and accelerate the redox reaction to form iron hydroxide and thereby cause serious damage to the equipment. However, even when alloys are operated dry, iron-con- taining alloys can be chlorinated or oxychlorinated when operating them at a temperature range above 160°C to 204°C, according to equations (2) and (3).⁷ HCl reacts in a gas phase reaction to form a metal halide and hydro- gen. Furthermore, some metal halides are volatile, espe- cially FeCl 3 , which sublimates at 120°C, contaminating the downstream unit equipment:
100
HCl:O @ 0.5 barg 0.25 0.5 1 2
95
90
85
80
75
70
(2)
Fe + 2HCl ↔ FeCl₂ + H₂
65
(3)
FeCl₂ + HCl ↔ FeCl₃ + 0.5 H₂
Van-’t-Ho
60
d ln ( k ) dT
RT ∆ H R
55
=
Various alloy metals (such as chromium) are prone to form volatile salts that can move through the unit, depending on the surrounding conditions. Once the metal halide deposits pollute the downstream section of the unit, the major issue is their strong hygroscopic nature (deliquescence). The for- mation of water as a product during the Deacon reaction cannot be circumvented. If water is flowing through those substances, it will be strongly absorbed until a liquid phase is formed that consecutively acts as a corrosion hot spot. The affected part of the unit must be cleaned or replaced to remove hygroscopic deposits. Plate design and experimental conditions The Deacon reaction was performed using three standard catalyst systems known by the art, and it is shown in the scheme of the reactor packing design in Figure 2 . Each single reactor packing consists of an inlet and outlet zone filled with corundum and a catalyst zone, which was placed in between, separated by glass wool layers. For a more comprehensive description, the reactors are divided into
50
200
250
300
350
400
450
500
Reaction temperature (˚C)
triplets. Different residence times for each catalyst have been realised by filling different catalyst amounts (triplet 1, 2, 3), including one reactor per material equipped with a quartz thermowell to record online temperature profiles. Triplets 4 and 5 represent experiments with a variation of inert dilution and duplicate positions to assess the reactor- to-reactor reproducibility. An inert position was considered to measure the feed composition. The catalysts have been screened at a reaction temperature ranging from 360°C to 410°C at an HCl:O₂ molar ratio of 0.125-2. In addition, sup- plementary runs at an elevated pressure of 3.5 barg have been processed. Figure 3 HCl equilibrium conversion vs reaction temperature at 0.5 barg and different HCl:O₂ molar feed ratios
100
90
Equilibrium conversion
360
360˚C
80
355
50m ± 2 K
70
60
380
50
380˚C
40
375
Catalyst A
30
Catalyst B
20
400
400˚C
10
Catalyst C
395
0
320
340
300
350
400
420
100 110 120 130 140 150 160 170 180 190 200
Reaction temperature (˚C)
Reactor length (mm)
Figure 4 HCl conversion vs reaction temperature of all catalysts at 0.5 barg, HCl:O₂=0.5 mole/mole and a residence time of 600 kg * s/m³ (left), online temperature profile for reactors 4, 7, and 10 at 360°C, 380°C, and 400°C (target temperature), 0.5 barg, HCl:O₂=0.5 mole/mole and 600 kg * s/ m³, processing at a maximum temperature difference of +/-2K (right)
31
PTQ Q4 2022
www.digitalrefining.com
Powered by FlippingBook