Gas composition & Chalanges

Average Associated/Flare Gas Composition

The gas that ends up in a flare stack is usually associated gas (from oil wells) or waste gas from processing units. It’s highly variable by basin, but industry data shows typical ranges:

• Methane (CH₄): 40–70%
• Ethane (C₂H₆): 5–15%
• Propane (C₃H₈): 2–8%
• Butanes & heavier (C₄+): 1–8%
• Carbon dioxide (CO₂): 1–10%
• Nitrogen (N₂): 0.5–5%
• Hydrogen sulfide (H₂S): trace to 5% (sour gas fields)
• Other inerts (argon, water vapor, etc.): <1%

👉 In practice: A “lean” gas is ~80–90% methane. A “rich” gas can be closer to 50–60% methane with heavier hydrocarbons mixed in.

What the Gas Leaving a Flare Looks Like

When gas is flared:

• Ideally, it burns completely → mainly CO₂ + H₂O.
• In reality, combustion is often incomplete → you’ll see:
  • CO₂ (70–90% of exhaust)
  • H₂O vapor (5–15%)
  • Unburnt CH₄ (0.5–5%)
  • CO (1–5%)
  • NOₓ, SOₓ, soot (traces)

So the flare plume is mostly hot CO₂ and steam with small but critical amounts of unburnt methane (high GWP greenhouse gas).

How Much of Flared Gas Is Methane (Unburned)?

Studies show flare combustion efficiency ranges 92–98%. That means:

• If your incoming gas is ~60% methane, then only ~1–5% of that methane slips unburnt through the flame.
• Example: Gas stream = 60% CH₄, 100 MSCFD. After flare: 92% burned → ~55 MSCFD methane destroyed, ~5 MSCFD slips through.

How Much Carbon Can Be Captured (via Pyrolysis)

Methane pyrolysis reaction:

CH4→C(solid)+2H2CH₄ → C (solid) + 2 H₂CH4​→C(solid)+2H2​

• 1 mole CH₄ (16 g) → 1 mole carbon (12 g) + 2 moles H₂ (4 g).
• So 75% of methane’s mass can become solid carbon, and 25% becomes hydrogen.

👉 In percentage terms:

• From the methane fraction only: up to 75% of its weight becomes usable solid carbon.
• If flare gas is ~60% methane, then roughly 45% of the entire gas stream’s mass could be carbonized (the rest being hydrogen, CO₂, inerts, heavier hydrocarbons).

Is composition predictable?

• Within one asset: Somewhat. A single separator or plant train tends to swing inside a band (e.g., “lean” vs “rich” days), but turnarounds, well changes, and upset routing cause step-changes.
• Across sites & flare events: Not very. Different fields, liquids loading, and emergency flares = bigger swings (heavies, CO₂, H₂S, N₂, water).

Why it matters for pyrolysis (and any H₂/C recovery)

• Heating value & Wobbe swings: change reactor duty, residence time, and carbon morphology.
• Heavies (C₂+): raise coke/soot rates; can plug or foul.
• CO₂/N₂/H₂O (diluents): lower partial pressure of CH₄ → lower throughput per kW.
• H₂S/mercaptans: poisoning/corrosion; sulfur-contaminated carbon product.
• Water & aerosols: quench hot zones, cause thermal shock.
• Oxygen ingress (rare but during upsets): safety hazard in hot reactors.

Practical approach (pre-process)

1. Inline gas analysis: micro-GC + TDL/IR for CO₂/H₂S + dew-point meter. Aim for 30–120s refresh.
2. Feed conditioning:
   • Knockout + coalescer + heat tracing (remove liquids/water).
   • Desulfurization (ZnO bed or amine slipstream) if H₂S > product spec.
   • Optional CO₂ trim only if it crushes capacity; not always required.
3. Buffering & blending: small surge vessel to smooth short spikes; blend rich + lean sources when possible.

Reactor control (real-time)

• Closed-loop duty control: adjust wall temperature/power to hit target CH₄ conversion at measured CH₄%.
• Residence-time control: modulate flow to keep Damköhler number in spec.
• Anti-foul strategy: periodic thermal pulses or nitrogen sweeps; staged reactors for isolation/cleaning.
• Set “operating envelopes”: define safe/efficient windows for CH₄%, H₂S ppm, dew point; auto-derate or bypass outside limits.

Post-process “tackle plan”

• Hydrogen train: membranes → PSA; adjust cut based on measured diluents to keep purity ≥ spec.
• Carbon handling: cyclone + filter; de-ashing/desulfurization if needed; granulate/briquette to stable SKUs; QC for sulfur, surface area, and color.
• Sulfur management: regenerate ZnO/amine; consider sulfur recovery if volumes justify.
• Analytics & QA: batch-tag carbon lots with feed gas assays; sell high-spec lots premium, down-spec to industrial uses.
• Fallback modes: when composition goes out of envelope, divert to flare or thermal oxidizer and resume when stable—protects equipment and product brand.

Data/controls layer

• Soft sensors / “virtual analyzer”: ML model maps analyzer + process signals → predicted CH₄% and foul rate; pre-emptive setpoint moves.
• Run rules: site-specific recipes (e.g., “Basin A rich gas: +20°C wall temp, –10% flow”).
• SLA language with operators: define minimum CH₄%, max H₂S, dew point, and notification timelines; price in a variability fee.

Bottom line

Composition will vary more than you think during real-world flaring. Design for it: measure fast, precondition smart, control tightly, and grade product after.

Future option

There is active research on plasma and photochemical H₂S splitting (aimed at producing green H₂ from sour gas fields). If you’re building modular flare-gas units, you could someday bolt on an H₂S-to-H₂ + S₂ module. But today it’s still lab/early pilot scale—not bankable like amine + Claus.

Design basis (assumptions)

• Flow: 0.25 MMSCFD (~10,400 scfh ≈ 295 Nm³/h)
• Pressure available to contactor: 80–150 psig
• Temp: 20–40 °C (heated/traced to avoid condensation)
• Composition envelope: CH₄-rich gas with H₂S 0.5–2.0 mol%, CO₂ variable; target <4 ppmv H₂S to protect pyrolysis & carbon quality
• Solvent: 40 wt% MDEA (selective for H₂S; let most CO₂ slip)

What that sizes to (practical, not lab-perfect)

Amine circulation (lean → rich Δloading ≈ 0.33 mol/mol):

• Theoretical need at 0.5–2.0% H₂S: ~0.3–1.0 gpm
• Reality: hydraulics, control stability, and swings demand a minimum.
• Specify: 3–5 gpm design circulation (gives you ~3–10× headroom for spikes and easy turndown).

Contactor (absorber):

• 8–10 in. diameter, 10–15 ft packed height (structured packing), ANSI 150/300
• Mist eliminator + coalescer upstream

Regenerator (stripper) + reboiler:

• 6–8 in. diameter, 10–14 ft packed trays/packing
• Reboiler duty: ~75–150 kW (0.25–0.5 MMBtu/h) at 3–5 gpm circulation
• Lean/rich exchanger (plate & frame), air cooler or small water cooler

Tanks & ancillaries:

• Lean/rich surge: 0.5–1.0 m³ each
• Solvent inventory: ~1–2 m³ total
• Knockout + coalescer + electric heat tracing on inlet
• Optional: slipstream ZnO guard if sour spikes exceed spec

Controls:

• PID on circulation, reboiler duty, and regenerator overhead temp
• H₂S analyzer (inlet/outlet), dew point, and differential pressure (packing health)

Turndown & swings:

• Stable 3:1–4:1 turndown at spec with this sizing
• Add a 2–5 min gas surge vessel (ASME) before the contactor to smooth flare slugs

Skid & footprint options (how small can we go?)

Single-skid “micro” package (tight sites):

• Target footprint: 10–12 ft L × 4–6 ft W × 9–10 ft H
• What’s on it: contactor, pumps, exchangers, small reboiler (electric or NG-fired), controls, cable tray
• Trade-offs: tighter maintenance access; regenerator column height may require a fold-down or removable stack

Standard 20-ft ISO modular (most practical):

• Two short skids that bolt together or a single 20-ft frame
• Full access, better noise/heat management, easy trucking & craning
• Easiest path to adding a ZnO guard or bigger reboiler later

Ultra-compact 10-ft skid?

• Possible if you outboard the reboiler (separate cube) and accept tighter service clearances
• Works for steady, cleaner gas; less forgiving for big upset swings

Indicative CAPEX & utilities (budgetary)

• New, small modular (3–5 gpm): US$650k–$1.1M all-in (mechanical + controls; install & utilities excluded)
• Refurb/used packages: US$300k–$700k depending on condition & metallurgy
• Power/heat: 75–150 kW reboiler duty (electric or NG), 3–7 kW pumps & controls
• Chemicals: initial 1–2 m³ MDEA solution; small annual makeup

Why this works for your pyrolysis train

• Keeps H₂S → <4 ppm (protects carbon value & downstream metallurgy)
• Lets CO₂ slip, preserving energy efficiency and simplicity
• Sized for control stability (not just stoichiometry), so you can ride flare-event variability without tripping the unit
• Modular path: start with the micro-train, bolt on a larger reboiler or ZnO guard if your basin turns out spikier than expected

‍