Circulating fluidized-bed reactors have traditionally been used for the high-temperature Fischer-Tropsch synthesis process. But the limitations of this method prompted Sasol Technology (Pty) Ltd. of Johannesburg to apply conventional, fixed fluidized-bed technology in its Fischer-Tropsch process.
A commercial-scale fixed fluidized-bed reactor has been successfully operating at Sasol One in Sasolburg, South Africa since May of 1989. C.J. Jones, B. Jager, M.D. Dry of Sasol reported on the technology at Achema 91 in Frankfurt.
The prime benefits of this technology are cost-less than half that of the older technology-improved thermal efficiency, higher throughput, greater flexibility, and significant savings on operating and maintenance costs.
BACKGROUND
Fischer-Tropsch synthesis has been practiced at Sasol since 1955. Originally the synthesis was performed in packed-bed reactors, which were subsequently developed into fixed-bed Arge reactors.
Circulating fluidized-bed reactors were then developed for the production of gasoline, diesel, and chemicals from coal. These reactors have a much larger volumetric capacity than the Arge reactors and produce lighter and more olefinic than Arge products.
Over the years, the two-stage coal-to-gas, gas-to-liquid process has been improved. The Sasol process is known today as Synthol.
In 1987, a project designed to produce transportation fuels from offshore natural gas in South Africa was announced. The project, called the Mossgas Project, selected the Synthol route mainly because of the proven nature of the process and its consequent low commercial risk (see accompanying box).
SYNTHOL PROCESS
Synthesis gas produced from natural gas by reforming, or from coal by gasification, is the feed to the Synthol process.
The gas has an H2/CO ratio of about 2. It is contacted with promoted iron catalyst at high temperature and pressure, and the resultant hydrocarbon products are condensed and separated.
One advantage of this process is its versatility,. The product slate can be tailored to local market needs by varying the special-purpose catalyst in the Synthol reactors, and by adjusting product work-up facilities to suit.
The fuel products can vary from 80% gasoline to approximately 70% distillate fuel.
The fuels so produced also meet normal refinery specifications and are environmentally acceptable without the need for complex processing.
The distillate fuels produced are of premium, low-emission quality, with essentially zero sulfur and very low aromatics content, claims Sasol. A high cetane number is also achievable.
Synthol Fischer-Tropsch synthesis produces predominantly olefinic primary products. In typical refinery processing, the C2 to C4 olefins are oligomerized to fuel components. However, at small additional cost, the ethylene and propylene can be extracted and purified to standard polymer-grade feedstock.
Both low-density and high-density polyethylene, as well as linear low-density polyethylene and polypropylene, are produced in South Africa from Fischer-Tropsch olefins. And almost all of the olefin feedstocks used in the country are produced from coal in Sasol's Fischer-Tropsch plants, says the company.
SYNTHOL REACTOR
The Fischer-Tropsch reaction is highly exothermic. Thus one of the major design characteristics of the Synthol reactor is the method of heat removal.
In the latest version of the circulating fluidized-bed reactors, heat is removed in banks of cooling coils in the reactor section by the generation of high-pressure steam.
The gaseous mixture of products and unreacted synthesis gas, together with the catalyst, passes through the top bend into the hopper section, where most of the catalyst drops out. More than 99% of the remainder is removed in cyclones, through which the gas stream passes before it exits the synthesis section to the downstream cooling train.
The cooling train consists of the following elements:
- Catalyst scrubbing system to remove catalyst fines carried over from the reactor and to condense the heavy oil fraction and separate it from the product stream
- Feed/effluent heat exchange
- Product cooling and condensation
- Separation of tight oil, water, and unreacted gas
- Feed compression.
In addition to the cooling train, a solids removal system is necessary to separate the catalyst particles from the heavy oil fraction.
REACTOR DEVELOPMENT
The catalyst employed in the Synthol reaction is iron-based. Because of its high bulk density, unique problems are encountered with catalyst fluidization.
Because iron is abrasive and high circulation velocities are used in the catalyst circulating system, reactor run-lengths may be limited by the need to inspect the reactor. With the modernization of the process in the 1970s, many of these limitations have been overcome to a degree, but not eliminated.
The reactor-train capacity was increased by a factor of about three from the original unit, to the equivalent of 6,500 b/d. Further scale-up was unlikely because the equipment in the reactor loop had reached maximum feasible size, and because of the dynamics of catalyst circulation.
The limitations of the circulating fluidized-bed reactor caused Sasol to embark on a program to develop an improved reactor system.
A 1-m diameter demonstration fixed fluidized-bed reactor was designed and erected at the Sasol One plant in 1983. This reactor was designed as a slipstream unit on one of the operating commercial reactors. The unit provided the information required to build a commercial plant.
The demonstration reactor had a capacity of about 3% of the next commercial size envisaged. This was considered adequate to allow risk-free scale-up.
Sasol realized that the investment involved in a Fischer-Tropsch plant is large and that the reactor is at the heart of a complex in which normal risk criteria may not apply. Consequently, the company erected a 5-m diameter full-size commercial plant using the fixed fluidized-bed reactor, designed after the reactor was proved in the demonstration plant.
The 5-m reactor was placed in parallel with one of the existing Sasol One reactors so that the two shared gas circulation and cooling train systems. It was successfully commissioned in May 1989 and has been in commercial operation since then.
The scale-up to the largest unit envisaged for a future plant-16,500 b/d-is now only 3-5 times. This will involve an almost negligible technical and financial risk.
FIXED-BED REACTOR
The limitations of the circulating fluidized-bed reactor are largely eliminated in a conventional fixed fluidized bed (Fig. 1).
The fixed-bed reactor consists of a vessel with a gas distributor, a fluidized bed containing the catalyst, cooling coils in the bed, and a system to separate the catalyst from the gaseous product stream.
The fixed-bed reactor is simpler and much smaller than the circulating-bed reactor. It is also self-supporting, as opposed to the circulating-bed reactor, which needs a complex support structure.
Sufficient free space is given above the fluidized bed to disengage most of the catalyst, and the remainder is completely retained and returned by porous metal filters in the top of the reactor.
Cyclones-traditionally used in circulating fluidized-bed reactors-were originally installed in the commercial unit at Sasol One. But they have since been replaced with the filters described above.
The use of porous metal filters allows further simplification of the reactor cooling train (Fig. 2). Because catalyst particles are retained within the reactor, a catalyst scrubbing system and associated solids removal system are no longer necessary.
Apart from the obvious capital cost savings, this also allows for a more efficient cooling train and an increase in overall thermal efficiency.
Although these filters have been successfully tested from an operational standpoint, a longer-term mechanical reliability test program is still in progress. The use of cyclones instead of filters could result in an increase in capital cost of about 5%.
Because the diameter of the fixed-bed reactor is much greater than that of the circulating-bed, it is possible to install 50% more cooling coil area in the fixed-bed reactor. This allows for greater conversion capacity, making it possible to operate at higher pressures and flowrates because reactor capacity (and reaction heat generated) increase in proportion to pressure.
The increased cooling coil area thus allows operation at higher specific throughputs.
The fixed, fluidized-bed reactor is also able to operate at higher catalyst-to-gas ratios, thereby allowing an increase in reactor fresh feed throughput while attaining the conversion performance of the circulating-bed reactor. A capacity of 16,500 b/d of syncrude in a single train is now considered feasible.
Because of the increase in reactor capacity, it is possible to install fewer Synthol reactor trains for equivalent feed and product flowrates, which substantially reduces capital cost. Operating and maintenance costs associated with the fixed-bed reactor are also lower than for the circulating bed.
The catalyst inventory for the two reactors is similar at start-up, but the fixed-bed reactor catalyst consumption is about 26% lower. This, combined with the fact that fewer trains are required, significantly reduces overall catalyst consumption.
The absence of a catalyst circulation system causes the pressure drop across the fixed fluidized-bed reactor to be about half that of the circulating-bed reactor. This allows for savings on feed compression costs.
And the increased fresh-feed throughput (or reduced recycle ratios) for the fixed-bed reactor leads to further savings on compression costs.
Experience with the commercial unit has shown that the new reactor requires less maintenance. In general, the fluidization is much less energetic and there are no erosion-prone components-such as bottom and top transfer bends, standpipes, and slide valves-needed for catalyst circulation.
ECONOMICS
Sasol completed an extensive study of the economic factors for a synfuels plant based on natural gas. A dry natural gas feed for the production of synthesis gas was assumed.
To take advantage of the obvious economies of scale, a plant utilizing three trains of Synthol reactors was used as the study basis. The plant would produce the equivalent of 50,000 b/d of sulfur-free syncrude from 410 MMscfd of natural gas.
The study also assumed that the Synthol production would be handled in a nearby oil refinery.
The capital cost (based on standardized location factors) of such a plant, including air separation, CO2 removal, and H2 purification units, is estimated to be $1.69 billion.
Excluded from the capital cost and subsequent economic evaluation are the following:
- Site-specific costs like taxes, duties, and surcharges
- Cost of land and servitudes
- Royalties and license fees
- General company over-heads
- Finance charges and interest during construction
- Future escalation, contingency, and overheads.
The expected reduction in the capital cost of the Synthol plant is evident from Fig. 3.
The economics of such plant are sensitive to the price of natural gas. Depending on plant location and actual natural gas cost, the Synthol process could be economic at crude prices above $23/bbl.
Copyright 1992 Oil & Gas Journal. All Rights Reserved.
