Recent increases in hydraulic

Novel processing ideas for a condensate refinery Synergised unit flow schemes for a condensate refinery aim for high value refined products DONALD EIZENGA, DAVID SHECTERLE and FRANK ZHU UOP, a Honeywell Company Recent increases in hydraulic fracturing operations have produced a significant amount of condensate liquids, particularly in the US. Conversion of these light liquids into finished products for sale at maximum value is being considered via new units or expansion of existing facilities. The quality of many condensates is such that minimal hydrotreating is required, but significant upgrading of the paraffinic is required to meet gasoline specifications. A condensate refinery is therefore expected to include reforming and isomerisation of along with and diesel hydrotreating. In this article, novel synergised unit flowscheme solutions will be discussed, which consists of three key ideas. Idea The goal of the alternate condensate fractionation unit (CFU) design was to reduce equipment count and save capital and/or to reduce energy consumption for opex savings, especially when upstream of combined and diesel hydrotreating (CHT). Typically and diesel are rigorously separated in the CFU, but they would need to be combined and then re-separated for a CHT unit approach, resulting in redundant fractionation. Innovative designs were considered to overcome the penalty of redundant fractionation. Options to either eliminate a column or to adjust the column operating conditions and draws were considered. Energy benefits were found by optimisation of the columns, and capital cost savings were identified. The energy benefits were not sufficient to overcome the cost of redundant fractionation with a combined hydrotreating approach. The lowest operating costs were achieved when the hydrotreating units are individually optimised and a three-column CFU design is employed, while alternate approaches may provide modest capital cost benefits. Idea The goal of the combined hydrotreating design was to minimise equipment count and thereby optimise the refinery economics. In the final analysis, the combination of and diesel in a single hydrotreating unit was found to have limited economic justification. Cost penalties were associated with reactor section pressure, catalyst volumes, and redundant fractionation of. Efficient combinations of shared auxiliary systems and novel combinations were proposed so that each unit could be individually optimised. Idea UOP explored novel solutions for using isomerisation with the UOP Platforming process to determine an economic optimum configuration for gasoline production in the proposed design. A number of options and synergies were evaluated to reduce capital costs and operating costs. Significant capital cost savings were achieved using innovative complex schemes relative to the base case separated unit configuration. A higher yield case was found to have highest net pres- www.eptq.com Processing Shale Feedstocks 06

Current shale plays Prospective shale plays Basins Stacked plays Shallowest/youngest Mid-depth/mid-age Deepest/oldest Figure US based oil and gas shale plays ent value (NPV), while a lower capital approach was evaluated, which may make sense when a customer is severely capital constrained. The market seeks the technology with minimal capital investment as installation in remote locations could be a significant factor. UOP proposes to handle these issues with construction of modular units, which have been optimised for capital and operating expenses in consideration of the feed properties and product quality targets. While the various technologies used for hydraulic fracturing of shale to produce natural gas have developed over the last 40 years, the rapidly escalating energy prices through 04 led to a boom in use of the technology, particularly in the United States. Technology advances significantly reduced the cost for obtaining gas and light condensate oils, which has sustained some production even as crude prices have fallen in 05. Projections suggest that extraction of such liquids may increase further and they will likely be processed for many years to come. US shale deposits are distributed over a wide range of the country in a number of formations (see Figure ). In the process of extracting gas from the wells, condensates and natural gas liquids (NGLs) are also extracted. These liquids include a full range of hydrocarbon constituents from propane to heavy residues that boil at and above 000 F (540 C). The liquids are therefore similar to other petroleum crudes, but are lighter and generally sweet. For example, crudes are often differentiated by API gravity, with heavy crudes averaging ~5 API, medium at ~5 API, light at ~5 API, and extra light at ~45 API. s average 55 API, but there is a wide range of compositions depending on the source, and the composition may change as the field matures. The question for gas producers, midstream companies, and refiners is how best to process these liquids. Certainly, adjustment to refinery operation is required, and possibly revamp of existing equipment and/or addition of new process units (for instance, via a modular expansion). Alternatively, new processing facilities could be developed for local markets, which may be placed near the feed sources in remote locations. properties In order to understand how best to process condensate liquids, a good understanding of the properties is required. One key attractive property in the current market environment is that they are relatively low cost and sold at a discount to crudes, in part owing to the major recent increases in fracking. The other key attractive property is that they are light liquids, which makes them ideal for refining into higher margin gasoline and diesel products. Recent estimates have predicted dramatic growth and sustained high levels of production of condensate; however, this has been impacted by the high volatility in crude oil market pricing. Legislation on the sale of crude oil and lightly processed condensates from the US is also having further impacts, and in 05 the US began trading light crude oil for heavy Mexican crudes to the mutual advantage of refiners in both countries. Processing Shale Feedstocks 06 www.eptq.com

s may contain up to 0%, and are typically high in both light and heavy. In fact, a survey of over 00 worldwide condensates showed that + light + heavy ranged from 0% up to over 90%, with a median near 60%. The remaining material is mostly in the distillate range, suitable for jet fuel and/or diesel fuel with appropriate processing. s are mostly quite low in the heavy residue range (boiling above 685 F, 60 C). For the study, a light case and heavy case feed were considered for a 5000 b/d condensate refinery (see Table ). The API gravity was 59.4 for the light case and 5.0 for the heavy case. Sulphur was low at 0-0 ppm compared with levels of up to % in more contaminated condensates and conventional crudes. Target products for the study included off-gas, which could be used as fuel gas for the complex,, gasoline (or blending component) with an octane target of 87, ultra-low sulphur diesel, and a low sulphur residue (for refinery FCC or hydrocracking feedstock or fuel oil). Ideas for condensate processing For the base case of the study, a small hydroskimming style condensate refinery was proposed (see Figure ). The objective of the study was to analyse opportunities that might optimise the overall complex. The original premise of the study was that reduction of equipment count and minimisation of capital would be the most important objective. Three parts of the study Study basis: processing objectives Feed Light case Heavy case, b/d 5 000 5 000 API/SG 59.4/0.74 5.0/0.77 Sulphur, ppm Products, BPSD Off-gas, MMBTU/h 7 5 Total 800 000 Gasoline (R+M)/ = 87 600 0 00 ULSD 700 9600 Low sulphur residue 000 000 Table will be summarised: Idea. Efficient CFU design Idea. Combined hydrotreating design Idea. Innovative complex schemes. Idea : efficient condensate CFU design An early assumption in the project was that equipment count and capital could be minimised by combining the and diesel hydrotreating units into a CHT unit. For commercial designs where traditional separate hydrotreating is employed, the CFU has feed CFU been designed with a three-column approach that achieves a rigorous split between process streams. For a CHT approach, rigorous separation of from diesel is not required and overall economics would be penalised by the redundant fractionation illustrated in Figure. The first question posed when reviewing the CFU for the Refinery was: Can we use just one column like a typical Crude unit? Various options were then explored with a target of optimising the design for expected use of downstream combined hydrotreating (CHT, discussed below) and to minimise capital and/or operating costs in general. The questions then became:. What is the best condensate fractionation unit (CFU) configuration?. Can costs be reduced if combined hydrotreating is used? Upon further examination of the crude column approach NHT DHT Figure refinery: a big picture Isom. Reforming To gasoline product Residue www.eptq.com Processing Shale Feedstocks 06

Light Heavy Stabiliser to splitter fractionator hydrotreating Combined hydrotreating to splitter Fractionator Commercialised configuration hydrotreating fractionator Diesel product Traditional hydrotreating Atmospheric residue Proposed reduced equipment hydrotreating configuration Diesel product Figure Commercial CFU and redundant CHT fractionation shown in the first schematic of Figure 4, issues quickly became apparent. While the crude unit is typically a single column, the overhead gas ( at low pressure) is typically compressed and then processed in a gas concentration unit. This requires a high capital off-gas compressor. Higher crude column pressure can be used to keep the overhead as totally condensing, but there are constraints on the column, whereby fouling is expected at the bottom of the column when high pressure leads to bottoms temperatures above about 75 F (85 C), and higher pressure is less efficient for diesel fractionation from residue. The study did show that use of steam stripping and adjustments to the column pressure could handle a higher and lower residue feedstock. Sufficient in the overheads of the column keeps the as liquid at moderate pressure (<50 psig), but then a second higher pressure (>0 psig) stabiliser column is required to separate the product. Processing condensate in a CHT unit carries a significant economic penalty with no product quality benefit. Surprisingly, the energy input requirements remained high for this crude column approach even when rigorous separation of and distillate range streams was not required. The heat exchanger network was more constrained, so capital savings were not substantial. The three-column CFU design approach (see Figure ) avoids the use of an overhead compressor. A separate distillate fractionator column allowed for higher pressure in the first column and lower pressure in the second heavy oil column to effectively avoid fouling. However, capital cost and operating costs were high to meet on-spec product cuts with rigorous separation that was not optimised for a downstream CHT. Two alternative configurations with relaxation of design targets for the product cuts were then developed, shown in the second two schematics in Figure 4. The configuration optimisation study included adjustment of the column pressures, product draws, minimisation of reflux rates, and optimised heat exchanger networks. Higher pressures where is processed along with avoids the use of compression, while adjustments to draw rates and locations shifted the compositions to avoid high temperature fouling. Product draws and locations along with a diesel pumparound are also used to optimise the process heat exchanger network. Lower pressure distillate fractionation with steam stripping maximises recovery while minimising the energy requirement. Minimisation of reflux rates shrinks the columns and energy, but this approach is only applicable for a down- 4 Processing Shale Feedstocks 06 www.eptq.com

A Crude column Stabiliser Stabiliser Atmospheric residue B C Light Stabiliser splitter Light Heavy stabiliser Light and heavy Fractionator Heavy distillate Fractionator Heavy distillate Atmospheric residue Atmospheric residue Figure 4 Alternative CFU: crude, three-column and two-column configurations stream CHT where rigorous splits are not required as and distillates are co-processed. Optimisation of the configuration did result in minimisation of energy input and capital reduction. One key capital reduction methodology proposed is to use a hot oil system, which minimises the number and size of the high cost fired heaters in the entire condensate refinery complex. Similarly, the use of a steam system can allow for stripping, which improves product recovery and may reduce or eliminate heater services. These are parts of an overall complex energy required by the CHT fractionator column (relative to separate hydrotreating units) was not yet resolved. Fuel gas savings were achievable in the CFU, and the alternative three-column and especially the alternative two-column approach showed reduced capital. It was determined that the crude column approach was not in fact optimised and did not provide any heat advantage. The three-column approach was good for high /low residue condensate processing, with potential minimum capital using an adjusted two-column configuoutside battery limits (OSBL) optimisation, and the adjustment of draw rates and locations in the CFU allowed use of hot oil exchangers rather than fired heaters in the alternate configurations. A summary of the CFU column analysis is shown in Table. It is interesting to note that optimisation within the CFU leads to lower temperature and distillate, which would require more energy in the CHT to reheat, and the question of whether the capital and energy savings in the CFU could compensate for the additional capital and www.eptq.com Processing Shale Feedstocks 06 5

Parameter Commercial Crude column Alt column Alt column Column pressure, psig 5-75 5-50 5-50 >0 Column pressure, psig <5 --- <5 <5 Column pressure, psig >0 >0 >0 --- Column stages Base 67% 67% 67% Exchanger duty Base 00% 7% 7% Heat required Base 0% 87% 87% Equipment capital Base 95% 90% 80% Table ration. To determine a clear winner, the entire complex must be considered. Once the overall analysis of CFU and CHT including redundant fractionation was considered, it was found that the energy savings in the CFU were offset by the higher energy costs of the downstream CHT. CFU options comparison Idea : combined hydrotreating design For the relatively easy hydrotreating requirements within the small condensate refinery study, it was initially proposed that the objective should be to minimise equipment by combining the feeds into a CHT unit (see Figure ). The expectation was that significant cost savings could be achieved. The high level concerns about this approach were whether the was being over-processed, and also whether the cost of redundant fractionation of the from the distillate feeds could be overcome. Additionally, it was questioned what the best approach might be within the diesel hydrotreating (DHT) unit to optimise the separated configuration. In the case of separate hydrotreating, each unit has optimised catalyst selection, pressure conditions, gas circulation, and fractionation to meet the product objectives. The and distillate products require only stripping to remove H S and light ends, which can be accomplished with minimal capital and operating expense. Interestingly, it was found that, even for small units, sometimes the addition of equipment such as a hot separator or additional diesel drying equipment can be more cost effective. While combined hydrotreating can significantly reduce equipment count (by about 0%), full fractionation is required to cleanly separate the diesel product from the. This fractionation requires a relatively high cost column and fired heater. The is now being processed at the more severe conditions required for ULSD production, and especially in the case of condensates where quantity is higher than distillates, this is a significant processing penalty. Parameter Separated Combined NHT & DHT hydrotreating Equipment capital Base 0% Installed capital Base 90% Heat required Base 44% Annual operating expenses Base 80% Table CHT options comparison Estimated capital costs for the approaches were found to be similar (see Table ). The combined approach had slightly higher equipment costs, but with lower equipment count the erected equipment cost estimate and plot area required was slightly lower. This was a surprising result, explained by the fact that savings in reducing a number of pieces of equipment were offset by higher costs for the larger, higher pressure reactor section exchangers, reactors, and the fired heater reboiler (more expensive equipment). The main penalty with the CHT approach is the additional heater duty and operating cost associated with the required /diesel fractionation. In combination with an optimised CFU approach, the CHT approach showed a marginal NPV advantage with lower capital but higher operating expenses. Equipment installation factor estimates and fuel costs have a strong effect on the conclusion, and they are somewhat speculative for study level estimates. In this case, a more traditional, commercialised route could be chosen to minimise the project risk and fuel costs, or the more novel approach could be taken to minimise capital cost. In cases where the processing requirement is more severe, or the ratio of to diesel was lower, it is expected that a CHT approach could show greater benefits. The properties of the condensate feeds must be well understood, and the optimisation approach must consider the overall complex rather than individual units. 6 Processing Shale Feedstocks 06 www.eptq.com

In the separated unit case, each individual unit could be better optimised, and shared auxiliary systems and more limited integration is also possible. One such novel approach is the combination of stripper services to reduce column overhead equipment shown in Figure 5. Idea : innovative complex schemes Due to the paraffinic nature of the study feed, both catalytic reforming and isomerisation processes are required to meet the gasoline octane target (87 R+M / ). A variety of isomerisation unit technologies and configurations is available, depending on the octane level required and any other limitations such as gasoline pool aromatic and/or benzene content. Isomerisation processes can be used with the UOP Platforming process unit. For typical refinery separated unit configurations, each unit may have a reasonable design which is best for operability and maintenance, but this may be a high capital approach. On the other hand, novel synergised solutions may be used to optimise the refinery economics by integrating the units, reducing equipment, and adjusting the process operating conditions (see Figure 6). Both product value and equipment capital costs need to be considered (see Figure 7). For example, a low capex approach may eliminate equipment, but result in lower yields of gasoline, higher production, or even going off-specification for certain feeds. On the other hand, a high yield approach may also DHT separator liquids H MUG feed NHT Hot reflux stripper Diesel product to drying Figure 5 Novel combined stripping approach Figure 6 Illustrative complexes C 5 + volume, LV% 0 4 5 High yield Figure 7 complex trade-offs be able to eliminate some equipment as well but require more frequent catalyst regeneration. In the system studied H net gas Light Heavy Operating condition Isom. Reforming C 5 + yield yield Low capital to processing separator liquid To gasoline H net gas 0 for the condensate refinery, it was found that the high yield approach had a payback of less than two years for the 8 6 4 0 volume, LV% www.eptq.com Processing Shale Feedstocks 06 7

additional capital costs relative to a lower capital approach. UOP s capabilities in kinetic modelling of catalyst performance and process simulation of a variety of flowscheme options were used to optimise the unit. An economic analysis considering capital and operating expenses, yields, and cycle length impact was carried out for several solutions. Significant capital cost savings were achieved using innovative complex schemes relative to the base case separated unit configuration. A higher yield case was found to have the highest NPV, while a lower capital approach was evaluated, which may make sense when a customer is severely capital constrained. Even for a small complex where equipment minimisation appeared to be a sound approach from a high level, catalyst and process conditions have a significant impact on project economics. Conclusion In the construction of new plants, two choices are generally available: stick built construction and modular construction. There are a number of reasons why modular may be preferred for the study case, including: 5 000 b/d capacity is ideal for modular scale Higher levels of quality control can be achieved in the design Shop fabrication and inspection lead to higher quality and improved schedule Modular is more suitable for remote locations where skilled labour is limited Significant schedule reductions and logistics savings are realised Projects can be executed with a fixed lump sum price. UOP and UOP Russell have significant experience in the design, construction, and delivery of modular units. This includes natural gas liquids separation units, isomerisation units, Merox units, and catalytic process units such as NHT and reforming. The recent expansion of shale fracking for gas and generation of condensate feedstocks has been a great opportunity and a challenge to produce high value refined products. In order to capture the value, a good understanding of the feed and processing technologies is required. As illustrated, optimising is not the same as minimising, and the effort should consider the whole effected complex. The use of innovative processing approaches and potential for condensate refineries represent value added opportunities. Use of a modular approach can bring significant advantages in project execution. Platforming is a trademark of UOP. Acknowledgement The authors would like to express sincere thanks to the following Honeywell UOP personnel who participated in the work described here: Joshua Kaye, Ron Long, Charlie Van Zile, Dick Hoehn, Dave Bachmann, Paul Steacy, Mark Benson, and Rich Russeff. Reference Photo: American Petroleum Institute; Patrick C Miller, API s Energy from Shale campaign features Bakken, thebakken. com, Jan 05. Donald A Eizenga is a Principal Optimization Engineer with Honeywell UOP. He has 9 years at UOP in R&D, field operating services, and engineering design and holds a BS in chemical engineering from Michigan Technological University. Email: Don.Eizenga@honeywell.com David J Shecterle is a Fellow and Isomerization Technology Specialist with Honeywell UOP. He has 4 years at UOP and holds over 0 patents and a BS in chemical engineering from the University of Wisconsin. Email: David.Shecterle@honeywell.com Frank (Xin) Zhu is a Senior Fellow and Leader of the Engineering Innovation Group with Honeywell UOP. Before joining UOP, he was a Research Professor at the Centre for Process Integration, University of Manchester, UK. He is the author of Energy and Process Optimization for the Process Industries by Wiley in 04. Email: Frank.Zhu@honeywell.com 8 Processing Shale Feedstocks 06 www.eptq.com