Techno-economic evaluation of organic solvent nanofiltration : recovering dewaxing solvent from lube-oil

Dorrington, Bella Ann (2020-03)

Thesis (MEng)--Stellenbosch University, 2020.

Thesis

ENGLISH ABSTRACT: To recover the solvent from dewaxed lube-oil, classical separations such as evaporation and distillation are typically used. In the past decade separations that consume less energy and utilities have been investigated. One such technology is organic solvent nanofiltration (OSN). OSN is the size-exclusion of species over a membrane. Many studies have shown that OSN separation consumes less energy than evaporation and distillation. However, few have compared the cost of OSN to classical separations. Therefore, the aim of this investigation was to simulate and compare the cost and energy demand of a conceptual OSN unit with classical separation for the recovery of solvent from dewaxed lube-oil. Cost was expressed as the annualised capital plus total operating costs per year. Energy demand is the total energy consumed per volume of solvent recovered. Max-Dewax was the first and largest OSN membrane unit to recover solvent from lube-oil. For this investigation, a conceptual OSN unit was developed based on Max-Dewax to treat a 480 m3/h feed containing 19% lube-oil, and 46% methyl-ethyl ketone (MEK) and 35% toluene as solvents. Membrane permeate rates for MEK, toluene and lube-oil were 360, 91 and 1 mm/h. Long-term membrane stability was based on a 28% average decline in flux over a membrane life of 24 months. At an operating pressure of 42 bar, the OSN unit could achieve a 51% recovery of solvent at a 99% purity and average flux of 11 L/m2/h. A sensitivity analysis revealed that specific membrane cost, membrane life, flux decline over membrane life, operating pressure and membrane permeate rates had the biggest influence on the cost and energy demand of the OSN unit. The solvent remaining in the concentrated product from the OSN unit was further recovered by evaporation and distillation. The complete solvent recovery unit is referred to as hybrid-OSN and could recover 98% solvent at a 99% purity. The simulated cost of the hybrid-OSN unit for a specific membrane cost of 3200 ZAR/m2 effective area was 127 ZAR/m3 solvent recovered per year. The energy demand was 177 kWh/m3 solvent recovered. Compared to a classical separation including only evaporation and distillation, the energy demand of the hybrid-OSN unit was 64 kWh/m3 less. However, the hybrid-OSN unit was 40 ZAR/m3 more than the classical separation. Process simulations revealed that for a membrane cost of 990 ZAR/m2 and higher membrane permeate rates for MEK and toluene of 540 and 137 mm/h, the energy demand of the hybrid-OSN unit was 140 kWh/m3 and the cost equal to the classical separation. Extending membrane life to 72 months at a membrane cost of 1600 ZAR/m2, the cost of the hybrid-OSN and classical separation were also equal. Moreover, for a membrane cost of 630 ZAR/m2 and 14% decline in flux after 24 months, the energy demand of the hybrid-OSN unit was 163 kWh/m3 and the cost was equal to the classical separation. Typical results from process simulations showed that under certain conditions, OSN membranes bare the economic and energy potential to assist classical solvent recovery units. However, the profitability of a hybrid-OSN unit largely depends on the cost of membrane modules and long-term membrane performance such as flux decline over membrane life.

AFRIKAANSE OPSOMMING: Om die oplosmiddel uit ontwasse smeerolie te herwin, word klassieke skeidings soos verdamping en distillasie tipies gebruik. In die laaste dekade is skeidings wat minder energie en utiliteite gebruik ondersoek. Een so ’n tegnologie is organiese oplosmiddelnanofiltrasie (OSN). OSN funksioneer deur middel van die grootte-uitsluiting van spesies oor ’n membraan. Baie studies wys dat OSN-skeiding minder energie as verdamping en distillasie gebruik. Min het egter die koste van OSN teenoor klassieke skeidings vergelyk. Daarom is die doel van hierdie ondersoek om die koste- en energievereistes te simuleer en ’n konsepsionele OSN-eenheid met ʼn klassieke skeiding vir die herwinning van oplosmiddel uit ontwasse smeerolie, te vergelyk. Koste is uitgedruk as die jaarlike kapitaal plus totale bedryfskostes per jaar. Energievereistes is die totale energiegebruik per volume oplosmiddel herwin. Max-Dewax was die eerste en grootse OSN-membraaneenheid om oplosmiddel uit smeerolie te herwin. Vir hierdie ondersoek is ’n konsepsuele OSN-eenheid ontwikkel gebaseer op Max-Dewax om ’n 480 m3/h voer, wat 19% smeerolie en 46% metieletielketoon (MEK) en 35% tolueen as oplosmiddels gebruik het, te behandel. Membraan deurlaatbaarheid vir MEK, tolueen en smeerolie was 360, 91 en 1 mm/h. Langtermyn membraanstabiliteit is gebaseer op ’n 28% gemiddelde afname in fluks oor ’n membraanleeftyd van 24 maande. By ’n toegepaste druk van 42 bar, het die OSN-eenheid ’n 51% herwinning van oplosmiddel by ’n 99% suiwerheid en gemiddelde fluks van 11 L/m2/h behaal. ’n Sensitiwiteitsanalise het ten toon gestel dat spesifieke membraankoste, membraanleeftyd, fluks afname oor membraanleeftyd, toegepaste druk en membraan deurlaatbaarheid die grootste invloed op die koste- en energievereistes van die OSN-eenheid gehad het. Die oplosmiddel wat in die gekonsentreerde produk van die OSN-eenheid oorbly, is verder herwin deur verdamping en distillasie. Die algehele oplosmiddelherwinningseenheid word na verwys as hibried-OSN en kan 98% oplosmiddel by 99% suiwerheid herwin. Die gesimuleerde koste van die hibried-OSN-eenheid vir ’n spesifieke membraankoste van R3 200/m2 doeltreffende area was R127/m3 herwin per jaar. Die energie-aanvraag was 177 kWh/m3 oplosmiddel herwin. Vergeleke met ’n klassieke skeiding wat slegs verdamping en distillasie behels, is die energievereiste van die hibried-OSN-eenheid 64 kWh/m3 minder. Die hibried-OSN-eenheid was wel R40/m3 meer as die klassieke skeiding. Prosessimulasies het bekend gemaak dat vir ’n membraankoste van R990/m2 en hoër membraan deurlaatbaarheid vir MEK en tolueen van 540 en 137 mm/h, die energievereiste van die hibried-OSN-eenheid 140 kWh/m3 was, en die koste gelyk aan die klassieke skeiding. Met vergroting van membraanleeftyd na 72 maande by ’n membraankoste van R1 600/m2, was die koste van die hibried-OSN en klassieke skeiding ook gelyk. Verder, vir ’n membraankoste van R630/m2 en 14% afname in fluks na 24 maande, was die energievereiste van die hibried-OSN-eenheid 163 kWh/m2 en die koste gelyk aan die klassieke skeiding. Tipiese resultate van prosessimulasies het gewys dat onder sekere omstandighede, OSN-membrane dieselfde ekonomiese en energiepotensiaal het om klassieke oplosmiddelherwinningseenhede by te staan. Die winsgewendheid van ’n hibried-OSN-eenheid is egter grootliks afhanklik van die koste van membraanmodules en langtermyn doeltreffendheid soos fluks afname oor membraanleeftyd.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/108159
This item appears in the following collections: