(421bj) Oil Recovery System From Rolling-Mill Process by Wet Scrubbing

Material recycle has received increasing attention as global concerns about environment and resources grow. In this study, recycling entrained rolling-mill oil in local ventilation air was focused. Cold rolling mills consume a lot of oil for cooling and lubricating the rolls during the operation. Rolling oil, composed of kerosene-derived hydrocarbons, is usually used circulating between a rolling mill and a filtration apparatus, however some portion of the oil entrained into the local ventilation air. Several processes like activated-carbon adsorption and thermal oxidation are available for removing the entrained oil from the vent air, but not for recycling. Combination of wet scrubbing and distillation is a proven process to recover and recycle the oil in the vent air [1].

A packed tower type scrubber is usually employed for this process, though the adoptions are limited to large-scale plants with tons of oil consumption per day. For smaller scale plants common in Japan, the system should be more inexpensive in terms of capital cost as well as maintenance cost. In this study, another type of scrubber was examined: a very simple-structured scrubber of a rectangular cross section having a scrubbing liquid pool at the bottom and several curbed plates which divide the scrubber into a “nozzle section”, a “gas-liquid mixing section” and a “gas-liquid separating section”. This type of scrubber is commonly used as a dust collector using water as the scrubbing liquid, and is sometimes called “water-film” scrubber. Since it does not require any packing materials or scrubbing liquid sprays, it is expected to be easily maintained and highly tolerant to dust.

2. Experimental

The availability of “water-film” scrubbers for recovering the entrained rolling oil in the vent air was evaluated by providing part of vent air from an actual rolling mill to a pilot scale scrubber with a 150L scrubbing liquid pool. The provided flow rate of the vent air was 1,000-2,400 m³/h.

The vent air introduced to the scrubber is accelerated in the “nozzle section”, in which the nozzle-like flow path formed between one of the curved plates and the liquid level of the pool. The accelerated gas flow splashes the scrubbing liquid and increases the gas-liquid contact surface forming a number of scrubbing liquid droplets in the “gas-liquid mixing section”. The oil vapor in the vent gas is absorbed in the scrubbing liquid and the oil droplets, some of which is condensed in the scrubber, is adsorbed in the liquid droplets or gets together enough heavy to be separated from the gas. In the “gas-liquid separation” section, the gas stream decelerates and the most of the droplets of the oil and the scrubbing liquid fall to the pool trapping some portion of rolling oil in the scrubbing liquid. To remove droplets that still remain in the gas, a demister unit is available.

For the scrubbing liquid, di-(2-ethyl)hexyladipate (DEHA) was selected since it was reported to be efficient in absorbing volatile organic compounds [2] and the boiling point is reasonably higher than the end boiling point of the rolling oil.

The recovery efficiencies of the pilot-scale scrubber with and without a demister were evaluated by measuring the rolling oil concentrations in the vent air at the inlet and outlet of the scrubber operating at various gas temperatures and oil concentrations in the vent air.

In the pilot experiment, the following data were measured:

- Temperatures of the vent air at inlet and outlet of the scrubber

- Temperatures of the scrubbing liquid

- Oil concentrations in the vent air at inlet and outlet of the scrubber

- Concentrations of absorbed rolling oil in the scrubbing liquid

As the oil vapor and oil droplets transfer into the scrubbing liquid by different mechanisms, 6 sets of experiments were arranged to evaluate the recovery efficiency of the scrubber for gas phase and liquid phase; 3 sets with a demister and 3 sets with no demister.

3. Results and discussion

The rolling oil concentrations in the vent air were 0.6-1.7 g/Nm3 at the scrubber inlet where the saturated vapor concentrations were calculated to be between 0.3 and 0.5g/Nm3 based on a simplification that the oil is a mixture of alkenes and the mixture ratio was modeled from a gas chromatography analysis. Since the oil concentrations in the air were higher than the calculated saturated vapor concentrations in all sets of experiment, it was confirmed that the entrained oil in the air surely exists in both gas and liquid phase. However the estimated gas-liquid fraction of the oil varied from 25:75 to 53:47 presumably depending on the rolling-mill operations or the ambient temperatures, significance of evaluating the recovery efficiency of oil vapor and droplets separately was ensured. Averaged gas-liquid fractions of the entrained oil and total recovery efficiencies were as follows:

 - With a demister     30:70, 82%

 - Without a demister  48:52, 46% 

Theoretical lower limitations of the oil concentration at the scrubber outlet correspond to the saturated vapor concentrations of the oil dissolved in the scrubbing liquid which dependent on temperatures of the scrubbing liquid and the dissolved oil concentrations.

The estimation of the saturated vapor concentration of the dissolved oil was based on the assumption that it conforms to Raoult’s law; the vapor pressure of the dissolved oil in the scrubbing liquid equals the vapor pressure of pure oil multiplied by the mole fraction of the dissolved oil in the scrubbing liquid. The oil was assumed to be a mixture of alkenes as previously explained. Assuming that all of the droplets are possible to be collected in the demister and the oil concentration in the vent air is possible to be as low as the saturated vapor concentration of the dissolved oil, the limitations were around 0.08 g/Nm3 and the theoretical total recovery efficiency was estimated about 94% for the case with a demister of which the gas-liquid fraction of the entrained oil was 30:70 in average and the dissolved oil concentration in the scrubbing liquid was about 20 mol-%.

The vapor recovery efficiencies were calculated to be 38-56% and 47% in average. They were evaluated by the following formula:

 R_v = ( C_inlet_v– C_outlet_v) / (C_inlet_v – C_ideal_v)

where

              R_v is the recovery efficiency of the vapor in the scrubber

              C_inlet_v is oil vapor concentration in the air at the scrubber inlet

              C_outlet_v is oil vapor concentration in the air at the scrubber outlet

              C_ideal_v is the theoretical vapor concentration in the scrubber

The recovery efficiencies of the oil droplets were evaluated by the data without the demister, based on the assumption that the vapor recovery efficiency is always 47%. They were calculated to be 47-72% and 58% in average, and it was confirmed that additional gas-liquid separation is vital for this type of scrubber.

In the case of which the gas-liquid fraction of the entrained oil is 40:60 and the dissolved oil concentration in the scrubbing liquid is 20 mol-%, the total recovery efficiency was evaluated 54% for the scrubber itself and 79% for the scrubber with a demister where the theoretical recovery efficiency is 92% assuming almost all of the droplets can be collected.  Though the demister is at a risk of pressure drop increase or blockage by fine dusts which are not trapped in the scrubbing liquid, the recovery efficiency of “water-film” scrubbing is reasonably high comparing with the theoretical limitation.

4. Conclusion

Pilot-scale experiments were carried out to examine the capability of “water-film” scrubbing to recover rolling oil entrained in local ventilation air.

The total recovery efficiency employing a demister was 79% where the theoretical efficiency was calculated to be 92% in the specific circumstances of this study. Though a demister is a vital unit and required to be tolerant to dust blockage, the recovery efficiency was found to be favorably high.

The recovery efficiencies of the scrubber for the oil vapor and droplets were evaluated to be 47% and 58% in average, respectively. This evaluation enables the estimation of the performance for various gas-liquid fractions which dependent on rolling-mill operations, the ambient temperature and scrubbing liquid temperatures, and further assessment will be done.

5. References

[1] Axel E. Barten, Aluminium foil rolling, MPT  International 6(1997)68-77

[2] M.-D. Vuong, A. Couvert, C. Couriol, A. Amrane, P. Le Cloirec, C. Renner, Determination of Henry’s constant and mass transfer rate ofVOCs in solvents, Chem.  Eng. J. 150(2009)426-430.