Apr-2011

Diverting low-sulphur heavy stocks for fuel oil production

Studies enabled a refinery to divert low-value product stocks into higher-value products and preserve crude throughput levels

Rajeev Kumar, Chithra V, Peddy V C Rao and N V Choudary
Bharat Petroleum Corporation Ltd, India

Article Summary

Burgeoning use of natural gas, including liquefied/compressed natural gas (LNG/CNG), as a cheaper fuel option has made low-sulphur heavy stocks (LSHS) a product in surplus since 2008–2009. In fuel-consuming industries such as power production, liquid fuels like naphtha and LSHS have faced significant competition from an increased use of natural gas.¹ However, fuel oils have maintained a better market as a result of steady local demand and export markets, reflecting their ease of transportation.

LSHS is normally produced by blending two straight-run refinery streams; namely, low-sulphur vacuum residues (LS-VR) and clarified oil (CLO), with LS-VR as more than 90 wt% of the blend (see Figure 1). This is one of the simplest options for utilising residues. It has been practised for years and no processing units are required. In a period of low demand, large quantities of LSHS occupy storage tanks, which puts pressure on the capacity of low-sulphur-processing crude distillation units for the continuous production of LS-VR. In these circumstances, it is essential to evacuate LSHS to sustain crude throughput and add value.

In the current refining climate, refinery configurations have been improved to cater for developments in the processing of low-cost opportunity crude oils, including high-sulphur feeds, high TAN crudes and heavy oils.² Additionally, residue upgrading techniques and hydrotreating options are applied to refinery bottoms to achieve higher margins. However, many refineries still do not have the best hardware configurations to match their processing needs, or they may suffer space constraints for residue upgrading processes, such as delayed coking, visbreaking and solvent deasphalting.³ Older crude distillation units (CDU) were designed to process only low-sulphur crude oils producing LS-VR streams. This also leads to the production of quantities of LSHS, which fills the storage tanks in a surplus market. At one of BPCL’s refineries, in Mumbai, there are three CDUs with a total capacity of 12 million t/y. The two older units (CDU-I and CDU-II) have a combined capacity of 6 million t/y and they process only low-sulphur crude oils. Thus, the continuous operation of these units leads to LSHS production. When LSHS is in low demand, it fills many storage tanks. In this scenario, the evacuation of LSHS is essential to sustain the crude throughput of these older units. The LSHS product specifications are shown in Table 1.⁴

Conversely, CDU-III has a capacity of 6 million t/y and processes high-sulphur crude oils, generating high-sulphur vacuum residue (HS-VR). These streams are used in the production of fuel oil. Fuel oil is normally produced by blending HS-VR, LS-VR and kerosene/high-speed diesel (HSD) to meet product specifications for flash point, kinematic viscosity, pour point and so on. Normally, two grades of fuel oil are produced for different applications; namely, 180 cSt and 380 cSt. The product specifications are shown in Table 1. This is, again, a simple residue utilisation scheme for which no complex processes or operations are needed. The existing correlation for the fuel oil blending scheme has always had some quality giveaways while meeting all the specifications. Both of the viscosity grades for fuel oil either fail to meet specifications or have quality giveaways. In order to address these issues, optimisation of the fuel oil production scheme and the development of new correlation models has been taken up at BPCL’s Corporate R&D Centre. A blending scheme for fuel oil is shown in Figure 2.

In this article, a modelling approach is proposed, exploring various possibilities for dealing with surplus LSHS. Absorption of LSHS into fuel oil and/or making a new low-sulphur fuel oil (LSFO) product were ideas that evolved into immediate solutions. New correlations have been developed for the prediction of critical specifications for LSHS, fuel oil and LSFO products. The model uses Aspen+ and crude oil management tools for simulations.

The immediate need to deal with LSHS by diverting it into the fuel oil pool was the first choice for development and implementation. Experiments were carried out at BPCL (R&D) with refinery samples to generate a database for estimating parameters for the model. The model has been developed and implemented at the refinery and proved helpful in minimising quality giveaways.

Following implementation of the model at the Mumbai refinery, the work received a Judges Special Award in a corporate competition for new ideas.

Methodology for diverting LSHS into fuel oil_⁵
During the processing of low-sulphur crude oils in refineries, volumes of LS-VR are produced in excess, hence more LSHS is produced than is required. During 2008–2009, when demand for LSHS suddenly fell away, production moved into surplus and tank storage of the product for long periods became difficult. Various possibilities have been explored in order to avoid its production to sustain crude throughput — for instance, producing LSFO and absorbing LSHS into fuel oil — to achieve continuous operation of the low-sulphur crude processing units.

For maximising the absorption of surplus LSHS into fuel oil for both product grades, a methodology has been developed on the basis that LSHS is simply a blend of two components, LS-VR and CLO, which are common constituents of fuel oil blends (see Figures 1 and 2). The concept of diverting LSHS into fuel oil is shown in Figure 3.

Production of LSFO⁵
In view of stringent environmental regulations and particular concern about air quality in metropolitan cities, there is growing customer demand for LSFO. The scheme for producing LSFO is similar to the fuel oil blending scheme (see Figure 2), but the specification for LSFO is slightly different from fuel oil with regard to sulphur composition (see Table 1). As with fuel oil, two grades of LSFO, 180 cSt and 380 cSt, are required. To meet the LSFO requirement, an alternative scheme was developed involving surplus LSHS. LSFO is an upgraded version of the fuel oil product but with the sulphur specification reduced to 2.0 wt%. The other specification parameters are similar. Since sulphur content is the cut-off property for LSFO product, the composition can be simulated and fixed for sulphur-limiting streams. Optimisation of the other streams can then be carried out to meet the other specifications.

In the present case, the possibility of diverting LSHS into LSFO product was explored. Since the LSFO product was new, optimisation of the blend composition was carried out as part of the study. This proved helpful in producing LSFO with minimum quality giveaways.