Corrosion is a hidden cost measured in trillions, but refineries that blend smart chemistry with relentless monitoring are cutting failures and maintenance. The crude distillation overhead — ground zero for HCl and NH4Cl — shows why an integrated program matters.
Corrosion is a hidden cost in refineries — globally estimated at ~$2.5 trillion/year (≈3.4% of GDP) (www.jm.com). Even modest improvements yield outsized savings: one analysis estimates that applying best-practice corrosion prevention in oil & gas could avoid 14–33% of damage costs (www.mdpi.com). World‑class refineries investing in proactive reliability — corrosion control included — spend ~20–25% less on maintenance than peers, according to Solomon Associates data (www.digitalrefining.com).
That gap is widening as many sites run aging units, often well past design life, on more aggressive feeds (e.g., high‑sulfur or high‑TAN crudes) (www.chemicalprocessing.com) (www.aiche.org). Heavier, altered charge stocks introduce more chlorides, H₂S, ammonia, and organic acids, heightening risk; industry surveys show ~76% of equipment failures occur when operating outside design/integrity envelopes (www.digitalrefining.com).
The crude distillation overhead is especially unforgiving: salt hydrolysis generates HCl and NH₄Cl that condense and attack carbon steel as soon as water falls out (www.mdpi.com) (www.ogj.com). In one field report, even a small fraction of overhead condensate — 3% of the flow — carried 1,800 ppm chloride and 2,000 ppm iron under acidic conditions (www.ogj.com). By contrast, well‑controlled overhead condensate aims for ~10–20 ppm chloride (and pH ≈6–7 with neutralizers added) (www.ogj.com). Roughly one‑third of crude‑column failures have been traced to HCl attack (www.ogj.com).
Overhead dewpoint chemistry and attack
In the overhead, acid dewpoint corrosion is driven by chloride salt hydrolysis and organic acids. Mixtures of MgCl₂ and CaCl₂ in the feed hydrolyze to HCl above ~200–300 °C, and even parts‑per‑billion ammonia in vapor can form NH₄Cl on cooling; because HCl is highly soluble, the first condensate film can have pH near 0, causing rapid, localized metal loss (www.mdpi.com).
Process‑side neutralization and inhibition
Control starts with desalting. Effective units should remove ~90% of NaCl and >60% of MgCl₂ from crude (www.ogj.com). Typical single‑stage desalters remove ~90% NaCl but only 40–50% CaCl₂ or MgCl₂, with the rest passing overhead; because MgCl₂ hydrolyzes readily, it causes a disproportionate downstream problem, so experience suggests >60% MgCl₂ removal is needed (www.ogj.com) (www.ogj.com).
Downstream of the desalter, caustic wash (NaOH) neutralizes HCl precursors. A field rule of thumb is ~1 lb NaOH per 1 lb of entrained salt, with dosing adjusted to hold overhead chlorides to ~10–20 ppm (www.ogj.com) (www.ogj.com). One refinery targeted 10–20 ppm Cl in a second‑stage drum by quilling caustic into the overhead piping (www.ogj.com), typically via accurate metering through dosing pumps.
Neutralizing amines (volatile bases) pick up the slack. Acyclic amines like morpholine, diethanolamine, or higher‑boiling alkanolamines are injected in overhead vapor or reflux lines; water‑soluble, they partition into the condensate and raise pH toward neutral (≈6–7), curbing acid attack (www.oilplusfz.com) (www.ogj.com). Industry practice is to maintain overhead drum pH around 5.5–6.5; one troubleshooting guide recommends “sufficient addition of neutralizing amine to control [the overhead] water boot at a pH of about 6” (www.ogj.com). These amines are selected for high water‑solubility and low salt‑forming tendency (www.oilplusfz.com), and are often implemented as packaged neutralizing amine programs.
Filming amines add a hydrophobic barrier on metal, mitigating naphthenic acid corrosion in hot sections (crude/vacuum towers) and protecting wet‑wet areas (e.g., reactor discharges). By forming a thin film, they slow iron dissolution and hydrogen embrittlement; industry notes they help prevent issues like Hydrogen‑Induced Cracking and “orange‑peel” naphthenic corrosion seen with some heavy crudes (www.oilplusfz.com) (www.digitalrefining.com). In practice, continuous injection at column overheads — in oil and aqueous phases as appropriate — is common (www.oilplusfz.com), typically using refinery‑grade corrosion inhibitors.
Field performance and target envelopes
Integrated programs couple process control with chemistry: good desalting (>90% NaCl, ≳60% MgCl₂ removal) and caustic dosing to a 10–20 ppm chloride target (www.ogj.com) (www.ogj.com), plus volatile amines in overhead vapors to keep condensing‑water pH ~6 (www.ogj.com). Field studies confirm the approach: a crude‑unit monitoring project reported that continuous dosing of demulsifier, amine neutralizer, and filming inhibitor at key locations kept condensate acidity, chlorides, and dissolved iron within safe limits (www.researchgate.net) (www.researchgate.net). Periodic sampling showed overhead condensation water remained near‑neutral with low chloride and iron, and corrosion coupons measured rates “within the specified limit” after treatment (www.researchgate.net).
Corrosion probes and coupon surveillance
Inhibitors demand verification. Traditional coupons and probes provide ongoing rate logs: weight‑loss coupons give average metal loss; LPR (linear polarization resistance) infers corrosion current; ER (electrical resistance) measures wall loss directly. ER probes can detect thinning on the order of tens of nanometers, respond rapidly to changes, and operate at high refinery temperatures (www.digitalrefining.com) (www.digitalrefining.com). Limits remain: probes sample a localized point and can corrode away; many sacrificial‑tip probes need replacement every ~2–3 years, while equipment turnarounds may be 5–7 years apart (www.chemicalprocessing.com). Even so, they are “important for active corrosion management as well as the verification and optimization of corrosion inhibitor use” (www.digitalrefining.com).
Non‑intrusive ultrasonic sensing
Permanently mounted, non‑intrusive ultrasonic thickness sensors — often wireless — are increasingly standard. Clamp‑on transducers can detect thickness changes as small as ~10 µm, and deliver real‑time trends versus manual UT that samples every months or years (www.chemicalprocessing.com) (www.chemicalprocessing.com). Many wireless packs run ~9 years on one battery, sending data via fieldbus or Wi‑Fi (www.chemicalprocessing.com).
At Preem refinery, 17 wireless ultrasonic probes were installed on a cracked overhead line. Data streamed to analysis software and a historian; analytics revealed a surge (~47 mils/yr) of corrosion tied to a high‑salt crude blend (www.automation.com). Engineers discovered a broken amine quill, fixed it, and watched the corrosion rate drop to near zero (www.automation.com). Anticipated failures — and a likely shutdown — were averted, and plant personnel “were able to implement an effective preventative maintenance strategy, saving time and money” (www.automation.com). It is a quantitative example of how real‑time probes can force‑multiply inspection programs.
RBI programs, IOWs, and code compliance
On top of online sensing, periodic NDT (non‑destructive testing) — visual, ultrasonic thickness, radiography, MPI/ACI, and eddy current — is executed per API/ASME codes (e.g., API 570/653 for piping and tanks), with intervals set by RBI (risk‑based inspection) per API 580/581. Vessels handling sour water or overhead condensate often draw higher scrutiny; measured corrosion rates feed remaining‑life models. Integrity operating windows (IOWs) now codify key variables like overhead acid dewpoint and pH; API 584 guidance treats falling pH in crude‑overhead drums as a “Critical Limit IOW” — an alarm‑and‑response scenario because “acidic pH in the overhead of a crude distillation tower can cause rapid corrosion” (www.aiche.org).
Economic outcomes and monitoring adoption
The payoff is measurable. Implementing corrosion monitoring and control can potentially save 14–33% of costs in refining (www.mdpi.com). In one analysis, increasing heavy‑crude use — at meaningful price savings — became feasible only because corrosion was carefully managed, boosting profitability by ~$7–10 million/year on a 300 kbd plant with no new failures (www.digitalrefining.com). U.S. adoption is visible: by 2020, about 43% of West Coast refineries had installed real‑time corrosion sensors (in the context of OSHA PSM and EPA/LDAR compliance driving investment in continuous sensing) (pmarketresearch.com). Consistent with that, world‑class operators with strong corrosion programs spend ~20–25% less on maintenance than peers (www.digitalrefining.com).
Integrated program: chemistry plus integrity
In sum, chemical treatments neutralize acids in critical streams like the overhead system — evidenced by target pH (~6) and chloride limits (10–20 ppm) (www.ogj.com) (www.ogj.com). Field data show such regimes keep corrosion at “under control” levels (www.researchgate.net) (www.researchgate.net). Continuous probes and smart analytics ensure any excursion is caught early — in the Preem case, a >1 mm/yr (>~39 mils/yr) rate was driven effectively to near zero after fixing a broken quill (www.automation.com) (www.automation.com). Coupled with scheduled NDT and RBI, this approach “verifies assets and integrity, optimizes corrosion mitigation and control, provide[s] vital input to inspection planning and asset maintenance” — enabling refineries to process cheaper feedstocks without undue risk (www.digitalrefining.com).
Local (Indonesian) experts likewise emphasize material selection and environmental control (e.g., desalting, pH) as core to corrosion prevention, aligning with these practices (journal.lemigas.esdm.go.id).
