Modelling and optimising bitumen hydrocracking

A rigorous hydrocracking model for the quoted catalysts is used for predicting hydrogen consumption and product yields to maximise operating revenues.

Dynamic Simulation & Consultancy

Viewed : 1626

Article Summary

The Canadian oil sands industry reached peak production of roughly 3.4 million barrels of combined mining and in situ bitumen at the end of 2018. Only 55% of these barrels is upgraded to sweet synthetic crude oil (SCO). Despite low oil prices and several months of lockdown during the pandemic, Canadian oil sands production remains resilient and maintained about 3.0 million b/d in March 2020.

Two business factors have been the driving forces behind the bitumen upgrading process in the Canadian oil sands industry. One factor is to recover the expensive diluent which is then recycled back to steam- assisted gravity drainage (SAGD) central processing facilities, and the other is to produce high value-added SCO to eliminate the impacts of the volatile market price spread between WTI and WCS on the SAGD or mining operations. So far there is still a lack of adequate capacity to upgrade most produced bitumen to SCO. The production of finished oil products from Canadian bitumen requires either new, integrated upgrading and refining facilities, or the revamping or debottlenecking of existing upgrading/refining plants. There are two primary bitumen upgrading technologies: coking and hydroconversion. Currently, coking based upgraders have predominated in the bitumen upgrading market. In this, asphaltene residue is converted to petroleum coke and hydrogen is produced by reforming natural gas. This coking based upgrading scheme has become the standard template for building an upgrader in Alberta.

The primary challenge with Canadian bitumen is the large percentage of bottoms to be dealt with, as well as the relatively high asphaltenes, sulphur, and metals content of those bottoms. To use fixed-bed hydrocracking technologies, a feed clean-up unit such as a solvent deasphalter (SDA) should be installed upstream of the hydrocrackers to reject the heaviest CCR and metals as a pitch stream. Recently most commercial applications on bitumen upgrading configurations have focused on selection of primary upgrading technology and how to deal with low-valued unconverted pitch/asphaltene. There is a growing trend in the industry to utilise asphaltene gasification for hydrogen production to eliminate coking waste and its on-site storage. Examples include Long Lake upgrader and the North West Surgeon refinery.1

A proposed bitumen upgrading scheme based on the above considerations is shown in Figure 1. The bitumen upgrading process is integrated with a SDA for primary upgrading, fixed-bed hydrocracking for second upgrading, and asphaltene gasification for hydrogen production. Once the asphaltene has been separated from the vacuum residue, the deasphalted oil (DAO), along with atmospheric gasoil (AGO) and vacuum gasoil (VGO), is easier to process in the hydrocracking unit. The heavy asphaltene or pitch can be completely converted to synthesis gas and hydrogen products through gasification. This configuration has no undesired or low-value coke byproducts and is therefore truly bottomless. Detailed research work on bitumen upgrading configurations will be published later. As a key hydroconversion process, this article mainly focuses on process modelling and optimisation of the hydrocracking unit and light end recovery (LER) system.   

Predictive hydrocracking model
Process description

Figure 2 shows the process flow diagram of a single stage AGO/VGO/DAO hydrocracking unit and LER unit. The purpose of the hydrocracking unit is to process AGO from the diluents recovery unit (DRU), VGO from the vacuum distillation unit (VDU), and DAO from the SDA, convert these feeds into SCO, and blend with C4 and C5+ from the LER unit into the final SCO products.

The AGO/VGO/DAO feed is mixed with recycled gas and make-up hydrogen and heated up by the effluent heat exchangers and feed furnace. The combined feed enters through two fixed-bed reactors in series — R1 of the hydrotreater and R2 of the hydrocracker, each consisting of three beds. Within the R1 reactor, a catalyst guard zone is utilised to trap metals present in the feed, while the high activity hydrotreating catalysts reduce the sulphur and nitrogen contents. The produced hydrogen sulphide and ammonia can then be removed from the recycle gas loop. Hydrotreated feed will be quenched with recycle hydrogen and then routed to the R2 reactor for hydrocracking processes. In the hydrocracker, long-chain molecules will be cracked to short-chain molecules. Both R1 and R2 reactors contain specially ordered fixed-bed catalysts.

After process heats are recovered and MP steams are generated from the reactor effluent, the effluent is routed to the high pressure separator (HPS) to separate recycle hydrogen stream, sour water, and hydrocarbon. Recycle hydrogen will be routed to an amine scrubber to remove sulphur and ammonia compounds prior to recycle. Compressed recycle hydrogen will be supplemented with make-up hydrogen from the gasifiers as required for the hydrocracker unit operation. Hydrocarbon products will be stripped of sulphur compounds and light hydrocarbon products. The stripper bottom products will become the primary SCO blending component. The stripper overhead distillate will be routed to the LER unit for further separation in the depropaniser and debutaniser to recover usable fuel gas and C4 and C5+ components. The C4 and C5+ components will be blended with the primary SCO to produce the final SCO products.

Predictive hydrocracking model
To perform a parametric sensitivity analysis and optimise the operation of the hydrocracker, a system model is needed for the hydrocracker and LER system. Fundamentally, system modelling of the hydrocracking processes consists of three layers of modelling scope2:
-    Hydrocracking kinetic model
-    Reactor model
-    Process system model

The key layer is the kinetic model which focuses on the kinetic analysis of complex reaction networks to study reaction mechanisms, catalyst activities, and deactivations, as well as the influence of reaction conditions. The reactor model quantifies the reactor performance — product yields, hydrogen consumption, utility usages. A process system model is used for the optimisation of unit or upgrader-wide operating conditions to maximise profits and enhance process safety. There are two types of refining reaction modelling: correlation modelling and first principles modelling. Correlation modelling is empirically a ‘cause and effect’ model that correlates one variable to another, whereas first principles modelling is based on the chemical reaction fundamentals of kinetics, mass, and heat balances, with consideration of the following factors of the petroleum species:
- The large number of (unknown) chemical compositions present
- Knowledge of true kinetic behaviour with the presence of thousands of system compounds is largely incomplete or unknown

Add your rating:

Current Rating: 4

Your rate: