(327d) Evaluation of Oxygen Separation Performance Featuring High Temperature Pressure Swing Adsorption With Perovskite-Type Oxygen Sorbent

As indicated in the previous report, it was predicted that oxygen and nitrogen separation from air, featuring high temperature pressure swing adsorption using oxygen sorbent based on perovskite-type oxide (La_0.9Sr_0.1Co_0.9Fe_0.1O_3-δ, LSCF1991), would demonstrate an extremely low electric power consumption rate.

In the present study, a small-scale testing apparatus was fabricated to measure the electric power consumption rate, as well as the oxygen production rate per loaded oxygen sorbent.

In this series of experiments, the adsorption pressure, desorption pressure, adsorption temperature, cycle time, etc. were varied, and the relationships between the afore-mentioned operational conditions and the electric power consumption rate and oxygen production rate per loaded oxygen sorbent were elucidated.

At the same time, assuming the future mass production of LSCF1991, an evaporation-to-dryness procedure of metal nitrates was adopted in this study to synthesize LSCF1991. The synthesized LSCF1991 powder was then formed into pellets as the oxygen sorbent for the small-scale testing apparatus. 

Based on the data obtained, the electric power consumption rate and the oxygen production rate per loaded oxygen sorbent were posited as the medium-scale unit for 1,000 m³N/h oxygen production.

To summarize the results of this experiment with a small-scale apparatus,

1)      As the oxygen adsorbed amount of LSCF is proportional to the logarithm of oxygen partial pressure, the desorption pressure of 1-5 kPa (relatively higher vacuum conditions) gives the lowest electric power consumption rate and the greatest oxygen production rate per loaded oxygen sorbent.

2)      With this isotherm, even when the adsorption pressure is increased, the oxygen production rate per loaded oxygen sorbent does show any significant rise. On the other hand, the electric power consumption rate increases with rising compressor power consumption. Therefore, it appears that the adsorption pressure is sufficiently low to compensate for the pressure loss.  

3)    The adsorption temperature should be set higher than 500 deg-C to maintain a practical oxygen production rate per loaded oxygen sorbent.

4)      At higher than 500 deg-C, as LSCF1991 shows a relatively high adsorption rate, and even during a shorter adsorption period such as 30 seconds, stable oxygen and nitrogen separation can be undertaken. For periods longer than 30 seconds, the oxygen production rate per loaded oxygen sorbent is inversely proportional to the adsorption time.

In conclusion, for this HTPSA-oxygen system, when a heat recovery ratio of 92% can be maintained, an electric power consumption rate of 0.2 kWh/m³N is predicted.

With a relatively rapid operation of 100 seconds per 1 cycle (i.e., adsorption period of 40 seconds, desorption period of 40 seconds, and other operations requiring 20 seconds), an oxygen production rate per loaded oxygen sorbent of 30 m³N-O₂/h/ton (500 kg/towerA2 tower system) can be expected.

Compared with cryogenic separation, which delivers the lowest electric power consumption rate at the present time (0.32 kWh/m³N), the HTPSA-oxygen system in this study has great potential to reduce the consumption rate in the near future.

The initial cost of the HTPSA-oxygen system in this study is the same as that for the PSA-oxygen system using nitrogen adsorbent, which is the representative medium-scale oxygen production unit available in the market.