Feb-2023

Fouling in VDU ejector systems

Review of critical ejector-condenser interplay, with an overview of fouling in VDU ejector systems, including a case study where fouling impacted a refiner's bottom line

Jim Lines
Graham Corporation

Article Summary

Ejector system fouling can cost refiners millions of dollars annually. It leads to lost yield or lower throughput, either of which affects a refiner’s bottom line. An ejector system is a combination of ejectors and condensers configured in series, typically ejector-1st intercondenser-ejector-2nd intercondenser-ejector-aftercondenser. Declining thermal efficiency or heat rejection is seen with progressive fouling in heat exchangers and intercondensers. However, the fouling also increases intercondenser pressure.

If intercondenser operating pressure rises beyond the discharge pressure capability of its preceding ejector, overall system performance breaks and vacuum distillation unit (VDU) pressure rises sharply, an undesirable outcome for any refiner. In VDU service, a breakdown to ejector system performance can mean column overhead pressure may be 25-35 mm Hg rather than, for example, at design of 15 mm Hg abs. When overhead pressure rises in such a manner, vacuum tower bottoms (VTB) increases while vacuum gas oil (VGO) cut is reduced.

Condensers within an ejector system serve to a) condense vapours discharging from a preceding ejector at a pressure within the discharge capability of that ejector, b) minimise the mass flow rate of vapours exiting the condenser that must be handled by a downstream ejector to align with the performance capability of that ejector (mass flow rate versus pressure performance curve), and c) in effectively accomplishing a) and b) will keep ejector system energy consumption efficient and vacuum distillation unit pressure within specification.

It is the ejector-condenser interplay that is so critical to a refiner meeting vacuum distillation operating objectives. When fouling surpasses the design basis, the consequence for a refiner can be significant, frustrating, and difficult to troubleshoot.

VDU ejector system
First-stage ejectors maintain vacuum column overhead pressure by evacuating non-condensible gases, hydrocarbon vapours, and steam present in the overhead of the distillation process. The overhead is compressed by the first-stage ejectors and discharged into the first intercondensers, where steam and hydrocarbon vapour are condensed. Non-condensible gases saturated with steam and hydrocarbons flow to the second-stage ejector, where again compression occurs with discharge to the second intercondenser. Condensation of vapours then occurs and saturated non-condensible gases flow to the third-stage ejector, where they are compressed to a pressure above atmospheric pressure and discharged into an aftercondenser.

Due to the large scale of a modern refinery, a VDU ejector system, such as one recent case at a 200,000 bbl/day Asian refinery, may have multiple ejectors and condensers in parallel at each stage. Medium-pressure steam is the energy source for the ejector compression of gases. Water typically is used to effect condensation, such as cooling tower water, river water, or seawater.
Fouling can cause condenser operating pressure to rise above the discharge capability of an ejector that precedes it. When that happens, a performance break occurs where the VDU column rises sharply well above desired operating pressure. An elevated VDU column pressure increases VTB, thereby reducing VGO yield. If one considers a $10/bbl discount for resid versus VGOs, a 1% loss of yield on a 200,000 bbl/day refinery is approximately $7 million per annum.

Heat transfer basics
Most commonly, condensation takes place on the shellside with water flowing tubeside. Several variables influence the shellside heat transfer rate, including:
-    Mass flow rate and MW of non-condensibles. The greater the mole fraction of non-condensibles, the lower the shellside heat transfer rate
-    Mass flow rate, composition (MW, boiling point, mole fraction) of hydrocarbons. The higher the mole fraction or lighter the hydrocarbon composition, the lower the shellside heat transfer rate
-    Proportion of flow that is steam. Higher mole fraction of steam will generally yield a higher shellside heat transfer rate
-    Operating pressure correlating to velocity
-    Condensate mass flux (pph/ft2) or condensate film thickness.
The complication that arises in establishing the shellside heat transfer rate is immiscible condensate formation, as water and hydrocarbons are immiscible. The mole fraction of non-condensible gases progressively increases as heat is rejected along the release curve, as does condensate film thickness.
The following condenser heat release curve example discusses the amount of water and hydrocarbon condensates formed. The amount of non-condensible gas is constant. As heat release progresses, the shellside heat transfer rate declines:

Heat release curve along with water plus hydrocarbon condensate formation example (as follows):

The tubeside heat transfer rate is generally a straightforward calculation of forced convection in the turbulent flow:
Overall heat transfer rate clean - Uclean=                  1
                     1         OD
                Hshellside +   ID*Htubeside+ Rwall

Fouled heat transfer rate - Ufouled =             1
                    1
                Uclean + OAFF

Design or working heat transfer rate - Uworking = Heat released                                 Area*LMTD
Terms:
Uclean — Overall clean heat transfer rate, Btu/hr ft² °F
Ufouled — Overall fouled heat transfer rate, Btu/hr ft² °F
Uworking or Udesign — result of Q/Area * LMTD, Btu/hr ft² °F
OAFF — overall fouling factor incorporating shellside and tubeside fouling factors, hr ft2 °F/Btu
Q — thermal duty or heat released, Btu/hr
Area — condenser heat transfer area, ft²
LMTD — log mean temperature difference, °F