Jul-2001

Cost efficient revamps in hydrogen plants

Revamping of existing equipment is often seen as the cheapest and most cost effective solution, but this calls for a thorough knowledge of the plant, its capabilities and its limitations

James D Fleshman, Foster Wheeler USA

Article Summary

Whether oil prices or processing margins are favourable or unfavourable, the pressure to get more hydrogen at lower cost and with fewer resources never seems to go away. Depending on the site and the amount of hydrogen required, the solution can range from a new plant to hydrogen purchased over the fence, to revamping an existing plant. Where an increase of 10–50 per cent of existing steam reforming capacity is required, revamping is often the cheapest and most cost effective way to do this.

Rather than starting from a particular technology or a particular solution, changes to the plant and the investment cost can be minimised by a thorough knowledge of the plant and its limitations. Hydrogen plants usually contain areas of flexibility or overcapacity which can be used to unload tighter sections of the plant.

Overcapacity may have been built into a unit due to uncertainties in design, such as variation in equipment performance or the need to process a variety of feedstocks. It may also be due to changes in technology or to overall economics, such as changes in energy cost or product value.

As capacity is increased in a hydrogen plant, limitations become apparent which are often due to problems in another area. The high temperature in the methanator may be due to flooding in the CO2 absorber, or a lack of draught  in the reformer may be due to poor performance of a combustion air preheater. Identifying the sources of these problems and correcting them can have a large payout.

Hydrogen plants fall into two groups, depending on the type of CO2 removal system. Plants built since the late-1980s tend to use pressure swing adsorption (PSA) for purification, while older hydrogen plants used wet scrubbing. Many of the problems encountered in revamps, particularly in the reforming section, are common to both types.

Figure 1 shows a PSA-based hydrogen plant, with reforming, a single high-temperature shift converter, gas cooling, and finally, purification by PSA. The product purity of a PSA plant is not linked to reforming conditions as in a wet scrubbing plant.

Figure 2 shows a wet scrubbing hydrogen plant. The most common configuration uses two stages of shift conversion, followed by wet scrubbing and methanation. A variety of scrubbing processes are used, typically based on an amine or on potassium carbonate. Product purity depends mainly on the amount of methane leaving the reformer.

Sources of capacity
The newer plants were generally built with smaller design margins than the older ones. This is due to a number of factors: as more experience was gained in design, and as steam reforming lost its high-tech aura, the designs became leaner. For example, hydrogen is now often regarded more as a utility than a refinery unit. New plants have relied more on engineering contractor standards, and were often bid as lump-sum turnkey projects where there was a strong incentive to reduce overcapacity.

While the older plants may have had excess capacity when they were first built, by this time many have already been revamped at least once, and the catalyst tubes may have been upgraded to newer materials. In either case, the “easy capacity” has generally been used up, and it is necessary to look more closely for extra production.

Each plant tends to have areas where this search for extra capacity will pay off, and areas where the limits are more fixed. A steam reforming plant contains three functional areas: reaction, heat transfer, and separation. They are linked by hydraulics or flow capacity. Each of these areas reacts differently to changes in throughput, and since they are all related, changes in one area affect the others. Within the range of interest for most revamps, reaction is typically only slightly affected by operating rate. Catalyst volume is also affected by heat transfer or pressure drop requirements, such as in the reformer, or by a requirement for a particular run length. Existing reactors are therefore often able to operate at increased rate.

The situation with heat transfer varies. The gas heat transfer coefficient increases with velocity, so existing exchangers are often adequate as the capacity is increased. Whether or not extra capacity is needed will depend on the service, on design margins, and on operating flexibility. Air coolers in a CO2 removal unit, for example, often need to be modified by adding surface or increasing fan horsepower. Heat exchangers are also relatively small items, and it may be attractive to replace heat exchangers for other reasons: to reduce system pressure drop for example.

Separation equipment tends to have fixed limits, and can be costly to change. For example, the CO2 absorber and solvent regenerator towers in a wet scrubbing unit are difficult to replace. Fortunately, it is often possible to make changes to internals, or to the solvent composition, to provide more capacity.

Hydraulic capacity – pressure drop – shows up in each of these areas. Reformer pressure drop, for example, often becomes limiting at high flow rates, requiring a change in catalyst, or providing another reason for larger diameter catalyst tubes.

Reformer
The reformer itself is a good example of how these areas fit together. The actual production of hydrogen occurs in the radiant section. When the operating rate is increased, reformer firing is increased along with it. Particularly with a Terrace Wall or a side-fired unit, the limit to capacity will generally not be apparent in the radiant section itself.