Case Study: Dairy Waste Upcycling

Introduction

In 2012 there was approximately 9.3 million head of dairy cows in the USA which produced 91 million metric tons of milk; dairy operations generated up to an estimated 20 million metric tons of manure solids, and while milk sales were valued at $35.3B, very minute value was gained from the manure. Less than 4% of the manure solids were processed through anaerobic digestion (AD) for resource recovery. However, there are several reasons why much of the dairy manure resources remain untapped. For one the anaerobic digesters are capital intensive, and the resource recovery operations can also distract from dairy operations, and AD biogas cannot compete with the price of natural gas which has further decreased 40% in 2008 to 2014. While manure resource recovery efforts have been limited thus far, regardless there is a pressing need to enhance these efforts in the future, as dairy manure processing has potentially significant environmental impacts. In 2011, greenhouse gas (GHG) emissions from the agricultural sector constituted an estimated 6.9% of the USA GHGs, of which 7.0% resulted from dairy manure management. Furthermore, manure nutrient management is also a challenge, as each ton of manure contains approximately 6.6 kg of nitrogen (N) as well as 1.1 kg of phosphorus (P); improper Polyhydroxyalkanoates (PHA) production has also shown that there is two-stages in which dairy waste processing can dramatically reduce GHG production by up to 60%.

Additionally, PHA is a valuable product by itself, it is easily stored and transported, and it also retains almost all of the carbon of the original feedstock, finally, PHA can further be produced using mixed microbial cultures. Furthermore, PHA production addresses many of the operational and environmental concerns of manure resource recovery, PHA bioreactors remove little N or P. Therefore, PHA production and AD must be coupled with nutrient capture technologies, such as algal production.

Co-fermentation or co-digestion of algal biomass with dairy manure is topical, and several researchers have addressed the associated benefits of combining dairy manure with algae cultivation in order to capture nutrients. A research team from SCI, has investigated algal growth on manure-derived wastewater, as a means to sequester the additional N and P and is further interested in algal biomass recycling as a means to enhance AD. Algae and manure wastes may also be expected to co-digest well, as both contain residual products of plant metabolisms; however, there are differences between the two which poses challenges. One of the challenges of integrating algae, PHA production, and dairy manure processing is that the sources of process variabilityhas to be understood further and quantified in order to provide optimal process control. Related to AD, algae also appeared to be potentially problematic due to its wide range of methane yields which have been reported for algae AD. Which has ranged from 0.09 up to 0.45 LCH4 g−1VS. Several reasons in regards to this are thought to be in account for the differences in methane yields, such as changes in algal composition, conversion conditions, AD inhibition, and algal pretreatment effects, but any additional sources of process variation are required to be identified in order to successfully achieve an optimal process integration.

Several of the individual resource recovery concepts have been investigated independently, however there are more detailed assessment of the influence of incorporating algal biomass into a manure–PHA resource recovery framework that has not yet been reported. Thus, the purpose of the research was in order determine the feasibility and relative productivity of algae grown on PHA-process effluent (PHA-algae).

Results

Fermenter performance The fermenter feed changes were directly affect the quantity and type of VFAs which were produced, however, there was no substantial changes that were observed. Which further confirmed the relative consistency of the substrates. When compared fermenter performance once algal augmentation commenced, there were only three notable performance differences that observed between the first fermenter (manure+algae) and the two manure controls. However the effluent pH was notably 0.14 lower in fermenter 1, which held no real consequence, VFA yield and speciation also appeared affected by algal augmentation. It was also observed that the mean VFA yield of PHA-algae was 11.5% higher than manure.).

When high hydrogen partial pressures are applied they can produce results that thermodynamically favor the accumulation of H-Ca and H-Va; however, the potential effect of hydrogen accumulation was found to not be entirely clear due tot he fact that fermenter 1 also produced higher levels of H-Ac and lower levels of H-Pr and H-Bu than the control fermenter. This observed pattern is generally associated with lower hydrogen partial pressures. The higher H-Ac of fermenter 1 is more likely a by-product of algae lipids which degraded using metabolic pathways such as the beta-oxidation pathway that can result in preferentially increase in the acetate concentration. Despite the cause, the production of the longer chain VFAs could enhance PHA production, which as a substrate can be used by bacteria in order to produce longer chain hydroxyalkanoic acids that can ultimately enhance the polymer characteristics.

Ultimately it was found that the 11.5% larger VFA yield from PHA-algae coupled with the more diverse VFA speciation which further confirms that PHA-algae will be a suitable feedstock for VFA production. Furthermore, the results suggest that a full-scale PHA-algae fermenter will be operationally stable, not easily overloaded, and will not require sensitive controls.

Regarding the observed pH difference in Fermenter 1 in contrast to the control fermenters, the effluent pH decrease of 0.14 was statistically and also chemically significant,due to the substantial alkalinity that was present in the dairy manure which means that the large composition changes were required in order to effect even small pH changes. Formic and lactic-acid concentrations were also tested in order to further determine whether they also accounted for the observed pH difference, but neither was present in significant concentrations. Which means when the mineral buffering of dairy manure was tested it computationally controled the fermenter pH which was found to be 6.136 and was close to the measured pH 6.22. In contrast, the fermenter 1 was observed to be more ‘dilute’ manure mineral components which had a Minteq pH of 6.014, which was also close to the measured pH 6.08, and as the computational pH difference that also accounted for the 87% of the observed pH difference, as well as mineral buffering dilution which was concluded to be the most likely explanation for the observed fermenter pH difference

Anaerobic digester It was observed that the AD performance was stable over the full assessment period, with only minute traces of VFA amounts that were detected in the effluent during both the initial stabilization as well as the experimental periods. AD results were also consistent with the previous pre-fermented manure investigations, and the biomethane production was also not impaired by the nitrogen-rich manure. They found that the mean manure VS loading to AD2 (manure+raw algae) was 5.3 gVS day−1 and that it was also slightly higher than the 5.0 gVS day−1 target, as the combined manure+algae VS overloading was only 5% higher than the AD0 and AD1 loading. However, it is not thought to have been significant in its impact on the results that were produced.

The enhanced productivity of algae was observed to be similar to the phenomenon observed in the fermenters. In both the experimental and control ADs which produced similar methane quantities but the methane concentration of the PHA-algae biogas systems was statistically higher. In their further interrogation of the data it suggested that the increased methane concentration appeared to be due to an unexplained decrease in CO2 production which was discovered not to been from the actual increased relative methane production.

It was also further observed and documented by the team that the missing CO2 increased methane yields were due to the digester liquor that was assumed to be supersaturated with methane, as for the decreased CO2 volume which directly affected the gas transfer coefficient. When the missing CO2 gas volume was theoretically restored, the Kl a and the transferred gas volumes increased proportionally.These similar theoretical methane production values suggest there was little difference actually realized between the methane yield of the different substrates, and that the biogas methane concentration was increased due to the CO2 absorbing characteristic of PHA-algae digester liquor. Furthermore, they observed that as the ‘missing’ CO2 represented about 20% of the CO2 volume which was released by the AD control. It was also found that it was not essential that the PHA-algae be highly digestible to serve a useful role in a manure/PHA resource recovery system. This is because there are benefits to algae proteins being recalcitrant to digestion. Ultimately the augmentation of AD with algal biomass appeared to have no significant effect on the methanogenic population.

Impact of freezing algae samples Algae that the team subjected to in the investigations had been frozen, and were thawed prior to being feed to the fermenter or AD. Freezing was hypothesized to enhance the biodegradability of the algae substrate for AD; thus, the effect of freezing was then evaluated. The investigations revealed that the algae sample storage had a substantial impact on the amount of sCOD which was released. The substrate that was thought to be potentially bio-available for VFA or biogas production. Frozen algae samples released between 43 and 235% more sCOD after a single freeze and thaw cycle. However, only the first freeze/thaw cycle had this marked effect, as additional freeze/thaw cycles made little difference.The inter-batch sCOD differences appeared to have ranged from 9.3 to 28% of the digestible VS, which suggested that the accurate algae performance evaluations need to assess the several different algae batches and pre-treatment methods even when the algae species is identical, and the culture conditions are similar.

Conclusion

The purpose of the research that was conducted by the team, and which was presented was to determine the feasibility and relative productivity of algae grown on PHA-process effluent. The results that were discussed, showed that introducing 10% PHA-algal biomass to fermentation and AD did not cause the process of disturbances or any statistically significant changes in the methanogenic taxa; algae augmentation produced ∼11% more VFA under fermentative conditions and ∼11% more methane under AD than dairy manure; and the AD-enriched microorganisms were also present in the normal dairy manure biota. Finally they found that the sources of variability needed to be further controlled in algae conversion tests.

**Study Conducted by: Simon A Smith,a Eric Hughes,a Erik R Coats,b* Cynthia K Brinkman,a Armando G McDonald,c Jeric R Harper,d Kevin Ferisd and Deborah Newbye**