Recently, I received a question from a practicing chemical engineer. It was interesting, because it focused on one of the basic decisions involved in a new refinery: "What is the best oil refining process pattern if we want to produce a maximum amount of gasoline when refining a heavy crude oil." We all know that crude distillation is the first major process step that happens when crude is introduced into the refining process, but how do you decide what comes next?The various licensing companies (UOP, IFP, Exxon Research, etc.) as well as most of the engineering and construction companies and many engineering consultants all have their own ideas, which (naturally) feature their own processing strengths and prejudices, and tend to ignore their weaknesses. All of them also have one or more computer programs that are supposed to determine the optimum refinery configuration for a given crude oil of combination or crude oils. They are even (supposedly) set to give the optimum return on investment using built-in investment costs that usually reflect the most recent projects attempted.
What you can do with the crude naturally depends upon what sort of crude you have. Responding to the question above, I assumed that it is not only heavy, but also contains a medium amount of sulphur (2-3%) and middle- to high-level of metals (particularly nickel and vanadium), and not an excessive level of organic nitrogen. In other words, nothing very exotic.
Gasoline Refinery
A more-or-less traditional flow scheme from the USA (which is where the gasoline refinery concept was born) for a medium to large gasoline refinery is shown below. This ignores the modifications necessary today to produce specialised gasoline.
Atmospheric distillation block
- Atmospheric distillation
- Full range naphtha HT => splitter => heavy naphtha to catalytic reforming and light naphtha directly or via C5/C6 isomerisation to gasoline
- Light distillate (kerosene) HT à burning kerosene and/or jet fuel
- Mid-distillate (gasoil) HT à Diesel fuel and light heating oil
Vacuum distillation block
- Vacuum distillation
- Delayed coker (a large one, so that essentially no vacuum residue remains)
Conversion block
- FCCU – feeding medium VGO and heavy VGO with a feed hydrotreater to reduce the sulphur and nitrogen contents
- The C3/C4 streams go to alkylation, catalytic polymerisation, and MTBE (except in places where MTBE is not allowed in gasoline)
- An AGO/LVGO hydrocracker (depends on the needs of the gasoline pool and the cost of hydrogen production) as a combined FCCU feed HT and a gasoline component producer (the economics should decide whether this is necessary or not)
Hydrogen production
- There will be a large volume of refinery gas and C3-C4's coming from the cracking units; these can be used for hydrogen production (steam-HC reforming), perhaps balancing gasoline component production against hydrogen requirements
Another option, if there is not enough light material after fuel and LPG product needs for steam-HC reforming, is a partial oxidation unit, using vacuum residue as the feedstock (this is effective but capital-intensive) instead of or in addition to the steam-HC reformer, and partially replacing the coker
Other process and utility units
- Steam production for various usages, including utility, process, and power generation
- Sulphur recovery – there will be large amounts of sulphur, produced as H2S, that will have to be converted to sulphur for safe handling
- Coke handling – if the refinery coke is not burned in the power plant, it will need to be handled for export from the refinery – one option is to calcine the coke to make it a more valuable product
What you will note above is that virtually all products are hydrotreated at some stage of the process. This is to make the distillate fuels (gasoline, kerosene for jet or burning fuel, gasoil for heating oil and Diesel fuel) compatible with today's worldwide low sulphur environmental requirements. The product slate should be roughly 55-65% gasoline by weight, 25-15% light distillates by weight, with the remaining ±20% going to coke, sulphur, maybe propane/butane, and own consumption (including catalytic coke produced and consumed in the FCCU). The volume yield of gasoline can be almost 100% on crude, depending on specifications and the degree of cracking utilised.
If the refinery is to be a (relatively) small one, then the best solution may be to build the entire atmospheric distillation train, but only the vacuum distillation, the delayed coker, and the hydrocracker in the vacuum distillation train as a high-conversion unit (primarily gasoline components), and leave out the FCCU and its downstream units. The hydrogen production facilities will still be necessary unless H2 is available from outside the refinery. In the future, if the refinery is expanded, an FCCU plus its downstream units can be added, and the hydrocracker can be converted to an FCC feed pretreater with lower conversion (50-60 %) but with a much greater throughput (1,5 –2 times original).
Distillate fuel refinery
But what to do if you want mostly distillate fuels instead of gasoline? Well, this is the "European solution" as opposed to the "American solution" for gasoline above. Typically, the distillate solution (high cetane number Diesel fuel, light heating oil, kerosene for household fuel and aviation turbine fuel) uses different choices of conversion equipment, in order to provide more middle distillates and less gasoline.
The earlier solution (1960's) was to use low severity FCC units, which produced more than 50% light cycle oil instead of a large volume of gasoline. Today, the light cycle oil is a problem because of it's aromatic and sulphur contents, not to mention stability.
The more current solution is a hydrocracker using (typically) one conversion stage to produce large amounts of desulphurised, low aromatic distillates (kerosene and Diesel fuel). Depending on catalyst type and feedstocks, the volumetric yield of distillates can be over 100%. Conversion of the vacuum residue to hydrogen and energy (steam and/or electricity) may take on a more important aspect, since the large hydrocracking process requires a lot of hydrogen. However, the steam reforming straight run naphtha to hydrogen is another H2 source, if gasoline is really unimportant. The refinery is somewhat less complicated than a gasoline refinery of the same size. The blocks below give an idea of what might be used in a distillate fuels refinery. The common components that remain more or less unchanged are shown in bold print.
Atmospheric distillation block
- Atmospheric distillation
- Medium + light naphtha HT à medium range naphtha to catalytic reforming and light naphtha directly or via C5/C6 isomerisation to gasoline
- Heavy naphtha / light distillate (kerosene) HT à burning kerosene and/or jet fuel
- Mid-distillate (gasoil) HT à Diesel fuel and light heating oil
Vacuum distillation block
- Vacuum distillation
- Delayed coker (a large one, so that essentially no vacuum residue remains)
Conversion block
- AGO/VGO hydrocracker optimised for distillate production
- HC medium naphtha to catalytic reforming (hydrogen production and gasoline octane), HC light naphtha to gasoline blending, naphtha sales, or H2 feedstock
Hydrogen production
- There will be a respectable volume of refinery gas and C3-C4's coming from the cracking units; these can be used for hydrogen production (steam-HC reforming), perhaps balancing LPG component production against hydrogen requirements
Another option, if there is not enough light material for steam-HC reforming, is a partial oxidation unit, using vacuum residue as the feedstock (this is effective but capital-intensive) instead of or in addition to the steam-HC reformer, and partially replacing the coker
Other process and utility units (sizes may be different, but functions are essentially identical)
- Steam production for various usages, including utility, process, and power generation
- Sulphur recovery – there will be large amounts of sulphur, produced as H2S, that will have to be converted to sulphur for safe handling
- Coke handling – if the refinery coke is not burned in the power plant, it will need to be handled for transport – one option is to calcine the coke to make it a more valuable product
LPG Refinery
There is a third type variant, although it is relatively rare, and that is the LPG refinery. The objective of the LPG refinery is to optimise the production of propane and butanes in order to satisfy local market needs for LPG as fuel and as petrochemical feedstocks. Usually, these refineries are based on light crudes or condensates as a starting point, since there is a large potential for wasted energy when converting heavy oils to LPG.
The special process unit configurations that can be used here are
- hydrocracking units that are optimised for LPG production,
- catalytic reformers that run in what normally would be a stable but overchlorided catalyst condition (excessive hydrocracking), and
- FCC units trimmed for maximum LPG production (with ZSM-5 cracking additive).
Naturally, in every case, there will be a number of options to consider before the refinery's final process design is set, and special solutions for unusual problems or local conditions are the rule, rather than the exception, in refinery basic process flow design.
The selection of the optimum process flow scheme generally takes place by making runs with a refinery economics model. The objective of the model is to product the optimum profit, taking into consideration the base feedstock (crude), additional feedstocks, the product requirements, and the prices for feedstocks, products, and utilities. The first runs are with most, if not all of the process units in the model "open", i.e. the economic model is allowed to optimise without considering which of the process units will be built. This will probably give a refinery with a large number of process units, some of which will be uneconomically small from an operating point of view. Later, the least attractive processes will be eliminated, and the capital cost will be factored into the calculation (this may have already been done in the first step). Finally, other considerations (site requirements and limitations, transport considerations for feedstocks and products, transport considerations for refinery equipment delivery to the refinery, market limitations, process complexity, licensing considerations, political considerations, etc.) will finalise the refinery process design.
This should give you an overview of the process that is termed "refinery process design". Flexibility and an open mind are necessary in order to come up with a refinery configuration that fits all needs.
Questions are welcome!
Posts
