STEAM REFORMING TECHNOLOGY

A long history in steam

reforming at Billingham
The Catalyst Research Centre in Billingham, UK, has been at the epicentre for key
developments in the modern steam reforming process since the 1930s. J. Brightling of
Johnson Matthey Process Technologies recounts how a fundamental understanding of steam
reforming and expertise for catalysts over many years at Billingham has been instrumental in
the journey to today’s leading steam reforming technology.

J
ohnson Matthey’s KATALCOJM catalysts The first patents for methane reforming Billingham reformer also included extensive
for steam reforming are developed at catalysts for the reaction between steam heat recovery of both the process (make)
its advanced Catalyst Research Centre and hydrocarbon date back to 1913. When gas and flue gas, including a high preheat
in Billingham, UK. This location has been at the Haber Bosch process for ammonia temperature enabling maximum use to be
the epicentre for key developments in the production was initially developed, the made of the available catalyst volume by
modern steam reforming process which can hydrogen for the process was produced by reducing the preheating zone at the top of
be traced back as far as the 1930s. Over reacting steam with coke. However, inter- the tubes. The net result of these develop-
the years, the fundamental understanding est later centred on how the steam reform- ments was that the space velocity was a
acquired for this technology by the busi- ing of methane could be used to produce a factor of 2-3 times higher than had previ-
ness has resulted in many industry firsts. more economic source of hydrogen. Steam ously been achieved. This top-fired design
For example, it has resulted in the leading reforming was first used to produce hydro- (Fig. 2) was the most advanced application
commercially available steam reforming cat- gen in Baton Rouge in 1931 by Standard of steam reforming at the time.
alysts, KATALCOJM QUADRALOBE™, which Oil1, and soon afterwards a plant was com- The designs and drawings from the
are used by market leaders in technology missioned at Billingham, UK based on Bill- Billingham plants were used for eight ord-
development. ingham’s own process technology (Fig. 1). nance ammonia plants first built on behalf
Today Johnson Matthey continues to This process introduced novel process of the US Government in 1941. The con-
invest heavily in research and develop- features developed specifically for advanc- tracting companies involved were Chemical
ment, ensuring that its products and ser- ing the catalyst system performance, such Construction Corp. building Ozark Works,
vices directly meet the needs of customers as feedstock desulphurisation and Raschig El Dorado, Arkansas (four furnaces), Jay-
throughout the world. ring catalyst shaped reforming catalyst. The hawk Works, Baxter Springs, Kansas (two
furnaces), Cactus Works, Etter, Texas
Fig 1: Methane-steam plant flowsheet (Billingham, 1936) (three furnaces) and M.W. Kellogg Co.
who build Sterlington, Lousianna (two fur-
naces) and Alberta Nitrogen Products Ltd,
Calgary (two reformers). It was this trans-
mission of know-how from Billingham that
made a significant contribution to the rapid
growth of steam reforming technology in
the North American ammonia industry.
In these early processes gas reforming
took place at low pressure and tempera-
ture (1-4kg/cm2, 730-800°C) and catalysts
such as ICI 22-6 with 18% SiO2 performed
satisfactorily2.
Since those early days there have been
many advances in reforming technology,
which have allowed plant efficiencies and
reliability to be improved, to support these
technology improvements, catalysts have
had to be greatly developed. Throughout

44 Nitrogen+Syngas  327 | January - February 2014

 STEAM REFORMING TECHNOLOGY

Fig 2: Furnace cross section this time the expertise for catalysts from reforming at higher pressures and temper-
(Billingham 1936) Billingham, has been at the forefront of atures (up to 21kg/cm2, 800-850°C) and
industry advances. the first single train ammonia plants were
built producing 300 t/d. Soon however it
Steam reforming catalyst was discovered that silica migrated under
these conditions and this led to low silica
requirements (<0.2% SiO2) precipitated catalyst sup-
The ideal catalyst has to combine a num- ports made in standard ring shapes.
ber of essential properties: In the mid-1960s the “modern single
● high activity for the desired reaction (i.e. stream ammonia plant” was established.
methane to hydrogen/carbon oxides); Between 1965 and 1975 approximately 58
● maximum selectivity for the preferred million t/a steam reforming based ammonia
reaction (i.e. not methane to carbon); plants were built of which the majority were
● good strength; >1000 t/d plants with reformers operating
● good gas flow properties – heat trans- at even higher pressures and temperature
fer/ pressure drop; (33-34kg/cm2, 790-870°C). However insta-
● stability of the above – resistance to bility of these precipitated catalysts caused
poisons and plant upsets. problems with both hot bands and increas-
There have been continued improve- ing pressure drop seen when the plants
ments to all of these properties, as illus- were operated at increased throughputs
trated by some of the significant changes and heat fluxes. The impregnated calcium-
that have occurred in gas reforming cata- aluminate ceramic catalyst KATALCOJM 57-3
lysts over the past 75 years. (ring) was therefore developed; the support
From 1955 until 1965 the on-going strength allowed smaller size pellets to be
trend for process intensification led to used resulting in increased activity. The use
of calcium aluminate2 was pioneered at Bill-
Fig 3: Calcium aluminate spinel ingham. This ceramic support material has
significant advantages in terms of stability
under plant operating conditions including
its high retained strength. The calcium in
the support is a basic additive that removes
acidic sites which promote carbon. Being a
stable spinel structure calcium aluminate is
an excellent support for holding well distrib-
uted nickel giving a high stable nickel surface
area and low activity deterioration (Fig. 3).

First shaped reforming catalyst, ICI catalyst The potash story – first alkalised
22-1.
reforming catalysts
Fig 4: Simplified reaction routes in reformer tube It had long been known3 that the incor-
poration of alkali into a steam reforming
inlet 400-500°C
catalyst promotes the absorption of water
cracking on to catalyst surface and this accelerates
(acid sites) hydrocarbon <650°C catalytic cracking (Ni) the carbon-steam reaction. Figure 4 shows
feedstock
4
>650°C thermal cracking 1 a simplified arrangement of the reaction
routes possible in different sections of a
intermediates 1 25% reformer tube4. Alkali (incorporated within
olefin polymerisation/dehydrogenation: olefins
carbon 5 parafins to the catalyst support) is an effective
CH4 cracking (acid sites) H2 carbon removing agent – this technology
2 steam cracking 50% was first used in naphtha based reforming
processes which were widely licenced by
intermediates 2
oxygenated species 1970 with over 500 reforming plants built
CH4 75% by leading engineering companies (e.g.
6 H2
Kellogg, Uhde, Foster Wheeler, Selas).
carbon steam reforming/ 3 Between 1975 and 1985 many plants,
removal product gas equilibration
CH4, CO, CO2, particularly North American gas based
C + H2O → CO + H2 H2, H2O CH4 + H2O CO + 3H2 (Ni)
outlet 700-850°C reforming plants, were pushed to higher
(alkali) CO + H2O CO2 + H2 (Ni) rates (up to 135% of design). Feedstocks
also tended to contain higher proportions of

Nitrogen+Syngas 327 | January - February 2014 47

 STEAM REFORMING TECHNOLOGY

Fig 5: Tube wall temperature improvements heavier hydrocarbons (C2-C5) and this combi-
nation of rate and feedstock resulted in the
reappearance of the hot band phenomenon.
KATALCOJM 25-3 KATALCOJM 25-4 KATALCOJM 25-4Q
In order to address this, a new lightly alka-
900 lised formulation KATALCOJM 25-3 (ring) was
developed. The catalyst structure contained
855°C
a low level of potash in the form potassium
850 835°C
beta-alumina (KAl3O8) that was incorporated
maximum TWT, °C

825°C into the calcium aluminate support structure

800
in a way that slowly releases alkali at the
required concentration to ensure continuous
750
carbon-free operations4.
This was a key breakthrough for the
700
industry, establishing a catalyst that was
ideally suited for handling variable gas
650 feedstocks up to LPG. It resolved operat-
0 10 20 30 40 50
ing issues where there had been problems
distance down tube, %
with hot bands or poor performance attrib-
utable to carbon formation from the heavier
hydrocarbons or methane cracking, particu-
Fig 6: KATALCOJM 57-4g (4-hole) vs KATALCOJM 57-4XQ (QUADRALOBE) larly in the more highly stressed furnaces.
The fundamental understanding of cata-
lyst chemistry has meant that the proven
solution (alkali incorporated into the cata-
lyst’s support chemistry) has changed little
since the 1970s. With the engineered loca-
tion of the Ni together with K in the support
KATALCOJM KATALCOJM structure of KATALCOJM 25-3, the product
57-4g (4-hole) 57-4XQ (QUADRALOBE) gave carbon resistance with no loss of activ-
Nominal OD, mm 19 19.7 ity compared to a non-alkalised product. Sam-
Nominal hole diameter, mm 5.5 5.5 ples of discharged catalyst also proved that
Length, mm 19.6-20.4 19.6-20.4 the alkali release was uniform and slow with a
Nominal TBD, kg/l 0.9 0.9 high reservoir remaining in the spent catalyst
Average mean horizontal (radial) During the next era (1985 to 2000) trends
crush strength, kgf >56 >100 in plant developments focused on ever larger
Relative pressure drop 1.0 0.76 more efficient plants such that by the year
2000, 2000 t/d had become the norm for
a new plant. Existing operators continued
Fig 7: Modern design tools and reformer monitoring devices to push limits by “asset sweating” seek-
ing longer times between overhauls. This
resulted in the development of more efficient
shaped catalysts such as KATALCOJM 25-4
(4-hole). By combining an understanding
of carbon formation and carbon resistance
together with computer modelling, it became
possible to offer customised catalyst load-
ings to provide tailored solutions maximising
both catalyst and tube life.
Beyond the millennium to the present
day the same familiar trends remain with
the focus on improving the strength and
gas flow properties for steam reforming
catalysts. Based on very well proven mate-
rials, the development of the QUADRALOBE
range in the 2000s enabled greater activity
and heat transfer capacity to be added to
tubes further improving efficiency or reac-
tion and reducing the tube wall tempera-
ture. The net result has been increased life
of both catalyst and tubes (Fig. 5).

44 Nitrogen+Syngas  327 | January - February 2014

 STEAM REFORMING TECHNOLOGY

Fig 8: CF Industries Donaldsonville The QUADRALOBE pellet benefited in

refined shape characteristics as a result
of the use of computer based design tools
(FEA, CFD) making it possible to offer
improvements in catalyst strength and
pressure drop as set out in Fig. 6.
These catalyst shape improvements
based on the sure foundation of the well
proven alkalised (KATALCOJM 25-4 series)
PHOTO: THYSSENKRUPP INDUSTRIAL SOLUTIONS

and non-alkalised (KATALCOJM 57-4 series)

support chemistry, when combined with
improved design and monitoring tools (Fig.
7), have allowed boundaries to be pushed
even further.
Today the largest single stream ammo-
nia plants use the Uhde Dual Pressure Pro-
cess with plants manufacturing 3,300 t/d,
operating in Saudi Arabia and being con-
structed by CF at Donaldsonville in the USA5.
As illustrated in Fig. 8, the steam reformer in
the foreground remains at the heart of this
Fig 9: The cumulative users of Johnson Matthey’s natural gas based huge complex underpinning the vital supply
reforming products of hydrogen for the ammonia process.
The catalyst in the reformer has a direct
impact on the plant rate, reformer tube life,
ring 4-hole QUADRALOBE™ methane slip and the risk of carbon for-
catalyst catalyst catalyst
mation, all of which can have a significant
and detrimental impact on reformer perfor-
alkalised mance and operation. Improvements in the
cumulative catalyst sales

non-alkalised catalysts can help to extend the operating

envelope in areas such as plant rate, run
length, operating severity and resistance
to plant upsets. For this reason KATAL-
COJM 25-series catalysts have been widely
adopted (Fig. 9). Understanding the chem-
istry of the catalysts used in this process
is fundamental to reliable and efficient
operation of these process plants.
KATALCOJM QUADRALOBE catalysts
75 78 81 1984 987 990 993 1996 1999 2002 2005 2008 2011 2013
are installed in ammonia, methanol and
19 19 19 1 1 1 hydrogen plants around the world (Fig.
year
10) offering excellent recovery from plant
upsets resulting in flexibility and longevity
in operation for its customers.  n
Fig 10: Number of charges of KATALCOJM QUADRALOBE steam reforming
catalysts installed around the world
References
1. Gard N.R.: “Thirty years of Steam reforming
– A review of ICI developments and experi-
ence,” Nitrogen No39 (1966).
2. Hicks T.C.: “Steam reforming catalysts - The
importance of choosing the correct support,”
ICI Catalysts paper 58 (1979).
3. Bridger G.W.: “Catalyst for steam reforming
of hydrocarbons”, AIChE Symposium Safety
in Ammonia (1975).
4. Hicks T.C.: “Steam reforming,” ICI Catalysts
Symposium BRASIM 85, Rio de Janeiro (1985).
5. Will A.: “Nitrogen expansion in North Amer-
ica,” Nitrogen+Syngas Conference, Berlin,
Germany (2013).

55 Nitrogen+Syngas  327 | January - February 2014