New gasification technology tested in Sweden

EnviRes co-developed with Ashland Petroleum Company, now Marathon Ashland Petroleum, LLC a new coal gasification process through laboratory scale testing.  This process employs direct coal injection into molten iron without oxygen or steam.  At molten iron temperatures, the carbon in coal rapidly dissolves in the metal.  Iron was selected because it has a relatively high solubility for carbon and because dissolved carbon in iron oxidizes to carbon monoxide before significant iron oxidation occurs.

Carbon-Iron Phase Diagram

In this process, most hydrogen associated with coal forms molecular hydrogen in the gas phase.  Nearly all nitrogen in the coal becomes molecular nitrogen and also enters the gas phase.  A substantial amount of the sulfur in the coal converts to hydrogen sulfide, also a gas.  The remaining sulfur is captured in the slag layer.  Oxygen contained in the coal becomes carbon monoxide and enters the gas phase.  Most inorganic constituents (ash) become molten slag that floats on the molten iron as in steel making.  The slag must be periodically or continuously tapped from the reactor.  Mercury present in coal volatilizes into the gas stream where it can be easily removed by downstream treatment.  This results in a product gas stream that can be up to 90% hydrogen as it leaves the reactor depending on the coal composition.  When the carbon content in the metal nears the solubility limit, coal feed is interrupted.  Oxygen is then injected into the metal to convert the dissolved carbon to carbon monoxide.  Oxygen injection reduces the carbon in the metal to the desired level.  Oxygen injection provides more heat than the endothermic coal injection step requires so some steam or other temperature-moderating stream must also be injected during oxygen injection to keep the temperature of the metal at the desired value.  If steam is used as the temperature-moderating agent, it reacts with carbon dissolved in the molten iron to form hydrogen and carbon monoxide.  

The purposes of this program were to perform large scale (up to 3 tons/h coal feed) testing of the process and demonstrate performance parameters that are required for commercial practice of this technology and to develop a Design Basis Memorandum (“DBM”) for the construction of a commercial scale plant.  In a DOE funded portion of this work, EnviRes elected to do testing at a metallurgical research facility (MEFOS) in Sweden because they had nearly all of the equipment necessary for testing in place.  MEFOS already had a trained staff that was very skilled in doing this type of work.  EnviRes was able to perform this testing for less than one tenth the cost of building the equipment from scratch and saved several years in accomplishing the work.  As MEFOS generates data in testing, EnviRes and MEFOS review it and provide sufficient data to Kvaerner to develop the DBM and to assess the economic impact of hydrogen produced by HyMelt on refinery economics.  Generating data for the design of a commercial plant was crucial to the success of this project.  Such data must be generated in large equipment to minimize the uncertainty of scale change.  EnviRes decided to perform this testing at MEFOS (a Swedish acronym, in English this translates to “The Foundation for Metallurgical Research”) in Luleå, Sweden.  EnviRes selected MEFOS for the following reasons:

• MEFOS has nearly all of the equipment necessary for these tests.

• MEFOS conducted several large-scale coal gasification tests using molten iron in the 1980’s.

• MEFOS has a complete staff of technically trained people necessary to perform these tests.

• MEFOS does contract research and does not retain any intellectual property from the tests.

• EnviRes performed a survey and could not find any other facility in the world that could meet or even come close to these criteria.

All gasification tests take place in what MEFOS calls the universal converter (see picture below).  The universal converter is essentially a modified, highly instrumented, basic oxygen furnace (BOF).  A water-jacketed hood where the product gases are combusted sits atop of the universal converter.  The coal feed lance, oxygen lance, sample probe and flux addition lines penetrate the hood just above the 45° bend.  Tuyeres located in the bottom (not in view) can inject oxygen and/or solids. Slag tapping and metal tapping are done in the pit to avoid spillage on the work floor.  The pit also helps contain accidental breakouts of metal from the vessel.  The control room lies behind and to the right of the universal converter.  Trunnions support the universal converter so that it can be tilted for receiving hot metal as well as for pouring slag and/or metal.

Left, the universal converter at Mefos. Right, charging the tilted converter with the molten metal.

Left, looking into the converter, showing the slag layer. Right, tapping the slag into the pit.

Testing performed in June, 2003, used separate, top entry (through the hood) lances, one to inject feed and the other to inject oxygen.  Lime, dolomite (to adjust the slag) and scrap (for cooling) could be added separately.  A bottom stirring tuyere used nitrogen.  A top entry sample probe collected both dust samples and gas for online analyzers.  Separate infrared devices continuously analyzed for carbon monoxide, carbon dioxide and hydrogen, a paramagnetic device determined the oxygen content.  A quadrapole mass spectrometer also analyzed the gas from the sample probe.  Each day MEFOS melted the metal to be used for testing in an electric arc furnace, they poured the hot metal into a transport ladle supported from a load cell for weighing and charged it into the universal converter.  MEFOS tilted the universal converter to its upright position and decarburized the metal by oxygen blowing to get the carbon content in the range of 0.5 w% to 0.7 w% and the temperature in the range of 1,550°C.  Samples of the metal and its temperature were taken during the oxygen blowing.  At the end of oxygen blowing, the vessel was titled to get a slag sample, a metal sample and the melt temperature.  Oxygen blowing typically requires 15 to 30 minutes.  The oxygen lance has an outlet velocity of 2.0 Mach.  MEFOS deemed samples taken in the tilted position to be more representative as a result of additional mixing during tilting.  MEFOS determined in earlier studies that a typical heat loss for the universal converter was approximately 30 Mj/min.  After sampling, MEFOS returned the universal converter to its upright position.  MEFOS then injected coal feed to generate a hydrogen rich stream.  

Temperature probes (left) and slag samplers (right) before the testing began.  Right, taking a manual temperature reading from the hot melt.

The full report of this project is available by clicking on the link below:

HyMelt Gasification Process Development Project (Large PDF File, 4.8 MB)

  Index to all Focus on ICCI Research articles