Inflection Point Engineering Knowledge Base

Crude Desalter Troubleshooting Guide

A field-diagnosis reference for electrostatic crude desalters. Covers symptom tables, root-cause identification, and operator interventions for two-stage AC/DC desalter systems.

1. Why Desalters Exist

Salts (mostly NaCl and MgCl2) in crude oil hydrolyze to HCl in the atmospheric tower overhead, driving naphtha overhead corrosion. Desalters remove salts as aqueous brine via water wash + emulsion breaker + electrostatic coalescence. Target: 1 PTB (pound of salt per thousand barrels) in desalted crude for conventional units, 0.5 PTB for premium runs.

2. Top 10 Symptoms and First-Order Root Causes

SymptomMost Likely CauseFirst CheckFirst Action
High salt in desalted crude (> 3 PTB)Emulsion breaker underdose or wrong typeEB dosage, brand, injection pointIncrease dose by 50%; test alternate chemistry
Grid current high / tripWater in oil phase (interface too high)Interface radar / RF probe readingDrop interface; add dehydrator before desalter
Grid current low / erraticLow water in oil; low conductivityWash water % and flow rateIncrease wash water to 5-8%
Emulsion layer (rag) growingFines, asphaltenes, or EB mismatchRag height trend; EB bottle testRag drain; switch to asphaltene-compatible EB
Brine turbid (oil carryover)Under-mix (low pressure drop at mix valve)Mix valve dP (target 10-25 psi)Close mix valve; reduce oil carryover to WWT
High BS&W in desalted crudeInsufficient retention timeFeed rate vs designCut feed; verify vessel level
pH excursion in desalter brineCaustic injection excursion or sulfide saltsCrude sulfur, injection rateStabilize caustic; target pH 6-8 brine
Iron sulfide carryoverOxidized corrosion products in crudeCrude source; tank historyIncrease EB; add sludge carrier to brine
Slop oil layer in brineEmulsion breaker saturated or wrongBottle test w/ alternate EBSwitch EB; limit slop accumulation
Atm tower overhead chloride highDesalter not removing organics; hydrolyzable ClOverhead water Cl, pH; crude TANRaise wash water; add neutralizer upstream tower; review caustic injection

3. The Four Desalter Variables You Actually Control

  1. Wash water rate (typically 3-8 vol% of crude). Too little = poor salt removal; too much = overwhelms coalescer and grid.
  2. Mix valve delta-P (typically 10-25 psi across valve). Higher = better contact but tighter emulsion that's harder to break.
  3. Emulsion breaker dose (typically 5-30 ppm wt on crude). Matched to crude type. Asphaltenic crudes need ester-based; paraffinic crudes may use polyol.
  4. Interface level (oil-water interface position). Too high = water carryover to tower; too low = oil in brine to wastewater.

Recommended tuning order when things are off

  1. Check interface level first. Correct to design (usually 40-60% from bottom).
  2. Verify wash water flow and quality (chloride < 20 ppm ideal; condensate or stripped water).
  3. Run bottle test with current + 2 alternate EB chemistries to verify chemistry match.
  4. Adjust mix valve dP only after (1)-(3); changes here are slow to observe.

4. Crude-Type-Specific Gotchas

Crude TypeTypical ChallengeMitigation
Heavy sour (Maya, Kuwait)Asphaltene emulsion, high saltTwo-stage desalter; higher dose; raise brine pH to 7-8
Light sweet (WTI, Brent)Low salt but hydrolyzable Cl from stabilizerVerify stabilizer overhead; increase wash water at upstream unit
Shale (Bakken, Eagle Ford)High paraffin wax; olefins; ironHeat desalter feed to >260F; EB match to wax chemistry
Opportunity crude (Doba, Kern)High TAN; naphthenic acid saltsRaise caustic injection; monitor overhead corrosion carefully
Synthetic bitumen (Albian, SU Hardisty)Very high asphaltene; water-in-oil stableSpecialized EB; expect rag buildup; plan rag drain schedule
Tight oil + produced waterDivalent cations (Ca, Mg) tougher to removeEDTA or similar chelant in wash water

5. When to Suspect a Mechanical Problem vs. Chemistry

6. Sampling & Monitoring Plan

Minimum sampling to run a healthy desalter:

7. Design Heuristics for New / Revamped Desalters

For salt removal target < 1 PTB on heavy sour crude, use two-stage desalters in series. Single-stage rarely achieves <1 PTB on > 1% salt feeds.
Size for 45-60 min residence time in the oil phase at MAX design feed; API 661 provides no desalter spec but vendor (Natco Cameron, Sulzer, Howe-Baker) sizing methods are well-established.
Include a rag-drain nozzle at the interface and an emergency dump line to slops. Sooner or later every desalter gets fouled and needs to be drained hot.

8. References