Jun-2017

Simplifying hydrogen production

Catalyst systems developed for the purification and reforming sections of the hydrogen plant target significant barriers to efficient production.

DIANE DIERKING and KEN CHLAPIK
Johnson Matthey

Article Summary

Hydrogen continues to be a vital element in achieving clean fuel requirements globally. Its strategic value in a refinery is strongly related to the reliability and availability of hydrogen to the hydroprocessing units dedicated to producing fuel products and components. This value can vary greatly depending on the hydroprocessing unit being supplied. For example, the value of hydrogen to a hydrocracking unit is 2.5 times greater than that in a hydrotreating unit. There are a number of barriers that exist in hydrogen production that affect reliability and yet they are accepted as part of day to day operation. Trying to address these online typically means a curtailment of production, which quickly results in losses in high valued hydroprocessing units upward of $12.50/1000 scf hydrogen. These barriers can be addressed in hydrogen plant design, but result in a more complex and expensive flowsheet to accommodate the additional catalyst volume, controls, and redundancies. Many refineries are currently limited in capital and so building additional hydrogen capacity is being postponed while near term increased production is still needed.

Johnson Matthey has been developing catalyst technologies that can address and remove barriers to simplifying hydrogen production. This article presents two novel catalyst technologies that address these barriers and for the longer term provide a reduced capital flowsheet, simplifying hydrogen production while ensuring high reliability in operation.

Reliability barriers
Common factors affecting the reliability of a hydrogen plant are poisoning of the catalyst, feedstock changes, variable throughput and heat transfer efficiencies, as well as the consequences of a plant trip. The following discusses these factors in more detail.

Several of the catalysts in a hydrogen plant flowsheet are sensitive to poisoning, especially by sulphur. In most cases, the sulphur is removed to very low levels by the purification section. However, upsets occur when the sulphur level spikes well above the design value, thereby allowing sulphur into the plant. If a pre-reformer, medium temperature shift (MTS) or low temperature shift (LTS) vessel is included in the flowsheet, the activity of these catalysts might be affected such that it is necessary to replace them on an emergency basis. An unscheduled turnaround on these units alone can cost $1 million with $3 million product loss for a 50 mmscfd plant.

Natural gas can be purchased from different suppliers. When sources are switched, feed composition changes can occur, as well as differences in the sulphur speciation. Should the feed gas become heavier, the steam-to-carbon (S:C) ratio will decrease, possibly leading to carbon/coke formation in the primary reformer. A steaming procedure can be effective in removing light carbon; however, in some cases the event is severe and the reforming catalyst must be replaced. While the purification section will handle many different sulphur species, some are more difficult to convert to hydrogen sulphide (H2S). Should these more difficult species be present, the sulphur might not be removed and can poison downstream catalysts. Steaming can cost $600000 in lost hydrogen production for a 50 mmscfd plant, while unscheduled replacement of the reforming catalyst can cost $1 million and $4 million of lost hydrogen production.

Some operators process refinery off-gas (ROG) streams in the hydrogen plant, either by design or as a cost-effective alternative to purchasing natural gas. These streams vary widely in composition, not only from site to site, but can change throughout the day at a given site. While blending an off-gas stream with natural gas can help diminish compositional changes, these streams still present a challenge to the operator in controlling the S:C ratio, thereby managing the avoidance of carbon in the primary reformer. While steaming the catalyst might help the catalyst recover, there are costs to lost production as noted above.

Traditional steam methane reforming catalysts are shaped materials that utilise a robust ceramic base that has a high geometric surface area (GSA) onto which the active metals are dispersed. The catalyst also influences the heat transfer within the primary reformer tubes. Heat must be transferred from the hot tube metal across a static gas ‘film’ at the tube wall and then further into the centre of the tube. The shape and random packing of pelleted catalysts affect the efficiency of this heat transfer. Smaller pellets provide the most points of contact at the internal tube surface, thereby breaking up the gas film and delivering a higher efficiency heat transfer; however, smaller pellets will add to the pressure drop across the primary reformer. Changes in throughput can challenge the heat transfer capability of the reformer, affecting the overall energy used to produce hydrogen as well as the hydrogen production.

As the reforming catalyst ages, the active nickel sites sinter and the catalyst becomes less active over time. This translates to higher tube wall temperatures (TWT) as less heat is removed for the endothermic conversion of hydrocarbons. As such, a catalyst with poor heat transfer properties will contribute more to increased tube temperatures over time. The catalyst in the primary reformer is often changed out not because of degradation in hydrocarbon conversion, but because TWTs in the furnace are close to or exceeding the design limit. The result is that catalysts with better heat transfer properties will have longer lives; this means for a 50 mmscfd plant the value of the additional hydrogen produced between catalysts change-outs could be upwards of $20 million/y.

The pressure drop across the catalysts installed in a hydrogen plant will increase over time, which can be due to breakage or fouling. The primary reforming catalyst contributes to the majority of the catalyst related pressure drop in the plant. Often, pressure drop growth is not uniform across the furnace, resulting in flow variation across the tubes; those with lower flow will become hotter. An increase in the pressure drop across the plant can result in a lower product pressure, thereby increasing compression cost, or can require a reduction in throughput if there is a limitation on pressure drop through the unit. If changing the catalyst has to be delayed until a turnaround can be taken, curtailment of production can cost $9 million/y for a 
50 mmscfd plant.

This article presents two catalyst technologies, Katalco 33-1 and Catacel SSR, that overcome these barriers to ensure high reliability in hydrogen plant operation. Catacel SSR addresses issues specific to the primary reformer, while Katalco 33-1 provides improvements in the purification section to help protect downstream catalysts.

Simplifying purification
The purification section of the hydrogen plant is the first line of defence to ensure the highest reliability. Considering the feed and poisoning issues that have been discussed, it is understandably difficult to simplify this section to cover all possibilities. The most common poisons present in natural gas feeds are sulphur compounds, which can include H2S, carbonyl sulphide (COS), mercaptans, organic sulphides and disulphides, and on rare occasions thiophenes. Common odorants being used in natural gases are tertiary butyl mercaptan (TBM), dimethyl sulphide (DMS), and tetrahydrothiophene (THT). In many situations, the only poisons present are sulphur compounds and the purification section is designed to remove these to an acceptable level. Less common poisons might be present, such as chlorides, as well as more unusual poisons, including mercury or arsenic compounds, for which the purification section design must be adapted. In a modern hydrogen plant, the conventional purification section design comprises between two and four stages. The first stage is hydrodesulphurisation (HDS) in which organic sulphur compounds are hydrogenated catalytically to H2S. Any organochlorine compounds react analogously to give hydrogen chloride (HCl). Where chlorides are present in the feed, the second stage uses a chemical absorbent to remove the HCl. The third stage is removal of H2S using absorbents based on zinc oxide (ZnO) which react to form zinc sulphide. The degree of desulphurisation achieved with ZnO beds is typically 30 ppbv to 50 ppbv, with end of run usually defined as a sulphur slip of 0.1 ppmv (100 ppbv). Some systems have a fourth stage to add a final polishing, or ultra-purification, step to remove sulphur to around 10 ppbv.