Petrobras implements $29 million refining-technology program

Thi Chang
Refining/Petrochemical Editor

The RFCC unit at the Capuava (Recap), Maua, Sao Paulo, refinery, to be completed in 1999, is presently under construction. Here, the crane is lifting the regenerator head (Fig. 2).

Acrylic cyclones at the Shale Industrialization Superintendence (SIX) facilities' pilot unit are useful in studying the fluid dynamics of the system and testing the performance of various cyclone conceptions (Fig . 3).

In a move to modernize its refineries and adapt them for heavier crudes, Petroleo Brasileiro SA (Petrobras) created the Strategic Refining Development Program (Proter) in 1994.

Proter develops and tests new technologies for application in Petrobras' refineries. Since 1994, the program has invested about $29 million in five key areas:

Application of new technologies in residue cracking
Maximizing liquid production in coking units
Reduction of hydrogen costs
Application of new technologies in hydrorefining
Development of nonconventional refining routes, such as biotechnology.

At its conception, Proter was expected to generate about $1 billion between 1996 and 2002. This benefit was mainly expected by an increase in the residue-conversion rate, a reduction in operating costs and capital expenditures, and the replacement of light imported oils by Brazilian heavy oils.

A breakdown of expected savings is shown in Table 1 [20,348 bytes].

Proter

Proter was developed in response to several changes in the Brazilian petroleum industry in 1994:

A growing supply of heavy oil from the Marlim oil field in the Campos basin, Rio de Janeiro. Marlim crude has higher nitrogen and metals contents than the usual crudes processed in Brazil (Table 2 [9,698 bytes]). Proter is focused on the development of advanced hydrorefining processes. Residue-cracking processes, and FCC catalysts to accommodate this crude.
An increase in the supply and demand for natural gas. Low natural gas prices have discouraged the use of fuel oil in industry furnaces and thus have reduced the demand for fuel oil. Petrobras' strategy, in this case, is to increase its diesel and gasoline production via both hydrorefining and residue-cracking processes to upgrade unwanted residual oil.
Low refining margins. In search of profits, Proter is reviewing existing processes to maximize gasoline and diesel production.
Stricter gasoline and diesel specifications. Petrobras began producing premium gasoline and low-sulfur diesel in 1997. Gasoline and diesel specifications for Brazil can be found in Table 3 [20,453 bytes].

All developments by Proter are the result of collaboration between Petrobras' refineries and its Research & Basic Engineering Center (R&D Center) in Rio de Janeiro. Petrobras has simultaneously been working with experts from universities, research centers, and companies in Brazil and abroad. These institutions are listed in Table 4 [29,849 bytes].

In Petrobras' strategy, Proter's budget was to be put to use in the following proportions:

Seventy-five percent to expand technology. This category includes modifications of current refinery processes
Fifteen percent to work on advanced developments. Petrobras intends to apply the latest world technologies where profitable.
Ten percent to develop innovative technologies. Along with implementing new technologies from outside of Brazil, Proter expects to develop new ones.

New and revamped FCCUs

Proter played a key role in designing three new fluid-catalytic cracking units (FCCUs) for atmospheric residue for Petrobras. Also, it revamped Petrobras' remaining 11 FCCUs.

Table 5 [8,269 bytes] shows the feedstock design of the upgraded and new units. Fig. 1 [99,764 bytes] shows Petrobras' patented residue fluid-catalytic cracker (RFCC) design, which it is installing in the three new FCCUs.

Two of the three new units are under construction and are expected to be operational in 1999 (Fig. 2). Basic engineering was recently completed for the third unit, which is expected to start up in 2002.

The refineries with new FCCUs are listed in Table 6a [39,968 bytes]. Total investments for these three units are about $700 million.

The three new Petrobras-patented RFCC units will process 100% of the atmospheric bottoms. Residue similar to vacuum residue will make up about half, or about 63,000 b/d, of the bottoms processed in these units. The cracking diverts the residue from the fuel-oil pool, which is of lower commercial value.

The three new RFCC units double Petrobras' vacuum-residue cracking capacity. Assuming a gain of $1.30/bbl for cracked-vacuum residue, these three units are expected to generate additional sales of at least $28 million/year. This estimate considers only the difference in the price of vacuum residue as a fuel oil and the products generated by cracking this fraction of the residue.

The remaining 11 FCCUs owned by Petrobras were modernized with alternate Petrobras FCCU technology (Table 6b). In combination, the 11 units process up to 419,000 b/d. Of that, 15%, or 63,000 b/d, is vacuum residue.

FCC feed nozzles

For its new and existing FCCUs, Proter developed a proprietary dispersion system. The system is designed to atomize the feed to improve contact between the feed and the catalyst.

Better contact facilitates feed vaporization and discourages thermal cracking inside the riser. Thermal cracking leads to products of lower value, such as fuel oil, fuel gas, and coke.

Feed nozzle designs were chosen among several prototypes and field tests. The feed nozzles optimize the use of steam at supersonic velocities producing an oil mist of extremely small droplets with a low-pressure drop.

Feedback from Petrobras' commercial units have been incorporated into the present design for continual improvement. Units with improved feed nozzles are listed in Table 6.

According to Albertino Machado de Carvalho, Proter's coordinator, gains in gasoline and LPG production, brought about by the new feed nozzles, represent about $0.15/bbl of feed. For a daily feed rate of 419,000 b/d, increased revenues correspond to about $21 million/year.

Including the three new FCCUs, yet to be installed, the estimated increased revenue is $27 million/year.

Closed cylones

The Petrobras Advanced Separation System (PASS), a closed-cyclone system for FCCUs, was developed from pilot models. Design development for PASS began in 1992 with an acrylic model at the Shale Industrialization Superintendence (SIX) facilities in Sao Mateus do Sul, Paran .

Carvalho points out that the proprietary termination device has performed well in three commercial units: the Betim (Regap), Minas Gerais refinery, the Duque de Caxias (Reduc), Rio de Janeiro refinery, and the Manaus (Reman), Amazones refinery. PASS will also be incorporated into the three new RFCCs to be installed within the next 3 years.

PASS avoids overcracking of products in the disengager. In addition, it reduces carryover of hydrocarbon to the stripper to very low levels.

Radiotracer measurements by Petrobras confirm this performance. In the previous FCCU (45,000 b/d) at the Duque de Caxias (Reduc), Rio de Janeiro refinery, the hydrogen content in the coke (lower hydrogen content means lower carryover) was 6-8%. After the PASS system was installed, radiotracers measured only 4.5-5.5% hydrogen in the coke.

Petrobras claims that PASS reduces coke and gas yields by as much as 25% and regenerator temperature by about 80° C. These properties increase the catalyst-to-oil (CTO) ratio and improves gasoline conversion as well as LPG selectivity.

Table 7 [13,553 bytes] shows the benefits of the PASS technology at the Betim (Regap), Minas Gerais refinery on the Regap II FCCU. For total combustion, the conversion is 4.1 wt % higher.

Besides PASS, Petrobras designs conventional cyclones for both reactors and regenerators. Petrobras has retrofitted or replaced 180 cyclones in 11 of its FCCUs with various cyclone technologies.

These modifications, according to the company, have resulted in reduced catalyst losses, increased inventory of catalyst fines, and lengthened onstream times. In one example, at the Reduc refinery, before the revamp in 1997, catalyst losses were 0.041 kg/bbl, or 184.5 kg/day (45,000 b/d FCCU). After the revamp, catalyst losses were reduced to 0.009 kg/bbl, or 44.1 kg/day (49,000 b/d FCCU).

SIX has a pilot-scale cold cyclone unit made of acrylic (Fig. 3). Innovations from advanced testing stages at the SIX will be incorporated into the new systems in 1999 and 2002.

FCC catalysts

Proter developed its own catalyst to help the FCCU efficiently take 100% of the atmospheric tower bottoms. The catalyst required outstanding hydrothermal stability in the presence of high metals contamination, mainly nickel and vanadium.

The new FCCU for the Capuava (Recap), Maua, refinery will run feed with high vanadium content. Proter developed a new catalyst family based on a stable and high activity Y zeolite. Proter claims the catalyst can tolerate vanadium contents up to 6,100 ppm.

In this refinery, Proter estimates a savings of $1.5 million/year in fresh catalyst. Potential benefits of this vanadium-tolerant catalyst are shown in Table 8 [13,028 bytes].

For the Mataripe (Rlam), Bahia refinery, a catalyst based on a premium matrix for maximum nickel tolerance and a modified USY zeolite was formulated. Catalyst design required resistance of nickel contents up to 21,000 ppm.

Table 8 shows performance of the nickel-tolerant catalyst from laboratory-scale tests. Based on these tests, the improved catalysts will allow refinery savings up to $25 million/year based on reduced fresh catalyst consumption for the same or better base-yield profile.

Riser optimization

A 40 b/d prototype unit located at SIX and several FCC pilot units located at Petrobras' R&D Center, Rio de Janeiro, have been useful tools for FCC technology developments.

Tests from these tools as well as feedback from commercial units are combined to improve riser design tailored to specific feeds, such as 100% residue feeds. Optimized contact time and suitable velocities through all riser sections provide minimal pulsation and efficient CTO contact.

Good riser design leads to better gasoline selectivities. All commercial Petrobras units have or are having their risers modified. Several modifications for improved riser design include:

Changes in the diameter reduce the contact time of new risers.
Higher velocities and new geometries avoid sudden changes in direction and provide efficient CTO contact.
Adequate steam injection allows homogeneous flux of the catalyst in the riser region before the flue-oil injection. Thus, better CTO contact is achieved.
New geometry at the riser bottom allows higher catalyst circulation and better control, improving the available pressure.
Mechanical improvements allow the use of carbon steel instead of more-expensive alloys.

Real effects of these riser modifications are shown in Table 9 [8,712 bytes]. The Cubatao (Rpbc), Sao Paolo, refinery produced 1.3 wt % less coke, and had a 6.0% increase in naphtha production and a 1.9 wt % increase in conversion after modifying the riser and feed nozzles.

With a capacity of 8,400 cu m/day, the additional income generated by the improved performance shown in Table 9 is about $7.8 million/year.

FCC simulator

Simcraq is a steady-state FCC simulator designed by Petrobras. It is used for simulation, process monitoring, planning, and optimization. It can be used as either an offline or a real-time tool.

The development of Simcraq started in 1995 and was ready for use in 1998. Simcraq is being used at Petrobras' R&D Center, 10 refineries, and SIX.

The simulator can handle configurations of most FCCUs. It can accommodate catalyst coolers, total or partial combustion, one or two stages of regeneration, and naphtha and oil risers.

The simulator's measurements are based on an adjustable model and a very efficient solver. During optimization routines, every constraint, including temperatures, pressures, flow rates, catalyst-circulation rates, product qualities, and equipment capacities, can be pushed to its limits to achieve the objective function.

Various optimization objectives include maximum profit, maximum feed rate, maximum naphtha production, and maximum octane-barrel. Economic parameters and costs can be updated in a straightforward manner.

Applied as an off-line or real-time optimizer, the system is expected to save $0.15/bbl of feed, or $27 million/year based on Petrobras' capacity.

Delayed coking

Petrobras developed a new process configuration for delayed coking which it believes will maximize distillate yields and reduce coke yield by 10%. At the Canoas (Refap), Rio Grand do Sul and the Paulinia (Replan), Sao Paulo refineries, this benefit translates to revenues of $6 million/year.

Petrobras' process uses ultra-low pressure in the coking drums with distillate recycle to the feed. Pilot-plant tests run by Petrobras revealed that lighter-product recycles, which have higher hydrogen contents, produce lower coke yields. Thus, Petrobras recycles distillate rather than heavy gas oil.

The new design includes double-fired furnaces with individual pass control. Optimum thermal flux rates, multiple steam and water injections, minimum residence times, and constantly rising temperature gradients maximize yields and run lengths.

A comparison of conventional and Petrobras' delayed coking technology is shown in Table 10 [26,866 bytes]. Typical delayed-coking feedstocks have a gravity of 8.4° API, concarbon residue of 21.2 wt %, and sulfur content of 1.0 wt %.

Petrobras performed the basic engineering for new coking units at the following four refineries:

Cubatao (Rpbc), Sao Paulo, 14,500 b/d coking capacity, started up in 1986
Betim (Regap), Minas Gerais, 20,800 b/d coking capacity, started up in 1994
Paulinia (Replan), Sao Paulo, 31,500 b/d coking capacity, started up in 1998
Canoas (Refap), Rio Grande do Sul, 12,600 coking capacity, started up in 2001

Thermal cracking

The Mild Thermal Cracking Process (CTB) and the HCTB (CTB with hydrotreatment) processes are patented by Petrobras. They are thermal cracking process to produce diesel from vacuum gas oil (VGO). Proter claims the following advantages:

High diesel yield, typically 35 vol %
High diesel quality, that is, a cetane number of about 45 and maximum sulfur content of 0.01 wt %
Low investment and operational costs. The HCTB process, according to Proter, requires less capital and power. For a 20,000 b/d unit, costs are estimated at $16 million/year for a paraffinic stock that does not require hydotreating, or $28 million for a nonparaffinic stock that requires hydrotreating.

In the thermal cracking process, the proportion of converted feed depends on two variables. The conversion increases with increasing temperature and increasing residence time. The higher the process temperature, however, the worse the diesel quality as a result of polymerization and condensation reactions that are accelerated with temperature.

The operating temperature in the furnace-cracking process must be high because the residence time is minimized to avoid coke formation, which plugs the coils. The HCTB process is a soaking thermal cracking process with two reactive zones.

Part of the conversion is achieved in a furnace with low residence time and moderate temperatures, and the final conversion is achieved in an adiabatic reactor (soaker drum) at low temperature with a high residence time.

A 1,900 b/d demonstration plant at the Cubatao (Rpbc), Sao Paulo, refinery ran from August 1994 to March 1996. During this time, Proter collected data and completed the development of the HCTB process. Table 11 [28,866 bytes] illustrates the diesel yield of the HCTB process from pilot plant tests.

Several feedstocks were used in the demonstration plant: VGO from Brazilian crudes, a paraffinic crude from the Bahia basin, and a naphthenic one from the Campos basin. When the process feedstock had a high paraffinic content, like the VGO from Bahia crude or mild hydrocracking residue, the diesel product was stable without hydrotreating.

In its second CTB application, Petrobras plans to erect a new unit in its Manaus (Reman), Amazonas, refinery by 2000.

Hydrogen recovery

Petrobras and GKSS, a German research center, are developing a new process for hydrogen recovery. It is a hybrid process using membranes and a pressure-swing adsorption unit.

The new process makes it possible to recover hydrogen from refinery offgas that has a hydrogen content of 15-25 mole %. Today, hydrogen recovery from refinery offgas is economically feasible for gas streams with a hydrogen content higher than 40 mole %.

Proter built a pilot plant in the Duque de Caxias (Reduc), Rio de Janeiro, refinery to further develop this process.

Proter's goal is to recover high levels of hydrogen from low pressure and low concentration hydrogen streams at competitive costs. With this process, the hydrogen content can be increased from concentrations as low as 15% to as high as 99%.

Hydrorefining

In consideration of the high nitrogen content of Brazilian oil from Marlim, Proter is developing novel ways to remove nitrogen in the hydrotreating process.

A pilot plant at SIX is studying the vacuum-residue hydroconversion using expanded-bed technology. The National Institute of Research and Environment (NIRE), Japan, is working with Petrobras to develop this technology further.

Also under development is microbiological desulfurization and denitrogenation of petroleum products. Proter is studying microorganisms that are biocatalysts in removing sulfur and nitrogen from crudes and products. The present hurdle is to develop a process that is efficient and economically feasible.

This subject is being studied in a joint project with universities and research centers in and outside of Brazil.