Alkaline pretreatment of herbaceous feedstocks such as corn stover prior to mechanical refining and enzymatic saccharification improves downstream sugar yields by removing acetyl moieties from hemicellulose. However, the relationship between transport phenomena and deacetylation kinetics is virtually unknown for such feedstocks and this pretreatment process. Here, we report the development of an experimentally validated reaction–diffusion model for the deacetylation of corn stover. A tissue-specific transport model is used to estimate transport-independent kinetic rate constants for the reactive extraction of acetate, hemicellulose and lignin from corn stover under representative alkaline conditions (5–7 g L −1 NaOH, 10 wt% solids loadings) and at low to mild temperatures (4–70°C) selected to attenuate individual component extraction rates under differential kinetic regimes. The underlying transport model is based on microstructural characteristics of corn stover derived from statistically meaningful geometric particle and pore measurements. These physical descriptors are incorporated into distinct particle models of the three major anatomical fractions (cobs, husks and stalks) alongside an unsorted, aggregate corn stover particle, capturing average Feret lengths of 917–1239 μm and length-to-width aspect ratios of 1.8–2.9 for this highly heterogeneous feedstock. Individual reaction–diffusion models and their resulting particle model ensembles are used to validate and predict anatomically-specific and bulk feedstock performance under kinetic-controlled vs. diffusion-controlled regimes. In general, deacetylation kinetics and mass transfer processes are predicted to compete on similar time and length scales, emphasizing the significance of intraparticle transport phenomena. Critically, we predict that typical corn stover particles as small as ∼2.3 mm in length are entirely diffusion-limited for acetate extraction, with experimental effectiveness factors calculated to be 0.50 for such processes. Debilitatingly low effectiveness factors of 0.021–0.054 are uncovered for cobs—implying that intraparticle mass transfer resistances may impair observable kinetic measurements of this anatomical fraction by up to 98%. These first-reported quantitative maps of reaction vs. diffusion control link fundamental insights into corn stover anatomy, biopolymer composition, practical size reduction thresholds and their kinetic consequences. These results offer a guidepost for industrial deacetylation reactor design, scale-up and feedstock selection, further establishing deacetylation as a viable biorefinery pretreatment for the conversion of lignocellulosics into value-added fuels and chemicals.

Rapid characterization of biomass composition is a key enabling technology for biorefineries—the ability to measure the chemical composition of biomass materials entering the biorefinery as well as the composition of key process intermediate streams would allow real-time process control and the development of robust models to predict process performance. The utility of near-infrared (NIR) spectroscopy for rapid characterization requires multivariate algorithms for building calibration models. The most prevalent algorithm used for building calibration models using NIR spectra is the linear modeling algorithm Partial Least Squares Regression (PLS). Nonlinear regression algorithms (which are typically more computationally intensive than linear modeling approaches) have gained popularity in recent years due to their ability to solve a wide variety of classification and regression problems and the dramatic increase in available computational resources. In this work, we demonstrate that a calibration model can predict the composition of corn stover process intermediate samples pretreated with three different treatments—hot water (HW), dilute acid (DA), and deacetylation followed by dilute acid (DDA). We quantitatively compare three different algorithms for building prediction models based on near-infrared spectroscopy—partial least squares (PLS), support vector machines (SVM), and random forests (RF). We demonstrate the utility of improving model performance by accounting for instrument performance variability using repeated measurements of standard materials (e.g., the “repeatability file” strategy) and investigate its performance with nonlinear regression techniques, and we discuss methods for quantifying the uncertainties of specific predictions among the three methods.

A simple and highly selective analytical procedure is presented for the determination of cellulosic glucan content in samples that contain both cellulose and starch. This method eliminates the unacceptably large compounding errors of current two-measurement methods. If both starch and cellulose are present before analytical hydrolysis, both will be hydrolyzed to glucose causing bias and inaccuracy in the method. To prevent this interference, the removal of starch prior to cellulosic quantification is crucial. The method presented here is a concise in-series procedure with minimal measurements, eliminating large compounding errors. Sample preparation consists of a starch extraction employing enzymatic hydrolysis followed by a simple filtration and wash. The samples are then subjected to a two-stage acid hydrolysis. The concentration of glucose is determined by ion exchange high-performance liquid chromatography with a Pb 2+ column and a refractive index detector. The cellulosic glucan content is calculated based on the initial dry weight of the starting material. Data for the native biomass materials studied show excellent reproducibility, with coefficients of variance of 3.0% or less associated with the method. This selectivity for cellulosic glucan by the procedure was validated with several analytical techniques such as liquid chromatography coupled with mass spectrometry (LC–MS), Raman spectroscopy, and nuclear magnetic resonance.

Feedstock variability that originates from biomass production and field conditions propagates through the value chain, posing a significant challenge to the emerging biorefinery industry. Variability in feedstock properties impacts feeding, handling, equipment operations, and conversion performance. Feedstock quality attributes, and their variations, are often overlooked in assessing feedstock value and utilization for conversion to fuels, chemicals, and products. This study developed and employed a multiscale analytical characterization approach coupled with data analytic methods to better understand the sources and distribution of feedstock quality variability through evaluation of 24 corn stover bales collected in 4 counties of Iowa. In total, 216 core samples were generated by sampling nine positions on each bale using a reliable bale coring process. The samples were characterized for a broad suite of physicochemical properties ranging across field and bale, macro, micro, and molecular scales. Results demonstrated that feedstock quality attributes can vary at all spatial scales and that multiple sources of variability must be considered in order to establish and manage biomass quality for conversion processes.

Early lignocellulosic biorefineries have been plagued with numerous issues that involve feedstock handling problems and variations in conversion efficacy that stem from feedstock variability and complexity in dimensional, physical, chemical, and mechanical attributes. Feedstock ash and moisture content vary considerably in corn stover harvested from farms for bioconversion, and their effects on preprocessing (grinding/milling) and subsequent chemical and enzymatic conversion to fermentable sugars is systematically explored here using pilot-scale hammer mill grinders and a chemical hydrolysis reactor. Corn stover with high ash content due to contamination from soil was found to (1) consume higher power during grinding resulting in reductions of processing rates and (2) produce a larger fraction of fines in the feedstock that were lost to dust mitigation systems causing higher mechanical wear rates. Corn stover feedstock coming from fields with a high residual moisture content resulting in bale degradation due to self-heating caused a more pronounced drop in preprocessing throughput due to grinder overloads and process upsets leading to equipment downtime. Conversion yield to sugars was not affected, although differences in fermentation performance on these sugar streams was not examined. The overall process throughput was only 40-70% of nameplate capacity due to preprocessing problems.

Production of biofuels, bioproducts, and bioenergy requires a well-characterized, stable, and reasonably uniform biomass supply and well-established supply chains for shipping biomass from farm fields to biorefineries, while achieving year-round production targets. Preserving and stabilizing biomass feedstock during storage is a necessity for cost-effective and sustainable biofuel production. Ensiling is a common storage method used to preserve and even improve forage quality; however, the impact of ensiling on biomass physical and chemical properties that influence bioconversion processes has been variable. Our objective in this work was to determine the effects of ensiling on lignocellulosic feedstock physicochemical properties and how that influences bioconversion requirements. We observed statistically significant decreases (p < 0.05) in the content of two major structural carbohydrates (glucan and xylan) of 5 and 8%, respectively, between the ensiled and non-ensiled materials. We were unable to detect differences in sugar yields from structural carbohydrates after pretreatment and enzymatic hydrolysis of the ensiled materials compared to non-ensiled controls. Based on this work, we conclude that ensiling the corn stover did not change the bioconversion requirements compared to the control samples and incurred losses of structural carbohydrates. At the light microscopy level, ensiled corn stover exhibited little structural change or relocation of cell wall components as detected by immunocytochemistry. However, more subtle structural changes were revealed by electron microscopy, as ensiled cell walls exhibit ultrastructural characteristics such as wall delimitation intermediate between non-ensiled and dilute-acid-pretreated cell walls. These findings suggest that alternative methods of conversion, such as deacetylation and mechanical refining, could take advantage of lamellar defects and may be more effective than dilute acid or hot water pretreatment for biomass conversion of ensiled materials.

The performance of a conventional laboratory near-infrared (NIR) spectrometer and two NIR spectrometer prototypes (a Texas Instruments NIRSCAN Nano evaluation model (EVM) and an InnoSpectra NIR-M-R2 spectrometer) are compared by collecting reflectance spectra of 270 well-characterized herbaceous biomass samples, building calibration models using the partial least squares (PLS-2) algorithm to predict five constituents of the samples from the reflectance spectra, and comparing the resulting model statistics. The prediction models developed using spectra from the Foss XDS spectrometer were slightly better than the prediction models developed using spectra from either the TI NIRSCAN Nano EVM and the InnoSpectra NIR-M-R2 as measured by the root mean square error (RMSECV) and the correlation coefficient (R2_cv) for “leave-one-out” cross-validation (CV). The models built from the two prototype units were not statistically significantly different from each other (p = 0.05). The Foss spectrometer has a larger wavelength range (400–2500 nm) compared with the two prototypes (900–1700 nm). When the spectra from the Foss XDS spectrometer were truncated so their wavelength range matched the wavelength range of the two prototype units, the resulting model was not statistically significantly different from the models from either prototype.

Switchgrass (Panicum virgatum L.) and giant miscanthus (Miscanthus x giganteus Greef & Deuter ex Hodkinson & Renvoize) are productive on marginal lands in the eastern USA, but their productivity and composition have not been compared on mine lands. Our objectives were to compare biomass production, composition, and theoretical ethanol yield (TEY) and production (TEP) of these grasses on a reclaimed mined site. Following 25 years of herbaceous cover, vegetation was killed and plots of switchgrass cultivars Kanlow and BoMaster and miscanthus lines Illinois and MBX-002 were planted in five replications. Annual switchgrass and miscanthus yields averaged 5.8 and 8.9 Mg dry matter ha−1, respectively, during 2011 to 2015. Cell wall carbohydrate composition was analyzed via near-infrared reflectance spectroscopy with models based on switchgrass or mixed herbaceous samples including switchgrass and miscanthus. Concentrations were higher for glucan and lower for xylan in miscanthus than in switchgrass but TEY did not differ (453 and 450 L Mg−1, respectively). In response to biomass production, total ethanol production was greater for miscanthus than for switchgrass (5594 vs 3699 L ha−1), did not differ between Kanlow and BoMaster switchgrass (3880 and 3517 L ha−1, respectively), and was higher for MBX-002 than for Illinois miscanthus (6496 vs 4692 L ha−1). Relative to the mixed feedstocks model, the switchgrass model slightly underpredicted glucan and slightly overpredicted xylan concentrations. Estimated TEY was slightly lower from the switchgrass model but both models distinguished genotype, year, and interaction effects similarly. Biomass productivity and TEP were similar to those from agricultural sites with marginal soils.

Biomass analysis is a slow and tedious process and not solely due to the long generation time for most plant species. Screening large numbers of plant variants for various geno-, pheno-, and chemo-types, whether naturally occurring or engineered in the lab, has multiple challenges. Plant cell walls are complex, heterogeneous networks that are difficult to deconstruct and analyze. Macroheterogeneity from tissue types, age, and environmental factors makes representative sampling a challenge and natural variability generates a significant range in data. Using high throughput (HTP) methodologies allows for large sample sets and replicates to be examined, narrowing in on more precise data for various analyses. This review provides a comprehensive survey of high throughput screening as applied to biomass characterization, from compositional analysis of cell walls by NIR, NMR, mass spectrometry, and wet chemistry to functional screening of changes in recalcitrance via HTP thermochemical pretreatment coupled to enzyme hydrolysis and microscale fermentation. The advancements and development of most high-throughput methods have been achieved through utilization of state-of-the art equipment and robotics, rapid detection methods, as well as reduction in sample size and preparation procedures. The computational analysis of the large amount of data generated using high throughput analytical techniques has recently become more sophisticated, faster and economically viable, enabling a more comprehensive understanding of biomass genomics, structure, composition, and properties. Therefore, methodology for analyzing large datasets generated by the various analytical techniques is also covered.

Core Ideas Harvest yield varies more across species than sugar content and accessibility. Harvest yield and sugar accessibility are both critical parameters for conversion. Near‐infrared/partial least square models are valuable for quickly evaluating biomass for bioconversion. Cool‐season (C3) perennial grasses have a long history of cultivation and use as animal forage. This study evaluated 15 cultivars of C3 grasses, when harvested in late June for increased biomass yield, as biofuel feedstocks using near‐ infrared spectroscopy (NIR) based partial least square (PLS) analysis. These grasses were grown near Iliff, CO, for three growing seasons (2009–2011). The carbohydrate composition and released carbohydrates (total glucose and xylose released from dilute acid pretreatment and enzymatic hydrolysis [EH]) were predicted for samples from the study using NIR/PLS. The results were analyzed from a biofuels perspective, where composition combined with harvest yield provided information on the carbohydrate yield available for biomass conversion processes, and released carbohydrate yield provided information on the accessibility of those carbohydrates to conversion methods. The range in harvest yields varied more among cultivars (2900 kg ha−1) than did the range in carbohydrate composition (56.0 g kg−1) or released carbohydrates (60.0 g kg−1). When comparing carbohydrate yield to released carbohydrate yield between cultivars, an efficiency as high as 87% release of available carbohydrates was obtained for pubescent wheatgrass [Thinopyrum intermedium (Host) Barkworth & D.R. Dewey ‘Mansaka’], with a low of 71% for hybrid wheatgrass [Elytrigia repens (L.) nevski ´ pseudoroegneria spicata (PURSH) A. Love ‘Newhy’]. Though hybrid wheatgrass had the lowest release efficiency, its high harvest yield resulted in release of more total carbohydrates than half the other cultivars analyzed. This suggested that harvest yield, carbohydrate release, and carbohydrate composition, togetherplay significant roles in biofuel feedstock evaluation.

Harnessing stem carbohydrate dynamics in grasses offers an opportunity to help meet future demands for plant-based food, fiber and fuel production, but requires a greater understanding of the genetic controls that govern the synthesis, interconversion and transport of such energy reserves. We map out a blueprint of the genetic architecture of rice (Oryza sativa) stem nonstructural carbohydrates (NSC) at two critical developmental time-points using a subpopulation-specific genome-wide association approach on two diverse germplasm panels followed by quantitative trait loci (QTL) mapping in a biparental population. Overall, 26 QTL are identified; three are detected in multiple panels and are associated with starch-at-maturity, sucrose-at-maturity and NSC-at-heading. They tag OsHXK6 (rice hexokinase), ISA2 (rice isoamylase) and a tandem array of sugar transporters. This study provides the foundation for more in-depth molecular investigation to validate candidate genes underlying rice stem NSC and informs future comparative studies in other agronomically vital grass species.

In this work, we examined the behavior of feedstock blends and the effect of a specific feedstock densification strategy (pelleting) on the release and yield of structural carbohydrates in a laboratory-scale dilute acid pretreatment (PT) and enzymatic hydrolysis (EH) assay. We report overall carbohydrate release and yield from the two-stage PT-EH assay for five single feedstocks (two corn stovers, miscanthus, switchgrass, and hybrid poplar) and three feedstock blends (corn stover-switchgrass, corn stover-switchgrass-miscanthus, and corn stover-switchgrass-hybrid poplar). We first examined the experimental results over time to establish the robustness of the PT-EH assay, which limits the precision of the experimental results. The use of two different control samples in the assay enabled us to identify (and correct for) a small bias in the EH portion of the combined assay for some runs. We then examined the effect of variable pretreatment reaction conditions (residence time, acid loading, and reactor temperature) on the conversion of a single feedstock (single-pass corn stover, CS-SP) in order to establish the range of pretreatment reaction conditions likely to provide optimal conversion data. Finally, we applied the assay to the 16 materials (8 feedstocks in 2 formats, loose and pelleted) over a more limited range of pretreatment experimental conditions. The four herbaceous feedstocks behaved similarly, while the hybrid poplar feedstock required higher pretreatment temperatures for optimal results. As expected, the yield data for three blended feedstocks were the average of the yield data for the individual feedstocks. The pelleting process appears to provide a slightly positive effect on overall total sugar yield.

Rice plants (Oryza sativa) accumulate excess photoassimilates in the form of non-structural carbohydrates (NSCs) in their stems prior to heading that can later be mobilized to supplement photosynthate production during grain-filling. Despite longstanding interest in stem NSC for rice improvement, the dynamics of NSC accumulation, remobilization, and re-accumulation that have genetic potential for optimization have not been systematically investigated. Here we conducted three pilot experiments to lay the groundwork for large-scale diversity studies on rice stem NSC. We assessed the relationship of stem NSC components with 21 agronomic traits in large-scale, tropical yield trials using 33 breeder-nominated lines, established an appropriate experimental design for future genetic studies using a Bayesian framework to sample sub-datasets from highly replicated greenhouse data using 36 genetically diverse genotypes, and used 434 phenotypically divergent rice stem samples to develop two partial least-squares (PLS) models using near-infrared (NIR) spectra for accurate, rapid prediction of rice stem starch, sucrose, and total non-structural carbohydrates. We find evidence that stem reserves are most critical for short-duration varieties and suggest that pre-heading stem NSC is worthy of further experimentation for breeding early maturing rice.

Residence time is a critical parameter that strongly affects the product profile and overall yield achieved from thermochemical pretreatment of lignocellulosic biomass during production of liquid transportation fuels. The residence time distribution (RTD) is one important measure of reactor performance and provides a metric to use when evaluating changes in reactor design and operating parameters. An inexpensive and rapid RTD measurement technique was developed to measure the residence time characteristics in biomass pretreatment reactors and similar equipment processing wet-granular slurries. Sodium chloride was pulsed into the feed entering a 600kg/d pilot-scale reactor operated at various conditions, and aqueous salt concentration was measured in the discharge using specially fabricated electrical conductivity instrumentation. This online conductivity method was superior in both measurement accuracy and resource requirements compared to offline analysis. Experimentally measured mean residence time values were longer than estimated by simple calculation and screw speed and throughput rate were investigated as contributing factors. A semi-empirical model was developed to predict the mean residence time as a function of operating parameters and enabled improved agreement.

Pretreatment is a critical step in the biochemical conversion of lignocellulosic biomass to fuels and chemicals. Due to the complexity of the physicochemical transformations involved, predictively scaling up technology from bench- to pilot-scale is difficult. This study examines how pretreatment effectiveness under nominally similar reaction conditions is influenced by pretreatment reactor design and scale using four different pretreatment reaction systems ranging from a 3 g batch reactor to a 10 dry-ton/days continuous reactor. The reactor systems examined were an automated solvent extractor (ASE), steam explosion reactor (SER), ZipperClave®Reactor (ZCR), and large continuous horizontal screw reactor (LHR). To our knowledge, this is the first such study performed on pretreatment reactors across a range of reaction conditions and at different reactor scales.

Biomass compositional methods are used to compare different lignocellulosic feedstocks, to measure component balances around unit operations and to determine process yields and therefore the economic viability of biomass-to-biofuel processes. Four biomass reference materials (RMs NIST 8491-8494) were prepared and characterized, via an interlaboratory comparison exercise in the early 1990s to evaluate biomass summative compositional methods, analysts, and laboratories. Having common, uniform, and stable biomass reference materials gives the opportunity to assess compositional data compared to other analysts, to other labs, and to a known compositional value. The expiration date for the original characterization of these RMs was reached and an effort to assess their stability and recharacterize the reference values for the remaining material using more current methods of analysis was initiated. We sent samples of the four biomass RMs to 11 academic, industrial, and government laboratories, familiar with sulfuric acid compositional methods, for recharacterization of the component reference values. In this work, we have used an expanded suite of analytical methods that are more appropriate for herbaceous feedstocks, to recharacterize the RMs' compositions. We report the median values and the expanded uncertainty values for the four RMs on a dry-mass, whole-biomass basis. The original characterization data has been recalculated using median statistics to facilitate comparisons with this data. We found improved total component closures for three out of the four RMs compared to the original characterization, and the total component closures were near 100 %, which suggests that most components were accurately measured and little double counting occurred. The major components were not statistically different in the recharacterization which suggests that the biomass materials are stable during storage and that additional components, not seen in the original characterization, were quantified here.

BACKGROUND:For biofuel production processes to be economically efficient, it is essential to maximize the production of monomeric carbohydrates from the structural carbohydrates of feedstocks. One strategy for maximizing carbohydrate production is to identify less recalcitrant feedstock cultivars by performing some type of experimental screening on a large and diverse set of candidate materials, or by identifying genetic modifications (random or directed mutations or transgenic plants) that provide decreased recalcitrance. Economic efficiency can also be increased using additional pretreatment processes such as deacetylation, which uses dilute NaOH to remove the acetyl groups of hemicellulose prior to dilute acid pretreatment. In this work, we used a laboratory-scale screening tool that mimics relevant thermochemical pretreatment conditions to compare the total sugar yield of three near-isogenic brown midrib (bmr) mutant lines and the wild-type (WT) sorghum cultivar. We then compared results obtained from the laboratory-scale screening pretreatment assay to a large-scale pretreatment system.RESULTS:After pretreatment and enzymatic hydrolysis, the bmr mutants had higher total sugar yields than the WT sorghum cultivar. Increased pretreatment temperatures increased reactivity for all sorghum samples reducing the differences observed at lower reaction temperatures. Deacetylation prior to dilute acid pretreatment increased the total sugar yield for all four sorghum samples, and reduced the differences in total sugar yields among them, but solubilized a sizable fraction of the non-structural carbohydrates. The general trends of increased total sugar yield in the bmr mutant compared to the WT seen at the laboratory scale were observed at the large-scale system. However, in the larger reactor system, the measured total sugar yields were lower and the difference in total sugar yield between the WT and bmr sorghum was larger.CONCLUSIONS:Sorghum bmr mutants, which have a reduced lignin content showed higher total sugar yields than the WT cultivar after dilute acid pretreatment and enzymatic hydrolysis. Deacetylation prior to dilute acid pretreatment increased the total sugar yield for all four sorghum samples. However, since deacetylation also solubilizes a large fraction of the non-structural carbohydrates, the ability to derive value from these solubilized sugars will depend greatly on the proposed conversion process.

Metabolomics is an emerging method to improve our understanding of how genetic diversity affects phenotypic variation in plants. Recent studies have demonstrated that genotype has a major influence on biochemical variation in several types of plant tissues, however, the association between metabolic variation and variation in morphological and physiological traits is largely unknown. Sorghum bicolor (L.) is an important food and fuel crop with extensive genetic and phenotypic variation. Sorghum lines have been bred for differing phenotypes beneficial for production of grain (food), stem sugar (food, fuel), and cellulosic biomass (forage, fuel), and these varying phenotypes are the end products of innate metabolic programming which determines how carbon is allocated during plant growth and development. Further, sorghum has been adapted among highly diverse environments. Because of this geographic and phenotypic variation, the sorghum metabolome is expected to be highly divergent; however, metabolite variation in sorghum has not been characterized. Here, we utilize a phenotypically diverse panel of sorghum breeding lines to identify associations between leaf metabolites and morpho-physiological traits. The panel (11 lines) exhibited significant variation for 21 morpho-physiological traits, as well as broader trends in variation by sorghum type (grain vs. biomass types). Variation was also observed for cell wall constituents (glucan, xylan, lignin, ash). Non-targeted metabolomics analysis of leaf tissue showed that 956 of 1181 metabolites varied among the lines (81%, ANOVA, FDR adjusted p < 0.05). Both univariate and multivariate analyses determined relationships between metabolites and morpho-physiological traits, and 384 metabolites correlated with at least one trait (32%, p < 0.05), including many secondary metabolites such as glycosylated flavonoids and chlorogenic acids. The use of metabolomics to explain relationships between two or more morpho-physiological traits was explored and showed chlorogenic and shikimic acid to be associated with photosynthesis, early plant growth and final biomass measures in sorghum. Taken together, this study demonstrates the integration of metabolomics with morpho-physiological datasets to elucidate links between plant metabolism, growth, and architecture.

BACKGROUND:Obtaining accurate chemical composition and reactivity (measures of carbohydrate release and yield) information for biomass feedstocks in a timely manner is necessary for the commercialization of biofuels. Our objective was to use near-infrared (NIR) spectroscopy and partial least squares (PLS) multivariate analysis to develop calibration models to predict the feedstock composition and the release and yield of soluble carbohydrates generated by a bench-scale dilute acid pretreatment and enzymatic hydrolysis assay. Major feedstocks included in the calibration models are corn stover, sorghum, switchgrass, perennial cool season grasses, rice straw, and miscanthus.RESULTS:We present individual model statistics to demonstrate model performance and validation samples to more accurately measure predictive quality of the models. The PLS-2 model for composition predicts glucan, xylan, lignin, and ash (wt%) with uncertainties similar to primary measurement methods. A PLS-2 model was developed to predict glucose and xylose release following pretreatment and enzymatic hydrolysis. An additional PLS-2 model was developed to predict glucan and xylan yield. PLS-1 models were developed to predict the sum of glucose/glucan and xylose/xylan for release and yield (grams per gram). The release and yield models have higher uncertainties than the primary methods used to develop the models.CONCLUSION:It is possible to build effective multispecies feedstock models for composition, as well as carbohydrate release and yield. The model for composition is useful for predicting glucan, xylan, lignin, and ash with good uncertainties. The release and yield models have higher uncertainties; however, these models are useful for rapidly screening sample populations to identify unusual samples.