Jul-2002

Improving hydrogen plant performance: Part I

A reliable and cost effective asset management strategy can increase hydrogen plant throughout as well as optimise operating conditions and energy efficiency

P V Broadhurst and P E J Abbott
Johnson Matthey Catalysts

Article Summary

The first step in meeting the increased need for hydrogen is to maximise production from existing refinery sources, including a dedicated hydrogen production plant based on steam reforming technology. In this case, capital expenditure is not required to deliver a moderate increment in hydrogen output from existing hydrogen plants. Optimisation of operating conditions to maximise throughput, combined with installation of optimised catalysts and absorbents to increase throughput and/or improve conversion are likely to maximise benefits of hydrogen production.
Optimisation of catalysts and absorbents and support/hold down media will be dealt with in Part II of this article in the Autumn PTQ.

Increasingly stringent fuel specifications in terms of aromatics and sulphur content in gasoline and diesel are driving up the need for hydrogen. In particular, EC legislation due for implementation in 2005 affects European refiners in this respect. Detailed modelling of differing hydrogen plant flowsheets has allowed general guidelines to be defined for optimising operating conditions, depending on the main drivers for the hydrogen plant operator. The main scenarios to be considered are:
— Maximum hydrogen output
— Minimised energy costs at a given production rate
— Integrity of hydrogen supply and assets at a given production rate.

A clear outcome from the study is that, if detailed recommendations are wanted, each plant situation is unique due to the site-specific nature of parameters such as hydrogen requirement, energy costs (fuel/feedstock/electricity), steam value/cost and asset management strategy. In addition to optimisation of the operating conditions, there are other low cost options for improvement of hydrogen plant throughput, reliability and cost effectiveness.

These options relate to:
— Use of higher activity catalysts to increase conversion leading to incremental hydrogen production
— Selection of lower pressure drop shapes/sizes of catalysts, absorbents and support/hold down media allowing increased throughput if the plant is pressure drop limited, or hydrogen re-compression requirements and costs can be reduced
— Choice of products, which better protect against catalyst poisoning to augment catalyst activity and life.

Optimising operations
The ways in which product selection can benefit hydrogen plant operation are reviewed using examples and case studies. The benefits of optimising operating conditions and catalyst choice will vary with each situation, but throughput increases of up to 10 per cent might be realised. Alternatively, similar savings in energy usage may result if existing hydrogen production rates are adequate.

 Of the options available for increasing hydrogen availability, steam reforming technology may be used to yield additional hydrogen in three ways. First, a new steam reforming hydrogen plant may be added. Second, existing steam reforming hydrogen plant capacity may be upgraded. Third, the current operation may be optimised within existing plant design constraints, which is the focus of the remaining discussion. In this case, capital expenditure is not required to deliver a moderate increment in hydrogen output from existing hydrogen plants.

Some reasons predicating the need for more hydrogen in refinery operations include:
— Use of more sour and/or heavier crude oils, requiring more hydrogen to desulphurise products and/or achieve the same product slate
— Changing local market requirements in terms of feed slate so that a greater proportion of the product slate is at the lighter end
— National and/or regional legislation to meet tighter specifications for environmental reasons. This directly links to the type of processes which consume hydrogen and which can be generally divided into three major categories: hydrotreating, hydrocracking and de-aromatisation.

Use of more sour crudes increases the amount of hydrotreating required. If the crude is heavier, more upgrading will be needed to produce the same amount and combination of lighter products for which increased hydrocracking is an option. Similarly, if local market conditions demand a shift in the product slate to lighter products, more hydrocracking is likely to be used. Finally, tighter specifications on products streams in terms of sulphur or aromatics will also drive the severity of hydrotreating and de-aromatisation processes.

Sources of hydrogen depend on the unit operations in and around the particular oil refinery flowsheet, and not every facility has every option. Most refineries have a hydrogen distribution system connecting the available hydrogen sources to the hydrogen-consuming processes.

This system has often been modified over time as the refinery has commissioned, de-commissioned or revamped various unit operations. When a refinery is running short of hydrogen, a logical first step is to critically assess the efficiency of the hydrogen distribution system. Changes to the distribution system may allow better use of the available hydrogen to narrow or close the hydrogen shortfall. If a hydrogen shortfall remains, then the options for adding capacity to the available hydrogen sources must be made if hydrogen needs are to be met.

Major refinery sources of hydrogen include hydrogen-rich off-gas derived from catalytic reforming units, “over-the-fence” type of hydrogen supply arrangement, hydrogen recovery processes and operation of an “on-purpose” hydrogen production plant.

The first two options have been well reviewed and discussed in the trade press and industry forums. The hydrogen recovery processes used to treat hydrogen-containing offgas from other unit operations include pressure swing absorption (PSA), membrane separation and cryogenic separation. Each has a relatively discrete set of operating conditions at which each technology is effective, so the technologies do not necessarily compete.