Technology and project development in the field of lignocellulosic-based fuels and chemicals continues, albeit at a slower pace due to the downturn in global petroleum prices which have served to undermine project economics. New lignocellulosic feedstocks made more readily available through the co-location of 2^nd Generation projects at 1^st Generation biofuels plants offer new project development opportunities. Integration of the biomass process technologies into the 1st Generation biofuel plant offers reduced capital expenditure and improved production economics. Pilot plant operations have proven to be essential in defining key physical and chemical properties of a new class of lignocellulosic feedstocks. In depth knowledge of the new feedstock’s chemical as well as physical properties, which often introduce unforeseen challenges during the technology development, piloting and scale-up phases of the project, is critical to the selection and design of commercial scale systems. A scale-down pilot plant approach was employed in the development of data required to mitigate risk associated with implementing the commercial-scale horizontal screw pretreatment reactor process technology and ancillary material handling and process equipment.

Many of the process bottlenecks experienced in nascent biorefineries are the result of feedstock handling failures due to the inherent chemical and physical variabilities of biomass energy crops such as corn stover (low temperature conversion) or pine residues (high temperature conversion). The interface between unit operations from the field to the conversion reactor throat is a hotspot for problematic operations of Integrated Biorefineries (IBRs).

Related research and development (R&D) efforts at 8 DOE national labs (ANL, INL, LANL, LBNL, NREL, ORNL, PNNL and SNL) have been coordinated into a Feedstock-Conversion Interface Consortium (FCIC). The FCIC goals are to identify and address the impacts that feedstock chemical, mechanical and physical variability have on supply logistics, storage handling, preprocessing and conversion equipment operation and process integration, so as to develop and validate improved integrated feedstock/conversion processes that increase the operational reliability. The DOE National Laboratories possess unique capabilities, which will be leveraged to establish results that surpasses what an individual lab or company can deliver.

This talk will give an overview of the R&D in five areas: 1) Feedstock Variability and Specification Development, 2) Feedstock Physical Performance Modeling, 3) Process Integration, 4) System-wide Throughput Analysis, and 5) Process Control and Optimization, which all aim to coordinate with industry advisors and stakeholders to deliver solutions with the intention of near term industry adoption. These advances will lead to a robust U.S. Bioeconomy enabling agricultural development, domestic job creation, energy security and the reduction of greenhouse gas emissions.

Pioneer biorefineries have encountered many challenges in quickly achieving plant design throughput and biofuel conversion yields. The difficulties are caused by many factors: variable physical and mechanical properties of biomass feedstock, lack of data on the impact of biomass properties on equipment and process performance, matching capacity and capability in equipment integration, sub-optimal equipment design and operation, and less-than-optimal integrated equipment and process control.

This presentation will discuss key impactful factors on major unit operations such as feedstock storage, conveying, size reduction, pretreatment reactor feeding and discharge, and will provide recommendations for R&D activities, improvements in equipment design and in-line instrument and sensors to provide data necessary for successful scaling up bioconversion technologies to commercial operation. Various pretreatment process modifications to potentially mitigate operational challenges that are commonly encountered will also be discussed.

There is a large interest in processes using second generation renewable biomass feedstocks to release sugars. To address the scale up challenges of this process, BPF has a state-of-the-art pilot scale system and is involved in many projects which design demo-scale or large-scale facilities.

The pretreatment pilot at the BPF is based upon a two-stage technology with high temperature and mild chemical treatment (acid/base). The pilot unit is a scaled-down model based upon pulp digesters of which hundreds of systems have been commissioned on industrial scale. The pretreatment is in this case typically followed by enzymatic conversion to sugars using enzyme cocktails which are being improved continuously.

Examples of design considerations for this type of pretreatment technology are:

• Type of feedstock and supply chain.
• The chosen severity conditions.
• The use of co-solvents such as alcohols or gases such as SO₂.
• The setup of the pretreatment process;

The above choices can give rise to several operational challenges, for example:

• Feeding a wet stream of biomass into a high pressure reactor, while maintaining the high pressure steam conditions in the reactor.
• The release of biomass from high pressure.
• The presence of residual dirt in the biomass.
• Solid liquid separations, pretreatment conditions and size reduction are relevant.

Piloting is essential to recognize and mitigate operational challenges before confrontation with expensive issues at large scales. The relation between design considerations and our practical experience on operational issues will be discussed in detail during the presentation.

Feedstock flexibility can lower operational cost for biorefineries, as it secures feedstock sourcing and increases competition on the supplier side. In Denmark, a full scale 2G bioethanol plant is expected to run primarily on wheat straw (WS). In this study, we investigate the compatibility of 16 alternative biomasses with a wheat straw based platform (different types of straw, sawdust, grass fiber, chaff, deep litter, and industrial wastes). All biomasses have been pretreated under conditions optimized for WS, i.e. hydrothermal treatment at 15% DM, 190°C, for 10 min, to mimic their addition to a WS based production line.

Untreated and pretreated biomasses are analyzed via strong acid carbohydrate analysis, comprehensive microarray polymer profiling, elemental analysis, water retention value, as well as enzymatic hydrolysis. The liquid fractions from the pretreatment are analyzed for monomeric and oligomeric sugar content, inhibitor level and fermentability.

The alternative biomasses vary in their carbohydrate content and accessibility, which reflect their intrinsic recalcitrant traits. These traits and relations to the conversion efficiency are discussed. Furthermore, ethanol potential is estimated and included to a number of techno-economic parameters. Based on the results, some biomasses are found compatible in a wheat straw biorefinery, while others must be disregarded or processed differently. For example, barley straw performs better, oat straw equal and rye straw is poorer, compared to WS, while these straw types have similar price and logistic systems. Contrary to this, biogas fibers are economically beneficial, but the carbohydrates are highly inaccessible even after hydrothermal pretreatment.

Steam explosion has been examined by many for the optimal fractionation of sugarcane bagasse within the biorefinery concept. Exogenous acid catalysts are usually applied to improve pretreatment efficiency but the benefits of using such catalytic systems are not always clear compared to auto-hydrolysis. In this study, experiments performed at the same combined severity factor (CSF) of 0.76 ± 0.02 were compared: three auto-hydrolyses (SEB), three catalyzed by dilute sulfuric acid (SEB/SA) and one catalyzed by dilute phosphoric acid (SEB/PA) that was used as reference. Steam explosion at the same CSF produced substrates with similar chemical composition, crystallinity index, rheological behavior and glucose yields by enzymatic hydrolysis at 15% total solids (TS) using Cellic CTec3, except for SEB/SA experiments that were carried out at lower temperatures and shorter times. Hence, CSF values were very useful to adjust the strength of auto-hydrolysis and acid-catalysis steam explosion, however, this tool was not valid for all range of temperatures, residence times, acid catalysts and acid concentrations employed in this study. In principle, better results were obtained by auto-hydrolysis but pretreatment had to be performed at higher temperatures and residence times. The composition of pretreatment water-solubles varied among different pretreatment conditions with regard to their contents in carbohydrates, aliphatic acids, furans and aromatic compounds. Finally, hydrolysis of SEB at 15% TS in a lab-scale bioreactor required the lowest power consumption compared to both SEB/SA and SEB/PA because these latter substrates were more heterogeneous and contained more fiber aggregates.

The synergy between large commodity based infrastructure and smaller niche production can be leveraged to the advantage of both, as evidenced in petroleum and wet corn mill refining. The current dry-grind ethanol industry is a large volume low margin business with only a limited number of products available. The operators of these facilities continue to increase efficiency and yields, using both conventional and cellulosic conversion technologies. This is prone to resulting in back end bottlenecks, allowing front end excess capacity. Through process intensification this over capacity of feedstock handling can be leveraged to begin producing small volume, high margin specialty products. Purity of media to perform bio-conversions can also be tailored depending on process requirements, resulting in similar opportunity costs regardless of purity. The combination of process intensification and co-location can have a significant impact on overall economic viability, allowing new processes to reach commercial production faster, cheaper.