Revamp of Slovnaft’s hydrocracking unit
A revamp to maximise middle distillate production and increase unit reliability and safety.The KHK hydrocracking unit at the Slovnaft refinery in Slovakia has been revamped in a joint effort by Haldor TopsÃ¸e and Slovnaft
OLE FREJ ALKILDE, Haldor TopsÃ¸e
FILIP STANCL, Slovnaft
Viewed : 6876
The unit is a full-conversion hydrocracker, which has been in operation since 1991. The nameplate capacity is 104 t/h, equivalent to 17 700 b/d. The â€¨unit is processing straight-run â€¨VGO from a Russian export crude blend.
Topsøe and Slovnaft initiated a revamp project with the primary objective of maximising middle distillate production. Additionally, unit safety and reliability were to be improved. This should all be done without costly modifications to major equipment items.
The revamp project was conducted in three stages. The first stage was a revamp study, where the current performance of the unit was analysed in detail and a range of possible improvements were identified. In the second stage, the most desired improvements were selected and the revamp design package was executed. In the third stage, the detailed design and actual unit modifications were performed.
The KHK hydrocracking unit at the Slovnaft refinery in Bratislava, Slovakia, is a single-stage full-conversion hydrocracker (see Figure 1). It has been in operation since 1991 and is licensed by Unocal and designed by Snamprogetti.
The high-pressure loop is configured in a cold separator layout with a cold separator pressure of 13.3 MPag. The unit was designed with three reactors, with three beds in the first reactor, two beds in the second reactor and three beds in the last reactor. Heat is supplied by both a recycle gas heater and a recycle oil heater. The recycle oil from the fractionation section is sent to the second reactor inlet when the unit is operating in recycle mode.
The fractionation section is configured with a debutaniser first then a fired heater reboiler. The debutaniser bottoms are sent to a main fractionation column with kerosene and diesel side strippers as well as kerosene and diesel pumparounds. The fractionator overhead is sent to a naphtha splitter (not shown in Figure 1) for the separation of light and heavy naphtha.
The original design of the unit allowed for three modes of operation: maximum diesel production, maximum naphtha production and once-through low conversion for unconverted oil production (steam cracker feed). Since diesel is foreseen to be the most desired product in the unit in the future, the overriding requirement for the revamp was to increase the diesel yield. The maximum naphtha yield could be reduced if required.
In the time leading to the revamp, several maintenance and mechanical problems became evident in the unit. Some of the problems were solved within the revamp project, while others were solved as regular maintenance items during normal service.
The unit has experienced one serious temperature runaway event. In March 2006, following a feed pump trip and subsequent restart assisted by a human factor, the catalyst bed temperatures ran away and within minutes exceeded measurable range at some catalyst bed points. Fortunately, the reactor skin temperatures did not reach critical values. The revamp also needs to address this serious operational risk to ensure adequate automatic safety procedures are in place to prevent such an event occurring again.
A simulation of the unit was established. This served as the basis for the following studies. A total of nine flow sheet cases were developed during the study, which detailed the performance of the unit.
Several areas of concern were identified during the study. First, the high-pressure loop operation was not sufficiently stable. This was most evident in recycle mode â€¨operation, where the recycle gas compressor was operating close to the surge line and the compressor instrumentation was not sufficient to determine exactly how close to the limit the compressor was operating.
Another serious concern with hydrocracking units is liquid maldistribution, which can lead to stagnant liquid pockets, reactor hotspots and ultimately temperature runaways. The existing distributor trays, quench zones and reactor temperature measurements were not performing as well as modern internals could.
The temperature response of the reactor system was slow, too. Temperature excursions in the reactor systems were difficult to control, and thermal feedback through feed/effluent and inter-reactor exchangers worsened the risk of temperature excursions leading to a temperature runaway.
Also, the high-pressure cold separator was not large enough to allow for efficient separation of water and hydrocarbons. This caused water slip-through to the debutaniser column. The debutaniser was experiencing several problems. The water slip-through from the cold separator caused bad operation and corrosion in the top of the column. Continuous methanol injection was required to control the formation of hydrates in the LPG line. The H2S removal in the column was also insufficient, resulting in too high a sulphur content in the light naphtha.
The capacity of the wash water pump was on the limit for protecting the high-pressure piping from corrosion due to NH2SH. The maximum rate of wash water was just able to keep the concentration in the sour water below 6 wt%, but there was no excess capacity for, for instance, treating heavier feed or handling a higher throughput.
Likewise, the kerosene pumparound piping and heat exchanger layout was limiting the heat recovery in the kerosene section. Also, the formation of heavy polynuclear aromatics (HPNA) in the high-conversion recycle mode could limit the maximum achievable overall conversion.
Finally, the catalyst volume was limiting the cycle length. The deactivation rate of the pretreatment catalyst was the main factor limiting the cycle length of the unit and, when the HDN activity of the pretreatment system became too low, the activity of the whole catalyst system was too low and EOR was reached.
The study considered two catalyst cycle length cases: either keeping the existing three-year cycle length or reducing the cycle length to two years while increasing the middle distillate yields.
Add your rating:
Current Rating: 4