Process optimization for polyhydroxyalkanoate (PHA) production from waste via microbial enrichment cultures

Polyhydroxyalkanoates (PHA) are compounds naturally produced by microorganisms, with many industrial applications, either as bioplastics or as precursors for production of chemicals. Until now, industrial PHA production was conducted with pure strains of bacteria fed with well-defined feedstocks, making the overall process non-economically feasible. The last decades research on PHA was devoted on producing them in open enrichment cultures using wastewater as substrate, and making the process continuous, decreasing the production costs. Laboratory research with well-defined VFA-based substrates enables high accumulation of PHA, up to 90wt% of the total biomass. After demonstrating the potential for PHA production via enrichment cultures, research was devoted on applying this research. Several wastestreams and operational conditions were used to test PHA production on pilot scale and maybe on short notice on industrial scale. Until now, it was shown that also when using fermented industrial wastewater (e.g. paper mill, food) a high cellular PHA-content could be achieved.
The object of this thesis was to tackle problems associated with PHA production when operating the process using wastewater and to make it feasible to apply the strategy universally.
In the first chapter general information about PHA (process- and material- based) are given.
In the second chapter leachate from the source separated organic fraction of municipal solid waste (OFMSW) was evaluated as a substrate for polyhydroxyalkanoates (PHA) production. Initially, biomass enrichment was conducted directly on leachate in a feast-famine regime. Maximization of the cellular PHA content of the enriched biomass yielded to a low PHA content (29 wt%), suggesting that the selection for PHA-producers was unsuccessful. When the substrate for the enrichment was switched to a synthetic volatile fatty acid (VFA) mixture -resembling the VFA carbon composition of the leachate- the PHA-producers gained the competitive advantage and dominated. Subsequent accumulation with leachate in nutrient excess conditions resulted in a maximum PHA content of 78wt%. Based on the experimental results, enriching a PHA-producing community in a “clean” VFA stream, and then accumulating PHA from a stream that does not allow for enrichment but does enable a high cellular PHA content, enables a high cellular PHA content, contributing to the economic feasibility of the process. The estimated overall PHA yield on substrate can be increased four-fold, in comparison to direct use of the complex matrix for both enrichment and accumulation.
The success of enriching PHA-producers in a feast-famine regime strongly depends on the substrate utilized. A distinction can be made between substrates that select for PHA-producers (e.g. volatile fatty acids) and substrates that select for growing organisms (e.g. methanol). In the third chapter the feasibility of using such a mixed substrate for PHA-production was evaluated. A sedimentation step was introduced in the cycle after acetate depletion and the supernatant containing methanol was discharged. This process configuration resulted in an increased maximum PHA storage capacity of the biomass from 48wt% to 70wt%. A model based on the experimental results indicated that the length of the pre-settling period and the supernatant volume that is discharged play a significant role for the elimination of the side population. The difference of the kinetic properties of the two different populations determines the success of the proposed strategy.
Double-limitation systems have shown to induce polyhydroxyalkanoates (PHA) production in chemostat studies limited in e.g. carbon and phosphate. In the fourth chapter the impact of double limitation on the enrichment of a PHA producing community was studied in a sequencing batch process. Enrichments at different C/P concentration ratios in the influent were established and the effect on the PHA production capacity and the enrichment community structure was investigated. Experimental results demonstrated that when a double limitation is imposed at a C/P ratio in the influent in a range of 150 (C-mol/mol), the P-content of the biomass and the specific substrate uptake rates decreased. Nonetheless, the PHA storage capacity remained high (with a maximum of 84wt%). At a C/P ratio of 300, competition in the microbial community is based on phosphate uptake, and the PHA production capacity is lost. Biomass specific substrate uptake rates are a linear function of the cellular P-content, offering advantages for scaling-up the PHA production process due to lower oxygen requirements.
In the fifth chapter, PHA accumulating microbial enrichment cultures were established in an anaerobic/aerobic sequencing batch reactor (SBR) with glucose as sole substrate. The effect of different solid retention times (SRT; 2 and 4 days) on PHA accumulation were investigated. The experimental results revealed that at both SRT conditions, glucose was first stored anaerobically as glycogen with energy generation from lactate fermentation. Subsequently lactate and glycogen were fermented to acetate and propionate in the anaerobic phase. At 2 d SRT operation, during the aerobic phase the fermentation products where rapidly sequestered by aerobic PHA accumulating microorganisms. When (limiting) nutrients were applied under aerobic conditions PHA formation occurred under anaerobic conditions. At a longer SRT of 4 days the fermentation products where already sequestered in the anaerobic phase into PHA by glycogen accumulating organisms (GAO). In all systems the glucose uptake rate was very fast (-2.7 C-mol/C-mol/h), making it the primary competition factor. Under the conditions tested direct conversion of glucose to PHA was not possible.
In the sixth chapter some recommendations and questions that remain unanswered are addressed. As suggested, process could be improved by using a continuous system which would include a settling tank for the removal of carbon that is slowly consumed and leads to growth of “inert” biomass. The possibility of operating the process at lower oxygen concentrations, or completely anoxically, is also discussed.