used to make chemicals such as ammonia and methanol, to refine petroleum, metals, and electronic materials, and to process food components. More than 32 million tons per year (t/yr) H2 (80 million kg/day) are produced using natural gas SMR. Hydrogen is also made today using partial oxidation and ATR.

The vast commercial experience based on this manufacturing capacity has led to many improvements in the technology, reducing costs and increasing efficiency. Perhaps the most important element is the tubular reactor in which the SMR reaction takes place. Progress has led to higher tube wall temperatures, better control of carbon formation, and feedstock flexibility.3 This progress in turn has led to lower steam-to-carbon ratios and improved efficiency. The water-gas-shift unit has also been improved, and now one-step shift can be employed to replace the former two-step operation at different temperatures. Finally, purification of the hydrogen product has been simplified by using PSA to remove methane, carbon oxides, and trace impurities in a single step. While designs today do not generally include CO2 capture, technology is currently available to accomplish this. Using a commercial selective absorption process, CO2 could be recovered for subsequent sequestration.

Progress has also been made in designing and building larger SMR plants. Currently, single-train commercial plants of up to 480,000 kg H2 per day (200 million standard cubic feet per day [scf/d]) are being built, and even larger plants can be constructed using multiple trains. Units as small as 300 kg/day are also being built.4 In many cases, the units built are one of a kind, with specific features to meet the requirements of a site, application, or customer. At least one company is fabricating commercial SMR hydrogen plants as small as 300 kg/day using components of fixed design, one of the elements of mass production.5

Partial oxidation utilizing natural gas is fully developed and used commercially. In most cases today, commercial units use feeds of lower value than natural gas, such as coal, coke, petroleum residues, or other by-products, because of economics. However, natural gas is a preferred feed for POX from a technical standpoint and can be used to generate hydrogen where competitive.

Oxygen-blown ATR with natural gas is used today in very large units that generate a mixture of CO and H2 for the Fischer-Tropsch process or methanol synthesis. This is attractive in part because the units can produce the hydrogen-to-carbon monoxide ratio needed in the synthesis step. Since the heat of reaction is added by combustion with oxygen, the catalyst can be incorporated as a fixed bed that can be scaled up to achieve further benefits of larger plant size in both the ATR and the oxygen plant that is required. ATR also offers benefits when CO2 capture is included. This is because the optimum separation technology for this design recovers CO2 at 3 atmospheres (atm), thus reducing the cost of compression to pipeline pressure (75 atm).

In summary, all three processes (SMR, POX, and ATR) are mature technologies today for the conversion of natural gas to hydrogen. SMR is less costly except in very large units, where ATR has an advantage. SMR is also somewhat more efficient when the energy for air separation is included. POX has the advantage of being applicable to lower-quality feeds such as petroleum coke, but this is not directly relevant to natural gas conversion.

Future Natural Gas Conversion Plants

Given the current interest in possibilities for a hydrogen economy and the current commercial need for hydrogen, significant effort is being focused on improving natural gas conversion to hydrogen. Improved catalysts and materials of construction, process simplification, new separations processes, and reactor concepts that could improve the integration of steam reforming and partial oxidation are being investigated. Catalytic partial oxidation is also under consideration. Since steam reforming and partial oxidation are mature technologies, the primary opportunities for improvement involve developing designs for specific applications that are cost-effective and efficient.

Several thousand distributed generators will be needed for the hydrogen economy, and it should be possible to lower the cost of these generators significantly through mass production of a generation “appliance.” Such appliances may be further improved by tailoring the design to the fueling application. For example, hydrogen would likely be stored at roughly 400 atm, and to the extent that the conversion reactor pressure can be increased, hydrogen compression costs would be reduced and efficiency improved. For distributed generators incorporating POX or ATR, suitable cost-effective methods for hydrogen purification need to be developed. Alternatively, in such cases there are potentially attractive opportunities to recover the oxygen needed with membranes and thus to lower the cost.

Other concepts are also in the exploratory research stage. These involve new or modified ways of providing the endothermic heat of steam reforming or utilizing the heat of reaction in partial oxidation.

New, lower-cost designs for distributed generation probably can be advanced to the commercial prototype stage in the next 5 to 7 years. Some of these improvements could be applicable to large plants.


The committee undertook cost studies as described elsewhere (in Chapter 5 and Appendix E) to identify the areas


J.R. Rostrup-Nielsen, Haldor Topsoe, “Methane Conversion,” presentation to the committee, April 25, 2003.


Personal communication from Dale Simbeck, SFA Pacific, to committee member Robert Epperly, April 30, 2003.


Dennis Norton, Hydro-Chem, “Hydro-Chem,” presentation to the committee, June 11, 2003.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement