S2S. Sales - Nomen

Process Safety and Environmental Protection
Fo
The S2S training & assessment tool for unexpected reactors: structure
and knowledge base
rP
Journal: Process Safety and Environmental Protection
Manuscript ID: PSEP-07-028
Manuscript Type: Full Paper

ee
Date Submitted by the

26-Feb-2007
Author:
rR
Complete List of Authors: Weve, Diederik; Shell Global Solutions International B.V., Chemical
Process Development
Sales, Jaime; Institut Quimic de Sarria, Universitat Ramon Llull,
Chemical Engineering Department
Nomen, Rosa; Institut Quimic de Sarria - Universitat Ramon Llull,
ev
Chemical Engineering; Universitat Ramon Llull
unexpected reactors, chemical hazard, assessment tool, reaction

Keywords:
triangle
ie
w
On
ly
IChemE Journals
Page 1 of 25 Process Safety and Environmental Protection
1
2
3
4
5
6 The S2S training & assessment tool for unexpected reactors: structure and
7
8 knowledge base.
9
10
11
12
13 Diederik Weve1, Jaime Sales2 and Rosa Nomen2
14
15
16
Fo
17 1
18 Shell Global Solutions International BV
19
20 Postbus 541, 2501 CM Den Haag, Nederland
rP
21
22 Tel: +31 70 377 5343 Fax: +31 70 377 5400
23
24
E-mail: Diederik.weve@shell.com
ee
25
26
27 URL: www.shell.com/globalsolutions
28
rR
29
30
31 2
32 Institut Quimic de Sarria, Universtitat Ramon Llull
33
ev
34 Via Augusta 390, 08017 Barcelona, Spain

35
36
iew
37
38
39 Abstract
40
41
42
43
On
44 The S2S assessment tool for unexpected reactors aims to identify how reactive
45
46 chemical hazards can exist in process units or plant equipment outside the reactors.
47
ly
48 The assessment tool can also be used as a training tool since it is based on simple
49
50
51 concepts, combining the following items:
52
53
54
55
56 • A reaction triangle that describes the factors necessary for triggering a
57
58 chemical reaction. The factors considered are temperature, concentration and
59
60 residence time.
1
IChemE Journals
Process Safety and Environmental Protection Page 2 of 25
1
2
3
4
• Process streams with physical and chemical properties that can be mapped on
5
6 the reaction triangle.
7
8 • Equipment or unit operations, which determine how process streams are
9
10
11 positioned and moved in the reaction triangle. The design of the plant should
12
13 ensure that these moves are safe. Equipment will have failure modes and can
14
15 exhibit side effects that cause a shift in the reaction triangle that enables a
16
Fo
17
18 reaction and may pose a hazard.
19
20
rP
21
22
23
The tool is presented as a computer program that generates discrete questions to
24
identify potential hazards and explains the likely mechanisms. It is applicable both to
ee
25
26
27 batch, semi-batch and continuous processes, and it assumes that the user has
28
rR
29
30 previously identified/characterized the reaction hazards through literature studies,
31
32 experiments or other S2S assessment methods.
33
ev
34
35
36
The tool can be found at the S2S web portal http://www.s-2-s.org/
iew
37
38
39
40
41
Keywords: unexpected reactors, chemical hazard, assessment tool, reaction triangle
42
43
On
44
45
46 Introduction
47
ly
48
49
50
51 Recent studies based on accident analyses (Sales et al., 2007) show that nearly 70% of
52
53 the reactive chemical incidents take place in equipment or units other than the
54
55
56 chemical reactor itself. When analysing such incidents in terms of causal effects
57
58 (Murphy et al., 2004) reaches the conclusion that 60% of the reactive incidents are
59
60 caused by inadequate recognition of reactive hazards. In the past many studies and
2
IChemE Journals
1
2
3
methodologies have attempted to highlight the hazards at the chemical reactor
4
5
6 (Nomen et al., 2004a, Nomen et al., 2004b), but many times the possible effects of
7
8 chemical runaways at other process equipment are left unnoticed. Methodologies exist
9
10
11
to determine reactive properties ranging from generally applicable to specialised for
12
13 process upsets (CCPS, 1991, Legget 2006). We therefore believe that the inadequate
14
15 recognition of reactive hazards outside reactors lies in the fact that the equipment
16
Fo
17
18 types and process steps are diverse. The diversity makes it difficult to detect patterns
19
20 and to transfer the knowledge to operators and engineers. Thus we set the task to
rP
21
22 develop a teaching concept based on commonality.
23
24
ee
25
26
27 The Safety2Safety (S2S) project is a European Commission sponsored network that
28
rR
29
aims at improving the safety of SMEs (Small and Medium sized Enterprises) that
30
31
32 perform chemical processes. The project deliverables are knowledge bases, training
33
ev
34 modules and assessment tools on areas like toxicity, fires & explosions and reaction
35
36
hazards. The project deliverables are freely available through the S2S web-portal
iew
37
38
39 http://www.s-2-s.org/. The project was active during the period 2002 - 2006.
40
41
42
43
On
44 This paper describes the actual knowledge that is contained in the S2S training and
45
46 assessment tool for unexpected reactors, including:
47
ly
48
49
50
51 • A training concept that helps recognizing how deviations from design
52
53 intent will result in reaction hazards.
54
55
56 • A method to generate general mechanisms for such deviations and channel
57
58 user input in user specific, decisive questions.
59
60
3
IChemE Journals
1
2
3
4
• A database containing the above knowledge in tables that was build from a
5
6 limited number of incidents
7
8 • A spreadsheet containing the logic programmed as Microsoft Excel
9
10
11 macros to use the database.
12
13
14
15
Training Concept
16
Fo
17
18
19
20 For a reaction1 to occur at any part of a chemical plant, a combination of 3 factors
rP
21
22
23
needs to be fulfilled:
24
ee
25
26
27 • Concentration of reactants.
28
rR
29
30 • Temperature of reactants.
31
32 • Time for reaction.
33
ev
34
35
36
This can be described in a triangle similar to the fire triangle as seen in Figure 1.
iew
37
38
39
40
41
42 <Figure 1>
43
On
44
45
46 The reason why these three factors are chosen is that they also represent the most
47
ly
48
49 important degrees of freedom in process operation and design. As with the fire
50
51 triangle, a reaction requires an appropriate combination of these three factors:
52
53
54
reducing the contribution of concentration, temperature or residence time in the
55
56 triangle will reduce the possibility of a reaction to take place.
57
58
59
1
60 The method implies the assumption of simple reactions such as A+B C and A
C+D
4
IChemE Journals
1
2
3
4
5
6 There are some special reactions that need a catalyst or high interfacial area between
7
8 phases (V/L, L/L, S/L or V/S) in order to obtain sufficient reaction speed. These
9
10
11
factors should be added to the Time-corner, in the sense that they ‘help’ the time
12
13 factor. They both determine the speed of the reaction and can be varied independently
14
15 of the concentration and temperature. More so, they are frequently used to reduce the
16
Fo
17
18 time for reactions to complete and could rightfully take the place in the reaction
19
20 triangle. The Time corner could then be replaced by the TIC corner
rP
21
22 (Time/Interface/Catalyst).
23
24
ee
25
26
27 Now let us apply the reaction triangle to equipment examples in a chemical process:
28
rR
29
30
31
32 Example 1: A reactor is evidently equipment somewhere near the middle of the
33
ev
34 triangle.
35
36
iew
37
38
39 Example 2: A feed pre-heater in a process with a fast reaction, for instance, can be
40
41 considered as equipment on the temperature-time side of the triangle. This means that
42
43
On
44 it is supposed to heat a substance to reaction conditions, and it cannot avoid

45
46 substantial residence time (typically 1-5% of the reactor time), so it should have a low
47
ly
48 concentration of one of two reactants to prevent unintended reactions. So, what is

49
50
51 needed to turn this equipment into an unintended reactor? The answer would be to
52
53 increase the missing axis of the triangle, which is concentration. This situation would
54
55
56
be generated by an unexpected event, maybe backflow from the reactor, maybe
57
58 product remnants from another process step in a multipurpose plant.
59
60
5
IChemE Journals
1
2
3
Example 3: A centrifugal pump for a thermally unstable liquid can be designed to
4
5
6 operate in the concentration corner of the triangle. Then, two things are needed to turn
7
8 the pump into an unexpected reactor: Temperature and Time. This situation can be
9
10
11
achieved by a blocked inlet or outlet (time) and a pump motor that keeps running and
12
13 generates heat (temperature).
14
15
16
Fo
17
18 Like the feed pre-heater and the pump, all unit operations have been conceptualized
19
20 with a specific task and are used with assumptions on their capabilities. Conceptually
rP
21
22 they can all be positioned on the sides of the reaction triangle, and will not pose
23
24
hazards unless they move towards the middle. There are several ways in which
ee
25
26
27 equipment can vary their position in the reaction triangle: there are outside / process
28
rR
29
failures (like backflow), equipment failure modes (e.g. pipe bundles in heaters leak
30
31
32 and foul) and equipment inherent side effects (e.g. pumps generate heat). Some of the
33
ev
34 equipment effects are specific to the product or process properties. For example, if a
35
36
thermally unstable product is involved, it will reduce the temperature margin for
iew
37
38
39 safety operation; a heater or a system using an evaporative coolant / diluent can
40
41 evaporate a product and therefore bring the equipment closer to the concentration
42
43
On
44 corner of the triangle.

45
46
47
ly
48 Like the fire triangle, the main factors that contribute to the hazard are easily
49
50
51 recognized and provide a simple concept to learn. Unlike the fire triangle which
52
53 serves as a tool to solve a single situation, the reaction triangle is meant as a complete
54
55
56
description of the process. It helps understanding how the process design intends to
57
58 avoid reaction hazards and how deviations can break the design intentions.
59
60
6
IChemE Journals
1
2
3
The goal of the method is therefore to provide a system that will allow the user to
4
5
6 identify the possible ways in which the equipment used in an installation could
7
8 ‘move’ process streams with their properties over the triangle map, hence generating a
9
10
11
possible hazardous situation.
12
13
14
15 General Mechanisms for Reaction Hazards
16
Fo
17
18
19
20 A non-reactor equipment item has the potential to move its process stream from a
rP
21
22 reaction triangle side towards the centre, thereby creating a reaction hazard. This
23
24
potential is not always recognised. However, the mechanisms involved in the move
ee
25
26
27 can be generalized. The benefit of generalization is that the knowledge can be applied
28
rR
29
in a wider range of cases than that from which the generalization derived. This is
30
31
32 especially important when analyzing incidents, where there is a desire to derive as
33
ev
34 much learning as possible from a few cases. Generalizations require deep

35
36
understanding of the physics and the design intent, skills that are not usually
iew
37
38
39 accessible in the team that performs the risk analysis. Once they have been developed
40
41 they can be applied by people with less experience.
42
43
On
44
45
46 The generalized mechanisms developed for the preparation of the assessment tool
47
ly
48 have been divided in three classes:

49
50
51
52
53 • Failure: equipment specific failure is also related to the way the equipment is
54
55
56
designed to accomplish its design intent. As an example, all coolers need a
57
58 cooling medium that can mix with the process fluid through leakage.
59
60
7
IChemE Journals
1
2
3
4
• Side effects: equipment inherent side effects are related to the way the
5
6 equipment is designed to accomplish its design intent. For instance, a pump
7
8 can also be considered a little mixer. Normally these side effects will focus on
9
10
11 physical phenomena such as phase behaviour, friction, turbulence, dynamics,
12
13 stability and driving forces like gravity, heat and pressure.
14
15
• External causes: mechanisms that are not caused by the equipment themselves
16
Fo
17
18 but by general events in the process. As an example, solar radiation may
19
20 sometimes heat a drum up to a decomposition reaction temperature.
rP
21
22
23
24
Mechanisms can be developed from scratch using theoretical and practical knowledge
ee
25
26
27 of equipment and chemistry behaviour but it is also possible to develop them based on
28
rR
29
30 an incident. The mechanism will subsequently be tested for its potential to occur in
31
32 other equipment. Sometimes it is found that the mechanism itself involves specific
33
ev
34 physico-chemical properties of the process or process materials. An example is shock-

35
36
sensitivity.
iew
37
38
39
40
41
Example 4: A decomposition reaction was found to be triggered by the presence of
42
43
On
44 rust in the bottom of a distillation column. The following generalizations can be

45
46 derived from this example:
47
ly
48
49
50
51 • Triangle generalization: it was a move towards the Time
52
53 (Time/Interface/Catalyst) corner.
54
55
56 • Physical generalization: particles can accumulate by settling out in a stagnant
57
58 zone with a low liquid velocity. The mechanism can therefore be expected in
59
60
vessels, tanks and heat exchangers.
8
IChemE Journals
1
2
3
4
• Process generalisation: There are many process materials that are incompatible
5
6 with rust and may decompose.
7
8 • Failure generalization: carbon steel being a common construction material,
9
10
11 rust formation should be expected.
12
13
14
15
Generating the ‘Decisive Question’
16
Fo
17
18
19
20 It is clear that having a set of generalized mechanisms involved in potential hazards
rP
21
22
23
does not constitute a training / assessment tool in itself, especially if a combination of
24
general mechanisms and factors is needed to pose a hazard. There are two problems
ee
25
26
27 that need to be addressed:
28
rR
29
30
31
32 a) Delivery of the information
33
ev
34 b) Accuracy of the method

35
36
iew
37
38
39 In the proposed method, both problems are handled through the so called ‘User
40
41
specific, decisive Question’.
42
43
On
44
45
46 a) Delivery of information.
47
ly
48
49
50
51 Since the objective of the tool is to train people in recognizing reactive hazards, the
52
53 entire mechanism will be split in just two parts: a condition and a consequence.
54
55
56
57
58 IF condition THEN consequence
59
60
9
IChemE Journals
1
2
3
The user will have to supply the ‘condition’ by answering a final Decisive Question in
4
5
6 order to learn the ‘consequence’ of the potential hazard. This forces the user in an
7
8 active role and regulates the flow of information. To stimulate thinking, the Decisive
9
10
11
Question should preferably hint to the mechanism involved. To limit the amount of
12
13 information, the user first selects from a list which equipment and process properties
14
15 he/she wants to consider.
16
Fo
17
18
19
20 Looking at example 4, if the user has selected the process property that rust may
rP
21
22 catalyse reactions and that he/she has a tank, then the following question will be
23
24
generated:
ee
25
26
27
28
rR
29
Is there equipment upstream that can form rust (e.g. made from carbon steel)?
30
31
32
33
ev
34 If answered positively the consequence will be displayed as:

35
36
iew
37
38
39 This equipment can accumulate rust and provide both a high temperature and a
40
41 long residence time and therefore has a potential for a reactive hazard.
42
43
On
44
45
46 b) Accuracy of the method.
47
ly
48
49
50
51 The set of generalized mechanisms for a standard set of equipment and process
52
53 properties are broad enough to describe most potential hazards, but would generate
54
55
56
too many false positives (identifying a potential hazard where the likelihood is very
57
58 low). The risk of many false positives is that the attention of the user is diverted from
59
60 the more realistic potential hazards. The accuracy can be improved by asking the user
10
IChemE Journals
1
2
3
to supply additional information on the likelihood themselves. The information can
4
5
6 come from the process design & control, construction materials, and so on, and
7
8 essentially it moves a conditional aspect from the entire mechanism into the decisive
9
10
11
question.
12
13
14
15 Looking at example 3, if the user has selected a pump and a thermally unstable
16
Fo
17
18 product, then the potential hazard can be described in the general mechanism as:
19
20
rP
21
22 IF product=unstable AND equipment=pump
23
24
THEN potential-hazard=”If the flow can be blocked then the temperature rise due to
ee
25
26
27 dissipation of heat may be sufficiently high to initiate a decomposition
28
rR
29
reaction.”
30
31
32
33
ev
34 With a User-specific, Decisive Question “Is it possible to block flow through this
35
36
equipment?” The user provides the unknown likelihood and the potential hazard
iew
37
38
39 becomes more relevant:
40
41
42
43
On
44 IF product=unstable AND equipment=pump AND condition=TRUE

45
46 THEN potential-hazard=”The temperature rise due to dissipation of heat during
47
ly
48 blocked flow may be sufficiently high to initiate a decomposition reaction.”

49
50
51
52
53 Building a database from incidents
54
55
56
57
58 A knowledge database was build using different sets of incident sources, both from
59
60 open literature and proprietary information. The main source of information used has
11
IChemE Journals
1
2
3
been the IChemE Accident database. A query was performed on the database as to
4
5
6 identify incidents with unwanted chemical reactions as triggering causes of the events
7
8 and yielded a total of 631 cases for analysis. From the public information of the
9
10
11
accidents reported to the MARS database 68 useful cases were identified. Incident
12
13 descriptions and summaries were interpreted and generalized using the concepts
14
15 described previously. Some cases are also used as examples in the tool.
16
Fo
17
18
19
20 The resulting database consists of tables for 16 equipment types (Table 1), 7 process
rP
21
22 properties (Table 2), 22 User Specific/Decisive Questions, 25 general mechanisms
23
24
and 61 incident examples. Matrices are used to assign process properties to general
ee
25
26
27 mechanisms and to equipment Matrices and Decisive Questions to General
28
rR
29
Mechanisms. The complete tables are downloadable from the S2S website.
30
31
32
33
ev
34 <Table 1>
35
36
iew
37
38
39 <Table 2>
40
41
42
43
On
44 Implementation of the assessment tool

45
46
47
ly
48 The S2S training & assessment tool for unexpected reactors has been implemented as
49
50
51 a Microsoft Excel® Worksheet included in the S2S web portal. It can be run either on-
52
53 line at the S2S web portal or downloaded for use off-line. The tool has the following
54
55
56
functionality:
57
58
59
60 • Simple user interface including time-stamped print.
12
IChemE Journals
1
2
3
4
• Executing the knowledge contained in the database.
5
6 • The knowledge database can be viewed / modified by specialists.
7
8 • Language Portability.
9
10
11
12
13 The tool requirements are limited to Excel 2000 with macros enabled option.
14
15
16
Fo
17
18 a) User interface.
19
20
rP
21
22
23 The tool developed is started from the S2S website by double-clicking on the
24
corresponding link. This will automatically launch Excel to display & run the tool as
ee
25
26
27 shown in Figure 2. The User interface is a main screen with access to the main
28
rR
29
30 functions: Help, Assessment & Print. The Assessment itself cycles through the
31
32 screens and displays tables to complete and results to view & print.
33
ev
34
35
36
<Figure 2>
iew
37
38
39
40
41
42
b) Execution of the knowledge.
43
On
44
45
46 The knowledge of the database is executed using Visual Basic Macros to perform the
47
ly
48
49 following logic steps.
50
51
52
53 1) Display an ‘Equipment Table’ and a ‘Process Properties Table’ in which the user
54
55
56 selects which equipment and process properties combinations are to be assessed.
57
58 2) It pre-selects possible mechanisms from the ‘Equipment-Mechanism Matrix’.
59
60
13
IChemE Journals
1
2
3
3) Using the ‘Mechanism-Properties Matrix’, it removes mechanisms that need a
4
5
6 process property that was not selected.
7
8 4) For the remaining possible mechanisms, it builds a list of ‘Decisive Questions’
9
10
11
from the ‘Mechanism-Question Matrix’.
12
13 5) A ‘Decisive Questions’ list is displayed, which the user must answer with Yes or
14
15 No. In order to facilitate the visualization of the hazards that the tool intends to
16
Fo
17
18 identify, different examples based on real accidents triggered by such hazards are
19
20 available for each of the ‘Decisive Questions’ generated.
rP
21
22 6) For each positively answered ‘Decisive Questions’, a table listing the relevant
23
24
mechanisms will be displayed as well as the equipment, process properties and
ee
25
26
27 mechanism type that are involved in the mechanism.
28
rR
29
30
31
32 The macros are triggered by jumping to the next or previous page. Thus the list of
33
ev
34 Decisive Questions and Mechanisms is adjusted dynamically to changes in the user

35
36
response.
iew
37
38
39
40
41 c) Customisation of the Database.
42
43
On
44
45
46 The database can be customized to change the language of the assessment and to
47
ly
48 change the knowledge database (modify or extend equipment, process properties,

49
50
51 mechanisms and so on). The excel sheet can be switched between a protected runtime
52
53 (=user) mode and an unprotected maintenance mode. The switch is possible with two
54
55
56
buttons that are located two pages down on the start screen. After the modifications,
57
58 the excel sheet can be saved under a new name.
59
60
14
IChemE Journals
1
2
3
The language of the tool can be changed by replacing the English texts in the Excel
4
5
6 tables. Furthermore, a new Assessment Case can be defined for a new potential
7
8 reactive hazard. The user must provide definitions for new elements to be included in
9
10
11
the Tables along with their relevant combinations for the new case. This is done by
12
13 completing the 2 matrices that define the equipment and process properties for which
14
15 the mechanism is valid: in the Equipment-Mechanism matrix, all Equipment that can
16
Fo
17
18 exhibit the mechanism is defined by a 1; in the Property-Mechanism matrix, all
19
20 Process Properties that are essential to exhibit the mechanism are defined by a 1.
rP
21
22 Table 3 shows the required information for each of the elements of a new Assessment
23
24
Case.
ee
25
26
27
28
rR
29
<Table 3>
30
31
32
33
ev
34 When developing an Assessment Case an effort should be made to identify all

35
36
equipment that could have a similar mechanism and reconsider modification of
iew
37
38
39 Mechanism & Decisive Question in order to combine them in a more general
40
41 mechanism. The advantage of fewer mechanisms is fewer questions for the same
42
43
On
44 number of equipment-mechanisms. The disadvantage of too little general mechanisms

45
46 is that the abstraction level increases and that the hazards may become too general.
47
ly
48
49
50
51 Conclusions
52
53
54
55
56
The S2S training and assessment tool for unexpected reactors can be a useful method
57
58 for SME’s engineers and operators to identify reactive hazards in plant equipment.
59
60 The combination of simple reactivity concepts included in a user friendly Excel based
15
IChemE Journals
1
2
3
system allows the user to perform an assessment of the chemical hazards that may be
4
5
6 encountered during different unit operations of an industrial process. Moreover, the
7
8 experience gained in the use of the tool and the possibility to adapt it to new
9
10
11
assessment cases gives the user the possibility to adapt it to the specific needs of a
12
13 particular establishment.
14
15
16
Fo
17
18 The combination of concepts described has proven to be flexible enough to cover a
19
20 wide range of incidents and process steps in an efficient knowledge representation.
rP
21
22 Thus it has the potential for maximising learning from a sparse number of typical
23
24
incidents.
ee
25
26
27
28
rR
29
It must be stressed that it is not the aim of the tool to cover all the hazards related to
30
31
32 reactivity in a process plant, but to provide basic guidance towards commonly
33
ev
34 unnoticed hazards. At the same time, it is not intended to provide specific safety
35
36
measures related to the hazards identified. It should be the user, by means of his / her
iew
37
38
39 knowledge of the particular process and plant possibilities, to come up with the most
40
41 feasible solutions in order to perform the process under safe conditions.
42
43
On
44
45
46 The assessment tool can be used by a person with some knowledge of the process
47
ly
48 chemistry and physics. However, in case any doubt should arise by the use of the
49
50
51 method, the advice of an expert in the field of chemical reactivity and process safety
52
53 should be asked. It must also be kept in mind that this method is a part of the S2S
54
55
56
package of assessment tools that covers other aspects of chemical reactivity, as well
57
58 as other hazards like fire or explosions. These tools should be used whenever
59
60 necessary; particularly the reactivity assessment tool should be used in case of doubt
16
IChemE Journals
1
2
3
regarding stability, compatibility and reactivity of the substances and mixtures
4
5
6 involved in the process.
7
8
9
10
11
References
12
13
14
15 Center for Chemical Process Safety (CCPS), 1991, Guidelines for hazard evaluation
16
Fo
17
18 procedures, (American Institute for Chemical Engineers, New York, USA).
19
20
rP
21
22 Legget, D., Rapid Identification of Reactivity Hazards in a Multiuse facility, Process
23
24
Safety Progress, 25(2): 108:115.
ee
25
26
27
28
rR
29
Major Accident Reporting System, MARS database, 2007, Major Accident Hazards
30
31
32 Bureau, Institute for the Protection and Security of the Citizen, European Commission
33
ev
34 - Joint Research Centre, Ispra, Italy.

35
36
iew
37
38
39 Murphy, J. and Holstrom, D., Understanding Reactive Chemical Incidents, Chemical
40
41 Engineering Progress, March 2004: 31-33.
42
43
On
44
45
46 Nomen, R. Sempere, J., Serra, E., Pey, A., Sales, J. & Ghinaglia, V., 2004, Check
47
ly
48 Cards for Runaway (CCR). An operative tool for the risk assessment of Highly
49
50
51 reactive systems performed in small and medium enterprises. Trans IchemE, Part B,
52
53 Process Safety and Environmental protection, 82(B1): 5-11 (a).
54
55
56
57
58
59
60
17
IChemE Journals
1
2
3
Nomen, R. Sempere, J., Sales, J. & Ghinaglia, V., 2004, Development and validation
4
5
6 of operative tools for the risk assessment of highly reactive systems: HarsMeth
7
8 version 2. Journal of Loss Prevention in the Process Industries, 17(5), 365-371 (b).
9
10
11
12
13 Safety 2 Safety web site, 2007, http://www.s-2-s.org.
14
15
16
Fo
17
18 Sales, J, Mushtaq, F., Christou, M.D. and Nomen, R., 2007, Study of major accidents
19
20 involving chemical reactive substances: analysis and lessons learned. Trans IChemE,
rP
21
22 Part B, Process Safety and Environmental Protection, 85(B3): 1-8.
23
24
ee
25
26
27 The IchemE Accident Database, 2007, Institution of Chemical Engineers, Davis
28
rR
29
Building, 165-189 Railway Terrace, Rugby, CV21 3HQ, United Kingdom.
30
31
32
33
ev
34
35
36
iew
37
38
39
40
41
42
43
On
44
45
46
47
ly
48
49
50
51
52
53
54
55
56
57
58
59
60
18
IChemE Journals
1
2
3
4
5
6 Concentration
7
8
9 C
10
11
12
13
14 Temperature H T Time
15
16
Fo
17 Figure 1. The reaction triangle.

18
19
20
rP
21
22
23
24
ee
25
26
27
28
rR
29
30
31
32
33
ev
34
35
36
iew
37
38
39
40
41
42
43
On
44
45
46
47
ly
48
49
50
51
52
53
54
55
56
57
58
59
60
19
IChemE Journals
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Fo
17
18
19
20
rP
21
22
23
24
ee
25
26
27
28
rR
29
30
31
32
33
Figure 2. First page of the S2S training & assessment tool for unexpected reactors.
ev
34
35
36
iew
37
38
39
40
41
42
43
On
44
45
46
47
ly
48
49
50
51
52
53
54
55
56
57
58
59
60
20
IChemE Journals
1
2
3
4
5
6 Vessels
7
8 Pipes
9
10
11 Valves
12
13 Shared systems
14
15
16
Pumps
Fo
17
18 Centrifuges
19
20 Filters
rP
21
22
23 Heat exchangers
24
Distillation
ee
25
26
27
28
Absorber/Adsorber
rR
29
30 Extractor
31
32 Dryers
33
ev
34
35 Mills
36
Drumming
iew
37
38
39
40 Flash / KO drum
41
42 Storage
43
On
44
45
46
47 Table 1. Equipment types included in the S2S training & assessment tool for
ly
48
49 unexpected reactors.
50
51
52
53
54
55
56
57
58
59
60
21
IChemE Journals
1
2
3
4
5
6 Shock sensitive
7
8 Thermally unstable
9
10
11 2 immiscible liquids L/L
12
13 Light solvent / diluent V/L
14
15
16
Rust catalyses
Fo
17
18 Pyrophoric
19
20 Water Sensitive
rP
21
22
23
24
Table 2. Process Properties included in the S2S training & assessment tool for
ee
25
26
27
unexpected reactors.
28
rR
29
30
31
32
33
ev
34
35
36
iew
37
38
39
40
41
42
43
On
44
45
46
47
ly
48
49
50
51
52
53
54
55
56
57
58
59
60
22
IChemE Journals
1
2
3
4
5
6 Assessment Element Information Required
7
8 Equipment A short (2 words max) description of the type, plus a help sentence
9
10 explaining details of the equipment where confusion might exist.
11
12 Process type A short (2 words max) description of the type, plus a help sentence
13
14 clarifying details of the Process, plus a multi-character abbreviation Id.
15
16 Mechanism A paragraph of text explaining how external causes, equipment failure or
Fo
17
18 equipment side effects can cause a move in the reaction triangle and
19
20 hence pose a hazard, plus a mechanism type number indicating the type
rP
21
22 of move in the reaction triangle, plus a Decisive Question number that
23
24 needs to be answered confirmative in order to make the hazard relevant.
ee
25
26 Decisive Question A question that can be answered by yes or no and -if answered
27
28 positively- displays to the user the mechanism of the hazard.
rR
29
30
31
32
33
Table 3. Summary of necessary information to create a new Assessment Case.
ev
34
35
36
iew
37
38
39
40
41
42
43
On
44
45
46
47
ly
48
49
50
51
52
53
54
55
56
57
58
59
60
23
IChemE Journals
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Fo
15
16
17
18
rP
19
20
21
22
ee
23
24
25
26
rR
27
28
29
30
ev
31
32
33
34
ie
35
36
37
w
38
39
40
On
41
42
43
44
45
ly
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IChemE Journals
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Fo
15
16
17
18
rP
19
20
21
22
ee
23
24
25
26
rR
27
28
29
30
ev
31
32
33
34
ie
35
36
37
w
38
39
40
On
41
42
43
44
45
ly
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IChemE Journals

S2S. Sales - Nomen

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

S2S. Sales - Nomen

Uploaded by

Copyright:

Available Formats

Process Safety and Environmental Protection

Manuscript ID: PSEP-07-028

Manuscript Type: Full Paper

Date Submitted by the

Chemical Engineering; Universitat Ramon Llull

unexpected reactors, chemical hazard, assessment tool, reaction

34 Via Augusta 390, 08017 Barcelona, Spain

44 it is supposed to heat a substance to reaction conditions, and it cannot avoid

48 concentration of one of two reactants to prevent unintended reactions. So, what is

44 corner of the triangle.

34 much learning as possible from a few cases. Generalizations require deep

48 have been divided in three classes:

34 physico-chemical properties of the process or process materials. An example is shock-

44 rust in the bottom of a distillation column. The following generalizations can be

34 b) Accuracy of the method

34 If answered positively the consequence will be displayed as:

44 IF product=unstable AND equipment=pump AND condition=TRUE

48 blocked flow may be sufficiently high to initiate a decomposition reaction.”

44 Implementation of the assessment tool

34 Decisive Questions and Mechanisms is adjusted dynamically to changes in the user

48 change the knowledge database (modify or extend equipment, process properties,

34 When developing an Assessment Case an effort should be made to identify all

44 number of equipment-mechanisms. The disadvantage of too little general mechanisms

34 - Joint Research Centre, Ispra, Italy.

17 Figure 1. The reaction triangle.

You might also like