts

Crude distillation unit (CDU) and vacuum unit (VDU) cuts (approximate boiling ranges):

• Gases and LPG (< 30 °C)
• Naphtha (30–180 °C) → petrol after reforming/isomerisation, petrochemical feed
• Kerosene/jet (150–270 °C)
• Diesel/gas oil (200–360 °C)
• Vacuum gas oil (360–550 °C) → FCC/hydrocracker; vacuum residue → coker, bitumen, fuel oil

Conversion and treating

Fluid catalytic cracking (FCC) cracks VGO into petrol and propylene; hydrocracking uses hydrogen for high-quality diesel/jet; delayed coking converts residue to lighter liquids and petroleum coke. Catalytic reforming raises naphtha octane (and makes hydrogen); hydrotreating removes sulphur and nitrogen to meet BS-VI/Euro-VI (10 ppm S). Complexity (Nelson index) measures conversion capability; complex refineries earn higher margins.

Midstream: gas, LNG and pipelines

Gas processing removes water (glycol dehydration), acid gases (amine treating for CO₂/H₂S), and recovers NGLs. LNG liquefies methane at −162 °C (600:1 volume reduction) for shipping; regasification terminals return it to gas. Pipelines are designed for hydraulics (pressure drop, pump/compressor stations), wall thickness, corrosion protection (coatings, cathodic protection), leak detection and pigging.

Petrochemicals and the transition

Steam crackers turn naphtha/ethane into ethylene and propylene for plastics; aromatics complexes make benzene, toluene and xylenes. The industry is shifting toward crude-to-chemicals, biofuel co-processing (HVO, SAF), green hydrogen to replace grey hydrogen, electrification of heaters, CCS on hydrogen and FCC units, and methane leak detection. Petroleum engineers also apply their skills to CO₂ transport and storage.

Formulas

• Mass balance: Σ product yields = 100% of crude (+ H₂ added, − fuel/loss)
• Pipeline pressure drop: ΔP = f·(L/D)·ρV²/2
• Swamee–Jain: f = 0.25/[log₁₀(ε/3.7D + 5.74/Re^0.9)]²
• Pump power: P = ΔP·Q/η
• Gross refining margin: GRM = Σ(product value) − crude cost ($/bbl)

Worked example: Diesel yield

Given: 200 000 bpd, straight-run diesel 27%

Find: Diesel bpd.

• 0.27 × 200 000

Answer: 54 000 bpd

Worked example: Pipeline velocity

Given: 200 000 bpd, D 0.6 m

Find: V.

• Q = 0.368 m³/s
• V = 0.368/0.2827

Answer: 1.30 m/s

Worked example: GRM

Given: Products worth $86/bbl crude, crude $78/bbl

Find: GRM.

• 86 − 78

Answer: $8/bbl

Refinery yields and crude pipeline hydraulics

# Refinery yields and pipeline hydraulics
crude_bpd = 200000
cuts = {"LPG": 0.03, "naphtha": 0.18, "kerosene/jet": 0.12, "diesel": 0.27, "vacuum gas oil": 0.25, "residue": 0.15}
for c, f in cuts.items():
    print(f"{c:15s} {f*crude_bpd:8.0f} bpd")
# Crude pipeline pressure drop (Darcy-Weisbach, Swamee-Jain friction)
import math
Q = crude_bpd * 0.158987 / 86400        # m3/s
D, L, rho, mu, eps = 0.6, 100e3, 860, 0.01, 4.5e-5
V = Q / (math.pi * D**2 / 4)
Re = rho * V * D / mu
f = 0.25 / math.log10(eps / (3.7 * D) + 5.74 / Re**0.9) ** 2
dP = f * L / D * rho * V**2 / 2
print(f"velocity {V:.2f} m/s, Re {Re:.0f}, f {f:.4f}, dP over 100 km {dP/1e5:.1f} bar")
print(f"pump power ~ {dP*Q/0.75/1000:.0f} kW per 100 km")
# LPG                 6000 bpd
# naphtha            36000 bpd
# kerosene/jet       24000 bpd
# diesel             54000 bpd
# vacuum gas oil     50000 bpd
# residue            30000 bpd
# velocity 1.30 m/s, Re 67164, f 0.0198, dP over 100 km 24.0 bar
# pump power ~ 1180 kW per 100 km

Straight-run distillation of this crude leaves 40% as heavy VGO and residue — conversion units (FCC, hydrocracker, coker) turn these into light products. A 24-inch pipeline moving 200 000 bpd loses 24 bar per 100 km, so pump stations are spaced every 100–200 km.

Common mistakes

• Wrong: Ignoring crude compatibility when blending. — Right: Asphaltene precipitation fouls equipment.
• Wrong: Undersizing pipelines to save capex. — Right: Balance capex against lifetime pumping cost.
• Wrong: Treating refinery CO₂ as unavoidable. — Right: Efficiency, electrification, green H₂, CCS reduce it.

Applications

• Refineries: Jamnagar, Paradip, Kochi — high-complexity conversion.
• Cross-country pipelines: Crude and product pipelines (e.g. Mundra–Panipat).
• LNG import terminals: Dahej, Kochi, Ennore regasification.
• Petrochemicals: Crackers and polymer plants.

Key takeaways

• Distil, convert, treat, blend.
• Complexity raises margins.
• ΔP = f(L/D)ρV²/2 sizes pipelines and pumps.
• Biofuels, green H₂ and CCS reshape refining.

Going further

• Linear-programming refinery models: Crude selection and product optimisation.
• Transient pipeline simulation: Surge and leak detection.
• Crude-to-chemicals: Maximising petrochemical yield.