Introduction
Biomass can be defined as a carbon-based material derived from living or recently living organisms. Biomass is a sustainable and renewable resource. In recent years, the use of biomass as a feedstock has been key to the development of biorefineries, several of which are either built, under-construction, or planned (
Demirbas, 2009;
Zhang et al., 2003). Biodiesel and bioethanol have become more attractive due to their renewable and environmental benefits (
Hansen et al., 2005;
Ma and Hanna, 1999).
This impressive development also needs to consider process safety, which has not been perceived to date as a potential barrier to sustainable development for biofuel production. The exponential growth of new biofuel industry faces several challenges, such as product yield, process efficiency, policies, and logistics. However, the issue of safety during biofuel production (
Casson Moreno and Cozzani, 2015;
Gómez et al., 2013;
Sovacool et al., 2015) is important. Although biodiesel and bioethanol production processes are relatively simple, the operational units still involve important risks; thus, accidents can evidently occur and be fatal. For example, biodiesel is a safe substance, but its production process can be dangerous because methanol causes plants to be vulnerable to fire and explosion (
Chimica and Federico, 2010;
Moss, 2010).
Recently, a number of significant accidents have occurred at biofuel facilities, resulting in the destruction of property, injury, and loss of life. For comparison, the accidents in the ethanol (
Calvo Olivares et al., 2015) and biodiesel industries (
Calvo Olivares et al., 2014) are listed with the accidents in conventional petroleum refineries. These data are plotted in Fig. 1 to indicate the number of accidents per billion barrels of fuel produced in biofuel plants (biodiesel and bioethanol) and petroleum refineries (gasoline and diesel). Therefore, while the absolute number of accidents in petroleum refineries is approximately the same as in biodiesel and ethanol plants, more accidents occur per barrel of biofuel produced than in conventional petroleum refineries. This finding can be attributed to one misconception in biofuel industries, wherein ethanol is regarded as a safe substance in prevalent opinion. Therefore, effective process safety managements are lacking, and accidents are more likely to happen in biofuel production.
Process safety management (PSM) programs have been implemented in many chemical process industries and companies around the world since the PSM of Highly Hazardous Chemical standard, 29CFR 1910.119, was implemented by the US Occupational Safety and Health Administration (OSHA) in 1992 (
Luo, 2010) to prevent major accidents. Accident risks have been evidently reduced in the chemical process by the application of PSM programs. Therefore, the misconceptions for biofuel must be seriously reconsidered. PSM programs are keys to improve the risk management in biofuel process.
This study aims to emphasize that sustainable production of biofuels cannot be achieved without appropriate consideration for process safety. Therefore, we present a case study of an ethanol distillation and dehydration process using hazard and operability study (HAZOP), which is one of the most important process hazard analysis (PHA) methods in PSM programs.
Methods
This study focuses on the distillation and subsequent dehydration process of a fuel ethanol plant to identify potential hazards at the level of process design and operating procedure. We analyzed the piping and instrument diagrams (P&IDs) of the plant and generated a list of required recommendations that are tabulated in a report formed by PSMSuite
®, which is a commercialized PSM software based on the prototype system PetroHAZOP (
Zhao et al., 2009). The process and the HAZOP results are presented in the following sections.
PSM standard and HAZOP study
The PSM standard, 29CFR 1910.119, which was implemented by the US OSHA in 1992, defined 14 elements, covering the entire lifecycle of a chemical plant and all aspects of the chemical process. Among the provisions defined by this standard, PHA is the most important. The purpose of PHA is to review a process design to identify hazardous scenarios and ensure that they are appropriately safeguarded.
Several PHA methods recommended in the PSM standard include the HAZOP study, a what-if/checklist analysis, fault tree analysis, and the failure mode and effect analysis. The HAZOP study method is most extensively applied in the chemical process industries (
Baybutt, 2015), especially in the petrochemical industries, due to its thorough and holistic nature. Although this method has been applied in other industries, it has been rarely reported in the biofuel industry. As a structured hazard identification method, HAZOP was first proposed by Imperial Chemical Industries in the UK to identify hazards in chemical plants in the 1960s (
CIA, 1997). HAZOP analysis assumes that hazards arise in a process plant due to deviations from design intents or from acceptable normal behavior.
In the HAZOP study, P&IDs are systematically examined by a group of experts (HAZOP team). The main objective is to stimulate the imagination of the review team, including designers and operators, in a systematic manner; thus, they can identify the abnormal causes and adverse consequences for all possible deviations from the normal operation that could arise in each section of the plant. The method systematically and critically identifies all the possible causes and consequences of each hypothesized process deviation in a formal and systematic approach. The methodology of HAZOP was described in a text by Kletz in 1999.
The concept of HAZOP involves the splitting of the process plant into sections and the systematic application of a series of questions to each section. The study team discovers the occurrence of deviations from the intended design and determines the consequences of the deviations from the viewpoint of hazards and operability.
HAZOP analysis can be used throughout the development of the project, from the feed to the operational phase. In the design phases, HAZOP is used to ensure that all potential risks and operability issues fall within the projects using the As Low As Reasonably Practicable (ALARP) criteria. In the operational phase, HAZOPs are commonly used as part of the management of change process. A basic HAZOP process flow is depicted in (Fig. 2).
Case study: Biofuel ethanol distillation and dehydration
The selected case study focuses on the distillation and subsequent dehydration section of a biofuel ethanol production process in a company in Henan province, China. The distillation and dehydration sections are the two final stages of the biofuel ethanol production process (
Ho et al., 2014), through which the water is removed from the feed stream, and the ethanol purity is raised from less than 10% to more than 99%. The distillation section is a double-effect differential pressure distillation process with four columns. Molecular sieve drying technology is adopted in the dehydration section. In this section, ethanol is passed through a molecular sieve bed with uniform pore sizes that preferentially adsorb water.
A simplified process flow diagram (PFD) of the distillation and dehydration process is shown in Fig. 3. For confidentiality reasons, details of the operating conditions and the process parameters are not presented herein; thus, the integrity of this work is not compromised. The feedstock mash, which contains approximately 10% ethanol, is first heated through heat exchanger E-18 and then split into two streams. One stream is fed into the negative pressure distillation line (starting from T-28), while the other stream is fed into the atmospheric distillation line (starting from T-11). In the negative pressure, the mash is first distilled in column T-28 (the first distillation column), and then the production stream is cooled and liquefied in T-32. In this manner, the purity of the ethanol is raised to more than 30%. Subsequently, the stream is fed to the rectifying column T-03, which purifies the ethanol to up to more than 90% under low pressure. In the atmospheric distillation line, the stream is distilled in column T-11, and the ethanol is purified to more than 40%. Then, after cooling, the stream is fed into the atmospheric rectifying column T-21, whose ethanol purity is raised to that of T-0.
After distillation, the two lines are fed into a buffer tank V-24 and subsequently fed to the adsorption column T-01, in which the water in the stream is reduced, and the ethanol content is raised again. Finally, the product is cooled and stored in the production tanks. The five tanks involved in the distillation operation, namely V-01A/B, V-24, V-16, and V-08, act as the column buffers, adjusting the feed in flow rate or the level of each column. The two distillation lines, which work under different conditions, are served by several heat exchangers to ensure efficient heat recovery and reduce energy consumption.
Level control loops are installed on each buffer tank and column, and every column has a flow control loop installed on its input stream line. A few of the heat exchangers and pipeline segments have pressure and temperature indicators with alarms. A complex subsystem program control is installed on the adsorption column T-01. This program control can automatically switch each column segment between the absorbing and recovering modes. In addition, other parts of the process are not highly automated; thus, human supervision is necessary. Also, the manual operations, such as the backup pump switch operations and the backup heat input operations, are often required. The main equipment and instruments of the process are listed in Table 1.
HAZOP methodology
The expert team of this HAZOP study comprised four people. The HAZOP team leader and the secretary were from the Chemical Process Accident Prevention and Emergency Response Research Center of Tsinghua University. The two other team members were from the biofuel ethanol plant: one is the manager of the plant, who is a skilled technologist, and the other is the manager of the plant safety department. Prior to the study, the team prepared the necessary documents for the process, including the process technical description, the process PFD and P&ID, the specifications of the process equipment and instruments, the material safety data sheets of the chemicals involved in the process, the up-to-date operating procedures, and the history and accident records at the plant. After understanding the design intent of the process, the team decided to divide the distillation and dehydration process according to the main equipment and streams. The process was divided into seven nodes as follows:
(a) First distillation column
(b) Outlet stream from the first distillation column
(c) First rectification column
(d) Second distillation column
(e) Second rectification column
(f) Production condensing
(g) Dehydration
Results and discussion
The total time of the HAZOP study was approximately 30 h. A total of 68 deviations of 42 parameters were analyzed, and 99 adverse consequences were identified. The consequences were classified, and the resulting statistics are presented in Fig. 4(a). The potential fire and explosion hazards account for the largest portion due to the nature of the chemicals within the process. The risk management standard of the company is shown in Fig. 4(b) (
Kang, 2009). The standard classifies potential accidents into four different risk levels, as follows:
Risk level I- the corresponding consequence hazard is slight.
Risk level II- the hazard is tolerable, but modifications must be performed to avoid the consequence.
Risk levels III and IV- the hazard is severe and highly severe, respectively. Therefore, appropriate counter measures must be recommended by the team to reduce the risk level.
The experts estimated the risk levels of 99 safety-related consequences using this risk matrix to clarify the harmful effects of such consequences. The distribution of all estimated risks is shown in Fig. 4(c). Although the biofuel ethanol distillation process is not highly dangerous, approximately one-third of adverse consequences with unacceptable risk levels still exist. The length of the entire HAZOP report reached more than 90 pages; therefore, it cannot be fully presented in this paper. Table 2 lists a section of the HAZOP study results, which show risk levels III and IV. In the table, all results with risk level IV are listed. However, Fig. 3 shows that the PFD is simplified, and a few auxiliary equipment is hidden. The HAZOP results with risk level III related with the hidden sections are removed from Table 2 to avoid confusion.
Through the HAZOP analysis, the HAZOP expert team proposed a total of 36 recommended action items. Table 3 lists the recommendations, and Table 4 shows the statistics for the types of actions. The plant owner accepted all the recommendations. Up to the time the authors completed this paper, the ethanol plant mentioned has implemented 90% of the action items raised by the recommendations, and the remaining 10% still remain under consideration. HAZOP analysis is a time-consuming process; thus, a smart software, which is referred as PSMSuite
®, was developed by the Tsinghua University research team to reduce the HAZOP team workload. PSMSuite
® provides a user-friendly software platform for knowledge management, data organization, history record browsing, and report generation. This software can effectively improve the work and reduce human errors during the HAZOP meetings. Figure 5 shows the main user interface of the software. PSMSuite
® has been developed based on the prototype system PetroHAZOP (
Zhao et al., 2009), which utilizes the case-based reasoning technology. Thus, the HAZOP analysis case based on the biofuel ethanol distillation and dehydration process has been established to analyze other biofuel ethanol production processes.
HAZOP is a well-known study and is utilized by the majority of chemical industries worldwide. This case study represents the majority of biofuels plants, which are currently operational, and shows the lack of awareness and understanding of hazards. Biofuel producers must adhere to standard safety practices.
The study focused on the biofuel ethanol distillation and dehydration process. Although this process includes a mature ethanol separation process, which does not contain many dangerous chemicals or extreme operating conditions, several unacceptable dangerous scenarios that were not disclosed prior to completion of the analysis still exist. The fire and explosion hazards account for approximately 56% of the identified adverse consequences. Moreover, the fire and explosions in chemical plants usually result in adverse environmental consequences. Therefore, recommendations to reduce the fire and explosion risks also provide benefits for a clean production.
Additionally, the production loss scenarios identified through this HAZOP study represent 22% of the consequences, thus resulting in feedstock waste and out-of-specification products. During the HAZOP study, we also identified a few situations with assigned low risk levels. Although these situations do not lead to major process safety hazards, they may still affect the efficiency of the plant. The adverse situations include waste of energy, loss of intermediate products, product degradation, and equipment or instrument damage. Recommendations were also raised for each of these situations during the one-week HAZOP meetings.
Although the methodology of the HAZOP study was presented to decrease the occurrences of major process safety hazards, the result of the HAZOP study reveals not only the major safety hazards but also all other operation, quality, and environmental risks. The bioethanol plant can comprehensively benefit from a thorough HAZOP study. Therefore, if possible, the authors suggest that all biofuel ethanol plants, including plants that are considered extremely safe, should undergo a thorough examination using the HAZOP method. Apart from the basic process control system and the combustible gas detectors installed in the site area, no other safeguards exist in the hazard prevention process. Through our preliminary study of the PI&D of the biofuel ethanol production process, the safeguards, such as pressure relief devices or safety instruments system that are commonly used in the petrochemical industry, were absent in this biofuel ethanol plant. Although the ethanol distillation process does not pose a high, the lack of hazard prevention consideration might still result in a few potential risks, which were identified by our HAZOP study.
Conclusions
Improving refinery safety is a goal strongly shared by governments, industries, workers, and communities. Unfortunately, the misconception regarding biofuel production safety issues leads to the majority of accidents due to lack of experience. The presented HAZOP study case indicates that although the bioethanol process is mature and does not contain dangerous chemicals or extreme operating conditions, unacceptable dangerous scenarios, such as fire and explosion, still exist. The case study reveals that potential high-level risks exist in the current biofuel process design and operating procedures, and these risks can be better controlled if they can be immediately identified. Immediately improving safety measures and working procedures and training personnel in this industry category is required. In addition, exchange of recorded information between industries and having access to a database can facilitate learning from the experiences of other companies and identification of the most probable safety tool to avoid accidents.
The Author(s) 2017. Published by Higher Education Press. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0)
PDF
(686KB)
