Refining Vacuum Tower
Vacuum Tower Troubleshooting Restores Vacuum and Extends High-Performance Operation
Two refinery cases show how pressure profiling, operating-data reconciliation and fouling-resistant internals resolved vacuum loss and declining wash-section performance.
CLIENTS: Confidential European and Asian refiners | APPLICATION: Refining vacuum distillation units
3 kPa
demister pressure drop identified
6 → 4 kPa
tower-top pressure recovery
13°C / 39°C
distillation-overlap improvementss
88.1 → 95.3
total heat removal,
M kcal/hd
Executive Summary
Vacuum-tower performance depends on more than the ejector or vacuum-pump package. Restrictions inside the tower, poor overhead injection design and fouling-prone wash-section internals can gradually reduce vacuum, product recovery and heat-transfer performance. Two GTI field investigations demonstrate how a disciplined review of operating data and internal hardware can distinguish the true root cause from more obvious but incorrect explanations.
In the first case, tower-top pressure doubled after a revamp even though the downstream vacuum-system pressures and load remained essentially unchanged. A temporary pressure measurement isolated a 3 kPa restriction across a plugged demister. Corrective operating actions avoided an emergency shutdown, and the turnaround inspection confirmed ammonium-salt deposits. In the second case, GTI redesigned fouling-prone trough distributors in a large vacuum tower. The revamp improved fractionation and heat removal while slowing the performance decline during the following operating cycle.
Two Cases, One Troubleshooting Method
Both investigations began with field data rather than assumptions: establish the baseline, reconcile system loading, locate where pressure or performance changed, connect the observations to a physical mechanism, and then select the smallest practical operating or hardware correction. This method prevented unnecessary work in the vacuum package and directed the wash-section revamp toward the features controlling run length.
Case 1: Vacuum Loss After a Capacity Revamp
A European vacuum distillation unit was revamped for higher capacity, including partial replacement of the tower internals and an upgraded charge heater. After start-up, an air leak at the water-ring vacuum-pump suction was found and repaired. Nevertheless, the tower-top pressure continued to rise over the next several months, increasing from the 3 kPa absolute design pressure to 6 kPa absolute. Vacuum-residue quality became unsuitable for bitumen production, and the refinery began preparing for another shutdown.
Pressure Profiling Isolates the Restriction
GTI evaluated the three possible cause categories: excessive vacuum-system load, equipment deterioration and excessive pressure drop. Sour-water flow from the hotwell showed essentially the same system loading before and after the revamp - 4.9 versus 4.7 t/h. The second- and third-stage suction and discharge pressures were also nearly unchanged, indicating that those stages were operating normally and that massive air leakage was unlikely.
-
Tower top (P1). Pressure increased from 3 to 6 kPa absolute after the revamp.
-
Second-stage suction (P2). Pressure remained at 8 kPa absolute.
-
Third-stage suction and discharge (P3/P4). Pressures remained approximately 21-22 and 110 kPa absolute.
-
Temporary measurement. A digital manometer connected at the ammonia-water injection point measured 3 kPa absolute downstream of the demister.
Pressure profiling located a 3 kPa pressure drop across the demister
Root Cause
The 6 kPa tower-top pressure and 3 kPa downstream measurement established that the lost vacuum was caused by the demister, not by the downstream ejector stages. The most likely deposit was ammonium salt. A simple 1-inch ammonia-water connection on the vertical overhead line allowed liquid to run down the pipe wall toward the demister, where ammonia could react with hydrogen sulfide in the offgas and form deposits.
Operating Remedy Avoids an Emergency Shutdown
Because ammonium bisulfide salts decompose at elevated temperature, GTI recommended increasing the tower-top temperature and adding stripping steam. The tower pressure stopped rising and later declined from 6 to 4 kPa absolute. This allowed the refinery to continue operating for approximately another six months until the planned turnaround instead of taking an emergency outage.
Turnaround inspection confirmed ammonium-salt deposits in the demister. The demister was removed, and the ammonia-water injection arrangement was redesigned to prevent recurrence. The case demonstrated that a single pressure reading at the tower top could not distinguish vacuum-package limitations from an internal tower restriction; the pressure profile was the decisive diagnostic.
Outcome: the refinery avoided an emergency shutdown and maintained operation until the scheduled turnaround.
Case 2: Extending Vacuum-Tower Wash-Section Run Length
A large Asian refinery vacuum tower - more than 15 m (50 ft) in diameter - experienced a recurring cycle of strong performance immediately after turnaround followed by declining recovery and fractionation as fouling progressed. The wet-mode tower contained four packed pumparound beds, a fractionation bed, a wash bed and stripping trays. Several design constraints made the service particularly severe.
-
Near-sonic transfer-line velocity. High feed momentum atomized liquid and produced substantial entrainment from the flash zone; major transfer-line modification was not practical.
-
Pumpdown and fractionation duties. Pumparound sections also received liquid from the bed above and therefore had to provide separation as well as heat transfer.
-
Low liquid loading. Gravity trough distributors required relatively small holes to achieve distribution across the specified operating range.
-
Progressive fouling. Distributor plugging and packing deposits increased pressure drop and reduced product recovery as the run progressed.
Vacuum-tower configuration (left) and fouled trough distributor observed during inspection (right)
How Distributor Fouling Reduced Performance
The wash, HVGO and HHVGO trough distributors used drip-point multiplication but still required holes of only 5-6 mm. Corrosion products, coke and other solids circulated through the pumparound system and entered the distributors. External strainers could not capture every particle, and the longer residence time in hot gravity distributors also encouraged coke formation inside the troughs.
Once a hole plugged, the associated packing area received less liquid and ran hotter. The dry region promoted additional coke formation, which reduced packing void area, increased pressure drop and further degraded fractionation and heat transfer. A small distribution defect could therefore grow into a self-reinforcing fouling cycle.
Distributor Redesign for Greater Fouling Resistance
-
Revised operating range. The specified turndown was narrowed from 40-120% to 70-120%, reflecting more than a decade of actual operation without substantially reduced charge rates.
-
Larger liquid openings. The distributor-hole diameter increased from 5-6 mm to 7-8 mm, improving resistance to particulate plugging.
-
Elevated hole locations. Raising the openings above the distributor deck created space for solids to accumulate before reaching the active holes.
-
Packing renewal. Fouled packing was replaced as part of the revamp, restoring open area and mass-transfer performance.
Multiple-stage trough distributor with elevated liquid openings
Verified Post-Revamp Results
The revamp produced immediate improvement in product fractionation and increased calculated total heat removal. The source paper reported the following before-and-after operating comparisons:
Performance measure
Before
After
Improvement
HVGO 5% - LVGO 95% overlap
-81°C
-68°C
13°C
HHVGO 5% - HVGO 95% overlap
-120°C
-81°C
39°C
Total pumparound heat removal
88.1
95.3
+7.2 M kcal/h
Long-term trends during the first two years showed that fouling and the associated loss of fractionation developed more slowly than during the previous operating cycle. The changes did not eliminate the severe-service fouling mechanisms, but they extended the period of higher-yield, higher-quality operation.
The result: better vacuum diagnosis, improved fractionation and a longer high-performance operating window.
Performance Under More Severe Operation
At the time of the paper, the revamped tower was operating at approximately 3.9-4.5 mmHg absolute overhead pressure. A leaking vacuum pump required a reduction in steam consumption from the original 4,200-4,500 kg/h to approximately 2,800-3,600 kg/h. Product recovery was maintained by increasing flash-zone temperature from 385°C to 395-399°C. Despite these more severe conditions, tower performance remained better than the middle-of-run performance following the prior turnaround.
Engineering Lessons
Measure the pressure profile, not only the tower top. Intermediate measurements separate vacuum-package limitations from internal restrictions and identify where the available vacuum is being lost.
Reconcile load before diagnosing equipment condition. Stable sour-water flow and interstage pressures demonstrated that the downstream vacuum equipment and gas load had not materially changed.
Use actual operating history to challenge design specifications. Unneeded turndown forced smaller openings and increased fouling risk; a realistic operating range enabled a more robust distributor design.
Design for fouling rather than assuming it can be eliminated. Larger and elevated openings provided greater tolerance for solids and coke, extending the useful operating period in severe service.
GTI's Integrated Troubleshooting Approach
GTI combines process-data reconciliation, site surveys, pressure profiling, internal-equipment assessment and operating-history review. The objective is to translate field evidence into a physical explanation and then into practical actions - immediate operating changes when a shutdown can be avoided, followed by targeted hardware modifications during the next turnaround.
Experiencing vacuum loss, wash-section fouling or declining VDU yields?
Let GTI Solutions evaluate your vacuum tower.