By Lev Serebrinsky
As we mentioned in Article 1 in the March 2007 newsletter on the subject of Description of Methodology, the Refinery Bad Actors List is the major tool for resolution of heat exchanger problems across the refinery. When the refineries started to apply the tools, it was noticed, however, that while Bad Actors vary dramatically from refinery to refinery, there are five heat exchangers which are in each Refinery’s Bad Actors List. We focus this article, therefore, on reliability recommendations to these Typical Bad Actors as follows:
AT OH Condensers in Crude Units cause significant Lost Opportunity Costs to the whole refinery and substantially increase Crude Unit Maintenance Costs when they are frequently taken out of service to repair internal leakage from tubes or for mechanical/chemical cleaning.
Experience has demonstrated that the root cause of 95% of the outages was shell side corrosion of the tubes. Therefore, a massive program of eddy current tube wall thickness measurements was taken followed by reliability analysis of the data. The analysis revealed a common pattern of corrosion development in all AT OHCs bundles. That was that, first, the corrosion is localized to certain areas of the bundles, and, second, that the corrosion strikes a certain group of tubes within the areas.
In addition, review of the maintenance history of these bundles showed
that the groups of tubes are attacked by corrosion in a certain sequence,
with a four to six month interval between the attacks. Further, the
inspection reports showed that the shell side of the bundles was fouled so
severely by the tube corrosion products that the condensers had to be
mechanically cleaned after the third group of
tubes was leaking. Furthermore, the maintenance practice was that after the fourth leakage the bundles were, as a rule, retubed due to concerns with their ability to make the next run.
Based on these studies, all AT OHC tube bundles were “re-engineered for reliability.” This re-engineering included:
To minimize the cost for the re-engineering, the recommendations were implemented during the re-tubing of existing bundles or during manufacturing of the replacement bundles. Since then there has been no unscheduled outages of any AT OH Condensers, and the bundle’s design life was increased from 4 to 12 years.
Hottest Crude Train Exchangers in Crude Units are responsible for Crude Unit High Maintenance Costs due to (aaa) external leakage from the Stationary Tubes Sheet Flange Joint and (bbb) costly mechanical cleaning of severely fouled tube bundles.
(aaa) Costly repair of external leakage from the Stationary Tubes Sheet Flange Joint. These exchangers have the highest temperature crude inside the tubes and the tower’s hottest side stream in the shell side. Their stationary tube sheet flange joint often leaks each time the tower’s side stream is temporarily lost due to a tower’s upset. When this happens, the temperature across the joint swings down and then up in the range of 350 - 400°F in a short period of time. In reliability engineering this is known as a “high temperature swing.” For all “regular” flange joints the temperature swing means nothing. But because the stationary tubesheet (STS) flange joint is unique (it has two gaskets, one tube sheet sandwiched between the gaskets, two flanges abutting the tubesheet, and the longest studs), the high temperature swing is critical.
That is that during the temperature swing of that magnitude, all six components of the STS flange joint experience the large and uncontrollable expansion-contraction movements which result in relaxation of the studs and loosening of the flange joint itself.
Reliability re-engineering of the joint consists of:
This complex approach eliminated the leaks totally.
(bbb) Costly mechanical cleaning of severely fouled tube bundles. The hot crude plugs the tubes end-to-end because the crude’s heavy hydrocarbons are solidifying and sticking to the tube inner wall at these exchangers’ operating temperature. Anti-fouling programs typically do not work, and many bundle mechanical redesigns did not work either. The problem was typically resolved completely by application of On-Line-Mechanical-Cleaning-Technologies (OLMCTs) which constantly scrubs the asphalt-like compounds from the tube internal wall on the run thus providing a between-the-cleanings run of these exchangers of up to five (5) years.
There are several OLMCTs available commercially (e.g., Spirelf Turbulence Promoters, Brush-and-Basket Technology, etc.). The selection of an individual OLMCT should be evaluated locally based on specific unit operating conditions, manufacturer’s experience, and cost.
Vacuum Tower Bottoms/Crude Heat Exchangers in Crude Units suffer from High Maintenance Costs due to coking up of the bundle while the heat exchanger is being steamed out as a part of preparation for unit shut down. The bundles are often stuck inside the shell cylinder so tightly that it could require several days of expensive work to pull the bundle and to clean the shell. In many cases, however, it was impossible to clean the coked-up bundles so that they had to be retubed on an overtime basis.
The problem was resolved by application of a multi-step procedure of washing the shell side using certain hydrocarbons, and by injecting brake-up-coke chemicals during preparation of the unit for the turnaround. It also sometimes required some redesign of circulation connections in the shells.
Feed/Effluent Exchangers in Catalytic Reformers suffer from external leakage of hydrogen from the STS Flange Joint and from the shell cover-to-shell flange joint during routine switch of the reactors during operation. For small size exchangers, reliability improvements made to gaskets and bolting were successful. For large size exchangers, however, gasket and bolting changes did not work reliably. The best industry experience for large exchangers was achieved using a welded-in stationary tubesheet and with a welded-on shell cover. This design has resulted in 16 years of leak-free, fouling-free run of all these heat exchangers.
Slurry/Feed Exchangers in FCCU suffer from end-to-end plugging of tubes by the catalyst’s slurry fluid.
The root cause of the problem is the low velocity of the slurry flow. Depending on local guidelines, the reliability-recommended velocities are 4 ft/s as a minimum, and 10-12 ft/s as a maximum (the maximum is limited by erosion concerns). These guidelines should be applied to all slurry circuit piping as well.
Some mechanical reliability features such as certain tube size, tube material, and the floating head pull-thru bundle design should be applied as well.
General. Remember that the exchanger reliability improvements noted above will not happen “automatically” for new exchangers just by meeting the relevant API, TEMA, and even most owner-company engineering standards. They must be specified as part of the technical specification that is prepared for exchanger services where experience has demonstrated them to be worthwhile.