"Contrary to the prior art teachings of uniform adsorbent bed
temperature during pressure swing air separation, it has been unexpectedly
discovered that these thermally isolated beds experience a sharply depressed
temperature zone in the adsorption bed inlet end..." (Collins,J., Air
Separation by Adsorption, US Patent 3,973,931, Aug. 1976)
The formation of a temperature "cold spot" in industrial O2PSa and
O2VSA plants (at times as much as -70°F) has been a source of mystery and
confusion in both academic and industrial communities. As reported by
Collins in 1976, the cold spot significantly decreased plant performance and in
many cases shut the plant down. It is reasonable to ask why O2PSA lab
experiments did not report this feature - the reason is that lab units, no
matter how well insulated, are not adiabatic to the extent of large diameter
industrial beds. Small heat leaks significantly affect the extent of the
"cold spot". The reasons for the formation of the cold
spot and its dependence on design and operating conditions was the basis for
this research project and is described below.
The O2VSA process is a good example of a bulk-gas adsorptive
separation. Since the adsorbents used for O2VSA adsorb the nitrogen (the
bulk component in air) and not the oxygen, and since adsorption is exothermic,
there is a substantial temperature swing in the adsorption beds. This
swing in temperature between adsorption and desorption can be as much as
10-15°C. Together with the temperature swing, the gas flows up and down
the bed will convect some of this energy through the bed. Since multiple
layered beds are usually employed (a guard bed is used to remove water and CO2)
in O2VSA, and since the adsorbents have different adsorptive capacities and
properties, it is common for unusual temperature profiles to establish
themselves within the adsorption beds.
In this project (completed in 1999), we investigated the mechanisms behind
the formation of temperature profiles in multi-layered VSA beds both
experimentally and theoretically. The pilot scale O2VSA unit was used to
measure the thermal profiles in the bed and this data together with detailed
numerical simulation to validate simple hypothesis as described in mechanistic
models. The figure below shows the temperature profile in a 2-layer bed at the
end of each step in a cycle at cyclic steady state
The prelayer occupies the first 0.3m of the bed and consists of inert
alumina. The main layer occupies the remaining portion of the bed and is
high capacity LiX zeolite. Since the prelayer is inert, the temperature
swing in the prelayer is very small, as seen in the figure. In the main
bed, the temperature swing is larger (5-10°C) relfecting the adsorption and
desorption occurring during the cycle. A pronounced "cold spot"
is seen at the interface between the layers.
In essence, the prelayer acts as a regenerative heat exchanger, cooled by the
desorption gas as it flows countercurrently over the prelayer and heated by the
incoming feed gas during the feed step. A balance is achieved when cooling
by the desorbed gas exactly matches heating by the feed gas as dictated by an
energy balance. This "cold spot" does not occur in single
layered beds since no "regenerative heat exchanger" is present.
It is important to note, however, that if the air to a single layered bed is not
dried, it will create a de facto prelayer since the moisture and CO2 will
contaminate the bottom few cm of the bed greatly reducing nitrogen adsorptive
capacity. A small prelayer is then inadvertently created again giving rise
to a "cold spot". Much more detail on the formation of the cold
spot and how to exploit it to our advantage in O2VSA design can be found in the
thesis of Simon Wilson and associated publications.
For more information on this project, contact Paul
Webley.