13 mm Hg abs and the first intercondensers were replaced with larger condensers to address greater heat rejection requirements. The refiner sought assistance to identify the root cause and develop a solution for achieving 13 Hg abs overhead pressure because 25 to 35 mm Hg abs was the current VDU overhead operating pressure. That was unacceptable as it led to >$18 million per year in lost VGO yield. The performance improvement engineer requested oper - ating data trends to assess what had been transpiring with the ejector system. From the trend data, it was not possible to deduce what was causing the elevated operating pres - sure that caused the first-stage ejector to operate in a ‘bro - ken’ condition. Broken ejector performance occurs when the back pressure the ejector must discharge to exceeds its maximum discharge capability. An engineer was dis - patched to the site to conduct a performance survey, speak with operators, and identify the root cause. A deep dive early in the design phase and prior to ejector system procurement can avoid a costly yield shortfall The performance survey identified that the first-stage ejectors were subjected to 107-110 mm Hg abs back pres - sure, while the maximum discharge capability without loss of compression shock wave was 100 mm Hg. Furthermore, the process side (shellside) pressure drop across the first intercondensers varied between 15 to 20 mm Hg. Measured values varied due to the unstable operation of the system. However, it was clear that the pressure drop across the shellside of the condenser well exceeded the expected pressure drop to 5 to 6 mm Hg. The operating pressures at the suction of the third-stage ejectors confirmed the non-condensible gas load was below design as the measured pressure was about 100 mm Hg below design. With low non-condensible gas mass flow rate load to a second- or third-stage ejector exiting its preceding condenser (non-condensible gases plus vapours of saturation) will be below design, thus enabling that ejec - tor to pull to a lower pressure. Focus was on the first intercondenser. Potential causes for elevated operating pressure and high pressure drop included: • Lower cooling water flow rate • Elevated inlet cooling water temperature • Excessive hydrocarbon loading from the vacuum column • Excessive non-condensible gas loading from the vacuum column • Fouling on the shellside, tubeside, or both. The engineer ruled out cooling flow rate or tempera - ture issues as both were within acceptable range. Non- condensible gas loading was ruled out, as previously noted, due to lower inlet pressure for the third-stage ejectors. Excess hydrocarbon loading could not be ruled out, nor could fouling.
The refiner had a shutdown planned within a few days, so the performance improvement engineer recom - mended that the first intercondenser bundles be pulled for inspection to observe the extent of fouling. The system had run for about 10 years since the revamp and installa - tion of new, larger first intercondensers to improve clean fuels production; therefore, some degree of fouling was anticipated. A long-term trend for the time-based relationship between condenser condensate temperature and cooling water inlet temperature inferred that a fouling issue was likely. The difference between condensate temperature and cooling water inlet temperature progressively increased with operating life. The longer the condenser was in ser - vice, the greater the difference became between the two temperatures, inferring fouling was suppressing the overall heat transfer rate. Bulk condensate, while imperfect, does correlate reason - ably to condenser operating pressure. Directionally, warmer condensate temperature correlates to higher operating pressure. The data highlighted that the condenser pres - sure was increasing with run time. The first intercondenser bundles were pulled, and severe fouling was observed on both the tube and shell sides. The performance improvement engineer identified the root cause of poor vacuum column overhead pressure with consequent loss in gas oil yield due to fouling on both the shellside and tubeside of the first intercondenser. This caused the working overall heat transfer rate to drop below the design basis, increasing condenser operating pressure beyond the discharge capability of the first-stage ejector. Conclusion Fouling is an important design parameter. Applying fouling factors into a design increases ejector system Capex. Failing to apply good fouling factor ‘judgment’ for ejector system design can result in significant economic loss. A thorough understanding of the fouling tendencies of both the process side and cooling water side is crucial. Mitigating tubeside fouling can be addressed via filtration, chemical treatment, backflushing, and overall good cooling water utility system management. On the process side, principally with regard to hydrocar - bons, knowing the anticipated range of crude feedstock or crude blends to be processed is important. Any prior opera - tional experience with such feedstocks will prove informa - tive with regard to fouling. Striking an appropriate balance between operational reliability to avoid an unplanned shut - down due to fouling and the initial capital cost of the ejector system and its installation must be addressed. A deep dive early in the design phase and prior to ejector system pro - curement can avoid a costly yield shortfall. Jim Lines recently retired from Graham Corporation, where he worked in engineering, sales and general management for 37 years. He con - tinues to provide technical and business management support to the company in an advisory role. He holds a BS degree in aerospace engi - neering from the State University of New York at Buffalo. Email: Jlines@graham-mfg.com
64
PTQ Q1 2023
www.digitalrefining.com
Powered by FlippingBook