Analyzing Potential Failures and Effects in a Pilot-Scale Biomass Preprocessing Facility for Improved Reliability

Abstract

This study demonstrates a failure identification methodology applied to a preprocessing facility generating conversion-ready feedstocks from biomass meeting conversion process critical quality attribute (CQA) specifications. Failure Modes and Effects Analysis (FMEA) was used as an industrially relevant risk analysis approach to evaluate a logging residue preprocessing system to prepare feedstock for pyrolysis conversion. Risk evaluations considered both system-level and operation unit-level assessments considering process efficiency, product quality, cost, sustainability, and safety. Key outputs included estimations of semi-quantitative risk scores for each failure, identification of the failure impacts, identification of failure causes associated with material attributes and process parameters, ranking success rates of failure detection methods, and speculation of potential mitigation strategies for decreasing failure risk scores. Results showed that deviations from moisture specifications had cascading consequences for other CQAs along with process safety implications. Failures linked to fixed carbon specifications carried the highest risk scores for product quality and process efficiency impacts. As increased throughput can be inversely related to meeting product quality specifications; achieving throughput and other material-based CQAs simultaneously will likely require system optimization or prioritization based on system economics. Ultimately, this work successfully demonstrates FMEA as a risk analysis approach for other bioenergy process systems.

1. Introduction

Biomass, characterized by its profusion and renewable nature, emerges as a promising energy reservoir capable of furnishing a sustainable substitute for conventional fossil fuels. The availability of biomass varies depending on the region, climate, and agricultural practices. Nevertheless, prevailing estimations suggest the existence of ample biomass reserves are posited to make a substantial contribution to global energy supply. For example, Errera et al. reported that primary bioenergy could supply 37% of the global energy demand by 2050 [1]. Demonstrating versatility, biomass engenders diverse utilities encompassing biofuel synthesis, electric power generation, thermal heating, and culinary applications, all conveying environmentally advantageous attributes [2,3].

Lignocellulosic biomass, particularly woody biomass, has emerged as a promising sustainable and stable energy source, offering an attractive alternative to fossil fuels. The utilization of this abundant and renewable resource through pyrolysis has garnered significant attention due to its potential to mitigate environmental concerns and promote energy security. One prospective avenue for the utilization of woody biomass in the domain of high-temperature conversion is pyrolysis. Pyrolysis is a thermally induced decomposition mechanism wherein materials are subjected to temperatures ranging between 400 °C and 600 °C in an oxygen-deprived environment [4,5,6]. In a seminal paper Bridgwater discusses the fast pyrolysis of biomass, including woody biomass, and its potential for producing liquid fuels and chemicals, while also addressing the technical and economic feasibility of the process [7]. Bridgwater et al. also provide a comprehensive overview of the pyrolysis process for biomass conversion, including woody biomass, and discusses the potential for producing value-added products, such as bio-oils and chemicals [8]. A critical review by Mohan et al. examines the pyrolysis of woody biomass and its potential for bio-oil production, highlighting the advantages and challenges associated with this process [9]. Researchers from the National Renewable Energy Laboratory (NREL) provide a detailed assessment of the techno-economic aspects of large-scale pyrolysis oil production from woody biomass, highlighting the potential for sustainable and stable energy generation [10].

Pyrolysis offers several notable advantages over other high-temperature conventional methods, such as combustion. It encompasses an expanded spectrum of viable biomass substrates, affords elevated energy conversion efficiencies, and yields an array of value-added derivatives. However, it is imperative to underscore that the caliber and uniformity of the biomass feedstock possess substantive implications for the efficacy and efficiency of the pyrolysis process [11,12,13].

To ensure and maintain the quality and consistency of the feedstocks, preprocessing is necessary prior to conversion unit operations. The selection of preprocessing methodologies hinges upon the intrinsic attributes of the biomass feedstock as well as specifications of the conversion process itself. Predominant among these preprocessing techniques are drying, separation, and size reduction. Biomass preprocessing encompasses additional strategies including densification, torrefaction, and chemical treatment [14,15,16,17]. Along with achieving process quality specifications, the facet of process safety assumes pivotal significance within the ambit of biomass preprocessing, given the handling and manipulation of materials harboring latent hazards (e.g., susceptibility to dust explosions, chemical exposure) if managed inadequately.

It should be recognized that the bioenergy industry not only strives to meet product and process quality standards while maintaining and enhancing safety, but also serves as a sustainable alternative to the conventional and environmentally detrimental fossil fuel-based petrochemical industry. This transition introduces additional challenges, including the need to produce fungible products within specific cost and sustainability constraints. Deviations from operational norms and expected specifications can lead to unforeseen impacts, potentially rendering the bioenergy alternative financially nonviable and jeopardizing long-term success. Consequently, it is crucial to thoroughly assess and quantify inherent risks. Dysfunction of suboptimal performance in any preprocessing unit can trigger disruptions across the conversion continuum, potentially escalating to a complete plant shutdown. Therefore, identifying operational risks inherent to each process unit and within a system of process units, and implementing necessary measures to mitigate these risks is of paramount importance. A proactive approach is essential to ensure the viability and success of this sustainable alternative industry.

Numerous tools designed to manage process risks have been suggested for both industrial and regulatory applications such as basic risk management facilitation methods (i.e., flowcharts and check sheets), fault tree analysis, hazard analysis and critical control points (HACCP), hazard operability analysis (HAZOP), preliminary hazard analysis (PHA), and failure mode and effects analysis (FMEA).

FMEA constitutes a structured risk management tool tailored to methodically recognize and evaluate the origins and ramifications of plausible failures and mitigation strategies within a given system. FMEA translates the experiential insights and empirical results offered by subject matter experts (SMEs) into a semi-quantitative risk priority number (RPN), facilitating a nuanced assessment of potential risks [18]. Multiple variants of FMEA have been developed, with a primary focus on enhancing comprehension and refining the design of many types of systems. This includes System FMEAs, Design FMEAs, and Process FMEAs. Among these, System FMEA assumes a preeminent role as it represents an elevated tier of analysis encompassing an entire system comprising of diverse subsystems. This application is primarily directed toward discerning and evaluating systemic shortcomings, encompassing matters such as system safety, integration, and interdependencies between subsystems, as well as their interactions with external systems and the human environment. System FMEA adopts a holistic stance, centering on the unique functions and intricate interplay characterizing the system as an integral entity. Moreover, it incorporates the assessment of failure modes associated with single-point vulnerabilities, wherein the failure of a sole component culminates in total system collapse. In fact, it has been observed that approximately 50% or more of system anomalies manifest at the interfacial junctures linking subsystems or components [18]. Consequently, comprehending and effectively addressing interfaces and integration emerges as an imperative prerequisite to establishing systems that are both secure and dependable. Furthermore, the systemic evaluation facilitated by System FMEA avails the opportunity to scrutinize alternate designs, capable of mitigating failure risks while concurrently enhancing dependability and cost-effectiveness [19].

Given these imperatives, this study is dedicated to demonstrating the use of FMEA on a preprocessing system design for comprehensive risk analysis applied to a bioenergy process. FMEA was applied using multiple stages to evaluate a system of preprocessing unit operations. In the first stage, FMEA interviews with SMEs covered the entirety of preprocessing unit operations, thus fostering comprehension at the system-level perspective and allowing for failure identification at interfaces and integration points. In the second FMEA stage, the approach was to delve into granular interviews with SMEs pertaining to each individual unit operation within the system, thereby affording an in-depth exploration into the specific failings pertaining to the individual constituents within the configuration. These two stages conferred varying degrees of information resolution, collectively facilitating an enriched comprehension of latent failures intrinsic to the system.

2. Scope and Methods

2.1. Overview of a Biomass Preprocessing System

The focal point of this investigation was to preprocess pine residue chips for pyrolysis, representing a high-temperature conversion pathway. Notably, this configuration retains a semi-theoretical nature, as the application of completely continuous processing has not been fully realized within the parameters of the research equipment that was used during the FMEA interviews. This conceptual configuration incorporates various constituent units within the confines of the Idaho National Laboratory (INL) Biomass Feedstock National User Facility. The throughput capacities of the system, based on unit operations, were within the 1–5 ton/h range. A schematic depiction of the preprocessing system configuration for generation of pyrolysis reactor feed from pine residue chips is presented in Figure 1.

The evaluated system represents a typical industrial-scale high-temperature conversion framework to generate a conversion ready woody feedstock for high temperature conversion meeting conversion quality specifications [20]. These CQAs were aligned with the specifications encompassing particle size distributions spanning 1.18 mm to 6.00 mm, fixed carbon contents surpassing specified threshold values (18% or 21%), moisture contents maintained at or below 10%, and ash content not exceeding 1.75% for a pyrolysis conversion system [21]. Initially, the biomass undergoes drying within a rotary dryer (RD) to reduce its moisture content to approximately 10–15 wt%. After this, air classification (AC) is applied, resulting in the segregation of a fixed-carbon-enriched stream, exemplified by the white wood-rich stream. This fraction is subsequently comminuted through a hammer mill (HM) using ½” mesh screen, and size classified into a final material with a particle size range between 1.2–6.0 mm using an oscillating screen (OS) system. This finely size classified material is then intended to be fed into the high temperature reactor using a screw feeder (SF). The FMEA interviews for the SF for a pyrolysis conversion system were used to inform the severity impact scores for the preprocessing system but were not included in the system evaluation.

2.2. Scope: FMEA Method

Failure Mode and Effects Analysis (FMEA) is a widely employed risk analysis technique, particularly when quantitative data are limited or scarce. FMEA’s primary advantage lies in its ability to systematically identify and prioritize potential failure modes, their causes, and effects, using a qualitative or semi-quantitative approach. This makes it well-suited for scenarios where comprehensive quantitative data are lacking. Several peer-reviewed studies have discussed the merits of FMEA in such situations [18,22,23]. In contrast, methods like Fault Tree Analysis (FTA) and Hazard Operability Analysis (HAZOP) are heavily dependent on quantitative data and statistical analysis, making them less suitable when such data are scarce. Similarly, Hazard Analysis and Critical Control Points (HACCP) and Preliminary Hazard Analysis (PHA) rely heavily on historical data and quantitative risk assessment, which may not be available in an industry like this one. While FMEA has its limitations and may not provide the same level of quantitative rigor as some other methods, its ability to systematically identify and prioritize risks based on qualitative or semi-quantitative assessments makes it a valuable tool when quantitative data are limited or unreliable.

For the purposes of this study, the FMEA approach was contoured to the system of interest to collect meaningful data from the interviews with SMEs. A generalized framework for conducting these interviews is presented in Figure 2. Within each interview session, the base questions were meticulously formulated to concentrate the discourse of the FMEA on distinct components of equipment and configurations operating under normative conditions.

Following the establishment of operational parameters and equipment demarcations, the diverse failure modes were identified. These failure modes were delineated as instances where the unit operation or process deviated from its intended performance or output. For each failure mode, a comprehensive assessment was undertaken, encompassing the repercussions of the failure (Severity), the causative factors underpinning the failure, the probability of the failure occurring (Occurrence), and the techniques governing the detection and/or prevention of the failure (Detection).

The computation of the risk score or risk priority number (RPN) for each failure mode is the product of the severity (S), occurrence (O), and detection (D) scores using Equation (1):

RPN = S × O × D (1)

Interpreting RPNs effectively helps prioritize issues within an FMEA to focus efforts on mitigating the most critical risks first. Higher RPNs indicate higher risk due to high severity, high likelihood of occurrence, and/or low likelihood of detection. These are typically prioritized for corrective actions. Lower RPNs suggest a lower risk and may not require immediate attention compared to higher RPNs. Typical implementation of this approach is often to set specific thresholds for RPN that dictate when to take action. For example, an RPN above 80 might require immediate action, while an RPN between 40 and 80 might prompt a review, and below 40 might be considered acceptable without immediate action. These actionable thresholds are specific for each target application and organization. It is worth noting that while useful, RPN should not be the sole factor in decision-making as it does not account for the potential interactions between different failure modes. Also, very high or low scores in one of the three factors can skew the RPN, potentially leading to an over- or underestimation of risk.

The scales for quantifying the Severity, Occurrence, and Detection of each failure were developed [18]. A five-level scale was used for this work. Each of the scales that guide the ranking procedure has been formulated with a level of generality, designed to assist the subject matter expert (SME) in assigning a ranking to the pertinent failure mode (see Table 1).

It is noteworthy that certain instances may witness the applicability of only specific segments within the criteria statements, contingent upon the severity, occurrence, or detectability of the failure. In such scenarios, the rationale underpinning the SME’s selection of a particular ranking is meticulously documented through the interview process.

The exhaustive compilation of fundamental interview questions is annexed within the Supplementary Information section (refer to FMEA Interview Form in the Supplementary Information section). It is noteworthy that while the operational modalities specific to research endeavors (namely, modes that extend beyond adhering to product specifications) did not serve as determinants for assessing the severity, occurrence, and detection of failures; they were nevertheless instrumental in the interviews by aiding in the identification of underlying causes for the failures under scrutiny.

Throughout the interview sessions, the rationale underpinning the assignments to each facet of the evaluative scale were also systematically recorded for prospective reference. In alignment with the principles of Quality by Design (QbD), each failure mode was contextualized alongside the identification of Critical Quality Attributes (CQAs), which could manifest as either the root cause of the failure (e.g., a deviation from particle size specifications) or elements adversely influenced by the failure (e.g., an equipment shutdown impacting the target throughput CQA specifications). Simultaneously, these impacts and CQAs were systematically classified into domains encompassing ‘Process Efficiency (PE)’, ‘Product Quality (PQ)’, ‘Economics (E)’, and ‘Sustainability (S)’, acknowledging the overarching relevance of these four categories within most operational frameworks. Impacts associated with process safety were also identified and captured during the interviews.

Moreover, alongside the identification of the effects and CQAs associated with a given failure mode, the causes of the failures were captured and associated with ‘Critical Material Attributes (CMAs)’ and ‘Critical Process Parameters (CPPs)’ as applicable. Where accessible, the SMEs documented the delineation of the ranges or thresholds pertaining to the CMAs and CPPs. The outcomes gleaned from the ongoing FMEA endeavors, as encapsulated by this study, essentially furnish a platform wherein each calculated RPN score for a specific failure mode finds association with distinct CQAs, CMAs, and CPPs. These then constitute a semi-quantitative hierarchy, permitting the comparative assessment of critical properties amongst diverse failure modes, pertinent to a given unit operation and/or system of unit operations. One to two SME interviews were used for each individual unit operation along with the system-wide evaluation interview. All interviews were performed with one SME at a time and one to three FMEA evaluators. It should be noted that for the system-wide interviews the SME was asked to focus on the feedstock of interest and defined system configuration. For the unit operation interviews, the SMEs were able to draw on experience outside of the system configuration to define failures and other FMEA elements (e.g., CMAs from other biomass materials).

3. Results and Discussion

The interviews conducted at a system-wide level primarily focused on failures pertaining to the targeted CQAs linked with the intermediate feedstock product, specifically intended for deployment within the high-temperature conversion framework. These included particle size distributions spanning 1.18 mm to 6.00 mm, fixed carbon contents surpassing specified threshold values (18% or 21%), moisture contents maintained at or below 10%, and ash content not exceeding 1.75%. It is noteworthy that within the domain of FMEA interviews, deliberations extended to include discourse on the impact of throughput and energy consumption; however, for the context of this research-scale system, a quantified specification for these parameters was not identified. A comprehensive summary of the outcomes gleaned from the system-wide FMEA interviews is documented in Table 2 and Table 3. It should also be noted that while the SMEs provided information to the best of their abilities, additional failures not identified are likely and represent some of the uncertainty of this approach.

Instances of deviation from the moisture content specification (>10%) primarily emanated from the RD unit operation, as anticipated. The repercussions of this failure were subjected to evaluation from the perspective of both PQ and PE. Considering that the RD functions as the inaugural unit operation within the system (as depicted in Figure 1), the FMEA interview process aptly captured cascading failures extending to subsequent downstream equipment units (see Table S1). In terms of PQ, elevated moisture content above the specified 10% threshold exhibited an adverse influence on conversion efficiency, attributable to the heightened moisture content in the ultimate product [24,25]. Additionally, the augmented moisture content exerted secondary effects on the fixed carbon and ash contents of the product, as delineated through secondary impacts [26,27]. These collateral effects emanate from the fact that the separation efficiency of the AC unit within the system is perturbed by unanticipated surges in moisture content. These potential separation efficiency vulnerabilities stemming from moisture content are meticulously outlined in the Supplementary Information section (see Table S2), where it is evident that variable moisture content contributes to the augmentation of bark content in the heavier (product) stream. Overall, the aggregate risk evaluation for PQ, attributed to moisture content exceeding 10% in both the in-process material stream and the final product, tallied to a score of 180. This score, the second highest within the system, effectively highlights the associated risk.

In the realm of PE, instances of moisture content exceeding the 10% threshold had a great impact on encompassing energy consumption and throughput CQAs. These effects predominantly stemmed from the decrease in HM performance, a manifestation of cascading failures precipitated by elevated moisture content. An observable correlation emerged, whereby every incremental 10% surge in material moisture corresponded to a twofold augmentation in HM energy consumption [28]. Furthermore, the throughput capacity of the HM exhibited heightened susceptibility to such influences, in comparison to other components within the system. Subsequently, within the framework of the isolated FMEA analysis pertaining to the HM, discernible reductions in throughput arising from escalating moisture levels were evidenced through occurrences of partial screen plugging failures (see Table S3). The RPN calculated for PE registered a slightly diminished value of 144, when contrasted with the RPN linked with PQ. This discrepancy is primarily attributable to the severity rankings for throughput and energy consumption, assessed as high (8) rather than very high (10). The OS also exhibited that throughput impacted CQAs resulting from escalated moisture content, realized through screen plugging and motor failure modes (see Table S4).

As moisture can be seen as contributor to failures for each unit operation and CQA specification throughout the system, a summary of the impacts of moisture at various level is provided in Figure 3 to demonstrate the use of FMEA for high-level system evaluation focused on a single CMA. During the interviews for each unit operation when moisture was identified as a CMA, multiple scenarios (10–40% moisture) were evaluated to better quantify the impacts of moisture. It can be seen here that the RD, AC, and HM all had the highest RPNs associated with incoming moisture levels that impacted the system’s ability to meet the product moisture specification, fixed carbon specification, and throughput. Scenarios where moisture levels of 40% and above were evaluated always had higher RPNs.

Analogous to the failures attributed to deviations from the moisture specification, failures connected with the fixed carbon specification underwent evaluation from both the PQ and PE perspectives. In terms of PQ, the affected CQAs encompassed fixed carbon concentration and ash content. These effects were instigated by the inadequate removal of bark material via the AC unit, leading to inefficient separation. Notably, the escalation in moisture content, which is known to influence separation efficiency in the AC unit, leads to cascading failure—a phenomenon interconnected with the RD, as detailed earlier. From the vantage point of the AC unit, the RPN score pertaining to the attainment of the fixed carbon specification emerges as the highest within the system, at a value of 192. This elevated score can be attributed to a severity score of 8, which could intimately be tied to failures within the high-temperature reactor and other undesirable attributes inherent to bark materials.

Within the sphere of PE, the escalation in the presence of inorganic components, particularly ash content, stemming from the departure of bark and/or extraneous particulates during the AC operation, had the potential to curtail the fixed carbon concentration. This increase in ash content carried the prospect of exacerbating wear and tear within both the HM and the SF. The consequence of this phenomenon resonated through the aspect of throughput, as it was liable to prolong the intervals required for hammer replacement in the HM. Furthermore, if unaddressed, this influence could extend to encompass changes in the particle size distributions emerging from the HM. The PE failure linked with the fixed carbon specification slightly trailed the PQ failure, registering a score of 72. This assessment hinged upon the categorization of the severity of downtime connected with HM maintenance, which was appraised as low (3).

The predominant factors contributing to deviations from the fixed carbon specification, as ascertained via the AC unit, encompass moisture content, as mentioned earlier, in conjunction with the dimensions of the tissue fractions, as discerned through the AC-specific FMEA analysis (see Table S2). In cases where the tissue fractions exhibit close dimensional proximity, notably when smaller white wood fragments are compared with bark and needles, the separation efficiency of the AC unit becomes less effective in distinguishing these fractions from each other. This observation underscores the significance of particle size relationships in influencing the AC unit’s efficacy. Furthermore, if substantial variations in harvesting equipment yield profound shifts in particle size distributions, the task of optimizing air velocities and feed rates can become a more intricate endeavor. This scenario brings to light the potential complexities introduced when attempting to align air speeds and feed rates with the altered particle size dynamics induced by diverse harvesting mechanisms.

The strategies devised to counteract failures linked to fixed carbon concentrations were primarily oriented toward improvements in detection mechanisms. The implementation of moisture sensors emerged as a prospective approach, enabling the automated adjustment of air velocities in response to an anomaly arising from the RD operation. Moreover, visual detection techniques geared towards distinguishing between white wood and bark could also be incorporated. Through the incorporation of sensor-based strategies, it was postulated that a discernible reduction in RPN scores could be attained. Specifically, the RPN scores for PQ-related failures could potentially diminish from 192 to 72, while PE-related failures could potentially see a reduction from 144 to 54.

Particle size-related failures were segregated into two distinct failure categories: (1) the excessive generation of oversize particles surpassing the maximum specification of 6 mm, and (2) the excessive generation of fines, which are characterized as particles smaller than 1.18 mm within the configuration of this particular system (refer to Table 2). In instances where an undue surplus of oversize particles was generated, this failure was associated with two specific pieces of equipment: the HM and the OS. The current system design, incorporating a ½” screen on the HM, is predicated on the presumption that particles exceeding 6 mm will be subjected to recycling and subsequent reprocessing. The assessment of excessive overs generation from the HM ensued when more than a third of the material underwent recycling at any given occasion. This phenomenon bore implications predominantly on throughput, which stands out as the principal CQA influenced by increased recycling rates. Positioned downstream from the HM, the OS is equipped with a dual array of screens. The upper screen, set at 6 mm for this configuration, serves as a conduit for trapping a significant portion of oversize particles, which are then rerouted for recycling to the HM. Meanwhile, the lower screen functions to expel fines from the product stream. This devised arrangement averts the inclusion of oversize particles within the final product. Failures within the OS primarily exert an impact on particle size, yet their contribution pertains more to the augmented generation of fines, rather than oversize particles. Throughout the comprehensive system-wide interview process, the elevated rate of recycling attributable to particles exceeding 6 mm was ascribed to a risk score of 108, coupled with a severity rating of 6 (moderate). This evaluation was rooted in the impacts for system throughput resulting from this operational anomaly. The escalation in fines generation stemmed from several distinct unit operations, encompassing AC, HM, and OS. The proliferation of fines holds the potential to exert dual consequences: (1) the diminishment of product yields due to the augmented discarding of material, and (2) implications for PQ whereby fines are integrated into the final feedstock product.

While the system-wide interview predominantly spotlighted the implications contingent on product volumes and yields, it is pertinent to consider a more pronounced level of severity with regard to the consequences for PQ. An assigned RPN of 72 reflects a lower magnitude compared to the oversize particle generation. Additionally, it is worth noting that there is a plausible economic dimension linked to escalated volumes of discarded materials, an aspect that warrants incorporation into a technoeconomic analysis (TEA) of this system. Nonetheless, the precise quantification of this economic impact was not feasible within the scope of the interview.

In the context of the HM, the escalated generation of fines was correlated with an elevated moisture content exceeding 20% and reduced feed rates. Moreover, the sizing of screens held substantial influence over the extent of fine generation for the entire system. It has been estimated that when employing ¼” screens, fine concentrations could attain levels as high as 20%. Conversely, when utilizing ½” screens, as implemented in the present system configuration, the magnitude of fine generation tends to range from 10% to 15%. The propensity for excessive fines generation within the current system configuration, under the stipulated parameters for the HM, was deemed to vary from remote to low, contingent upon the moisture content of the materials. While the risk score associated with this circumstance was relatively modest (9), it is imperative to recognize that an escalation in fine concentration can exert a tangible impact upon the drag chain conveyors (see Table S5). Fine material is susceptible to becoming entrapped within the moving parts of the conveyor, thereby triggering wear and tear that ultimately culminates in mechanical failures. An increased prevalence of such failures could potentially be observed within conveyor systems subsequent to HM operations, where larger concentrations of fines are likely to accumulate.

The underlying purpose of the oscillating bottom screen resides in the extraction of fines from the ultimate product stream. Nevertheless, an assortment of infrequent-to-low-likelihood failures (rated between 1 and 3) were discerned, carrying the potential to result in the accumulation of fines within the final product (see Table S4). This occurrence can materialize through the combination of escalated fine content and elevated moisture, facilitating the presence of fines within the final product. While the probability of this event transpiring in materials with a moisture content of 10% was deemed remote, the eventuality of a failure in the upstream RD is low-likelihood (rated as 3). This particular failure scenario, characterized by screen plugging in the bottom screen of the OS, underwent evaluation from both a PE and a PQ perspective. The resultant RPN values were computed as 180 and 108 for the PQ and PE impacts, respectively, in the context of a scenario involving a 30% moisture content. An additional remote likelihood failure attributed to the OS could be plugging of the top screen, bearing the potential for either material loss or the amplification of fines within the final product. However, it is noteworthy that this During the interview, it was noted that the occurrence of a plugging event in the top screen is likely indicative of a preceding plugging event in the bottom screen.

Among the array of CQAs subject to scrutiny within the system, adherence to ash content specifications exhibited the lowest RPN scores, amounting to 90 and 80 for their respective impacts on PE and PQ. The pivotal unit operations intertwined with the fulfillment of ash content specifications encompass the AC and the OS. Within the PE context, the recognition materialized that suboptimal separation of high inorganic content components, notably bark, needles, and soil, from the AC could potentially failure is not prone to transpire in instances characterized by low moisture levels (ranging between 10 and 15%). Importantly, when the upper screen of the OS becomes obstructed, there is a higher likelihood of recycling on-spec material through the system, resulting in overprocessing and the subsequent escalation of fine generation incurring increased wear on the HM, subsequently impinging upon throughput by diminishing HM processing efficiency. This sequence of events might further result in unplanned downtimes for hammer replacement. PQ repercussions stemming from an elevated ash content in the final product were predicated on failures associated with separation inefficiencies in the AC concerning the extraction of higher ash tissue fractions, along with OS malfunctions concerning the removal of fines, presumed to contain higher concentrations of inorganic species.

In the holistic assessment, the likelihood of these failures manifesting was deemed to be low (rated at 3) and remote (rated at 1) in relation to their impacts on both PE and PQ, specific to the processing of pine residue materials under low-moisture conditions. The AC interview brought to light failures linked with heightened concentrations of bark in the heavy (product) fraction (see Table S2). The propensity for such failures was observed to be more pronounced in elevated moisture conditions, as duly documented during the interview, wherein the particle sizes of white wood and bark exhibited closer proximity.

In terms of mitigation strategies, the integration of in-line moisture and ash sensors holds promise in diminishing the occurrence of ash-related failures while simultaneously increasing their detectability. Upon inclusion of these sensors, the projected RPN scores of 18 and 48 were computed for ash-related failures associated with PE and PQ, respectively (see Table 3).

The instances of throughput deviations in this system have been recognized as secondary failure consequences, contingent upon the previously delineated CQAs. These include increased moisture levels (due to RD failures), fixed carbon specifications (entailing failures originating from AC separation that affects HM performance via increased ash content in the stream), particle size specifications (manifesting as excessive overs that necessitate recycling and discarding of greater amount of fines, consequently impacting system throughput specifications), and ash specifications (likewise, emanating from AC separation failures impacting HM throughput performance). The efficacy of throughput is, to a degree, dependent on meeting these aforementioned CQA specifications. In general, these auxiliary specifications tend to wield greater influence in research contexts from which the majority of experimental data are derived. In contrast, within an industrial setting and within a continuous system, throughput is likely to assume a higher degree of significance. The system-wide interview further encapsulated potential failures germane to throughput. Given that throughput is predominantly managed via the system’s feed rate, it is plausible to conceptualize the system in terms of potential secondary failures that might transpire when throughput specifications take precedence.

It was determined that when throughput resulted in lower than optimal feed rates for the HM than the potential for generation of additional fines was identified. The current configuration, with the AC positioned upstream of the HM, features a distinct throughput capacity. This aspect suggests that in a continuous system, there might be an elevated tendency for fine generation (as expounded in the particle size CQA section). The throughput-associated failure linked to fine production from HM garnered a PQ severity ranking of 180. In contrast, the HM-specific FMEA interview underscored that excessive feed rates could elevate the production of overs. As elucidated earlier, this increase in overs negatively impacts the overall system throughput, owing to amplified recycling.

An additional throughput-related failure, not directly addressed within the prior discussion, could potentially be attributed to the AC unit. The AC unit may encounter failures in its separation efficiencies, affecting the extraction of specific fractions such as bark, needles, and dirt. Conversely, it can also experience shortcomings in extracting the appropriate amounts of white wood from the light and medium discard streams. This particular failure mode was elucidated within the individual AC unit operation FMEA, subsequently pinpointed as one of the most critical risks afflicting the unit’s performance (see Table S2). The loss of white wood concentrations in the discard fraction is more likely to occur when the input chips are smaller or contain elevated concentrations of branches and twigs, coupled with excessively high air speeds. Such a scenario can necessitate supplementary reprocessing or result in material loss, both of which are seen as negative impacts on the system’s throughput.

Given that a substantial portion of the identified throughput failures stemmed from unanticipated elevations in moisture content, the primary mitigation strategy recognized involves the implementation of in-line moisture sensors. This technological integration would facilitate automated adjustments to temperature and air flow within the RD, optimize feed rates across the entire system, and regulate air speed for the AC—ultimately mitigating throughput failures. These measures were projected to yield reduced RPN scores of 27 and 90 for their respective impacts on PQ and PE, as depicted in Table 3.

This style of detailed, integrated testing and data are scarce in literature as they can easily be considered sensitive or confidential information to developing and established biorefineries. Nevertheless, recent work within the US DOE national laboratories has demonstrated some of these key relationships [29]. Klinger et al. found the impacts of ash content and moisture were investigated through feedstock preprocessing and fast pyrolysis. In this study it was shown that, for example, if the RD failed to meet the moisture specification prior to size reduction, the HM observed a direct throughput deration of approximately 25% while simultaneously consuming 60% more mass-specific energy per unit of feedstock. This significant throttling of throughput is indicative of necessary process adjustment to avoid process downtime. On the other hand, it was observed that material prepared with higher moisture typically also required more operation interventions in the conversion stages to sustain operations. In addition, the overall performance due to higher fixed carbon and inorganics in residue feedstock, or inability of the AC to effectively separate detrimental tissues, was also readily observed with a substantial decrease in yield for high FC and ash feedstock. Overall, it was found from the experimental results that the particle size (mean and the related generation of fines) and the moisture content were the most significant predictors (both negative relationships) of on-stream performance and achievement of nameplate capacity.

The results presented through the FMEA interviews presented in this research touch on some of the key challenges facing second generation biorefineries including:

• Feedstock variability;

• Feedstock flowability;

• Equipment downtime;

• Lack of equipment performance data;

• Defined feedstock specifications.

These challenges arise due to the inherent heterogeneity and complexity of biomass feedstocks, as well as the relatively nascent stage of development in biorefinery technologies compared to conventional petroleum refineries.

Feedstock variability refers to the differences in physical, chemical, and compositional properties of biomass feedstocks, which can vary significantly based on factors such as plant species, growing conditions, harvesting methods, and storage conditions. This variability can have a significant impact on the efficiency and performance of biorefinery operations, as the conversion processes are often optimized for specific feedstock characteristics. Failing to account for feedstock variability can lead to inefficient processing, lower product yields, and increased operational costs. Tao et al. investigated the impact of feedstock variability on the performance of a biomass pretreatment process [30]. The authors found that variations in biomass composition, particularly the lignin content and structure, significantly affected the efficiency of the pretreatment process, highlighting the importance of considering feedstock variability in biorefinery operations.

Feedstock flowability is another critical challenge, as biomass feedstocks often have poor flow properties due to their fibrous nature, irregular particle shapes, and tendency to form clumps or bridges. Poor flowability can lead to issues such as feed line blockages, uneven distribution, and inconsistent processing rates, ultimately affecting the overall efficiency and throughput of the biorefinery. Dai and Grace discuss this challenge through an investigation into the flow properties of various biomass feedstocks and proposed strategies for improving their flowability [31].

Equipment uptime/downtime and lack of equipment performance data are closely related challenges. Biorefineries often face challenges in maintaining consistent equipment operation and achieving desired levels of uptime due to the complex nature of biomass processing and the relatively new technologies involved. Additionally, there is often a lack of comprehensive performance data for biorefinery equipment, making it difficult to optimize operations and identify areas for improvement. Chandel et al. analyzed the operational performance of a biorefinery plant and identified potential bottlenecks and areas for improvement. They highlighted the importance of collecting and analyzing equipment performance data to enhance operational efficiency and reliability [32].

Defining feedstock specifications is another crucial challenge, as it involves establishing clear guidelines and standards for the physical, chemical, and compositional characteristics of biomass feedstocks that are suitable for specific biorefinery processes. Establishing well-defined feedstock specifications can help mitigate the challenges associated with feedstock variability and ensure consistent processing and product quality. Ilic et al. proposed a comprehensive framework for defining biomass feedstock specifications based on various factors, including feedstock characteristics, conversion processes, and product quality requirements [33].

In summary, the challenges of feedstock variability, feedstock flowability, equipment uptime/downtime, lack of equipment performance data, and defining feedstock specifications are critical issues facing pioneer biorefineries. Addressing these challenges through an objective risk-based framework such as FMEA can lead to technological advancements, and the establishment of industry standards that are crucial for the successful commercialization and widespread adoption of biorefinery technologies.

4. Summary and Future Directions

The FMEA methodology facilitated comprehensive assessments of the capacity of two distinct system design configurations to adhere to CQA specifications pertinent to their respective target conversion processes. Employing FMEA, discernible risks associated with achieving CQA specifications were identified, alongside ancillary repercussions or failures interlinked with the risks rooted in CQA specifications. This multifaceted approach granted a more holistic understanding of the system’s dynamics. The FMEA process included both system-wide evaluations and individual equipment-level assessments, thus encompassing failures not exclusively tied to system-level CQA specifications. Such failures bear significance when evaluating the applicability of the studied unit operations across diverse system configurations. The key findings from this study are listed below:

• Deviations from moisture specifications through the RD exhibited potential far-reaching consequences, cascading across multiple CQAs pertinent to both PQ and PE—manifesting as variations in moisture, fixed carbon, and ash contents. The resultant RPN scores were computed as 180 for PQ and 144 for PE, subsequently impacting downstream equipment.

• Failures linked to fixed carbon specifications carried the most substantial risk scores of 192 and 144 for PQ and PE impacts, respectively, originating from separation efficiency shortcomings in maximizing white wood concentration through the AC unit. These risk scores could be substantially ameliorated through the incorporation of in-line sensors.

• While the system design exhibited competence in achieving particle size specifications for the final product, it is imperative to factor in secondary failures linked with elevated fines volume discard and throughput deficiencies originating from excess overs generation—prompting RPNs of 72 and 108, respectively.

• Analogous to fixed carbon, ash specifications are presumed to be met through the heightened removal of bark, needles, dirt, and fines—effectuated by the AC unit and OS. However, detection ranking for ash specifications is currently considered highly uncertain (10). To consistently meet ash specifications, the integration of in-line sensors would be indispensable.

• The intricacies of throughput failure are intrinsically linked to secondary, cascading failures stemming from other failure modes implicated in meeting moisture, fixed carbon, particle size, and/or ash specifications. As increased throughput can inversely impact PQ specifications, harmonizing throughput and product material CQA, e.g., fixed carbon, specifications necessitate systematic optimization or prioritization grounded in economic considerations.

It should be noted that the failures and risk scores identified for this system design do not necessarily represent all failures for an industrial scaled system. The results of this study are limited by the scale and operational parameters available for this analysis.

The interview process revealed certain limitations in the current FMEA approach. Notably, most FMEA SMEs were well-acquainted with equipment operation and could articulate the ramifications pertinent to PQ and PE. Economic and sustainability implications were occasionally identified by these SMEs without quantification due to the SMEs’ familiarity with these factors. It is important to acknowledge that the existing severity assessment scale is not ideally tailored to evaluate the economic and sustainability ramifications of failures. The TEA along with life cycle analysis (LCA) will likely offer more robust tools for quantifying these repercussions, given that numerous product and process-based failures ultimately translate into economic and sustainability risks.

The FMEA approach, while widely used in various industries, has several inherent limitations that the authors acknowledge. One of the primary limitations is its subjective nature, as the severity, occurrence, and detection ratings are based on expert judgment and experience, which can vary among individuals [18]. Although this paper reflects a quasi-static snapshot analysis of a preprocessing system, FMEA is a continuous improvement framework that can adapt to dynamic changes in the processes as they appear [34]. Additionally, while FMEA focuses on individual failure modes, the authors recognize the importance of addressing the interactions or cascading effects between multiple failures, which is why a system-wide FMEA was also performed in this work [35]. Furthermore, FMEA lacks a quantitative risk assessment methodology, relying instead on semi-quantitative RPNs that may not accurately reflect the true risk levels [22]. Moreover, FMEA can be time-consuming and resource-intensive, especially for complex systems with numerous components and failure modes.

Uncertainty is a significant issue that can affect the reliability and accuracy of FMEA results. Uncertainties can arise from various sources, including imprecise data, subjective expert judgments, and inherent variability in system parameters [36], yet the FMEA approach proves highly valuable in scenarios where quantitative data are limited, as is the case in the present study. Failure to account for uncertainties can lead to inaccurate risk assessments and suboptimal decision-making [37]. As we progress and gather more quantitative data, we plan to mitigate these uncertainties using several approaches. Firstly, sensitivity analyses can be performed to evaluate the impact of input parameter variations on the FMEA results [38]. Secondly, fuzzy logic or probabilistic methods can be incorporated to quantify and propagate uncertainties through the FMEA process [39]. Additionally, Bayesian networks or evidential reasoning techniques can be integrated with FMEA to provide a more robust and quantitative risk assessment framework [39]. Furthermore, continuous monitoring and updating of FMEA data can help capture dynamic changes in the system and reduce uncertainties over time [22].

The case study presented here demonstrates the potential for an FMEA approach to be applied more broadly to bioenergy processes. The standardized methods for assessing process risk considering process efficiencies, product quality, economic impacts, sustainability, and safety simultaneously allows for identification of significant facility issues before they occur. The approach suggested in this study represents a proactive methodology to help ensure the viability and success of the bioenergy industry.

Author Contributions

Conceptualization, L.V.-M., R.M.E., N.S. and J.L.K.; methodology, R.M.E., P.H.B., L.V.-M. and T.B.; investigation, R.M.E., P.H.B., L.V.-M. and T.B.; data curation, R.M.E., P.H.B., L.V.-M. and N.S.; writing—original draft preparation, R.M.E. and N.S.; writing—review and editing, R.M.E., N.S., J.L.K., L.V.-M., P.H.B. and T.B.; supervision, R.M.E. and J.L.K.; funding acquisition, R.M.E. and J.L.K. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

Acknowledgments

The authors would like to acknowledge the expertise provided by the subject matter experts providing the input for the FMEA interviews: Neal Yancey, Jeffrey Lacey, Jordan Klinger, Kristian Egan, Corey Landon, Mark Small, Anthony D’Andrea, Quang Nguyen, Stephen Kanyid, and Cody Scheer from Idaho National Laboratory.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. High-temperature preprocessing system configuration.

Figure 2. Failure modes and effect analysis (FMEA) interview flowchart.

Figure 3. Moisture impacts on CQAs for each system unit operation.

Table 1

Severity, occurrence, and detection guideline scale for FMEA approach.

Mode	Guide Word	Rank	Criteria
Severity	Minor	1	None to minor disruption to production line. A small portion (<5%) of product may have to be reworked online.
	Low	3	Low disruption to production line. A portion (<15%) of product may have to be reworked or lost. Minor annoyance exists (e.g., increased operational oversight).
	Moderate	6	Moderate disruption to production line. A small portion (>20%) of product may have to be reworked or lost. Some inconvenience exists (e.g., continuous operational oversight).
	High	8	High disruption to production line. A portion (>30%) of product may have to be reworked or lost. Process may be stopped, and customer would be dissatisfied.
	Very high	10	Major disruption to production line. Close to 100% of the product may have to be scrapped. Process unreliable and failure occurs without warning. The customer is very dissatisfied. May endanger to operator the equipment.
Occurrence	Remote	1	Failure is very unlikely. No failures associated with similar processes.
	Low	3	Few failures. Isolated failures associated with similar processes.
	Moderate	6	Occasional failures associated with similar processes.
	High	8	Repeated failures. Similar processes have often failed.
	Very high	10	Process failure is almost inevitable.
Detection	Almost certain	1	Process control will almost certainly detect or prevent the potential cause of subsequent failure mode.
	High	3	High chance that the process control will detect or prevent the potential cause of subsequent failure mode.
	Moderate	6	Moderate chance that the process control will detect or prevent the potential cause of subsequent failure mode.
	Remote	8	Remote chance that the process control will detect or prevent the potential cause of subsequent failure mode.
	Very uncertain	10	There is no process control. Control will not or cannot detect the potential cause of subsequent failure mode.

Table 2

Failure Mode and Effects Analysis for the high-temperature system-wide configuration.

Failure	Impacts	CQAs	Severity	Causes	CMAs	CPPs	Occurrence	Detection Methods	Detection	RPN
Excessive overs production (>6 mm) Unit Operations: HM, OS	Reprocess material through recycling	Throughput (PE) Particle size (PQ)	6	Higher moisture HM screen size	Moisture	Screen size (HM) Mill speed (HM) Feed rate (system)	3	Energy consumption sensors (HM)	6	108
Excessive fines production (<1.18 mm) Unit Operations: HM, AC, OS	Material loss Conversion efficiency	Yield (PE, E) Particle Size (PQ)	3	HM screen size AC light fraction removal Moisture impacting particle size in HM Moisture impacting AC separation efficiency	Moisture content	Screen size (HM) Mill speed (HM) Feed rate (system)	3	Current readings for automated and manual in-process adjustments to feed rate Manual adjustment to pneumatic assist air flow	8	72
Deviation from fixed carbon specification (<18 or 21%) Unit Operations: AC	Conversion efficiency impacted through high amounts of anatomical fractions with lower fixed carbon	Fixed Carbon (PQ) Ash Content (PQ)	8	Inefficient separation of bark, needles, dirt/fines from the white wood in AC	Moisture content	Air speed (AC) system configuration (RD position)	3	Visual detection by trained observer—manual adjustment to air speed	8	192
	Increased inorganic material not removed through AC causing increased equipment wear	Throughput (PE)	3	Inefficient separation of bark, needles, dirt/fines from the white wood in AC	Inorganic species	Air speed (AC) system configuration (RD position)	3	Visual by trained observer—manual adjustment to air speed	8	72
Deviation from moisture specification (>10%) Unit Operations: RD, HM	Impacts to HT conversion efficiency Decrease/impact separation efficiency in AC Decrease separation efficiency in OS increasing fines in the final product Indirectly higher moisture impacts the efficiency of AC separation leading to high ash content and lower fixed carbon	Moisture (PQ) Fixed Carbon (PQ) Ash Content (PQ)	10	Initial moisture content of the material high (>40%) Feed rate through RD too high Chip size or grind size too large (>1″) Shredded material	Moisture content (>40%) Particle size (>1″) Particle Shape (non-chips)	Feed rate (RD) Outlet Temperature (RD) Airflow (RD) Rotational Speed (RD)	3	Offline moisture measurement to set initial feed rate and temperature and recycle number Observations by trained observer—manual adjust feed rate Automated inlet temperature adjustments based on outlet temperature readings	6	180
	HM energy increase (2X for every 10% increase in moisture) Decrease in system throughput from most equipment	Energy Consumption (PE) Throughput (PE)	8	Initial moisture content of the material high (>40%) Feed rate through RD too high Chip size or grind size too large (>1″) Shredded material	Moisture content (>40%) Particle size (>1″) Particle Shape (non-chips)	Feed rate (RD) Outlet Temperature (RD) Airflow (RD) Rotational Speed (RD)	3	Offline moisture measurement to set initial feed rate and temperature and recycle number Observations by trained observer—manual adjust feed rate Automated inlet temperature adjustments based on outlet temperature	6	144
Deviation from ash content specification (>1.75%) Unit Operations: AC, OS	Increased wear on equipment (HM especially)	Throughput (PE)	3	Inefficient separation of bark, needles, dirt/fines from the white wood in AC (moisture, physical properties of material) Higher bark content	Moisture content (>20%) Higher bark contents	Feed rate (AC) Air speed (AC)	3	No in-line controls for meeting specific ash specifications	10	90
	Impact on conversion efficiency	Ash Content (PQ)	8	Inefficient separation of bark, needles, dirt/fines from the white wood in AC (moisture, physical properties of material) Higher bark content Failure with OS	Moisture content (>20%) Higher bark contents	Feed rate (AC) Air Speed (AC) Screen size (OS)	1	No in-line controls for meeting specific ash specifications	10	80
Throughput lower than target Unit Operations: HM, AC, RD	Target material volume not achieved Impacts on conversion process (in a continuous system)	Throughput (PE) Yield (PE)	3	Initial moisture—longer drying times and/or more drying cycles. Moisture (>10%) impacts HM throughput. Chip size and harvest method impact drying time Ash impact hammer mill mostly Too much air speed on AC removes too much material.	Moisture Particle size Ash/soil	Outlet temperature (RD) Air Flow (RD) Feed rate (System) Air Speed (AC)	3	Level sensors in system conveyors	6	54
	Lower throughputs due to decreased feed rates can generate more fines through milling process More material discarded	Particle Size (PQ) Particle Size Distribution (PQ) Yield (PE)	10	Initial moisture—longer drying times and/or more drying cycles. Moisture (>10%) impacts hammer mill throughput. Chip size and harvest method impact drying time	Moisture Particle size Ash/soil	Outlet temperature (RD) Air Flow (RD) Feed rate (System) Air Speed (AC)	3	Level sensors in system conveyors Energy consumption (HM)	6	180

Table 3

FMEA mitigation strategies for the high-temperature system-wide configuration.

Failure	Severity	Occurrence	Detection	RPN	Mitigation(s)	Severity	Occurrence	Detection	RPN
Excessive overs production (>6 mm) Unit Operations: HM, OS	6	3	6	108	In-line particle size analyzer (in-process) Replace HM with rotary shear mill; more experimental data to support optimal screen combination for meeting particle size specifications (Implemented)	6	3	3	108
Excessive fines production (<1.18 mm) Unit Operations: HM, AC, OS	3	3	8	72	System reconfiguration to move RD to after milling step. Would product more overs; less fines might be removed through AC unit.	3	TBD	8	TBD
Deviation from fixed carbon specification (<18 or 21%) Unit Operations: AC	8	3	8	192	Moisture sensor for detecting materials higher than 10% moisture (In-process) Visual detection for identifying “non-white wood” Carbon concentration sensor	8	3	3	72
	3	3	8	72	Moisture sensor for detecting materials higher than 10% moisture (In-process) Visual detection for identifying “non-white wood” Carbon concentration sensor	6	3	3	54
Deviation from moisture specification (>10%) Unit Operations: RD, HM	10	3	6	180	Replace HM with rotary shear mill; impact of moisture not as significant—(implemented) After RD in-line moisture sensor (in-process)	10	3	3	90
	8	3	6	144	Replace HM with rotary shear mill; impact of moisture not as significant—(implemented) After RD in-line moisture sensor (in-process) Tarping or covering material; rain prevention Particle size sensor and mass sensor	8	3	3	72
Deviation from ash content specification (>1.75%) Unit Operations: AC, OS	3	3	10	90	In-line moisture (and ash) sensors after RD, AC and OS	3	1	6	18
	8	1	10	80	In-line moisture (and ash) sensors after RD, AC and OS	8	1	6	48
Throughput lower than target Unit Operations: HM, AC, RD	3	3	6	54	In-line moisture sensor	3	3	3	27
	10	3	6	180	In-line moisture sensor	10	3	3	90

Note: TBD: To be determined. As this is a theoretical mitigation strategy, the SME did not have a good idea of the potential reduction in failure occurrence.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/en17112516/s1, FMEA Interview Form, Table S1: Failure Mode and Effects Analysis for the rotary dryer processing pine residue chips to 10% moisture. Table S2: Failure Mode and Effects Analysis for the air classifier separating anatomical fractions of pine residue chips. Table S3: Failure Mode and Effects Analysis for hammer mill processing pine residue chips. Table S4: Failure Mode and Effects Analysis for the oscillating screen processing pine residue chips. Table S5: Failure Mode and Effects Analysis for drag chain conveyors.

References

1. Errera, M.R.; Dias, T.A.d.C.; Maya, D.M.Y.; Lora, E.E.S. Global bioenergy potentials projections for 2050. Biomass Bioenergy; 2023; 170, 106721. [DOI: https://dx.doi.org/10.1016/j.biombioe.2023.106721]

2. U.S. Energy Information Administration. Biomass Explained. Available online: https://www.eia.gov/energyexplained/biomass/ (accessed on 22 May 2024).

3. Yogalakshmi, K.N.; Sivashanmugam, P.; Kavitha, S.; Kannah, Y.; Varjani, S.; Adish Kumar, S.; Kumar, G. Lignocellulosic biomass-based pyrolysis: A comprehensive review. Chemosphere; 2022; 286, 131824. [DOI: https://dx.doi.org/10.1016/j.chemosphere.2021.131824] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34388872]

4. Omoriyekomwan, J.E.; Tahmasebi, A.; Dou, J.; Wang, R.; Yu, J. A review on the recent advances in the production of carbon nanotubes and carbon nanofibers via microwave-assisted pyrolysis of biomass. Fuel Process. Technol.; 2021; 214, 106686. [DOI: https://dx.doi.org/10.1016/j.fuproc.2020.106686]

5. Qiu, B.; Tao, X.; Wang, J.; Liu, Y.; Li, S.; Chu, H. Research progress in the preparation of high-quality liquid fuels and chemicals by catalytic pyrolysis of biomass: A review. Energy Convers. Manag.; 2022; 261, 115647. [DOI: https://dx.doi.org/10.1016/j.enconman.2022.115647]

6. Yaashikaa, P.; Kumar, P.S.; Varjani, S.J.; Saravanan, A. Advances in production and application of biochar from lignocellulosic feedstocks for remediation of environmental pollutants. Bioresour. Technol.; 2019; 292, 122030. [DOI: https://dx.doi.org/10.1016/j.biortech.2019.122030] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31455552]

7. Bridgwater, A.; Peacocke, G. Fast pyrolysis processes for biomass. Renew. Sustain. Energy Rev.; 2000; 4, pp. 1-73. [DOI: https://dx.doi.org/10.1016/S1364-0321(99)00007-6]

8. Bridgwater, A.V. Review of fast pyrolysis of biomass and product upgrading. Biomass Bioenergy; 2012; 38, pp. 68-94. [DOI: https://dx.doi.org/10.1016/j.biombioe.2011.01.048]

9. Mohan, D.; Pittman, C.U., Jr.; Steele, P.H. Pyrolysis of wood/biomass for bio-oil: A critical review. Energy Fuels; 2006; 20, pp. 848-889. [DOI: https://dx.doi.org/10.1021/ef0502397]

10. Ringer, M.; Putsche, V.; Scahill, J. Large-Scale Pyrolysis Oil Production: A Technology Assessment and Economic Analysis; National Renewable Energy Lab. (NREL): Golden, CO, USA, 2006; 93.

11. Lan, K.; Park, S.; Kelley, S.S.; English, B.C.; Yu, T.H.E.; Larson, J.; Yao, Y. Impacts of uncertain feedstock quality on the economic feasibility of fast pyrolysis biorefineries with blended feedstocks and decentralized preprocessing sites in the Southeastern United States. Glob. Chang. Biol. Bioenergy; 2020; 12, pp. 1014-1029. [DOI: https://dx.doi.org/10.1111/gcbb.12752]

12. Ore, O.T.; Adebiyi, F.M. A review on current trends and prospects in the pyrolysis of heavy oils. J. Pet. Explor. Prod.; 2021; 11, pp. 1521-1530. [DOI: https://dx.doi.org/10.1007/s13202-021-01099-0]

13. Rezaei, H.; Yazdanpanah, F.; Lim, J.C.; Lau, A.; Sokhansanj, S. Woody feedstock pretreatments to enhance pyrolysis bio-oil quality and produce transportation fuel. Biomass for Bioenergy-Recent Trends and Future Challenges; IntechOpen: London, UK, 2018.

14. Tumuluru, J.S. Biomass Preprocessing and Pretreatments for Production of Biofuels: Mechanical, Chemical and Thermal Methods; CRC Press: Boca Raton, FL, USA, 2018.

15. Tumuluru, J.S.; Wright, C.T.; Kenney, K.L.; Hess, R.J. A technical review on biomass processing: Densification, preprocessing, modeling and optimization. Proceedings of the 2010 ASABE Annual International Meeting; Pittsburgh, Pennsylvania, 20–23 June 2010; Volume 1.

16. Tumuluru, J.S.; Yancey, N. Conventional and advanced mechanical preprocessing methods for biomass: Performance quality attributes and cost analysis. Biomass Preprocessing and Pretreatments for Production of Biofuels; CRC Press: Boca Raton, FL, USA, 2018; pp. 15-49.

17. Yancey, N.; Wright, C.T.; Conner, C.; Hess, J.R. Preprocessing Moist Lignocellulosic Biomass for Biorefinery Feedstocks; Idaho National Lab. (INL): Idaho Falls, ID, USA, 2009.

18. Stamatis, D.H. Failure Mode and Effect Analysis; Quality Press: Welshpool, WA, USA, 2003.

19. Pillai, D.S.; Rajasekar, N. A comprehensive review on protection challenges and fault diagnosis in PV systems. Renew. Sustain. Energy Rev.; 2018; 91, pp. 18-40. [DOI: https://dx.doi.org/10.1016/j.rser.2018.03.082]

20. Burli, P.H.; Hartley, D.S.; Thompson, D.N. Woody Feedstocks 2021 State of Technology Report; Idaho National Laboratory (INL): Idaho Falls, ID, USA, 2021; 51.

21. Burli, P.H.H.; Damon, S.; Thompson David, N. Woody Feedstocks 2022 State of Technology Report; INL/RPT-22-69474 Idaho National Laboratory (INL): Idaho Falls, ID, USA, 2022.

22. Liu, H.-C.; Liu, L.; Liu, N. Risk evaluation approaches in failure mode and effects analysis: A literature review. Expert Syst. Appl.; 2013; 40, pp. 828-838. [DOI: https://dx.doi.org/10.1016/j.eswa.2012.08.010]

23. Sankar, N.R.; Prabhu, B.S. Modified approach for prioritization of failures in a system failure mode and effects analysis. Int. J. Qual. Reliab. Manag.; 2001; 18, pp. 324-336. [DOI: https://dx.doi.org/10.1108/02656710110383737]

24. Fonseca, F.G.; Funke, A.; Niebel, A.; Dias, A.P.S.; Dahmen, N. Moisture content as a design and operational parameter for fast pyrolysis. J. Anal. Appl. Pyrolysis; 2019; 139, pp. 73-86. [DOI: https://dx.doi.org/10.1016/j.jaap.2019.01.012]

25. Demirbas, A. Effect of initial moisture content on the yields of oily products from pyrolysis of biomass. J. Anal. Appl. Pyrolysis; 2004; 71, pp. 803-815. [DOI: https://dx.doi.org/10.1016/j.jaap.2003.10.008]

26. Mariyam, S.; Alherbawi, M.; Pradhan, S.; Al-Ansari, T.; McKay, G. Biochar yield prediction using response surface methodology: Effect of fixed carbon and pyrolysis operating conditions. Biomass Conv. Bioref.; 2023; pp. 1-14. [DOI: https://dx.doi.org/10.1007/s13399-023-03825-6]

27. Yildiz, G.; Ronsse, F.; Venderbosch, R.; Duren, R.v.; Kersten, S.R.A.; Prins, W. Effect of biomass ash in catalytic fast pyrolysis of pine wood. Appl. Catal. B Environ.; 2015; 168–169, pp. 203-211. [DOI: https://dx.doi.org/10.1016/j.apcatb.2014.12.044]

28. Thompson, D.N.; Hartley, D.S. Particle Scale Impacts on Deconstruction Energy of Pine Residues; Idaho National Lab. (INL): Idaho Falls, ID, USA, 2022.

29. Klinger, J.; Carpenter, D.L.; Thompson, V.S.; Yancey, N.; Emerson, R.M.; Gaston, K.R.; Smith, K.; Thorson, M.; Wang, H.; Santosa, D.M. et al. Pilot Plant Reliability Metrics for Grinding and Fast Pyrolysis of Woody Residues. ACS Sustain. Chem. Eng.; 2020; 8, pp. 2793-2805. [DOI: https://dx.doi.org/10.1021/acssuschemeng.9b06718]

30. Tao, L.; Tan, E.C.; McCormick, R.; Zhang, M.; Aden, A.; He, X.; Zigler, B.T. Techno-economic analysis and life-cycle assessment of cellulosic isobutanol and comparison with cellulosic ethanol and n-butanol. Biofuels Bioprod. Biorefining; 2014; 8, pp. 30-48. [DOI: https://dx.doi.org/10.1002/bbb.1431]

31. Dai, J.; Grace, J.R. Biomass granular screw feeding: An experimental investigation. Biomass Bioenergy; 2011; 35, pp. 942-955. [DOI: https://dx.doi.org/10.1016/j.biombioe.2010.11.026]

32. Chandel, A.K.; Garlapati, V.K.; Singh, A.K.; Antunes, F.A.F.; da Silva, S.S. The path forward for lignocellulose biorefineries: Bottlenecks, solutions, and perspective on commercialization. Bioresour. Technol.; 2018; 264, pp. 370-381. [DOI: https://dx.doi.org/10.1016/j.biortech.2018.06.004] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29960825]

33. Ilic, D.; Williams, K.; Farnish, R.; Webb, E.; Liu, G. On the challenges facing the handling of solid biomass feedstocks. Biofuels Bioprod. Biorefining; 2018; 12, pp. 187-202. [DOI: https://dx.doi.org/10.1002/bbb.1851]

34. Shahin, A. Integration of FMEA and the Kano model. Int. J. Qual. Reliab. Manag.; 2004; 21, pp. 731-746. [DOI: https://dx.doi.org/10.1108/02656710410549082]

35. Xu, K.; Tang, L.C.; Xie, M.; Ho, S.L.; Zhu, M.L. Fuzzy assessment of FMEA for engine systems. Reliab. Eng. Syst. Saf.; 2002; 75, pp. 17-29. [DOI: https://dx.doi.org/10.1016/S0951-8320(01)00101-6]

36. Liu, Y.; Tang, Y. Managing uncertainty of expert’s assessment in FMEA with the belief divergence measure. Sci. Rep.; 2022; 12, 6812. [DOI: https://dx.doi.org/10.1038/s41598-022-10828-2]

37. Carlson, C.S. Understanding and applying the fundamentals of FMEAs. Proceedings of the 2014 Annual Reliability and Maintainability Symposium; Colorado Springs, CO, USA, 27–30 January 2014; pp. 1-35.

38. Gul, M.; Yucesan, M.; Celik, E. A manufacturing failure mode and effect analysis based on fuzzy and probabilistic risk analysis. Appl. Soft Comput.; 2020; 96, 106689. [DOI: https://dx.doi.org/10.1016/j.asoc.2020.106689]

39. Yucesan, M.; Gul, M.; Celik, E. A holistic FMEA approach by fuzzy-based Bayesian network and best–worst method. Complex Intell. Syst.; 2021; 7, pp. 1547-1564. [DOI: https://dx.doi.org/10.1007/s40747-021-00279-z]