Integrated Data (i-Data), Mining and Utilization ... - ACS Publications


Integrated Data (i-Data), Mining and Utilization...

1 downloads 83 Views 4MB Size

Subscriber access provided by University of Newcastle, Australia

Article

Integrated Data (i-Data), Mining and Utilization Approach for Effective Flare Management Strategies Ahmed AlNouss, Monzure-Khoda Kazi, Fahd Mohammed, and Fadwa Eljack Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.6b01774 • Publication Date (Web): 19 Feb 2017 Downloaded from http://pubs.acs.org on February 20, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 1. Total CO2 emissions in MMTA as per consumption and flaring of NG 262x138mm (96 x 96 DPI)

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2. Sources of Abnormal Process Losses 338x190mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 2 of 48

Page 3 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 3. Integrated Approach of Data Mining, Analysis and Utilization for Flare Reduction (i-Data) 254x190mm (96 x 96 DPI)

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. Process flow diagram for the Ethylene plant 406x265mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 4 of 48

Page 5 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 5. Frequency of Ethylene plant incidents over 10 years duration 330x165mm (96 x 96 DPI)

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. i-Data Flowchart – Level 1 1122x466mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 6 of 48

Page 7 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 7. Total flared amount per month for each of the ethylene plant incidents 258x167mm (96 x 96 DPI)

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 8. Total flared amount per year for each of the ethylene plant incidents relative to the mean value 952x1270mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 8 of 48

Page 9 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 9. i-Data Flowchart – Level 2 1294x365mm (96 x 96 DPI)

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 10. Data utilization from i-DATA for systematic integration of flare and water management tools 254x190mm (200 x 200 DPI)

ACS Paragon Plus Environment

Page 10 of 48

Page 11 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 11. i-Data Flowchart 1295x1350mm (96 x 96 DPI)

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 48

Integrated Data (i-Data), Mining and Utilization Approach for Effective Flare Management Strategies Ahmed AlNoussa, Monzure-Khoda Kazi a, Fahd Mohammeda, and Fadwa Eljacka* a

Qatar University, Department of Chemical Engineering, College of Engineering, Doha, Qatar, P.O. Box-2713

Abstract. Upset emissions occur during plant startup, shutdown, maintenance, malfunction and flaring incidents. A wide range of these upsets cannot be managed by standalone control systems; plant personnel intervention is necessary sometimes. The methods needed to assist plant personnel in order to control and prevent abnormal process operations are gathered under abnormal situation management. Abnormal operations that lead to flare have significant economic, environmental, and safety impacts. Flaring is necessary for managing process upsets, however, it leads to the emission of greenhouse gases (GHG) and volatile organic compounds (VOCs), causing negative social impacts and local transient air pollution. In addition, excessive flaring results in energy and raw material losses. These are valuable commodities that must be sustained. Therefore, flare minimization during normal and abnormal operational situations has great environmental, industrial, and societal benefits. It is not possible to quantify the impacts without understanding the properties and magnitude of these upsets. Such analysis requires extensive amount of historical data. There are large sets of design, operational and flaring data readily available;

however, the challenge when it comes to flare mitigation is in using them effectively and in a timely manner. In this paper, a systematic approach to collect, analyze and utilize historical flaring data based on current industrial practices is presented. An ethylene base case study along with its historical process and flaring incidents data is used to demonstrate the significance of using and integrating data within developed flare management strategies. In the presented case, design and historical process data are used to assess the environmental impacts of abnormal incidents; and to identify underlying causes and indicators that lead to process upset, i.e. abnormal situations. The data sets are utilized within an optimization algorithm to identify design alternatives to mitigate process incidents and reduce its root causes. The paper highlights the challenges that are faced by environmental agencies in terms of data utilization and documentation.

Keywords: Abnormal situation management (ASM), Data Integration, Flaring Data, Flare Reduction, Ethylene Plant, Process Upset.

1. Introduction 1.1 Flaring

Flaring is a common and vital operational activity in industrial facilities and is usually intended for safety purposes, for disposal of waste gases, or for management of abnormal situations. It serves an important role in protecting operators, owner, plant and environment. In addition to the economic losses associated with ineffective combustion of hydrocarbon streams, flaring contributes significantly to the emission of greenhouses gases and pollutant precursors. According to World Bank statistics, the volume of flared streams is at least 108 billion cubic meters per annum which can contribute up to 11% of the total GHG emission sources, including flaring and accidental releases or abnormal operation.1, 2

1

ACS Paragon Plus Environment

Page 13 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Gas flaring has been linked to chronic health problems and destruction of the environment and is a huge contributor to air pollution.3,

4

Moreover, comparing the emissions coming from natural gas (NG)

consumption versus emissions coming from NG flaring, we find that in 2008 emissions due to flaring accounts for 10% of the average flaring due to consumption (see Figure 1).5, 6 This 10% is equivalent to 4.5 million metric tons per year (MMTA) of CO2 emissions and 675,000 Wobbe Index (WI). This WI is the best representation of energy losses in a system and the main indicator of the interchangeability of fuel gases and is usually defined in the specifications of gas supply and transport utilities.7 Flaring is an incredible waste of valuable natural resources that can be directly transformed into profits on the bottom line.3, 4

CO2 Emissions

2008

2009

2010

2011

2012

Consumption (MMTA)

42.05

43.94

46.92

57.06

74.07

Flaring (MMTA)

6.95

7.66

5.41

1.08

1.36

Figure 1. Total CO2 emissions in MMTA as per consumption and flaring of NG.

5, 6

Such economic and environmental issues are of key concern to the oil and gas producing nations; hence researchers and companies are working simultaneously to find solutions for flare reduction. In addition, numerous protocols, international agreement and steps such as the Kyoto protocol, the United Nations Environment Programme (UNEP), and the World Bank Global Gas Flaring Reduction (GGFR) program have been initiated to mitigate the impact of industrial flaring. This is essential to avoid dangerous anthropogenic interference with the climate system.8

Countries such as United States (US), France, and Saudi Arabia have attempted different alternatives to reduce the amount of gas flared in their country.9-12 The US was able to reduce their emission level by 5.8% between 2008 and 2009.5, 6, 10 Developing courtiers like Qatar in the Gulf have been able to reduce industrial flaring rates from 7.7 MMTA to 1.4 MMTA in a 3 year span, in spite of increased industrial activity. Yet, 2-3% of the overall NG consumption still accounts for 1.5 MMTA CO2 emissions which is equivalent 45 Thousands Wobbe Index. Although, governments and companies have had success in 2

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 48

reducing flare gas with significant investments, global gas flaring has remained largely stable over the past fifteen years in the range of 140 to 170 billion cubic meters (BCM) annually, which is equivalent of 300 million tons CO2 emissions.13-15 Thus, the flaring problem is still present and the techniques and strategies must be developed and implemented to mitigate the causes.5, 6

Recent articles and reports show different approaches have been suggested for effective flare reduction and utilization such as: utilization of renewable and alternative energy, process efficiency improvement, modernization and conversion of old techniques, waste streams recovery, combining heat and power through process integration, and innovative design solutions for flare gas utilization.10-12, 16, 17 The concept of providing credits for carbon reduction along with viable flare reduction solution can be the added incentive that push industries for adopting enhanced flare management, recovery and utilization strategies.18

In order to comply with new regulations and reduce socio-economic and environmental impact of flaring, there is a need to enhance and improve flare reduction management practices and design. The main challenge faced by statistical and research institution is the lack of readily available information about the flaring incidents and its properties. The lack of both data availability and flare management industrial practices in literature have hindered the academic (non-industrial) research community from developing suitable mitigation and management strategies when it comes to flare.19-26

The right questions to be asked here are; how much of these flares can be recovered? What are the proper techniques to do this? And is there a systematic way in which date related to the historical flare incidents can be collected, analyzed and utilized in such a way to yield the correct mitigations?

1.2 Abnormal Situations

It is common practice in process operation to flare under normal situations as a safety precaution in order to protect the operator, the owner, and the plant facility. It is standard operation procedure to flare during upsets that occur in plant operation, such as equipment malfunction, off-spec production, depressurizing gas processing equipment, startup, or emergency shutdowns.17 The loss that occurs in the course of converting an input raw material into finished products is known as process loss. Process loss does include both the normal process loss; expected or anticipated loss prior to production, and the abnormal process loss; the loss realized over the normal loss. Weight losses, shrinkage, evaporation, rusting and others are the examples of normal loss. Normal loss increases the cost of production of the usable goods 3

ACS Paragon Plus Environment

Page 15 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

realized. Whereas, abnormal loss arises because of abnormal working conditions, bad working condition, carelessness, rough handling, lack of proper knowledge, low quality raw material, machine breakdown, accident and others.27 Figure 2 shows the sources of abnormal process loss as percentage of potentiality ranging across various sites.28

Figure 2. Sources of Abnormal Process Losses.28

Preventable losses from the unexpected process disruptions that is called “Abnormal Situations” charge the U.S. economy at least $20 billion annually.29 Distributed control systems are used in the petrochemical plants to control thousands of process variables in order to prevent these unexpected upsets.29,

30

On the other hands, the human role is essential in monitoring these advanced automated

control systems in order to make precise and timely control decisions that needs situational awareness.31-34 Managing these abnormal situations has its significant impact on the process industry economy, safety and environment.17,

35

Remarkable efforts have been made in the areas of process monitoring and

troubleshooting to manage the failures occurring in industrial plant operations. As a result, theoretical methods along with prototype software systems have been provided in specific aspects of Abnormal Situation Management (ASM).29 However, these methods and systems with the function of individual decision-making have not reach the ultimate aim of ASM, which is simply to recover the process plant from an ongoing abnormal situation to a safe operational situation at a globally minimum cost.36 The presence of flare history and its analysis results can be used to predict process behavior and reduce the operator response time during abnormal conditions.

4

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 48

Conventionally, ASM has been addressed through responsive operational strategies. Singh and coworkers have carried dynamic simulation studies on ethylene facilities. They reported flare reduction can be reached as high as 75% by simply informing control room on better control parameters set points to help the system equilibrate quickly. Thus process control systems have advanced drastically in response to industry’s demand for higher efficiencies, yet, control systems alone are still unable to eliminate these unexpected process disruptions.37, 38 The research presented here aims to focus on the development of a new approach to optimal management of process facilities based on simultaneous design and operational optimization. It requires that we identify key flaring and venting sources, causes and consequences of process upsets that result in flaring and the use of recent process design and optimization tools.17

Knowledge gained from studying ASM systems has made it clear that contribution of operation personnel is definitely critical. Thus, the collaboration between integrated operation individual must be supported in any developed ASM system.36 Moreover, the importance of flare reduction (FM) is a worldwide struggle. It directly contributes to limiting the environmentally harmful gaseous emissions by reducing the emitted amounts of CO2 and other GHGs.39, 40 A systematic approach must be followed in order to mitigate the abnormal situations. Better use of historical incidents data and root cause analysis, would give better chances to avoid future causes of process malfunction and the ability to suggest proper process improvements.

36

In this paper, this

approach will be presented and illustrated using a simplified ethylene process to show the environmental and economic impacts.

1.3 Problem Statement

The management of industrial upsets coming from abnormal situations needs cause and consequence analysis; however, it is not possible to quantify these impacts without proper understanding of the properties and magnitude of these upsets. On the other hand, the analysis of the properties and magnitude of the upset emissions needs extensive amount of historical data.41

Given any industrial process with known operational and design data, historical flare data, and utility requirements, it is desired to develop a systematic approach to collect and analyze process and flaring data for the purpose of enhancing the flare management system during abnormal situations. The collection of these data might require collaboration amongst various departments (teams) within the company in order to track, document and sort data. Later, the analysis of such large amounts of data 5

ACS Paragon Plus Environment

Page 17 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

requires that we leverage the expertise within the key departments in the company, e.g. technical, process, environment, maintenance, etc. The results of the analysis are key to finding suitable solutions for flare mitigation. The developed systematic approach will answer the following questions:

1. What type of data is needed to mitigate flare during abnormal process operations? 2. What is the time frame of historical data required in order to get sufficient information for the analysis? 3. Type of analysis required to determine suitable alternatives for flare minimization. 4. Which department(s)/team(s) within the plant must be involved in the data collection and analysis? The developed system approach will be illustrated by studying an ethylene base case along with its historical process and flaring incidents data used to identify underlying causes and indicators that lead to process upsets, i.e. abnormal situations. However, the beauty of this approach is that it can be applied to any industrial process in which flaring data, basic design and operational data are known. The systems approach will look into the impacts of these flares and the analysis of the causes, consequences and magnitudes. Later, suitable strategies to reduce flaring during abnormal situations are recommended. The recommendations came as a result of integrating the data and its analysis as part of a flexible optimization algorithm.42

2. Flaring Data 2.1 Importance

To continue achieving high levels of productivity growth, industrial companies need to control large datasets to drive efficiency across the extended enterprise and to design and market higher-quality products. Industries already have a significant amount of digital data that requires proper storing; and the amount of data generated will continue to grow exponentially. These data come from various systems including, for example, computer-aided design, computer aided engineering, computer-aided production, design and operational data, flaring and incidences data, standards of operations (SOPs) in the normal and abnormal situations, and across organizational boundaries in end-to-end supply chain data.43, 44

Hence, large sets of data are needed as means to an end; managing huge data is of great interest to production and operation work. Being able to predict future performance based on historical results, or to 6

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 48

identify sub-par production zones, can be used to shift assets to areas that are more productive.45 Moreover, analysis of historical flare data plays an important role as the basis for any proper mitigation measure that can lead to ASM and thus the reduction of flare. Data analysis is also needed to assess the environmental and economic impacts. The industrial performance in terms of managing environmental emissions in sufficient and economical way has direct relations with keeping track of the historical normal and abnormal flare data in order to mitigate the causes. From here, comes the importance of providing a systematic way to deal with flare data in order to utilize optimum mitigation measures for process operations.19-23, 41, 46 In the following sections, the key types of data and the acquisition techniques needed/available from the industrial processes will be highlighted. The section will emphasize the role of key departments or teams within the plant in supplying and analyzing these data.

2.2 Data Types and Records

Industrial plants are grouped into upstream, intermediate and downstream plants based on products. The operations within these plants can be grouped as, normal operations and abnormal operations. Hence, there is a wide range of data sets needed for the operations, maintenance and advancement of these varying types of systems. For industrial facilities, it is highly important to keep record of any single data to ensure safe, continuous and profitable operations. Keeping track of all these data is very difficult in absence of sufficient systematic sorting mechanism.43, 44, 46, 47

The key data that are relevant to setup a systematic FM and ASM approach can be classified into two big categories, process and flaring data. Process data refers to design and operational sets, and flaring refers to normal and abnormal historical incidents data. Design and operational data are normally used to keep the plant functioning for continuous operations. An example of this type of data is operating conditions, which usually includes temperatures, pressures, compositions and flowrates. Flaring and incidents data are normally used to track incidents, its causes and to report emissions. Flared streams, stream composition, duration of flaring incident, frequency, consequences and locations, are examples of this type of data. The continuous flare stream data is collected along with process operational data using the company’s online monitoring systems. The flare data of process upsets during abnormal situations are usually collected on an incident basis, which is not automated. The environment department in the company probably monitors these data and its management depends on internal policies. It should be noted that upset process flare data is documented for reporting purposes. Keeping record of these data is valuable. It allows for the easy tracking of historical incidents and could lead to finding suitable

7

ACS Paragon Plus Environment

Page 19 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

preventive alternatives.41,

46

Table 1 summarizes the important data types for the flare minimization

approach followed in this paper along with its uses and acquisition. 8, 20, 41, 46

Table 1. Data Types, Uses and Acquisition

Type of Data8

Uses20

Design Data

Gain insights about process productivity, and individual unit functionality. Obtain process feed and utility requirements.

Operational Data

Observe daily parameters of the running plant. Track of the normal operations failures and malfunctions which includes flare due to slip in production rates.

Example8

Acquisition41

• Molecular Properties • Conversion Factors

The access to this type needs confidentiality agreement with the internal policy of the company. Might not be available in the open literature. These data are often proprietary.

• Flowrates • Pressure • Temperature

Accessible through the bisystem of any company and usually the plant personnel; technical and operational teams, has great knowledge about them form daily practice. These data are usually stored automatically in the plant database and big portion of it can be found in literature.

8

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Historical Flare Data

Utility Requirements

Study process upsets and trace back root causes of incidents. Tackle process modifications and optimization strategies and enhance operational, environmental and economic performance.

Conduct preliminary/full cost analysis and evaluate the heat integration strategies based on performance criteria.

• • • •

Frequency Duration Flared Amounts Compositions

• Process utility demand • External utility demand

Page 20 of 48

Lack of data in literature made it difficult to conduct any global historical flare incidents studies. These data are usually collected manually by the planet personnel, especially the environmental team, and are stored in datasheets that are not well shared with other technical teams in order to be used for process enhancement and flare mitigation.

The modeling of the plant operation using simulation software can gives approximate information of this type of data. Usually, these data are well documented in the company and can be accessed by any plant personnel using the bi-system. The operational and technical teams are the responsible for this type of data and any modification or retrofitting in the plant must be followed with re-estimation for the utility requirements.

The collaboration between varying departments within the same company is very important to the process of collecting, sorting, and sharing the data. Technical, operational and environmental departments, and research and development teams (R&D) are essential units to ensure continuous operations and improvements.41, 43-46

Key data needed for the analysis of process or abnormal flares can be stored automatically through the plants’ online operational system or documented manually by plant personnel. The online plant operational system usually keeps record of the normal and abnormal operational data, documenting 9

ACS Paragon Plus Environment

Page 21 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

deviation that the process undergoes. The plant technical, environmental, maintenance engineers or operators manually document data and information related to safety inspections for isolated equipment, malfunction of control system, and non-continuous flares due to planned or unplanned trips.41, 46

During a process upset, the designed controlled systems are the first to respond and in some cases, the control operators manually intervene as a secondary response to get the process back to normal operations. In such cases, the online system continuously records deviations in flows and operating conditions. The control operators later supplement the upset incident data manually. They document the corrective actions implemented in response to the process upset.

Such information plays a vital role

when it comes to analyzing the deviation observed in the online process variables and it helps in determining the cause and consequence of such upsets. To complete their analysis, these engineers can access the online operational system to record frequency and duration of the upsets.41, 46

The main challenges in data acquisitions are (1) availability of data in a timely manner; and (2) the integration of manually documented flaring data that the online system cannot document (e.g. corrective actions taken by control operators and cause of process upset) between the various departments (technical, environmental, operational and maintenance).24-26, 41, 43-54

The miscommunication between departments can results in the loss of important analysis parameters and may create discontinuity in the works. From these two points, it comes the importance of defining a collaboration mechanism between the industrial plant teams to address the standard procedure to be followed in the collection of both normal and abnormal flaring data, and in the analysis and utilization of these data.41, 46, 48, 49 The proposed approach in this paper looks at integrating data collection and analysis from various sources with the key objective of developing suitable mitigations approaches to flaring reduction. A case study is also presented to illustrate the impacts of using this integrated data approach.

2.3 Challenges in Data Mining

The challenges faced by industry in addressing flare reduction and the development of mitigation strategies are often reflected through the inability to provide data that are relevant, timely and of the highest quality. The acquisition of data (design, operational, historical flare and utility requirements) is essential in performing suitable and efficient analysis in order to identify flare mitigation techniques.8 Situational assessments that have been conducted highlight the following issues: 24-26, 41, 43-54

10

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Page 22 of 48

The first challenge comes from the fact of low priority given to data and statistics from industries and government and even if the priority is given to data collection, the proper analysis and utilization is missing. In addition, the absence of coherent implementation plans plays a role in stopping any data processing initiatives.



Another challenge is the lack of agreed coordination mechanism within the same company. Key teams have to closely collaborate in order to integrate documented data and analysis results for the purpose of mitigating flares.



Lack of internal comprehensive harmonization is another major challenge that prevents the collaboration between key sections within the industry from accomplishing their goals. This can be found in the lack of standard definitions, methodologies of data mining procedures, methods of analysis and utilization of data and common classification systems.



Lack of effective data utilization is an important issue when it comes to data mining. Choosing the correct and most efficient utilization strategies helps in collecting the suitable data and perform the appropriate analysis. The lack of this can lead to losses in time and efforts in addition to inappropriate conclusions.

According to study done by IDC Energy Insights on the biggest challenges faced by the Oil and Gas companies when it comes to dealing with huge data, it showed that around 75% of the companies in the study agree that there is a problem with huge data that lack a solution. The study suggests that the first step towards improving data quality is by preprocessing of the data to identify variation and the proper analysis and utilization of these data in operation and production.50-52

There are also challenges that face the industrial and academic research communities as well as statistical agencies (such as environmental protection agency (EPA) in the US, and ministries of environment (MoEs) elsewhere), that aim to address the rising global flaring concerns, such as: 24-26, 41, 43-54



Existing gaps between users and producers sometimes drives users to approach local industries to fill data gaps, which in most of the cases cannot be generalized to other local or international processes. These data usually will be used to serve and evaluate specific needs that might not help accomplishing certain global solutions.



One of the major issues when it comes to flare reduction is the lack of flaring incidents data in literature. This makes the international and regional comparison very difficult due to the absence of relevant data sources. Then, the prediction of common future behavior is difficult and the forecasting of possible incidents is impossible. 11

ACS Paragon Plus Environment

Page 23 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

To address some of the challenges related to data, this paper propose an integrated data collection and utilization approach that both industries and research agencies can benefit to mitigate flare causes and reduce consequences.

3. Integrated Data Mining & Utilization Approach (i-Data) for Flare Reduction It has been stated clearly, that one of the challenges facing industry when it comes to mitigating flare during abnormal situations is the ineffective use of readily available large amounts of data sets. Hence, there is a need to create an awareness about the disconnect that exists between data acquisition and its utilization.24-26, 41, 43-54 In this paper, we propose a framework to facilitate communication of acquired data between various departments in the company. This approach came about as a result of collaboration between our research group and an industrial partner. As mentioned earlier, the lack of flaring incidents and flare management practices in the literature is an obstacle.

40

Hence, the industrial support to this

work provided an opportunity to develop this integrated Data approach and it allowed for the generation of potential alternatives to flare management

The i-Data approach developed for ASM is presented in Figure 3. The methodology consists of three major levels: data mining/acquisition, data analysis and data utilization. The foundation for this framework relies on inter department collaboration within the company in order to integrate and share the data, between those who mine it and those that analyze and use it for the generation of alternative mitigation methods.



Level 1: Data Acquisition. This level is critical. The main challenge in data acquisition relates to flare management, is the data related to flaring incidences is currently documented manually by different departments with the company. The key types of data that were mentioned previously, included the design and operational data and the historical incidents data, which are generally collected by the Technical, Operational and Environmental teams (TOE). The i-Data proposes that the environmental team take lead in documentation and reporting of upset flaring data. They are requested to include the following key parameters: frequency of flaring incident during the report period, duration, flared amounts, flared components, and location of the flared streams. The operational and technical teams are responsible to report the data that are related to normal operations such as flowrate, temperature and pressure of plant streams. This data is automatically

12

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 48

stored. Any deviations from design parameters will also be evident. The Research and Development team (R&D) or the task force responsible for flare management strategies is requested to acquire the above mentioned data from the Environment and Technical departments. •

Level 2: Data Analysis. This level deals with data management and analysis that provides an insight into the process dynamics during abnormal process operations. In this level, the Research and Development team acquires and sorts the data from the level 1, in order to prepare it for analysis. Depending on the company’s directive, various level of data analysis might be required. The analysis can be completed within the company based on available expertise, or can be contracted externally. The level of analysis might also be dictated by the regulatory agency, if required. The types of analysis may include: causes and consequences analysis, statistical analysis of the historical incidents, and upset root cause analysis. The data statistics and results of the analyses are prepared the R&D team.



Level 3: Data Utilization. In this level the R&D team studies and utilizes the collected data and derived analysis results to: (1) quantify the possible impacts and (2) suggests possible process and design performance enhancements. Again in this level, it is critical that members from the environment and technical, and maintenance/operations departments are consulted. Potential ways to utilize data include: quantifying environmental and economical impacts. The analysis results might show there is a need to enhance control system performance, derive potential design modifications, optimize process operations, or consider energy integration alternatives. The output from this level provides the decision makers a set of potential mitigation routes with their respective impacts.

The next section illustrates i-Data approach through an ethylene case study. Our research group has utilized the ethylene case study and its data to generate water-energy alternative for the management of normal and abnormal situations.8, 42

13

ACS Paragon Plus Environment

Page 25 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Industrial & Engineering Chemistry Research

Figure 3. Integrated Approach of Data Mining, Analysis and Utilization for Flare Reduction (i-Data)

14

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

4. Ethylene Process Case Study The importance of studying the ethylene plant comes from the fact that; eleven of the 17 petrochemical plant losses and six of the ten plastic and rubber plant incidents involved ethylene, polyethylene and polypropylene process units.55 The ethylene base case study considered has a 900,000 tons per year capacity which lies as an average value of the worldwide ethylene plants capacities.56, 57 While an ethylene plant will be illustrated here, the developed approach can be applied to any industrial process with known flaring data as explained earlier. The statistics shown here for ethylene plant could be collected similarly for other processes and three steps approach developed in this paper can be applied to draw potential mitigations to reduce flaring in these processes. The mitigations may differ from one process to another based on the causes of the flare upsets, quantities, duration and frequency.

4.1 Ethylene Plant

The feedstock of this case study is ethane rich natural gas which is initially fed into the cracking furnace. The effluent from the cracking furnace is sent to the quench tower where it is cooled and partially condensed. The quench tower overhead vapor product is then forwarded to the cracked gas compressors (CGCs), and the bottom stream is sent to waste water treatment plant. After that, the cracked gas enters the CO2 removal section where most of the CO2 content is removed. The cracked gas is then dried and sent to the chilling train to reduce its temperature rapidly before being fed to the deethanizer column (DeC2). The top product from the column which is mainly C2 and lighter components are further compressed, heated and sent to the acetylene reactor to convert all acetylene into ethylene. The bottom stream for the column consisting of C3 and heavier components is directed to the recovery section, which includes depropanizer (DeC3) and debutanizer (DeC4) columns to separate the C3 and C4 components, respectively. The acetylene reactor effluent is then cooled and fed into the demethanizer column (DeC1) to separate methane from C2 components. The top product from the column is treated to separate hydrogen, which will be used as a reactant in the acetylene reactor, from methane. The bottom product is sent to the final column, C2 splitter, where the main product, ethylene, is separated and the rest of components are circulated back to the upstream and combined with the feedstock.17, 39, 40

15

ACS Paragon Plus Environment

Page 26 of 48

Page 27 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

The ethylene plant is a front-end based technology, where the deethanizer unit leads the separation section, and its overhead product is fed to the acetylene hydrogenation unit. 900,000 tons of high grade ethylene per annum is produced. This base case was developed with the following assumptions: 17, 39, 40 •

Conversion and yield of the cracking furnace is 87.6% and 67.1% respectively. The selectivity is 76.60% based on yield and conversion. Feed compositions to furnace are mainly ethane and steam. The gas to steam ratio in the feed is 3.



Conversion in Acetylene reactor is 100%.



In sweetening unit, H2S content is decreases from 18 ppm to 0 ppm.

P-54

Flare F Reducing Valve

25

Ethane Recycle

4 Feed (Sour Gas)

Flare A

Steam

1 7 5 Gas Cooling

Mixing Valve

Cracking Furnace

8.2

8.1

6

2

1st Stage comp.

Quench Tower

cooling

2nd Stage comp.

8.5

8.4

8.3

Cooling

3rd Stage comp.

Cooling

9

Desulfur 3

10

H2S

Flare G

11

CO2

Flare B 13

H2O

CH4

22

Flare C

16

Flare D Flare E

4th Stage comp. 18

14

Ethylene Heater

Dryer

19

Cooler

21

17

Cooler CO2 Removal

23

20

15

12

24

DeEthanizer

Acetylene Hydrogenation

DeMethanizer

Ethylene Splitter

C3+ Sepration

Figure 4. Process flow diagram for the Ethylene plant

Figure 4 presents the developed process flow diagram for the studied ethylene plant.17 The bold-dashed arrows define major flaring sources in the process.

4.2 Flare Identification and Sources (Level 1 in i-Data Framework)

Plant trips and process upsets are one of the major flare activities in any process, in addition to start-ups and shutdowns. Data related to process flaring events and the affected process 16

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

streams/units has to be collected manually. It is not part of the automated data collection system. Hence, meetings with technical, environmental, and operational personnel are needed to complete the data set for any process. The base case plant data is provided for a 10 year period from 2004 – 2014. For the ethylene plant presented, 6 major flaring trips were identified. These along with start up and shutdown events are defined as abnormal process operations. The hypothetical flaring incidents and their frequency over 10 years duration are presented in Figure 5.

12

10 10 8

Frequency

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 48

6

5 4

4

3

3 2

2 0 Off-Spec

PRC Trip

Acytelen Reactor Trip

Electric Problem

CGC Trip

Utility Failure

Figure 5. Frequency of Ethylene plant incidents over 10 years duration

For the base case ethylene process, seven flaring locations are identified, which include: 17, 39, 40



Top of the quench tower (Flare A). Effluent gas from the cracking furnace will be diverted to the plant flaring system after quench tower and before CGC suction during abnormal situations where compressors are no longer able to receive it.



Top of deethanizer column (Flare B). When DeC2 overhead product does not meet the the acetylene reactor feed requirements. The product will be diverted to flare to prevent catalysis deactivation.



C2 reactor effluent (Flare C). Flaring from this location contributes to the majority process upsets occurs at ethylene plants. Since very low contaminant of acetylene is required at the outlet, C2 reactor suffers a lot of upsets. The effluent of the reactor during abnormal situation can produce off-spec product and cause the system several 17

ACS Paragon Plus Environment

Page 29 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

hour of delay in recovery. During these situations, the reactor outlet stream should be sent to flare directly.



Top (Flare D) and Bottom of demthanizer column (Flare G). Similar to the DeC2 flare location, when DeC1 overhead or bottom products do not meet the hydrogen treatment system and C2 splitter feed requirements, respectively.



C2 splitter overhead (Flare F) and bottom products (Flare E). The final products are high-purity ethylene from the C2 splitter overhead and ethane rich from bottom. The customer requirements must be met for the ethylene product, thus, off-spec products with lower purity than required must be sent to flare. Similarly, the ethane rich product quality can affect the performance of cracking furnace. Hence, any deviation must be controlled otherwise abnormal situation can occur and the stream will be diverted to the flaring system.

Table 2 lists the ethylene plant base case flare sources (streams A-G) and their operating specification. For each flaring incident or trip, one or more flaring location (stream) will be sent. The causes and consequences of each of the 6 major flare incidents determined for our base case process are listed in Table 3 along with their active flare locations. It should be noted that causes/consequence listed are meant as a representative sample for the base case; and do not include causes related to human and instrumentation error.

The acquisition of these data is highly important to complete the first level of the i-Data approach. Data acquisition and validation was achieved through various meetings with members from environment, technical and maintenance and operations departments; as they are responsible to compile and maintain such data. These data sets will be analyzed in level 2 of the framework.

42

Table 2. Flare stream description. Label

Description

Temperature [K]

Pressure [Psia]

A

Top of quench tower

318.00

23.00

B

De-ethanizer overhead

223.70

334.70

C

Acetylene hydrogenation outlet

355.70

464.00

D

De-methanizer bottom product

265.30

461.40

E

Off-spec Ethylene

244.60

270.00

F

Ethane Recycle

244.60

270.00

18

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

G

De-methanizer overhead

191.80

Page 30 of 48

460.00

Table 3. Flare Incidents causes and consequences

Active Flare Locations on PFD

Flare Incident

Causes*

Consequences

Off-Spec Product

* Spike in CO concentration * Catalyst Deactivation in Acetylene reactor

* Isolation of C2 reactor * Isolation of C2 splitter * Reactor feed is sent to flare

C, D & E

(Propylene Refrigeration Cycle) PRC Trip

* Loss of steam supply * Seawater failure * Power failure

* CGC trip * Furnaces are depressurized * Reactors are bypassed

All

Acetylene Reactor Trip

* Spike in CO concentration * Upset in the furnace * Runaway reaction

* Off-Spec product

Electric Problem

* Cut in power supply * Motor turbines failure

* All furnace gradual shutdown * Complete shutdown

All

CGC Trip

* Loss of steam supply * PRC problem

* Complete Plant Shutdown * Compressor Damage

All

Utility Failure

* Fuel failure * Ethane feed failure * Cooling water failure * Sea water failure * Electric power failure * Steam failure

Complete Shutdown

All

C&D

4.3 Design and Operational Data (Level 1 in i-Data Framework)

In addition to the flaring data, the design and operational data plays an important role in flare mitigation. They are essential to develop process models that reflect our process and they are necessary later when it comes to quantify the environmental and technical performance of the process. The data includes the process stream mass/molar flowrates, compositions and their operating temperature, and pressure. Table 4 shows a sample spreadsheet that summaries the 19

ACS Paragon Plus Environment

Page 31 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

necessary design/operational data for our base case process. The acquisition of the design and operational data is easier due to their availability through the automated data management system. Figure 6 shows the main steps that industry can follow to develop the necessary data sets.

20

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Page 32 of 48

Table 4. Base case streams table Stream No.

1

2

3

4

5

6

12

18

20

23

24

25

Description

Sour Gas Feed

Sweet Gas

Removed H2S

Furnace Steam

Furnace Feed

Furnace Effluent

Dryer Feed

Acetylene Reactor Feed

Acetylene Reactor Effluent

Ethylene Splitter Feed

Ethylene Product

Recycled Ethane

Component mass flow H2

80.9

80.9

80.9

79.5

137.1

137.1

137.1

137.1

0.0

C2H2

17.5

17.5

17.5

C2H4

1068.7

1068.7

1068.7

1087.5

900.0

196.1

196.1

194.8

194.8

191.6

C3H6

17.5

17.5

C4H4

4.8

4.8

23.9

23.9

8.0

8.0

C6H6

19.1

19.1

C8H8

3.2

C10H8

0.0

CH4

0.0

C2H6

1389.9

1389.9

1581.5

C4H6 C5H6

103 Tons/yr

H2S

0.5

CO2

0.1

0.1

H2O

2.9

2.9

1393.4

1392.9

K

298.1

Pressure

Psia

Enthalpy Flow

MMBTU/hr

Total Mass Flow Temperature

0.5

1.6

0.0

900.0 191.6

1.6

0.1

0.1

524.2

2.9

527.2

12.6

0.5

524.2

1584.5

2108.8

1587.7

1499.0

1499.0

1091.6

900.0

191.6

298.1

298.1

412.7

263.8

1200.0

318.0

249.5

355.7

265.3

244.6

244.6

23.0

23.0

23.0

50.0

23.0

23.0

514.0

464.0

464.0

461.4

270.0

270.0

-384.5

-384.4

0.0

-679.7

-446.4

8.9

93.0

73.3

101.0

71.3

154.0

-62.0

21

ACS Paragon Plus Environment

Page 33 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 6. i-Data Flowchart – Level 1

4.4 Magnitude Flaring Incident Analysis (Level 2 in i-Data Framework)

The main findings of this case study are best illustrated in terms of flaring magnitude and the processes energy losses and emissions. The data developed for the base case was for the total flared amount per month for each of the 6 flaring incidents. This information will be utilized to quantify the magnitude of each incident.

After collecting the basic design and operational data and the flaring data for any industrial process, certain calculations and analysis can be applied to understand the magnitude and frequency of plant upsets and to quantify the environmental and energy impacts. The average annual flaring rate due to abnormal situations for the base case ethylene plant is 1.8x107 kg per year of mainly hydrocarbon components. Figure 7 shows the annual flared amount for each individual incident in the 10-year time span and Figure 8 shows the flared amount for each upset annually and how it can be compared with the average annual flaring rates. Figure 9 represents the main steps to be followed for data analysis (level 2 of i-Data).

22

ACS Paragon Plus Environment

Page 34 of 48

6

5 Off-Spec

Flared Amount (kg)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

x 100000

Industrial & Engineering Chemistry Research

4

PRC Trip

3

Acetylene Reactor Trip Electric Problem

2 CGC Trip 1 Utility Failure 0 2004

2005

2006

2008

2009

2010

2012

2013

Time Figure 7. Total flared amount per month for each of the ethylene plant incidents

23

ACS Paragon Plus Environment

2014

Page 35 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 8. Total flared amount per year for each of the ethylene plant incidents relative to the mean value

24

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 36 of 48

4.5 Data Utilization (Level 3 in i-Data Framework) - Quantification of Economic and Environmental Impact

One way of quantifying the process economic losses from abnormal situations is by the estimation of the energy losses. In turn, energy losses are estimated using the Wobbe number or Wobbe index, which is a proper representation of the energy content of combusted hydrocarbons. The following generic equation can be used to estimate the Wobbe index of certain hydrocarbon gas on a stream gas line:16 W T  =

 





(1)

Where H T  is the volumetric calorific value at T and d is the relative density under

metering reference conditions. WI is used to estimate the combustion energy output of fuel gases as a function of their composition. High WI values indicate the presence of heavy hydrocarbons, and low WI values are indicative of non-combustible fuel components. To avoid confusion with the volumetric heating value of the fuel gas, WI can be reported as a dimensionless quantity. It can have units such as BTU per SCF or MJ per m3.

The CO2 emissions are used to assess the environmental impacts of the base case process. The estimation of carbon emissions is determined using flared stream amount, specifications and flare duration data fed into the in-house available GHG calculator. The flaring incident, associated process stream compositions, along with flare duration are used to approximate the environmental losses during abnormal situations. It should be noted, the values reported here are based on the assumptions of 12 hour per year average flaring rate for each incident. The equations and steps used to estimate these values are outlined by Mohammed and coworkers.16 Table 5 shows a summary of the main emissions and energy losses for the ethylene plant case study.

Table 5. Results of CO2 emissions and WI for ethylene case using average 12 hours per year flare rate for each incident

Incidents

Parameter

Unit

Off-Spec

PRC Trip

Acetylene Reactor Trip

Electric Problem

25

ACS Paragon Plus Environment

CGC

Utility

Page 37 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

CO2

Tons/yr

WI (Energy loss)

2.90E+04

5.77E+04

2.65E+04

7.50E+05

2.02E+05

3.46E+05

8.66E+02

1.73E+03

7.93E+02

2.24E+04

6.04E+03

1.04E+04

The economic impacts of flaring are assessed using CO2 tax and credit estimations as published by Mohammed and co-wrokers.16 Governmental and Regulatory environmental agencies in some countries like Australia, British Columbia and in the Europe United (EU), have already began implementing some taxes on industrial CO2 emissions.58-63 Table 6 shows CO2 tax prices in some focal countries. Using these as a basis and the base case flared amounts, compositions and duration, the CO2 taxes were estimated for all the flare incidents, see Table 7. Kazi et. al in 2014 presented the key parameters used in quantifying the economic impacts of this base case.8

Table 6. CO2 tax prices in select countries

Country / State Australia British Columbia Europe

Year

Price per ton CO2-e ($)

2014 / 2015

23.88 58, 59

2012

27.60 60, 61

18 May 2014

6.64 62, 63

Table 7. Estimated CO2 tax based on flaring incidents for ethylene base case process

Incidents Parameter

Unit Off-Spec

PRC Trip

Acetylene Reactor Trip

Electric Problem

CGC

Utility

Australia

$MM/yr

0.69

1.38

0.63

17.90

4.82

8.27

British Columbia

$MM/yr

0.80

1.59

0.73

20.70

5.57

9.56

Europe

$MM/yr

0.19

0.38

0.18

4.98

1.34

2.30

The analysis results at this stage (Level 3 of i-Data) can be summarized as follows: 1. Magnitude of flaring incident were presented and compared with the relative mean flared amount. 26

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 48

2. CO2 emissions from the 6 flare incidents of the Ethylene base case were calculated and used to quantify the environmental losses of the case study. 3. Wobbe Indices associated with the 6 flare incidents were calculated and used to address the energy losses for the case study. 4. CO2 tax prices for various countries along with the estimated CO2 emission were used to quantify the economic losses for our case study.

Figure 9. i-Data Flowchart – Level 2

4.6 Data Utilization (Level 3 in i-Data Framework) – Derive Flare Mitigation Technique

To illustrate the use of data and analysis in this base case, we have developed and published an extendible multi-objective optimization framework to further analyze potential energy integration alternatives (i.e., COGEN, TMD) as a mean to mitigate flaring during abnormal operations.8, 16, 42 The optimization framework integrated the analysis data to assesses the impacts of different energy alternatives for flare mitigation.

The proposed optimization framework is based on a combination of stochastic method (genetic algorithm) and linear programming (LP). The core of this optimization approach is formulated as a multi-objective optimization problem to determine the produces amount of required heat and power, while fulfilling process constraints. The objective function was expressed in terms of total annualized cost which was calculated by the following equation:

Total annual cost TAC = Annual operating cost Annual fixed cost # Annual income if any # Carbon tax savings

(2)

The total annual cost objective function described by the formulation in equation (2) can be expressed as following: 27

ACS Paragon Plus Environment

Page 39 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Objective function: TAC = 344444454444446 1.3 × F- × C./01 × H2 34444444444444544444444444446 1411 431 # ξ 16131 γ × [email protected] 344444444444444444444454444444444444444444446 DE 789:; FGG/B1 8-0HBIJGK 7LMI

k k @ × C /HJG0 58.5 × AR 1115 × W DE 3444444445444444446 3444444445444444446 @ × COLJ10H 344444444444444444454444444444444444446 789:; DE FGG/B1 .JS0 7LMI W ^

(3)

#P H62 # V ×4454 WV0HR0BI0 34 4446 # YZ Z e\ C4 304× 54 [- # e[J]- fJ]- a × CIBS 344444444454444444446 DE 34444444544444446 J_` ]_` 789:; 7BHLG BS bBJGKM

FGG/B1 WGXLR0

Where, the annual operating cost includes total raw materials cost and total utility cost. The annual fixed cost reflects the capital cost of the proposed mitigation tools. The annual income considers the revenue from co-products, heat utility savings, power generation, wastewater treatment savings, and income from permeate. Lastly, the environmental cost in terms of CO2 tax savings was considered for total emissions. The details of the objective function are available in our previous publications by Kazi and coworkers.8, 42

To accomplish such multi-objective optimization, the above-mentioned necessary information is obtained from i-DATA. The data and analysis results provided by i-Data are integrated with other raw material and utility costs data (e.g. feed cost, fuel price, electricity cost and other utility cost) to calculate annual operating cost. Historical flare data and utility requirement data collected by the i-DATA such as Wobbe index of each flare streams, flow rate limit, temperature and pressure are used to generate implicit process constraints.

Thus, the proposed optimization framework uses the historical flare, design, operational and utility requirement data obtaining from i-DATA as the basis of the framework. The data provided the necessary information to the optimization formulation and it was also used as input in to the GHG calculator to work simultaneously with proposed optimization methodology. The optimization framework and the scope of i-DATA utilization along with GHG calculator are mapped out in Figure 10.42

The final Pareto fronts resulting from the optimization framework help the end user in making informed operating point decisions depending on economic, energetic, and/or environmental trade-offs. The optimization framework came about as a holistic approach to assessing alternative FM strategies and it is an excellent tool for industry, limited only by the availability and management of process and flaring data. 28

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 10. Data utilization from i-DATA for systematic integration of flare and water management tools.42

Level 3 of i-Data framework can be reassembled and reproduced for any type of chemical/petrochemical processes. The methodology followed is typically the same with little difference in terms of type of potential mitigation alternatives to be compared and assessed. Figure 11 represents the main steps to be followed in the execution of the three levels of the iData Framework.

29

ACS Paragon Plus Environment

Page 40 of 48

Page 41 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Figure 11. i-Data Flowchart

30

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

5. Replication of i-Data A detailed step by step procedure to reproduce the work to other processes and case studies is presented in Figures 3 and 11. The steps included in the flowchart and the types of data required can be applied to other refineries, chemical and petrochemical processes.64, 65 The first level of iData; “Data Mining”, proposes that a task group to be created who take the responsibility of acquiring the key parameters for developing potential flare management strategies. The R&D team can take the lead in this task group with the help of TOE teams. The key data parameters; operational, design and historical flare data, needed for this level has been highlighted previously in section 2.2. In order to properly prepare the data for the next levels, a historical operation time period needs to be defined where all process deviations occurred are recorded and used to determine flaring incidents type. The type is usually classified based on the source and causes of the incident and key components and properties of the flared streams. Tables 2, 3 and 4 shows a sample of the classification of the flaring incidents. The second level of i-Data; “Data Analysis”, highlights some important statistical and qualitative analysis that can process the raw data into more sensible data in terms of environmental and economic impacts. First, the flare events that have been classified in level 1 are to be mapped to the active flare locations in the process flow diagram. This will help in tracking the flare events and identify the proper mitigations for the associated streams. Next, statistical analysis can be used in quantifying the magnitude and frequency of each flare incident by looking back in the historical time period and identifying the flared amount, duration and occurrences. The results of the analysis are used in the following level of i-Data, Level 3 “Data Utilization”. In this level, the environmental and economic impacts are assessed using most recent worldwide-developed indicators. These indicators could be the Wobbe index, CO2 – equivalent emissions, Nitrogen Oxides (NOx) emissions, Carbon tax estimation, and other indices that can be used to quantify the potential process losses and impacts. The user selects the indicators that fit their objectives. Level 3 guides the user about the utilization of the developed data and their analysis results in order to derive potential flare mitigation techniques. Number of techniques can be suggested, assessed and compared to come up with the optimum solution for flare management. For the presented case study in this paper a multi-objective optimization formulation, similar to the one in Figure 10, was developed to evaluate the performance of the assessed techniques. The details of the method have been presented by Kazi and co-workers.8,

42

The R&D expertise within the

31

ACS Paragon Plus Environment

Page 42 of 48

Page 43 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

company can develop other techniques. The i-Data framework can be replicated to any process with potential historical flare incidents that can be gathered and quantified.

6. Conclusion Abnormal Situation Management for the purpose of flare minimization is a difficult subject, but the potential for improved plant safety and profitability is also significant. ASM strategies and practices improve operator performance, and reduce incidents during abnormal situations. Availability of process and flaring data is necessary to the development of suitable FM strategies. Due to the uncertain nature and the lack of automated data gathering systems for abnormal flaring incidents, there is a need to develop methodical approach to data management. The paper presents i-DATA as a framework that guides industry on data mining, analysis and utilizations. A case study for an ethylene plant was used to demonstrate how its database was developed; and through the use of i-Data suitable FM strategies were generated.

The environmental and

economic impacts were assessed using an optimization algorithm that integrated i-DATA into its methodology. The results showed potential savings of 1.41 MM tons of CO2 emissions per year and 39 MM$ in CO2 tax credit (using 2012 British Columbia Carbon Tax rate), respectively.

32

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Author Information Corresponding Author: * Email: [email protected] Notes The authors declare no competing financial interest. Acknowledgment This paper was made possible by NPRP grant No 5-351-2-136 from the Qatar National Research Fund (a member of Qatar Foundation). The statements made herein are solely the responsibility of the authors. References 1. Bott, R., Flaring: questions+answers. Canadian Centre for Energy Information: 2007. 2. The World Bank Group. Oil, G., Mining and Chemicals Department Global Gas Flaring Reduction Initiative - Report on consultations with stakeholders; 0821356992; 2002. 3. Petrogas Systems, Gas Flare Reduction Technology. http://petrogassystems.com/technology/natural-gas-processing-and-dew-point-control/antiflaring/ (2017), 4. Haley, J., Pollution. Greenhaven Press: 2003; Vol. 27500. 5. Qatar International Energy Data and Analysis; U.S. Energy Information Administration: 2015. 6. International Finance Corporation IFC Annual Report; 0074-6061; International Finance Corporation: 1960. 7. Hobré Instruments B.V., General Information Wobbe Index and Calorimeters. http://www.hobre.com/files/products/Wobbe_Index_General_Information_rev.1.pdf (2016), 8. Kazi, M.-K.; Mohammed, F.; AlNouss, A. M. N.; Eljack, F., Multi-objective optimization methodology to size cogeneration systems for managing flares from uncertain sources during abnormal process operations. Computers & Chemical Engineering 2015, 76, 76-86. 9. Anomohanran, O., Estimating the Greenhouse Gas Emission from Petroleum Product Combustion in Nigeria. Journal of applied sciences 2011, 11, (17), 3209-3214. 10. Anomohanran, O., Determination of greenhouse gas emission resulting from gas flaring activities in Nigeria. Energy Policy 2012, 45, 666-670. 11. Jegannathan, K. R.; Chan, E.-S.; Ravindra, P., Biotechnology in biofuels-A cleaner technology. Journal of Applied Sciences 2011, 11, (13), 2421-2425. 12. Sharma, V.; Marano, D.; Anyanwu, C.; Okonkwo, G.; Ibeto, C.; Eze, I., Solar cooling: A potential option for energy saving and abatement of greenhouse gas emissions in Africa. Singapore J. Sci. Res 2011, 1, 1-12. 13. Davoudi, M.; Rahimpour, M. R.; Jokar, S. M.; Nikbakht, F.; Abbasfard, H., The major sources of gas flaring and air contamination in the natural gas processing plants: A case study. Journal of Natural Gas Science and Engineering 2013, 13, 7-19.

33

ACS Paragon Plus Environment

Page 44 of 48

Page 45 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

14. Homssi, A. M., Flaring Reduction and Emissions Minimization at Q-Chem. In Proceedings of the 3rd Gas Processing Symposium, Elsevier: Oxford, 2012; Vol. 3, pp 58-63. 15. AlGhanim, N.; Khraisheh, M.; Benyahia, F., Flare Reduction Options and Simulation for the Qatari Oil and Gas Industry. In Proceedings of the 3rd Gas Processing Symposium, Elsevier: Oxford, 2012; Vol. 3, pp 7-14. 16. Mohammed, F. M.; Kazi, M.-K.; AlNouss, A. M.; Eljack, F. T., Tracking of GHG Emissions and Tax Implication During Normal/Abnormal Situations – Ethylene Process Base Case Industrial Application. In Proceedings of the 4th International Gas Processing Symposium, Elsevier: Oxford, 2015; pp 251-260. 17. Kamrava, S.; Gabriel, K. J.; El-Halwagi, M. M.; Eljack, F. T., Managing abnormal operation through process integration and cogeneration systems. Clean Technologies and Environmental Policy 2014, 17, (1), 119-128. 18. He, Y.; Wang, L.; Wang, J., Cap-and-trade vs. carbon taxes: A quantitative comparison from a generation expansion planning perspective. Computers & Industrial Engineering 2012, 63, (3), 708-716. 19. Flaring and Venting Reduction Guideline; BC Oil Gas Commission: 2011. 20. Parameters for Properly Designed and Operated Flares; U.S. EPA Office of Air Quality Planning and Standards (OAQPS): 2012. 21. McEwen, J. D.; Johnson, M. R., Black carbon particulate matter emission factors for buoyancy-driven associated gas flares. Journal of the Air & Waste Management Association 2012, 62, (3), 307-321. 22. Oil Industry International ExplorationProduction Forum, Environmental Management in Oil and Gas Exploration and Production: An Overview of Issues and Management Approaches. UNEP/Earthprint: 1997; Vol. 2. 23. Guidelines on Flare and Vent Measurement; The Global Gas Flaring Reduction partnership (GGFR) and the World Bank: 2008. 24. Issues and Challenges in Data Collection - Suggestions for Measuring the Creative Industries - A Regional Perspective. WIPO - CARICOM Meeting of Experts on the Creative Industries and Intellectual Property: 2006. 25. Venkatasubramanian, V., Abnormal events management in complex process plants: challenges and opportunities in intelligent supervisory control. Proceedings FOCAPO: 2003; p 117ff. 26. McCoy, B. J.; Fischbeck, P. S.; Gerard, D., How big is big? How often is often? Characterizing Texas petroleum refining upset air emissions. Atmospheric Environment 2010, 44, (34), 4230-4239. 27. Welukar, R.; Harichandan, D.; Kulkarni, M.; Bandgar, P. K.; Arunachandan, P.; Kannan, V. S. Process Costing; University of Mumbai: 2011. 28. Bullemer, P. T.; Dal Vernon, C. R., Managing Human Reliability: An Abnormal Situation Management Historical Perspective. 2015. 29. Bullemer, P. T.; Cochran, T.; Harp, S.; Miller, C., Managing abnormal situations II: Collaborative decision support for operations personnel. ASM Consortium 1997. 30. Cochran, E. L.; Miller, C.; Bullemer, P. In Abnormal Situation Management in petrochemical plants: can a Pilot's Associate crack crude?, Aerospace and Electronics Conference, 1996; Proceedings of the IEEE 1996 National: 1996; pp 806-813 vol.2. 31. Wilkins, D. E.; Lee, T. J.; Berry, P., Interactive execution monitoring of agent teams. J. Artif. Intell. Res. (JAIR) 2003, 18, 217-261. 32. Yang, A.; Lu, M. L., A formulation of the collaboration mechanism for integrated abnormal situation management. Computers & Chemical Engineering 2000, 24, (2–7), 539-544. 34

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

33. Singh, A.; Li, K.; Lou, H. H.; Hopper, J. R.; Golwala, H.; Ghumare, S.; Kelly, T., Flare minimisation via dynamic simulation. International journal of environment and pollution 2007, 29, (1-3), 19-29. 34. Yang, X.; Xu, Q.; Li, K., Flare Minimization Strategy for Ethylene Plants. Chemical Engineering & Technology 2010, 33, (7), 1059-1065. 35. Bullemer, P.; Nimmo, I. In A Training Perspective on Abnormal Situation Management: Establishing an Enhanced Learning Environment, Proceedings of the 1996 AICHE Conference on Process Plant Safety, Houston, TX, 1996; Citeseer: 1996. 36. Yang, X.; Xu, Q.; Li, K., Safety-Considered Proactive Flare Minimization Strategy under Ethylene Plant Upsets. Chemical Engineering & Technology 2011, 34, (6), 893-904. 37. Buddaraju, D. Performance of control room operators in alarm management. Jawaharlal Nehru Technological University, 2011. 38. Eljack, F. T.; El-Halwagi, M. M.; Xu, Q., An Integrated Approach to the Simultaneous Design and Operation of Industrial Facilities for Abnormal Situation Management. In Computer Aided Chemical Engineering, Mario R. Eden, J. D. S.; Gavin, P. T., Eds. Elsevier: 2014; Vol. Volume 34, pp 771-776. 39. Noda, M.; Nishitani, H., Plant shift ability evaluation for abnormal situation management. Asia-Pacific Journal of Chemical Engineering 2011, 6, (3), 503-508. 40. Dinh, H.; Zhang, S.; Xu, Y.; Xu, Q.; Eljack, F.; El-Halwagi, M., Generic Approach of Using Dynamic Simulation for Industrial Emission Reduction under Abnormal Operations: Scenario Study of an Ethylene Plant Start-up. Industrial & Engineering Chemistry Research 2014, 53, (39), 15089-15100. 41. Management of continual improvement for facilities and activities: A structured approach. Nuclear Power Engineering Section, International Atomic Energy Agency: 2006. 42. Kazi, M.-K.; Eljack, F.; Elsayed, N. A.; El-Halwagi, M. M., Integration of Energy and Wastewater Treatment Alternatives with Process Facilities To Manage Industrial Flares during Normal and Abnormal Operations: Multiobjective Extendible Optimization Framework. Industrial & Engineering Chemistry Research 2016, 55, (7), 2020-2034. 43. Nedelcu, B., About Big Data and its challenges and benefits in manufacturing. Database Systems Journal 2013, 4, (3), 10-19. 44. Kaisler, S.; Armour, F.; Espinosa, J. A.; Money, W. In Big data: Issues and challenges moving forward, System Sciences (HICSS), 2013 46th Hawaii International Conference on, 2013; IEEE: 2013; pp 995-1004. 45. Baaziz, A.; Quoniam, L., How to use Big Data technologies to optimize operations in Upstream Petroleum Industry. International Journal of Innovation-IJI 2013, 1, (1), 19-25. 46. An Integrated Approach to Safety: Defense in Depth; Automation & Control Solutions, Process Solutions, Honeywell: 2010. 47. Abnormal Situation Solutions; Emerson Process Management Power & Water Solutions, Inc., 2016. 48. Davis, J.; Edgar, T.; Porter, J.; Bernaden, J.; Sarli, M., Smart manufacturing, manufacturing intelligence and demand-dynamic performance. Computers & Chemical Engineering 2012, 47, 145-156. 49. Smart Process Manufacturing: An Operations and Technology Roadmap; Steering Committee, Engineering Virtual Organization, Smart Process Manufacturing: 2009. 50. Rawat, N.; UPES, D., Big Data Analytics in Oil & Gas Industry. International Journal of Scientific & Engineering Research 2014, 5, (5). 51. Nicholson, R. Big Data in the Oil & Gas Industry; IDC Energy Insights: 2015.

35

ACS Paragon Plus Environment

Page 46 of 48

Page 47 of 48

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

52. Feblowitz, J., The big deal about big data in upstream oil and gas. Paper & presentation, IDC Energy Insights 2012. 53. A Study on Energy Efficiency Index in Petrochemical Industry; The Department of Alternative Energy Development and Efficiency (DEDE), Ministry of Energy, Thailand: 2006. 54. Manyika, J.; Chui, M.; Brown, B.; Bughin, J.; Dobbs, R.; Roxburgh, C.; Byers, A. H., Big data: The next frontier for innovation, competition, and productivity. 2011. 55. Olivo, J., Loss prevention in a modern ethylene plant. Journal of Loss Prevention in the Process Industries 1994, 7, (5), 403-412. 56. Meyers, R. A. e., Handbook of petrochemicals production processes. McGraw-Hill Education LLC: New York, N.Y., 2005. 57. Petrochemical Processes Handbook. Hydrocarbon processing 2010. 58. Jotzo, F., Australia's carbon price. Nature Clim. Change 2012, 2, (7), 475-476. 59. Applying an Equivalent Carbon Price to Synthetic Greenhouse Gases. In Department of Sustainability, E., Water, Population and Communities, Ed. 2012. 60. June Update - Budget and Fiscal Plan - 2013/14 – 2015/16. In National Library of Canada Cataloguing in Publication Data: 2013. 61. Duff, D. G., Carbon Taxation in British Columbia. Vt. J. Envt. L. 2008, 10, 87. 62. Carbon Emissions Analysis. http://in.investing.com/commodities/carbon-emissionsopinion/ 63. Agostini, P.; Botteon, M.; Carraro, C., A carbon tax to reduce CO2 emissions in Europe. Energy Economics 1992, 14, (4), 279-290. 64. Emam, E. A., Gas Flaring in Industry: An Overview. Petroleum & Coal 2015, 57, (5), 532555. 65. Ismail, O. S.; Umukoro, E. G., Global impact of gas flaring. Energy and Power Engineering 2012, 4, 290-302.

36

ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research

FOR TABLE OF CONTENTS ONLY (i-DATA) INTEGRATED APPROACH: DATA MINING, ANALYSIS AND UTILIZATION FOR FLARE REDUCTION Select a certain historical period (t) of process opera on

*The bigger your historical data period, the be er you understand how your process behave and the be er you es mate the impacts

*Data is usually available in the automa c system

Level 1

Process Flow Sheet (PFD) Stream Name or #, Flowrates, Composi on (xi), T, P

Historical Opera onal Data

Level 2

Data Analysis

Data Mining

Design & Opera onal Data

Gather Process and Abnormal Opera ons Data

Iden fy me period (dura on) where devia ons in process opera ons occurred. *indicate flaring incident

Cause & Consequence Analysis

Map each flare event to ac ve flare loca on PFD

Sta s cal Analysis

Magnitude of each flare incident/type

Determine flaring incident type

*Data is usually determined by several mee ngs with the Environment and Opera on Teams

*This can be done by simula ng the system dynamically or mee ng with the Environment and Opera ons teams

Root cause of incident

Frequency of each flare incident/type

Data Input

Level 3

Quan fy the Environmental and Economic Impacts

Data U liza on

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 48 of 48

Derive Flare Minimiza on/ U liza on Technique(s)

Assess Impacts of Proposed Technique(s) on Flare Reduc on

Wobbe Index Formula on

GHG Calculator

Energy Losses

CO2/CO2-e Emissions

Carbon Tax

Economic & Environmental Impacts

*R&D Exper se Required

Energy & Integra on Dynamic Simula on

Waste Minimiza on

Enhance Control Opera ons

Op miza on Tools

37

ACS Paragon Plus Environment

….