Abstract
Polyhydroxyalkanoates (PHAs) are bio-based, biodegradable thermoplastics that may be substituted for conventional fossil-fuel based, non-biodegradable plastics in numerous applications. Currently, PHAs are produced under axenic conditions with pure cultures and refined feedstocks at a cost several times that of conventional plastics. To reduce these production costs, an alternative process using mixed microbial consortia and waste feedstocks (termed feast-famine PHA production or aerobic dynamic feeding) has been intensely researched over the past three decades. Fueled by the process’ economic promise and success at bench scale, the transition to pilot- and full-scale is underway; however, the process’ complexity and unreliability with carbon-limited waste feedstocks continues to pose challenges.The aim of the research presented herein was to contribute to the scale up of the feast-famine PHA production process on both a practical and theoretical level. First, prior bench-scale studies of PHA production from dairy manure were translated to a pilot-scale implementation and investigated for overall performance, stability, and polymer properties characterization. A consistent, economically viable PHA mass fraction was achieved despite less stringent operational control. Second, the sequencing batch reactor (SBR; which serves as the primary reactor configuration for the process) was generalized and a calculable expression for the average net specific growth rate of a functional phenotype during a non-zero duration division of the SBR cycle was derived. This equation was then compared with the predictions from the conventional approach to calculating this quantity (the inverse of the solids retention time) and was found to be more accurate. Third, the derived relationship was applied to the feast-famine PHA production process to develop a new heuristic to aid in process design and characterization, the average net specific growth rate of the non-PHA producing bacteria during the feast phase. The commonly used heuristics were compared with a large performance database aggregated from literature and found to be poor predictors of process performance. When the new average net specific growth rate of the non-PHA producing bacteria during the feast phase heuristic was compared with the same database, a clear correlation emerged. Finally, the results were summarized and considered in the context of the current state of feast-famine PHA production process.