8

Relative ranking

Chapter outline
8.1 Definition ................................................................................................................................. 71
8.2 Description ............................................................................................................................... 71
8.3 Resource requirements ........................................................................................................... 73
8.4 Timing ...................................................................................................................................... 74
8.5 Advantages, disadvantages and uncertainties ..................................................................... 74
Advantages ...............................................................................................................................74
Disadvantages ...........................................................................................................................74
Uncertainties .............................................................................................................................75
8.6 Applications for relative ranking methods ............................................................................ 75
8.7 Example of the Dow FEI [1] .................................................................................................... 75
8.8 Conclusion ............................................................................................................................... 78
References ...................................................................................................................................... 78
Further reading .............................................................................................................................. 78

8.1 Definition
Relative ranking is a structured method of analysis leading to a numerical grading of the main
hazards associated with each section of a plant or process. These grades may be used in a num-
ber of ways, for example to identify low hazard options or to select sections that require further
hazard analysis. The hazards of concern usually include fire, explosion, airborne releases of
toxic substances and in some cases releases of materials dangerous for the environment.

8.2 Description
The most widely used ranking method for fire and explosion is the Dow Fire and Explosion
Index (FEI), augmented by the Dow Chemical Exposure Index (CEI) for airborne releases
of toxic substances and both are described here. Detailed guides are available for both the
FEI [1] and the CEI [2] and further accounts appear in standard references [3, 4], which
also mention some other ranking methods. A specialised index has been developed
and tested for offshore oil and gas installations [5] and the basic methods can be extended
if required but these are not covered here.
The Dow FEI, first published in 1964, is now in its seventh edition. It is widely used,
both within the Dow Chemical Company and by others, primarily to give a relative value
for the potential of process unit losses due to fire and explosion hazards. Dow has recently
extended the method into a software version, which is used internally.
A Guide to Hazard Identification Methods. https://doi.org/10.1016/B978-0-12-819543-7.00008-2 71
© 2020 Elsevier Inc. All rights reserved.
72 A Guide to Hazard Identification Methods

The guide provides the necessary detail, formulae, graphs and methodology for an
experienced engineer to carry out an assessment. The initial output from the FEI is a value
for the index for that unit of the plant. The index by itself may be of direct use, for example
the Dow Chemical Company requires that all facilities are required to carry out a Process
Hazard Analysis (PHA) on a 3-year cycle and whenever the leadership of a facility changes
or a significant new project or development is planned. When the FEI within the PHA pro-
cess reveals an index of 128 or higher, a Layer of Protection Analysis (LOPA) is required. In
addition, technology new to the company always requires a HAZOP. The method may be
taken further to get estimates of the realistic maximum losses for the units, expressed in
terms of an Actual Maximum Probable Property Damage (Actual MPPD), Maximum
Probable Days Outage (MPDO) and Business Interruption (BI). In estimating these values,
an area of exposure and a damage factor are also obtained. These later estimates apply
well for fire cases, but if an explosion is a required case, Explosion analysis using PHAST
or TNO Multi Energy methods are often applied.
The first step in applying the FEI to a plant is to select the key units—known as pertinent
process units—within the plant. The selection is based on the individual unit operations
such as unloading system, storage, tank, reactor, distillation column, etc. type of material
within the unit, the quantity of material, the process conditions, the cost of equipment,
the presence of items critical to operation and any history of problems. A process seldom
requires a detailed analysis for more than 3 or 4 units. A material factor (MF) is derived
for the materials contained in each unit using standard NFPA flammability and instability
rankings for the dominant material that is present. The guide provides an extensive list
of these and a means of estimating the MF for unlisted materials or for mixtures.
Two penalty factors, one for general process hazards (F1) and the other for special pro-
cess hazards (F2), are then established. The general process hazards cover exothermic and
endothermic reactions, material handling and transfer, enclosed process units, access,
and drainage and spillage control. Special process hazards are based on the toxicity of
the materials being processed, low- and high-pressure working, operation near a flamma-
ble range, dusts, low-temperature operation, the quantity of material in process, corrosion
and erosion, leakage and a few other unusual aspects such as the use of fragile process
containing equipment such as glass, hoses, expansion joints. The analyst determines
whether or not a feature applies and, if so, chooses an appropriate penalty factor following
the guidance provided or, for pressure and quantity, using the given empirical correla-
tions. The guide includes examples showing how these factors should be estimated.
The penalties are then totalled to give F1 and F2 and the product of these with the MF
gives the value for the FEI. Based on the extensive experience of using this method the
unit is then classified in one of five bands, from light to severe. A useful effect of using
the index during projects is that modifications can be assessed from an Inherent Safety
aspect, since a reduction in index usually accompanies an Inherently Safer design.
In a full analysis, the next step is to select and assign values to loss control credit factors.
These categories are process control, material isolation and fire protection. Features
such as emergency shutdown, operating procedures, other hazard analyses, interlocks,
 Chapter 8 • Relative ranking 73

fire-cladding, etc. earn credit factors, which are usually in the range of 0.90–1.0 with a few
below 0.9. An equation is given to derive the overall Loss Control Factor from the relevant
credit factors. Apart from their use in the calculation of the Actual MPPD, the derivation of
these factors provides a valuable record of the safety features within system. These must
be maintained during the plant lifetime if target safety levels are to be achieved. The final
step of the analysis is now straightforward. An area of exposure and a damage factor are
derived from the MF, F1 and F2. These are used in combination with the replacement
value of equipment and materials in that area to calculate the remaining factors (Base
MPPD, Actual MPPD, MPDO and BI). The relationships used in the final calculations rely
upon historical incident data and some judgement is necessary in the application to each
unit. The radius of damage is useful in determining separation of key process units to
avoid ‘domino’ effects.
The Chemical Exposure Index (CEI) provides a ranking method for the acute toxicity
health hazard potential from hypothetical chemical releases on a plant. As with the FEI,
the documentation is available through the AIChE [2].
Initially various possible release scenarios are considered using experience-based rules
as to likely ruptures and failures. An airborne quantity is calculated (actually a vapour gen-
eration rate) based on the type of release. Possibilities include gases, flashing liquids and
liquids forming pools, including partial flashing if appropriate. This is combined with an
emergency exposure level (ERPG-2 [6] or equivalent), which, over 1 hour, should not cause
irreversible health effects or impair the ability to take protective action. The value of the
CEI is then obtained by formula as are hazard distances—an indication of the distance
within which the ERPG is exceeded.
The latest issue of the AIHA Guide to ERPG values should always be consulted. If no
value is available, the CEI Guide gives advice on how to proceed. Detailed examples illus-
trate the main release scenarios and a containment and mitigation checklist is provided to
help the review process. The Dow Chemical Company requires units having a CEI above a
threshold value of 200 to be subjected to a further risk review. Whilst the CEI does not pro-
vide an absolute measure of risk or precise values of hazard distances it is a well-based
method of ranking different scenarios across a wide range of processes, materials and pos-
sible incidents. It can be used for screening purposes, to identify the main sources of toxic
hazards and to help with the mitigation of them, to compare different options or to show
the possible effects of a modification. The only difficulty in its use occurs when there is no
published ERPG. When ERPG2 concentrations are predicted to exceed fence-line
distance, deeper study is indicated (e.g. PHAST) and LOPA.

8.3 Resource requirements

Manpower: The method can be used by an individual, although there are advantages in
having two or more involved, and it is always advisable to have the results reviewed by
an independent person. It is possible for an untrained, but experienced, engineer to apply
the methods although training or initially working with an experienced assessor is helpful.
74 A Guide to Hazard Identification Methods

When the data are available and a site visit has been made for existing plant, an experi-
enced user can normally complete the assessment on a unit in 1–2 hours.
Data: The analyst will need a site plan and a minimum of a block diagram of the plant
showing the main items of hardware, including the actual or intended layout; overall
design and operating data; a list of all the materials, quantities, full material and safety
data sheets; the appropriate guide and recording forms.

8.4 Timing
For a new project the technique would always be used at an early stage, either when
alternative routes are being explored, when different scales of operation and the possible
process methods are being evaluated and, after a final decision, before the detailed design
work in undertaken.
For existing plant it may be used to review the operation, especially after a period
of operation, or whenever a change of process or modification of the plant is
proposed.

8.5 Advantages, disadvantages and uncertainties

Advantages
• Relatively simple with limited need for plant and process information.
• Can be applied at an early stage in a project before a detailed P&ID is available.
• Indicates the main areas with hazard potential.
• The methodology can be adapted to suit company needs.
• The method supports the drive to inherently safer design, since it is clear that a process
unit with a lower index is inherently safer since it prioritises the principles of
moderation, reduction, etc.

Disadvantages
• The methods are limited to the hazards of fire, explosion and chemical exposure.
• Skilled interpretation is needed. In Dow, this is carried out by the Process Safety
Engineer function.
• Detailed features of the design, which may lead to hazardous events are not
highlighted.
• The methods usually need to be used in conjunction with other hazard identification
and assessment methods.
• The method was developed for use with mainstream chemical activities. It cannot be
used directly in some others such as the off-shore oil and gas industry.
• It does not explicitly cover environmental problems.
• Little account is taken of human factors.
 Chapter 8 • Relative ranking 75

Uncertainties
• Although the starting point (the MF) is precise, many of the subsequent factors involve
a judgement by the analyst. Differences in the final result of 10% can occur. It is also
possible to have differences in the sectioning of the plant.
• The levels for action are based upon the use of the technique on many plants over
many years. Where this basis is not available or not applicable it is difficult to interpret
the numerical values.

8.6 Applications for relative ranking methods

The calculations from the Dow FEI can be used
• as a guide to the selection of fire protection methods;
• to focus attention on key areas or items of equipment;
• for the comparison of alternative process routes;
• to encourage use of the inherent safety;
• to evaluate the effects of modification;
• to act as a filter before the application of more detailed hazard study methods;
• to follow trends within a site over time;
• to help in planning layouts and separation distances.
Indices are particularly valuable in the early stages of a project due to their simplicity and
speed of use and also as a screening tool for the application of other hazard review
methods in the case of both new and existing plants. For multisite companies they can
provide a means of standardisation across varied plants, processes, management and
regulatory regimes.
The CEI can be used in many of the above ways, particularly in directing attention to
the most significant scenarios and helping to ameliorate these and as an aid to effective
Emergency Response Planning.

8.7 Example of the Dow FEI [1]

The application of the Dow Fire and Explosion Index to a Butadiene unloading operation.
The facility is an outdoor unloading station where railcars of 20,000 kg of 1,3-butadiene are
unloaded into a storage tank. The connection to the inlet of the storage tank is via an unload-
ing arm, which replaced the armoured hose used previously. The are two connections
(1) to the bottom outlet of the railcar
(2) to the top vapour space of the railcar
The first of these is to allow the butadiene to flow to the storage tank. The second is con-
nected to a compressor which compresses butadiene vapour from the top of the storage
tank to provide the driving pressure for the unloading operation. The overall operation is
76 A Guide to Hazard Identification Methods

described as ‘in process’ rather than ‘in storage’ because the case involves several oper-
ating steps from the positioning of the railcar, connection, transfers, disconnection and
isolation.
The operation is outdoors, with good access and a modern drainage system to a
containment pond.
In the sections below each entry into the FEI worksheet in Table 8.1 is discussed and
explained.
Material factor: The material factor for Butadiene is 24 (from the NFPA ratings in the
appendix of the manual).
1. General Process Hazards: Exothermic or Endothermic Reactions
No reactions take place, so no penalty is applied.
Material Handling and Transfer: This operation clearly falls into this category.
A penalty of 0.85 is applied (Flammability rating 4).
Enclosed or Indoor Process Units: No penalty is applied for an outdoor operation.
Access: In this example, the access for emergency response is good, so no penalty is
applied.
Drainage and Spill Control. This area is well drained to a containment system. In view
of the fact that Butadiene has a high vapour pressure, drainage is not deemed to be a
major factor. No penalty is applied.
Applying the penalties with the base factor of 1.00 results in a General Process Hazard
(F1) of 1.85.
2. Special Process Hazards:
A. Toxic Materials: Butadiene has a NFPA toxicity rating of 2, so a penalty of 0.4
applies.
B. Subatmospheric Pressure: This question relates to subatmospheric pressure,
which could cause air to be drawn into the system. The operating pressure is in
excess of 300 kPa, so no penalty applies.
C. Operation In or Near the Flammable Range: This operation is significantly above
the upper flammable limit. No penalty applies.
D. Dust Explosion: This is not relevant. No penalty.
E. Pressure: The operating pressure is 300 kPa, the relief system pressure is 450 kPa. The
system calculated penalty is 0.12 when the multiplier for LPG type materials is applied.
F. Low-Temperature Operation: No penalty is applied.
G. Quantity of Flammable Material—IN PROCESS: is 20,000 kg and the Heat of
Combustion is 10,700 kcal/kg. This produces a penalty of 1.38.
H. Corrosions and Erosion: The material of construction is stainless steel. No
corrosion is experienced with this material in butadiene service. No penalty is
applied.
 Chapter 8 • Relative ranking 77

Table 8.1 FEI worksheet [1] for butadiene unloading example.

78 A Guide to Hazard Identification Methods

I. Leakage, Joints and Packing: There are seals on the compressor and on the loading
arm connections. A penalty of 0.1 is applied.
J. Use of Fired Equipment: This refers to operations like fire burners and furnaces. In
this example, the nearest such operation is at least 300 m distant. No penalty is
applied.
K. Hot Oil Heat Exchange System: There is no heating in this operation. No penalty is
applied.
3. Special Process Hazard (F2) calculates at 3.00.
4. The Fire and Explosion Index calculation is carried out. F1 × F2 × Material Factor
yielding the result is 133. In Dow Chemical risk range ranking, this would result in
study such as their form of HAZOP and Layer of Protection Analysis.

8.8 Conclusion
This example demonstrates the ease with which an FEI analysis can be made for a routine,
although potentially hazardous, operation can be assessed. The result, using the Dow
Chemical risk range ranking, is that a further, more detailed, method of hazard identifi-
cation and assessment should be applied. It also indicates that it can be used both for
a new process operation as well as for existing plant.

References
[1] Dow Chemical Company, Fire & Explosion Hazard Classification Guide, seventh ed., American Insti-
tute of Chemical Engineers (AIChE), New York, USA, 1994 (rev 1997).
[2] Dow Chemical Company, Chemical Exposure Index Guide, second ed., American Institute of Chemical
Engineers (AIChE), New York, USA, 1993 (rev 1997).
[3] S. Mannan (Ed.), Lees’ Loss Prevention in the Process Industries: Hazard Identification, Assessment
and Control, fourth ed., ISBN: 0123971896, 2012.
[4] Guidelines for Hazard Evaluation Procedures, third ed., CCPS, ISBN: 978-0-471-97815-2, 2008.
[5] F.K. Crawley, M.M. Grant, Concept risk assessment of offshore hydrocarbon production installations,
Trans. IChemE 75 (Part B) (1997) 157.
[6] Emergency Response Planning Guideline as issued by the American Industrial Hygiene Association.
The 2016 list covers 149 compounds. It is periodically updated.

Further reading
[7] F.I. Khan, T. Husain, S.A. Abbasi, Safety weighted hazard index, Trans. IChemE 79 (Part B) (2001) 65.