The demand for middle distillates

Revamp cat feed hydrotreaters for flexible yields Revamping a cat feed hydrotreater to a flexible mild hydrocracker can be the most attractive economic option for adjusting the gasoline to diesel ratio DAVID SCHWALJE, LARRY WISDOM and MIKE CRAI Axens North America The demand for middle distillates from US refiners has increased steadily since 2, to current production levels totalling over 6.6 million b/d 1 as a result of two main market conditions: the US enjoys a feedstock advantage over European and South American refiners thanks to the light tight oil (LTO) boom; and increasing demand from Mexico and South America for transportation fuels. In Q1 of 216, the top three countries of export for middle distillate fuels refined in the US were Mexico (11.8 million barrels per month), Chile (9.8 million barrels per month), and Colombia (8.8 million barrels per month). 2 Low crude prices have reduced revenues for vertically integrated, state-owned refiners, which has contributed to major project delays and increased uncertainty in those regions. Diesel exports to Latin America should therefore continue, with demand for on-road diesel expected to grow at an annual rate of 2.2% through 22. 3 In the near term, the tight oil feedstock advantage over European refiners is expected to tighten thanks to the lifting of the US crude export ban; however, demand in South America and Mexico will remain strong due to project delays 4 and demand growth. In the long term, middle distillate demand growth will continue as a result of the commercial transportation sector and consumption growth in developing regions. Higher worldwide crude prices would result in increased domestic demand from the exploration and production sector as producers in Diesel-gasoline spread, $/BBL 15 5 5 15 2 Jan 2 Figure 1 US monthly diesel-gasoline spread Jul 2 Jan 211 Jul 211 Jan 212 Jul 212 Jan 213 Jul 213 Jan 214 Jul 214 Jan 215 Jul 215 Jan 216 the shale plays return to pre-215 output. At the same time, consumption of gasoline may decrease as a result of increasing numbers of electric vehicles and improving fuel economy standards. As a result, many US refiners have either planned or commissioned projects to shift their diesel to gasoline ratio (D/) to favour middle distillates, as refiners with a high D/ ratio have benefited from a diesel-gasoline finished product spread that has, on average, favoured diesel since 2. Projects have included new brownfield, high conversion hydrocracking units, reconfiguration projects, and the planned shutdown of multiple FCC units. However, product markets remain elusively volatile, and the incentive to produce diesel in lieu of gasoline does not exist throughout the year, nor in all regions. Short-term volatility in seasonal diesel-gasoline spreads has historically favoured gasoline during the summer driving season. This trend was especially apparent in the summer of 215 when low pump prices resulted in increased domestic gasoline demand. This was followed immediately by a mild winter, which reduced domestic middle distillate consumption. These phenomena, coupled with a reduction in diesel demand from E&P operations, resulted in a sustained period of high gasoline margins in 215-216. Figure 1 charts the US average dieselgasoline spread since 2. 5 The five-year trend has been clear and remarkably consistent, favouring middle distillates during the winter months, gasoline during the summer months, and diesel on the whole. The seasonal swing has historically been in the range of $-2/bbl. The profitability of the refinery is therefore fluctuating seasonally, not only based on crude acquisition costs but also with each www.eptq.com Revamps 216 3

Property WTI Brent Arab Lt. Maya Athabasca Bit. ravity, API 39 38.6 33.4 22 8.4 S, wt%.27.29 1.79 3.56 4.92 Distillation, vol% Naphtha 34 3 25 17 2 Middle dist. 29 29 28 22 14 VO 21 29 3 27 34 Residue 14 9 15 33 5 D/ ratio.85.94 1.12 1.28 7. Table 1 refiner s unique ability to rapidly shift its D/ ratio to favour the most profitable transportation fuel. There is a distinct financial incentive for a refiner to maintain flexibility of its product slate in response to market conditions. Current practices for product slate flexibility A number of strategies are available to refiners looking to shift their fuel balance. In practice, the preferred option for each refiner varies greatly as a function of various site-specific factors such as budget constraints, regional product demand, configuration, and crude source. The most common options include: Modifying the refinery s crude selection Modifying crude unit and product fractionator cut-points Reducing FCC severity to increase light cycle oil (LCO) production Installing new high conversion hydrocracking capacity with high diesel selectivity Conversion of gasoline yield, wt% 85 8 75 7 65 Crude properties Revamping existing cat feed hydrotreater (CFHT) units to mild hydrocrackers (MHC). These options range from low or no cost solutions such as shifting product cut-points to high cost and long lead-time solutions such as the construction of a new hydrocracking unit. This article focuses on the conversion of existing CFHT units to MHCs as a low cost, fast track solution. Modifying crude selection The modern refiner typically has a wide selection of available crude sources and shifts between crudes based on both economic incentives and processing constraints. The relative content of gasoline-boiling and middle distillate-boiling material in crudes varies greatly and can be a profitable first step in adjusting D/ ratio. Table 1 summarises the distillations of five major crudes processed in US refineries. However, crudes with higher percentages of middle distillates are also typically higher in sulphur, Conversion asoline Slurry 2 18 16 14 12 6 55 5 8 6 45 4 HYK 4 2 35 11. 11.5 12. 12.5 13. Feed H 2 content, wt% Figure 2 FCC performance with increasing pre-treatment aromatics saturation Slurry yield, wt% nitrogen and aromatic content, which limit their appeal for the refiner within an existing configuration. Changing crude diet affects the entire refinery; as a result, many refiners can only blend in small amounts of opportunity crudes up to known limitations. Modifying crude unit and product fractionator cut-points For a given crude blend, the refiner can also make adjustments in their distillation cut-points, both in the crude unit and in downstream product fractionators. On the lighter end of the boiling range, operators must be cognisant of flash point issues, low heavy naphtha cetane, and hydraulic limitations. On the heavy end, diesel end-point cannot typically be increased substantially due to the diesel product specifications for sulphur, end point, or seasonal cold flow properties. Reducing FCC severity to increase light cycle oil production While attractive on paper from a distillate yield perspective, operating at lower severity to increase LCO production for upgrading to ULSD is costly due to the loss in gasoline yield and increase in low value fuel oil production. As a result, many refiners choose to operate at a more typical gasoline oriented severity and undercut the gasoline to boost distillate yield. Installing new hydrocracking capacity with high diesel selectivity In the last five years, there have been multiple major hydrocracking projects announced in the US with some complete and some in various stages of engineering and construction. However, the installation of new high pressure, high conversion hydrocracking units is both capital and schedule intensive, requiring a long period between initial inception and unit start-up. Revamping the CFHT to MHC service The last option discussed here is to revamp the existing CFHT to increase run vacuum gas oil (VO) conversion while also improving 4 Revamps 216 www.eptq.com

CFHT mode (low conversion) Hydrotreating HDS+HDN+HDA Targets: FCC feed S, N, H low conversion Figure 3 CFHT and MHC catalyst configurations the quality of the FCC feed. The majority of operating FCC units in the US are equipped with upstream hydrotreaters with a wide range of operating conditions. More recently designed units have higher pressures (>12 psig) and lower space velocity (LHSV <1.) to maximise aromatics saturation and the associated volume swell in addition to the classic contaminant removal functions. It is these mid to higher pressure units that are ideal for revamping to MHC service. Revamping to perform mild hydrocracking in the CFHT has the following benefits: Reduction of nitrogen content of the unconverted FCC feed and consequently greater activity on the FCC catalyst reater hydrogen addition (aromatics saturation) of the FCC feed and consequently higher FCC gasoline yield, higher LP olefins potential for propylene and alkylate, and an overall increase in FCC liquid product yields (see Figure 2) Decrease in FCC feed rate, which allows the refiner to either feed external VO to fill the FCC or increase FCC severity to boost LP olefin yields for either propylene recovery or to keep existing alkylation units full Increase in diesel-selective conversion to shift the refinery D/ ratio to favour middle distillates Increase CFHT fractionator flexibility to route CFHT heavy diesel either to the diesel pool to maximise diesel production or to the FCC to maximise gasoline production Mild hydrocracking mode (increased conversion) Hydrotreating HDS+HDN+HDA Very high activity Hydrocracking HT+HCK Increased conversion Amorphous or zeolite cracking catalyst Higher resistance to N slippage Increased selectivity to middle distillates Relatively low cost and quick implementation schedule when compared to a new unit. The scope and feasibility of the revamp will, of course, vary from location to location, but a general guideline is that CFHT units operating above 12 psig can be attractively revamped to maximise conversion, with straight run and cracked gas oil feed conversions in the range 25-5% possible in existing units, with only minor modifications to the reaction section. CFHT revamps consist of four phases: upgrade of the catalytic system, an engineering evaluation of the existing equipment, reactor internals, upgrade of the safety systems, and assessing the refinery-wide impact including FCC performance. Catalytic system The optimisation of a catalytic system in pure CFHT mode is focused on improvement of FCC feed quality mainly through sulphur and nitrogen removal, as well as consequential hydrogen addition. In some instances, onpurpose enhanced hydrogen addition is targeted when operating conditions and hydrogen supply are favourable. The ultimate catalyst system is selected depending on the specific unit targets and is typically comprised of one or more CoMo, NiMo or tri-metallic (CoMoNi) hydrotreating catalysts. During typical CFHT operation, there is always some inherent conversion of gas oil to diesel boiling range material through mechanisms of the various hydrotreating reactions. The amount of conversion, mostly achieved by boiling point shift, is typically in the 5-15% range. enerally, the use of CoMo or CoMoNi catalysts, which classically favour hydrodesulphurisation (HDS) over hydrodenitrification (HDN) and hydrodearomisation (HDA), results in conversions towards the lower end of the scale. The opposite is true of NiMo catalysts, which provide superior hydrogenation functions including increased HDA. Systems utilising hydrotreating catalysts in CFHT mode are usually very diesel selective. The additional increase of gas oil conversion from CFHT to MHC service typically introduces the use of cracking catalyst downstream of the hydrotreating catalyst. And, as one of the facets of a revamp is to use the existing catalyst system volume, the amount of hydrotreating catalyst must be reduced to allow room for the cracking catalyst. As a result, the choice of hydrotreating catalyst and its selected volume must consider final product quality targets as well as the sensitivity of the cracking catalyst to the presence of organic nitrogen, which will affect activity and consequently conversion and diesel selectivity. To achieve a satisfactory catalyst cycle in MHC service, a very high activity NiMo hydrotreating catalyst is typically used followed by cracking catalysts that have a high tolerance to nitrogen slip from the hydrotreating section. Axens has thoroughly studied the impact of organic nitrogen on catalyst activity and operational stability by pilot plant testing and through industrial feedback from their catalyst applications. Axens HYK 7-series zeolite cracking catalysts have achieved high selective conversion with nitrogen slip above ppm. Allowable slip for amorphous catalysts is even higher. Catalyst configurations for both CFHT and MHC operating modes are shown in Figure 3. It is important to note that some hydrocracking catalysts have very 6 Revamps 216 www.eptq.com

HDK amorphous catalyst HYK zeolite catalyst Diesel selectivity at high temperature +++ ++ Constant product quality through the cycle + ++ High activity (lower WABT for equal conversion) + +++ Deactivation rate + ++ Product slate flexibility ++ ++ Nitrogen slip tolerance +++ ++ Table 2 high hydrotreating functionality, thus allowing excellent operational flexibility during periods of low severity operation (maximum gasoline operation). Some requirements that drive the selection of the hydrocracking catalyst system are as follows: Overall conversion level Selectivity of diesel versus gasoline Product quality (for instance, cold flow property improvement). For example, in the case of the current revamp study considering improvement of the D/ ratio, it is clear that a cracking catalyst with a very high selectivity towards diesel production is preferred. It is therefore very important to have access to a large portfolio of hydrocracking catalysts that encompasses a wide range of selectivity and activity, allowing for a large variety of unit objectives. One of the most important elements of selecting the cracking catalyst involves the comparison between an amorphous cracking catalyst (Axens HDK Series) and a zeolite based cracking catalyst (Axens HYK Series). Each catalyst type has its own distinct advantages (see Table 2). Engineering evaluation Revamps include an evaluation of the following sections with the extent of revamping dependent on existing equipment limitations and site-specific conditions: Revamping of the reactor heater is seldom required since the cracking catalyst is loaded into the last reactor beds, and the bed temperature profile can be adjusted as needed using vapour quench control. Evaluate the reactor pressure Cracking catalyst comparison drop and tray hydraulics. Modern reactors are critical for revamping older CFHT units. As a result of increased HDS and HDN, the wash water facilities and reactor effluent air cooler (REAC) are evaluated. In addition, the duty requirement in the REAC is likely to increase, although this increase typically does not require cooler modification. Hydrogen consumption will increase, and the make-up compressors must be evaluated for the new operation. In many cases, revamps and the use of spare machines prevent the need for entirely new machines. The recycle compressor must be checked based on the new reactor circuit hydraulics and recycle gas composition. The top section of the product stripper is evaluated to ensure adequacy of the trays under the increased liquid and vapour loads. A new product fractionator is required for those units that are only equipped with a steam stripper. For units with an existing fractionator, the column, its trays and associated equipment must be checked for the increase in kerosene and diesel production. The addition of a new product fractionator along with its associated equipment is typically the highest cost revamp item. However, as many units feed substantial (8-12 vol%) diesel boiling material to the CFHT as a result of poor separation in the crude unit, the fractionator cost can be justified by recovery of feed diesel in addition to the cracked material. In some cases with high diesel margins and high feed diesel contents, the recovery of feed diesel alone can justify the investment. Recovering middle distillates contained in the feed can increase the apparent conversion of the revamped CFHT unit above 5%. Safety systems Hydrocracking units inherently demand a higher level of monitoring than classic hydrotreaters, primarily as a result of the higher reactor operating temperatures and the associated risks of reaction runaway. As a result, these units utilise a greater number of temperature measurement points per reactor bed, reduced spacing between measurement levels, and automatic emergency depressurisation facilities. MHC revamps often require upgrading of the safety systems. In most cases, additional measurement points can be added to the existing reactors without the need for new nozzles. Modern flexible thermocouples allow for multiple sensing points to be present in one sheath such that additional measurement levels can be fitted into existing nozzles, including the catalyst withdrawal nozzles. Supports for new levels can be spot welded (low temperature) to the vessel wall without sacrificing the mechanical integrity of the wall or cladding. The existing CFHT is typically equipped with manual depressurisation facilities. The revamp may therefore require the installation of automatic depressurisation facilities triggered by a high reactor temperature. EquiFlow reactor internals Making use of all of the available volume in the existing reactor(s) is critical to safely maximising conversion in revamps of older units. Particular attention must be paid to reactor systems containing first generation reactor internals, as poor distribution not only reduces catalyst activity but, more importantly, can lead to channelling and hot spots in the hydrocracking section. Axens has developed the latest generation of EquiFlow Trickle-bed distributors and Hy-Quench-XM trays, which are the culmination of over 2 years of continuous research and development. 8 Revamps 216 www.eptq.com

Axens reactor internals have two distinct features: EquiFlow Trickle-bed trays employ secondary redistribution plates below a chimney tray to ensure nearly ideal vapour/liquid distribution throughout the bed. This tray allows for tailoring design for each application under a wide range of operating conditions such as those experienced in a flexible MHC (see Figures 4 and 5). The following plots (see Figures 6 and 7) outline the high performance of the Trickle-bed trays in hydrocracking service throughout the course of a two-year cycle over a wide range of operating conditions. In all beds (height of over 23ft), the ratio of radial ΔT to axial ΔT remained consistent and outstanding throughout the run. The Hy-Quench-XM system has been completely redesigned employing an integrated quench box, mixing zone, and distributor tray, resulting in a compact system with higher thermal efficiency over a wider range of operating conditions than the previous generation. The thermal efficiency, which is defined as the ΔT of the outlet fluid divided by the ΔT of the inlet fluids, was significantly higher for the new Hy-Quench-XM design compared to the previous generation, achieving greater than 99% efficiency even at low velocities. The new design offers a more compact quench zone, allowing an increase of up to 32in of catalyst loading volume for larger (15ft) reactors. For a recent revamped hydrocracker for a Far Eastern plant, the space reduction of each quench box resulted in a % increase in catalyst volume for both the pretreat and cracking catalyst sections on a reactor of 15ft 6in diameter. This substantial gain in volume is critical for revamping CFHT to MHC. Refinery-wide impact: a case study In response to market conditions, Axens performed a case study to quantify the economic and operational incentive to revamping a typical CFHT unit. Axens used its database of operating CFHT units in North America to define a typi- amma-ray tomographic cross-section comparison of tray technologies Classic chimney tray distribution (Figure 4) Figure 4 Classic chimney tray distribution Radial T / Axial T, % 9 8 7 6 5 4 3 2 cal CFHT unit that would be a candidate for MHC service, which suggested that the average North American CFHT operates at 15 psig with a space velocity of 1. hr -1. The average cycle length was slightly under three years with a sulphur target of 2 ppm. Many of the units process a percentage of cracked gas oils, either from residue hydrocrackers or cokers (average of 25% cracked feedstocks). A refinery-wide case study was EquiFlow EquiFlow uniform distribution (Figure 5) Figure 5 Equiflow uniform distribution 2 3 4 5 6 7 8 Days on stream Figure 6 Hydrocracker temperature distribution 1-efficiency (1-η) 1.1 Figure 7 Hy-Quench-XM thermal efficiency Velocity Level 1 Level 2 Level 3 Level 4 22 2 18 16 14 12 8 6 4 Current New Axial T, ºC performed in which the study refinery processes a 3/3/4 blend of WTI/WTS/WCS crudes for a total crude throughput of 128 b/d (blend API = 23.3, S = 2.6 wt%, N =.23 wt%). The simplified refinery configuration is shown in Figure 8. The gasoline pool consisted of treated straight run naphtha, reformate, alkylate, and post-treated FCC gasoline. The diesel pool came entirely from the diesel hydrotreater stripper bottoms. For simplicity, the www.eptq.com Revamps 216 11

CFHT feed properties 128 NAPH 24 DSL 26 CDU/VDU VO 4 NHT 5 CFHT CFHT DHT FCC P+ CCR Alkylation D Property CFHT feed Flow rate, BPSD 5 % Coker O 2 ravity, API 19.1 S, wt% 3. N, ppm 2 CCR, wt%.4 Aromatics, wt% 55 Feed diesel (65 F-), vol% 12 Distillation, vol% IBP 465 3 718 5 793 7 871 EBP 72 Table 3 VR 38 Delayed coker Numbers in red XX = MBPSD Base CFHT operating conditions Figure 8 Base refinery configuration refinery produced a full range diesel without any jet product. Modelling of the crude and delayed coker units resulted in a 5 b/d CFHT unit with operating conditions summarised in Table 3. A high nitrogen feed was selected to show a more conservative case, as nitrogen severely inhibits the hydrocracking activity of the catalyst acid sites. The operating conditions chosen for the CFHT based on the operating unit database are summarised in Table 4. The base case unit was Max diesel mode Max gasoline mode Temp and operating temperatures (48 F+, 25 C+ FCC feed). Case 2 The revamp case: the unit was revamped to a flexible MHC and operated in either maximum diesel (conversion) mode or maximum gasoline mode, according to seasonal demands. A robust product fractionator was added in order to allow the refiner to swing the FCC feed cut-point: Maximum diesel (conversion) mode: the MHC maximised VO conversion with high selectivity towards middle distillate production. The fractionator ensured recovery of a full-range diesel product from the CFHT, which resulted in an FCC feed IBP of 65 F (34 C). Minimum conversion (maximum gasoline) mode: the catalyst load remained the same as for the maximum conversion mode; however, reactor severity was reduced to minimise VO conversion. Cut-points were adjusted in the new fractionator to maximise FCC feed (48 F+, 25 C+ FCC feed). Case 3 Same as the Case 2 maxidesigned with a product steam stripper and no product fractionator. Consequently, the cut-point of the gas oil product was 48 F+ (25 C+), and heavy diesel was cracked to light gasoline and LP in the FCC unit. A refinery-wide evaluation including modelling all of the process units was then conducted for three CFHT operating cases (see Figure 9): Case 1 The base case: current operation of the typical CFHT unit maximising HDS and minimising conversion via catalyst selection Cut Yield FCC IBP: 65ºF D CFHT Fractionator FCC MHC Figure 9 Study cases Conversion Cut-point FCC impact Temp FCC IBP: 48ºF Cut Fractionator D FCC LP Yield LP Property Operating conditions Reactor Inlet Pressure, psig 15 LHSV (overall), hr -1 1. Recycle gas ratio, SCFB 2 Cycle length, months 36 as oil product S, ppm 2 as oil product N, ppm 7 FCC feed TBP cut 48 F+ Table 4 12 Revamps 216 www.eptq.com

mum diesel operation with the addition of external VO being purchased to fully utilise the existing FCC and alkylation units. The external VO was assumed to be originating from a light, sweet crude such as WTI or Bakken (API=27.4, S=.4 wt%). Table 5 summarises the results of the case study. Standard FCC gasoline and LCO cut-points were used and remained consistent for all cases. It is also important to note that the reformer severity was varied from case to case in order to achieve a constant gasoline pool octane (RON = 92). An economic analysis was conducted based on these cases. The basis for product pricing was EIA data on 214 monthly average refinery pricing for US ULSD and regular gasoline. 6 The analysis accounted for the increased hydrogen and utility costs associated with the revamped operation and adjusted operating conditions on a monthly basis to maximise either gasoline or diesel, based on the most advantageous pricing. VO was assumed available at a 2% premium to WTI. Table 6 summarises the results and assumes the construction of a new product fractionator. The analysis demonstrates the quick payback that can be achieved as a result of unit flexibility even when moderate capital investment is required. Figure compares the relative composition of the transportation fuel pool for each case. The results of the study confirm the increase in profitability provided by unit flexibility. In particular, the extremely high selectivity of Axens MHC catalyst increases middle distillate production without leaving the refiner with the problem of processing large amounts of low octane hydrocracked naphtha. This, coupled with high HDS, HDN and HDA activity of the hydrocracking pretreatment catalyst at low severity, allows the refiner to maximise gasoline production without sacrificing FCC feed quality. Finally, a well-designed product fractionator allows for large adjustments in the gas oil product IBP to swing between FCC Property Case 1 Case 2 Case 2 Case 3 Base Case Max. Diesel Max. asoline Max. Diesel Revamp performed? No Yes Yes Yes Reactor severity Base High Low High MHC VO conversion, wt% % 27% 13% 27% MHC product yields Naphtha, vol% 1 4 1 4 Diesel, vol% 36 11 36 FCC feed, vol% 9 63 89 63 Purchased VO, BPSD - - - 13 228 FCC operation Feed rate, BPSD 44 934 31 76 44 681 44 934 Feed S, ppm 2 6 19 16 Feed SPR.911.879.94.882 FCC feed IBP, F 48 F+ 65 F+ 48 F+ 65 F+ FCC gasoline yield, v% Base +5.4 +1.6 +4.2 Refinery trans. fuel yield, v%feed 92.9 93.9 93.1 95.7 Refinery gasoline prod., BPSD 7 2-84 +57 +84 Refinery diesel prod., BPSD 49 169 +9656-32 +11 134 Refinery D/ ratio.7.95.69.8 Table 5 gasoline production and MHC diesel production as needed. The ULSD challenge: integrated hydrotreating and HyC- While revamping the CFHT to a MHC provides an excellent increase in middle distillate production, it is Case study results Revamp economics Case 2 Case 3 Revamp result Revamp result Annual revenue increase, $ MM/yr 45 354 Operating costs increase, $ MM/yr 7 257 Profit increase, $ MM/yr 38 97 Revamp capex, $MM 3 3 Simple payout, months < <4 Table 6 Production, BP/D 16 14 12 8 6 4 2 ULSD SR gasoline 49 169 7581 15 266 17 634 29 719 Figure Case study pool composition not a solve-all solution. When moderate pressure and limited space velocity of an existing unit do not permit the production of ULSD directly, the MHC diesel sulphur is typically in the range of 3 ppm to as high as 2 ppm. More severe cetane specifications (CARB or Alkylate Reformate 119 369 12 618 119 621 58 825 7581 12 898 18 651 22 664 48 849 7581 15 168 17 752 3 27 FCC gasoline Total pool 135 696 6 33 7581 17 525 18 668 31 62 Base Max diesel Max gasoline VO import www.eptq.com Revamps 216 13

VO CO DAO H 2 LCO LCO Figure 11 HyC- process flow scheme Feed sulphur Feed density Euro diesel, for example) can also limit the ability of the MHC to produce ULSD directly. In these Polishing section Fuel gas naphtha FCC feed ppm ULSD 5. 4.5 4. 3.5 3. 2.5 2. 1.5 1..5 2 3 4 5 6 7 8 9 Days on stream Before revamp.96 Current.95.94.93.92.91.9 2 3 4 5 6 7 8 9 Days on stream Figure 12 Commercial results cases, the refiner has classically had three options: 1. Treat the diesel in existing hydrotreating units to produce ULSD 2. Sell the low sulphur product to the marine gas oil or export markets 3. Blend the off-spec diesel into the ULSD pool, depending on the quality of the other pool sources. For refiners with access to export channels, producing ULSD is not critical to profitability. Many of the Latin American countries with the highest diesel import demands have specifications that can be achievable in revamped MHCs with higher space velocities. For refiners who must produce ULSD and are constrained by existing diesel hydrotreating capacity, Axens has developed and commercialised its HyC- process, which meets the ULSD challenge via the integration of a polishing reactor into the existing MHC unit. In HyC-, the MHC diesel polishing section receives the entire make-up H 2 required for both the polishing and MHC sections and is operated in a once-through configuration. The patented, integrated flow scheme requires less equipment than a new diesel hydrotreater unit, improving project payout times. In most cases, the existing make-up compressors can be revamped for the new scheme. Other advantages of the scheme include: Reduced polishing reactor size as a result of low vapour flow rates (once-through H 2 ) High hydrogen partial pressure, resulting in maximum aromatics saturation (cetane improvement), producing high quality diesel blendstock. This option provides a long term solution compared to cetane additives Improved utility consumption as a result of heat integration between the two sections. Because the polishing section is operated at high pressure, it can be used to co-process other difficult refining streams such as LCO or coker gas oils (HyC-+ process) for maximum cetane improvement. Operating units processing cracked feedstocks currently produce Euro V ULSD (cetane number >5). The HyC- concept has been 14 Revamps 216 www.eptq.com

commercially demonstrated for over years. Commercial case study One such HyC- unit was constructed in 25 at a Southern European refinery. The initial design of the unit specified loading % hydrotreating catalyst in the MHC portion of the unit, which treated straight VO and aromatic extract to achieve VO conversion of approximately 15 wt%. The integrated polishing section processed diesel from the MHC section, LCO, straight run heavy diesel, and visbreaker naphtha to produce high cetane, Euro V ULSD. In response to the shifting marketplace, the refiner worked with Axens to revamp the unit in 2 to maximise conversion to middle distillates and reduce FCC feed sulphur, while maintaining the unit cycle length of at least 31 months (see Figure 12). Axens developed the project in three phases: a safety study, a process study including catalyst pilot testing, and an engineering design package. The revamped unit started up in 213 and, despite a lower API and higher sulphur VO feedstock, the unit achieved 3-4 wt% VO conversion with extremely high selectivity to middle distillate (:1 cracked diesel to naphtha ratio by volume) via the use of amorphous HDK 786 MHC catalyst protected by HR 54 CoMoNi pre-treatment catalyst. Conversion was adjusted throughout the run based on the refiner s changing economic considerations. When diesel production was maximised, conversion was increased above 4% (see Figure 13). In short, the revamp was able to achieve the refiner s objective of increasing VO conversion while also increasing the cycle length to 36 months, as a result of the catalyst improvements. Euro V diesel production was continuous throughout the run. Summary Of the numerous options refiners have to achieve flexibility in their diesel to gasoline product ratio, revamping the CFHT to a flexible Net conversion, wt% 45 4 35 3 25 2 15 5 +15 to 2 pts 2 3 4 5 6 7 8 9 Days on stream Figure 13 Conversion increase and catalyst loading at MOH MHC can in many cases be the most attractive balance between rate of return, capex, and project implementation schedule. Increasing conversion of VO without the need for a costly new unit not only increases diesel production but also improves FCC performance, allowing the refiner the option to process excess external VO in the FCC, increase LP production, and overall increase hydrogen addition (volume swell), thus improving overall refinery economics. Since the unit conversion and product cut-points can be adjusted on the fly to meet market demands, a flexible MHC revamp project provides the refiner with a flexible and profitable option independent of unpredictable diesel and gasoline market forces. The addition of an integrated diesel polishing section to the MHC unit is an attractive solution for refiners short on middle distillate hydrotreating capacity. References 1 US Energy Information Agency, Weekly Refinery Net Production data for Diesel and Kerosene, www.eia.gov/dnav/pet/pet_pnp_ wprodr_s1_w.htm (accessed 11 Jul 216) 2 US Energy Information Agency, Distillate Fuel Oil Exports by Destination, www.eia.gov/ dnav/pet/pet_move_expc_a_epdxl_eex_ mbbl_m (accessed 11 Jul 216) 3 Axens projection. 4 Sotolongo K, USA: Refiner to the World, FUEL, Jun 215, 26-33. 5 Wisdom L, Bonnardot J, Dorbon M, Paouri E, Before revamp Current Reactor loading High activity CoMoNi HR 54 Cracking catalyst HDK 786 Axens 215 AFPM paper (AM-15-23), Extend Your Hydrocracking Operating Envelope without Compromising Safety or Yields. 6 US Energy Information Agency, U.S. Refiner Petroleum Product Prices, www.eia.gov/dnav/ pet/pet_pri_refoth_dcu_nus_m.htm (accessed 2 Jan 216) David Schwalje is a Business Development Manager with Axens North America, responsible primarily for bottom of the barrel technologies including residue and distillate hydroprocessing and hydrocracking. He joined Axens in 24 as a process engineer and project manager in the areas of VO hydrotreating, hydrocracking, lubes, and mid-distillate hydrotreating prior to joining the Axens Technology roup. He holds a BS in chemical engineering from the University of Delaware and is a licensed Professional Engineering in the State of New Jersey. Larry Wisdom is a Senior Executive with Axens in charge of marketing the heavy ends technologies in North America. During his 4- year career he has co-authored more than 35 papers on heavy oil upgrading and awarded two patents. Prior to joining Axens, he worked for Hydrocarbon Research Inc. and FMC Corporation and graduated from the University of Kansas with a BS in chemical engineering and a MBA in marketing. Mike Craig is a Senior Hydroprocessing Technologist for Axens North America and is responsible for catalyst and licensing activities involving distillate hydrotreating, gas oil hydrotreating and hydrocracking technologies. He has over 2 years of experience in the hydrocarbon processing industry including process design, R&D, licensing and technical assistance and holds a BS in chemical engineering from The Cooper Union. www.eptq.com Revamps 216 15