Mar-2012

Heat exchange reforming in gas synthesis

An exchange reformer design offers a very high reforming temperature and the option to reduce primary reformer size

HENRIK SURROW LARSEN, Haldor Topsøe

Article Summary

Steam reforming is the dominant technology for producing hydrogen or hydrogen-rich synthesis gas, for example for ammonia production. Feedstock may range from natural gas to kerosene.

The most recent reforming technology from Topsøe is the Haldor Topsøe Exchange Reformer (HTER). This operates in series or parallel with another reformer(s) and draws the necessary heat of reaction from the effluent gas from this source. It has already been commercialised in combination with an autothermal reformer at Sasol’s large facility in Secunda, South Africa, for conversion of coal to liquid fuels and other refined hydrocarbon products. The implementation of this technology at Sasol Synfuels has previously been described at length.¹

HTER in combination with a tubular reformer for a 206 000 Nm3/h hydrogen plant is in the design phase and will become reality within a few years. Application of this technology in ammonia plants in combination with a tubular reformer and a secondary reformer is also well suited. This application will be described in more detail in this article, including how HTER for ammonia plants can increase the reforming capacity in existing plants by up to 25% or, for a given capacity for a new plant, how it reduces the size of the primary reformer.

The recent development of the technology for ammonia offers a very high reforming temperature, resulting in low methane slippage without operating at a high steam-to-carbon ratio (S/C). A low methane slippage is crucial for ammonia plants as the methane will otherwise end up in the ammonia synthesis, affecting the rate of reaction. The overall S/C in the ammonia plant with both primary/secondary reformer and HTER is kept at the same level as for conventionalplants with only a primary/ secondary reformer. The low methane slippage is obtained with this technology, without increasing the outlet temperature from the primary reformer, because the required process air is still introduced in the secondary reformer (which leads to a lower methane slip from the secondary reformer), whereas only part of the feedstock passes through the primary and secondary reformer.

First industrial experience
Sasol Synfuels operates a very large facility in Secunda, South Africa, for the conversion of coal to liquid fuels and other refined hydrocarbon products. One of the important process steps on the way to the final products is the reforming section, in which a methane-rich gas is converted to synthesis gas on the basis of autothermal reforming using Topsøe’s burner technology. Sasol operates 16 parallel autothermal reformers (ATRs), each with a capacity equivalent to synthesis gas production for a 900-1025 t/d ammonia plant.

Sasol Synfuels is actively pursuing opportunities to increase its reforming capacity, in particular opportunities that allow an increase in capacity without an increase in oxygen consumption. This is achieved with HTER, whereby the sensible heat from the ATR effluent is effectively utilised for steam reforming rather than for raising less valuable steam. In this way, a significant capacity increase can be achieved and, at the same time, the carbon and hydrogen efficiencies of the process are improved. As a third and very important benefit, the hydrogen/CO/CO2 ratio of the synthesis gas can be controlled to match the stoichiometry of the downstream processes much better. Figure 1 is a sketch of the reformer section of the Sasol Synfuels plant prior to the revamp.

The layout is quite conventional, with the high-level heat of the synthesis gas being cooled in a steam-generating waste heat boiler placed downstream of the oxygen-fired ATR. One of the limitations to the process in the Sasol plant is the waste heat boiler inlet temperature as well as the duty. Both are stressed to or beyond the original design capacity and, in order to protect the boiler, steam is used to quench the gas before the entrance to the boiler. This is obviously not optimal from an energy perspective.

A more efficient process would be feed-effluent heat exchange at high temperatures, reducing the oxygen required inside the reformer or, more attractively, increasing the total synthesis gas product flow by driving an additional endothermic reforming reaction with the high-level heat. This process concept is shown in Figure 2, where the HTER taking care of the additional reforming is installed in parallel with the ATR. This revamp scheme was chosen for the Sasol Synfuels project.

The revamped unit was started up early in 2003 and an extensive, full-scale, 22-month-long demonstration run was performed at Sasol to prove the viability of the concept. The ATR/HTER pair has been in operation since and meets all expectations.

Since the initial startup, there have been no unforeseen stops related to the HTER, and it has been in continuous operation with the exception of planned shutdowns and the period of bypass operation. This has led to a high availability factor of 97%.

The unit has been shut down on a number of predetermined occasions for inspection. The main purpose of these inspections has been to verify the mechanical integrity given the high material temperatures experienced in this piece of equipment and to examine it for signs of corrosion, such as metal dusting. It has been concluded from these inspections that the reformer is performing well and the projected lifetime of the internals has been confirmed. As an additional benefit, these inspections have served to confirm its maintainability, and they have given valuable experience in shutting down and restarting the ATR/HTER pair.

During the test run, it was shown that the predicted capacity increase and conversion of the revamped unit was reached — in fact, there was some additional capacity in the unit compared to the expected 33% capacity increase. Likewise, pressure drops have been found to be stable and well within anticipated values.

Since this new equipment was built into the plant as a full-scale industrial unit, with no prior pilot-scale or side-stream operating experience, the operators were anxious to learn how the reformer pair behaved under industrial conditions with whatever upstream and downstream fluctuations that could be anticipated. It was found that the ATR/HTER pair was easily operable, but slightly more sensitive to fast cut back of the feed or the oxygen than the standalone ATRs.