PETRONAS TECHNICAL STANDARDS

DESIGN AND ENGINEERING PRACTICE

MANUAL

FILTERS FOR REMOVING PARTICLES FROM

HYDROPROCESSER FEEDS - TYPE SELECTION
AND DESIGN RULES

PTS 31.27.21.10
FEBRUARY 2010

© 2010 PETROLIAM NASIONAL BERHAD (PETRONAS)

All rights reserved. No part of this document may be reproduced, stored in a retrieval system or transmitted in any form or by any
means (electronic, mechanical, photocopying, recording or otherwise) without the permission of the copyright owner.
 PTS 31.27.21.10
February 2010
Page 2

PREFACE
PETRONAS Technical Standards (PTS) publications reflect the views, at the time of
publication of PETRONAS OPUs/Divisions.
They are based on the experience acquired during the involvement with the design,
construction, operation and maintenance of processing units and facilities. Where
appropriate they are based on, or reference is made to, national and international
standards and codes of practice.
The objective is to set the recommended standard for good technical practice to be applied
by PETRONAS' OPU's in oil and gas production facilities, refineries, gas processing plants,
chemical plants, marketing facilities or any other such facility, and thereby to achieve
maximum technical and economic benefit from standardisation.
The information set forth in these publications is provided to users for their consideration
and decision to implement. This is of particular importance where PTS may not cover every
requirement or diversity of condition at each locality. The system of PTS is expected to be
sufficiently flexible to allow individual operating units to adapt the information set forth in
PTS to their own environment and requirements.
When Contractors or Manufacturers/Suppliers use PTS they shall be solely responsible for
the quality of work and the attainment of the required design and engineering standards. In
particular, for those requirements not specifically covered, it is expected of them to follow
those design and engineering practices which will achieve the same level of integrity as
reflected in the PTS. If in doubt, the Contractor or Manufacturer/Supplier shall, without
detracting from his own responsibility, consult the owner.

The right to use PTS rests with three categories of users:

1) PETRONAS and its affiliates.
2) Other parties who are authorised to use PTS subject to appropriate contractual
arrangements.
3) Contractors/subcontractors and Manufacturers/Suppliers under a contract with users
referred to under 1) and 2) which requires that tenders for projects, materials
supplied or - generally - work performed on behalf of the said users comply with the
relevant standards.

Subject to any particular terms and conditions as may be set forth in specific agreements
with users, PETRONAS disclaims any liability of whatsoever nature for any damage
(including injury or death) suffered by any company or person whomsoever as a result of or
in connection with the use, application or implementation of any PTS, combination of PTS
or any part thereof. The benefit of this disclaimer shall inure in all respects to PETRONAS
and/or any company affiliated to PETRONAS that may issue PTS or require the use of
PTS.

Without prejudice to any specific terms in respect of confidentiality under relevant

contractual arrangements, PTS shall not, without the prior written consent of PETRONAS,
be disclosed by users to any company or person whomsoever and the PTS shall be used
exclusively for the purpose they have been provided to the user. They shall be returned
after use, including any copies which shall only be made by users with the express prior
written consent of PETRONAS.
The copyright of PTS vests in PETRONAS. Users shall arrange for PTS to be held in safe
custody and PETRONAS may at any time require information satisfactory to PETRONAS in
 PTS 31.27.21.10
February 2010
Page 3
order to ascertain how users implement this requirement.
 PTS 31.27.21.10
February 2010
Page 4

TABLE OF CONTENTS
1.  INTRODUCTION ...................................................................................................... 5 
1.1  SCOPE ..................................................................................................................... 5 
1.2  DISTRIBUTION, INTENDED USE AND REGULATORY CONSIDERATIONS ......... 5 
1.3  DEFINITIONS ........................................................................................................... 5 
1.4  SYMBOLS AND ABBREVIATIONS .......................................................................... 7 
1.5  CROSS-REFERENCES ........................................................................................... 7 
1.6  SUMMARY OF MAIN CHANGES SINCE LAST EDITION ........................................ 8 
2.  PURPOSE OF FEED FILTERS ................................................................................ 9 
3.  FEED FILTER SELECTION - OVERVIEW ............................................................. 10 
4.  DETERMINING THE FILTER RATING ................................................................... 11 
4.1  FILTER RATING PROCEDURE ............................................................................. 11 
4.2  DIRT CONCENTRATION IN THE FEED ................................................................ 12 
4.3  PARTICLE SIZE DISTRIBUTION OF THE DIRT .................................................... 14 
4.4  DIRT CHARACTERISTICS SUCH AS MORPHOLOGY, STICKINESS AND
ELEMENT COMPOSITION .................................................................................... 15 
4.5  DIRT PRODUCED BETWEEN FILTER AND REACTOR ....................................... 16 
4.6  PARTICLE SEPARATION CHARACTERISTICS (GRADE EFFICIENCY) OF
THE REACTOR AND ITS INTERNALS .................................................................. 17 
4.7  DIRT HOLDING CAPACITY OF THE REACTOR AND ITS INTERNALS ............... 18 
4.8  DIRT HOLDING CAPACITY TO BE SPARED FOR BY-PASSING ......................... 19 
4.9  PARTICLE SEPARATION CHARACTERISTICS (GRADE EFFICIENCY) OF
THE FILTER ........................................................................................................... 20 
5.  TYPES OF FEED FILTERS AND SELECTION OF MANUFACTURERS ............... 22 
5.1  TYPES OF FEED FILTERS .................................................................................... 22 
5.2  SELECTION OF VENDORS ................................................................................... 23 
6.  DATA AND INFORMATION TO BE SUBMITTED .................................................. 26 
6.1  RESPONSIBILITIES ............................................................................................... 26 
6.2  SPECIFICATION OF PROCESS CONDITIONS..................................................... 26 
6.3  DATA AND INFORMATION TO BE SUBMITTED WITH THE TENDER ................. 26 
6.4  DATA AND INFORMATION TO BE SUBMITTED AFTER AWARD OF ORDER .... 28 
6.5  VESSEL REQUIREMENTS .................................................................................... 28 
6.6  PIPING AND VALVES ............................................................................................ 28 
6.7  INSTRUMENTATION AND CONTROL .................................................................. 29 
7.  LOCAL COST DATA TO BE SPECIFIED BY PRINCIPAL...................................... 30 
7.1  GENERAL .............................................................................................................. 30 
7.2  FIXED COSTS ........................................................................................................ 30 
7.3  FILTER LIFE TIME AND DEPRECIATION ............................................................. 30 
7.4  VARIABLE COSTS ................................................................................................. 30 
8.  EVALUATION OF VENDOR PROPOSALS ON BASIS OF MINIMUM TOTAL
COST ..................................................................................................................... 32 
8.1  RECEIVED DATA ................................................................................................... 32 
8.2  DESIGN PHILOSOPHY OF VENDORS ................................................................. 33 
8.3  DESIGN FACTORS ................................................................................................ 34 
8.4  OPTIMUM SIZE ...................................................................................................... 37 
8.5  MINIMUM COST AS FUNCTION OF RATE OF SOLIDS REMOVAL ..................... 39 
9.  FILTER SELECTION .............................................................................................. 41 
10.  COST / BENEFIT / RISK ANALYSIS ...................................................................... 42 
11.  SITE TESTS AND COMMISSIONING .................................................................... 43 
12.  REFERENCES ....................................................................................................... 44 
 PTS 31.27.21.10
February 2010
Page 5
1. INTRODUCTION

1.1 SCOPE
This new PTS specifies requirements and gives recommendations for the selection and
design of feed filters for hydroprocessing units. Feed filters remove dirt particles from the
liquid feed which would otherwise foul the hydroprocessing reactors or other equipment.

1.2 DISTRIBUTION, INTENDED USE AND REGULATORY CONSIDERATIONS

Unless otherwise authorised by PETRONAS, the distribution of this document is confined to
companies forming part of PETRONAS group and to Contractors nominated by them. This
PTS is intended for use in oil refineries.
If national and/or local regulations exist in which some of the requirements may be more
stringent than in this PTS, the Contractor shall determine by careful scrutiny which of the
requirements are the more stringent and which combination of requirements will be
acceptable as regards safety, economic and legal aspects. In all cases the Contractor shall
inform the Principal of any deviation from the requirements of this document that is
considered to be necessary in order to comply with national and/or local regulations. The
Principal may then negotiate with the Authorities concerned with the object of obtaining
agreement to follow this document as closely as possible.

1.3 DEFINITIONS

1.3.1 General definitions

The Contractor is the party which carries out all or part of the design, engineering,
procurement, construction, commissioning or management of a project or operation of a
facility. The Principal may undertake all or part of the duties of the Contractor.
The Manufacturer/Supplier/Vendor is the party which manufactures or supplies
equipment and services to perform the duties specified by the Contractor.
The Principal is the party which initiates the project and ultimately pays for its design and
construction. The Principal will generally specify the technical requirements. The Principal
may also include an agent or consultant authorised to act for, and on behalf of, the
Principal.
The word shall indicates a requirement.
The word should indicates a recommendation.
 PTS 31.27.21.10
February 2010
Page 6

1.3.1 Specific definitions

Separation The rate at which solids in the feed are retained by the filter.
(filtration)
It depends on the characteristics of the filter and the particle size
efficiency
distribution and nature of the solids.
Grade The fraction of particles of a certain size retained by the separator.
efficiency
It is a function of particle size, and the curve obtained by plotting the
grade efficiency versus particle size is called the grade efficiency curve.
Many types of separators have grade efficiency curves which run from
100 % separation for relatively large particles to 0 % separation for
relatively small particles. Particles between the minimum and maximum
sizes have a probability of being collected by the filter equal to the
value of the grade efficiency, which can be read directly from the graph.
The grade efficiency curve can be seen as a characteristic curve of the
separator, although it will depend on the nature of the solids (free
flowing, sticky). It is commonly used to calculate separation efficiencies
of continuously operated separators such as cyclones. The
representation of the separation characteristics of a feed filter is
complicated by the fact that the separation is a transient process
depending on the cumulative amount of solids collected. The grade
efficiency is expected to shift to smaller particles as the cumulative
volume of retained particles increases. For the purpose of selecting a
filter, “average” grade efficiencies should be used.
Filter rating A relative measure of the size of particles retained by the filter,
expressed in microns.
It is usually only used consistently for one type of filter supplied by the
same Vendor. One single Vendor may use different ratings for different
types of filters having the same particle removal efficiency. Different
Vendors of the same type of filter may use different ratings for the
same particle removal efficiency. Addition of the removal efficiencies for
the specified filter rating makes it possible to compare the same type of
filters from different Vendors. Different types of filters should be
compared on the basis of the entire grade efficiency curve.
The term absolute filter rating is used by a number of Vendors to
indicate the particle size removed with an efficiency of 99.98 %. Other
Vendors associate the term “absolute” with the invariance of the pore
structure of the filter medium (some types of filter media release
already collected particles beyond a certain pressure drop (“shedding”)
due to rearrangement of the pore structure; pore structures can be
made “invariant” by gluing, coating or sintering the particles making up
the filter medium). However, more factors, such as the flow, fluid, type
and size distribution of the test dust and the dust concentration, have
an influence on the rating of a filter.
Dirt holding An important parameter used for separators without continuous
capacity removal of the dirt.
and
For separators suffering pressure drop increase due to fouling with dirt
potential
particles, the actual dirt holding capacity is the amount of dirt the
dirt holding
separator has collected at the maximum pressure drop increase across
volume
the separator that is recommended to trigger the change-out of the filter
elements. This amount can be expressed both as mass and as volume
and is expected to be a function of the flow and the properties of the
separated dirt such as dry bulk density and permeability. The potential
dirt holding volume is the maximum net volume (volume corrected for
the volume used by internals) in the separator available for holding dirt.
The volume based actual dirt holding capacity is always smaller than
the potential dirt holding volume.
 PTS 31.27.21.10
February 2010
Page 7

Filtration The surface area of the filter medium used for filtration in parallel flow,
area independently of the porosity of the medium and the number of stacked
medium layers.
For multilayer pleated filters the unfolded surface area of a single layer
should be used although a part of this area may be inaccessible for
flow. In that case an effective filtration area can be defined as the
filtration area accessible for fluid flow, but this may be difficult to
quantify.
Besides the filtration area, the open area is used for slotted and wire
screen filter surfaces as the smallest cross sectional area accessible
for flow. It is often expressed as a percentage of the filtration area.
Filtration The time needed for the pressure drop to increase (at given flux rate,
time fluid viscosity, dirt characteristics and dirt concentration) from the clean
pressure drop to the maximum pressure drop recommended for
initialising cleaning of the filter or replacement of the filter elements.

1.4 SYMBOLS AND ABBREVIATIONS

Unless stated otherwise, all symbols used in this PTS are expressed in the units given
below:
D internal diameter of vessel or large pipe m
(DHC) dirt holding capacity kg or m3
e filter efficiency -
F mass flow rate kg/s
K friction loss factor -
∆P pressure difference Pa
Re Reynolds number -
t time s
v velocity m/s
η dynamic viscosity Pa.s
ρ density kg/m3

Subscript:
b bulk

Superscript:
* potential

1.5 CROSS-REFERENCES
Where cross-references are made, the number of the section or sub-section referred to is
shown in brackets. All documents referenced in this PTS are listed in (12.).
 PTS 31.27.21.10
February 2010
Page 8
1.6 SUMMARY OF MAIN CHANGES SINCE LAST EDITION
This PTS is a revision of the PTS of the same number dated MARCH 2001; the following
are the main, non –editorial changes
Old Section New Section Subject
2. 2. paragraph 1 Included Hydrocracking to plants that use
filters
2. 2. paragraph 2 Included reducing catalyst deactivation rate as
one of the benefits of installing feed filter
3. 3. Added 3 more measures to be considered
before considering adding feed filter.
4.2.3 4.2.3 Include Additional method for Toluene
Insolubles analysis: ASTM D-4072
5.1 5.1 Additional filter type added. To clarify between
liquid assisted backwash and gas assisted
backwash
5.1 5.1 Table 1. Added performance of gas assisted
backwash to the comparison table
6.2 6.2 Include the statement “This will be performed
with support of vendor”, when specifying solids
loading and particle size distribution.
6.2 6.2 Include materials specification as an item to be
defined by the contractor based on applicable
Materials selection standards PTS 30.10.02.11
and PTS 30.10.02.13
6.5 6.5 Included PTS 80.45.10.12 and PTS
80.45.10.11 as additional standards.
Updated PTS 80.45.10.12 title to : Design of
Pressure Relief, Flare and Vent Systems
6.5 6.5 Added : gas assisted backwash (i.e.backflush)
under vessel requirement filter housing
specification
12. 12. Added reference
PTS 30.10.02.13: Non-Metallic Materials
Selection and Application
PTS 80.45.10.12: Emergency Depressuring
and Sectionalizing
PTS 80.45.10.11: Overpressure and
underpressure-Prevention and Protection
 PTS 31.27.21.10
February 2010
Page 9
2. PURPOSE OF FEED FILTERS

Intermediate oil products in a refinery are often contaminated with various types of solid
particles (dirt or dust) such as sand, clay, salts, coke, corrosion products, surfactants, oil
oxidation products and oil polymerization products, which may foul up equipment and
require cleaning and thus cause early expensive stops of equipment or a whole processing
unit. Hydroprocessing and Hydrocracking plants use filters in the liquid feed (“feed filters”)
to remove the harmful dirt. Normally cartridge or candle types of filters are used, but
sometimes other types as well such as deep beds of granules.
The purpose of a feed filter is normally to extend the unit’s run length to the life of the
catalyst by reducing the fouling rate of the reactor and/or catalyst deactivation rate or to
extend the standing times of equipment such as coalescers and heat exchangers. The filter
should be allowed to pass as many particles as can be accumulated and passed through
the reactor until the next catalyst change-out. A feed filter is not necessary if the dirt holding
capacity of the reactors is large enough to hold all the retained particles until the next
scheduled shut down. It is also pointless to install absolute filters to trap particles that are
so small that they will pass through all equipment. Hence, the filter should remove only the
harmful fraction of the dirt. Absolute filtration of sub-micron particles is costly and usually
not necessary. Fouling of evaporators by sub-micron particles can usually be reduced more
efficiently by changing the design of the evaporator than by removing the sub-micron
particles from the feed with a filter.
Translating the purpose of the feed filter into specifications requires the knowledge of the
grade efficiency characteristics of the equipment to be protected (coalescer, heat-
exchanger, reactor), the grade efficiency characteristics of the filter and the solids loading.
This should not be left to the filter Vendor, since he has no knowledge of the grade
efficiency characteristics of the reactor, which is dependent on other fouling abatement
measures (graded layers, top bed filters, etc.).
 PTS 31.27.21.10
February 2010
Page 10
3. FEED FILTER SELECTION - OVERVIEW

Before considering application of a feed filter, the following measures should be

considered:
• Reducing the dirt contamination of the feed by improved upstream process control and
housekeeping;
• Achieving a better distribution of the dirt over the reactor internals and the length of the
catalyst bed.
• A use of graded materials with high void volume where possible such as Hydrotreaters
reactors top. This includes use of improved and updated proprietary distributor for
homogenous distribution of the feed
• If cracked feedstocks are used, storage of feedstocks should be under nitrogen blanket.
This includes limiting the storage time for the cracked feedstocks as much as possible.
• Use of antioxidants and anti-fouling additives, provided these are compatible to
downstream system.

Optimum selection of feed filters should take the following into account:

- process conditions

- dirt characteristics
- required performance efficiency
- purchase cost
- installation cost
- operating costs (e.g. cartridge disposal and replacement)
- depreciation

Each of these steps is discussed in more detail in subsequent sections.

 PTS 31.27.21.10
February 2010
Page 11
4. DETERMINING THE FILTER RATING

4.1 FILTER RATING PROCEDURE

Determining the rating of the feed filter requires the following parameters to be known:
• dirt concentration in the feed;
• particle size distribution of the dirt;
• dirt characteristics such as morphology and stickiness;
• dirt produced between filter and reactor;
• particle separation characteristics (grade efficiency) of the reactor and its internals;
• dirt holding capacity of the reactor and its internals;
• dirt holding capacity to be spared for by-passing;
• particle separation characteristics (grade efficiency) of the filter.
If these parameters are known, then the required filter rating can be derived as follows:

This procedure relates the filter requirements to the dirt separation/holding characteristics
of the reactor (similar procedures apply for protecting the coalescer or the heat exchangers
if these are limited by fouling). A reactor with a high dirt holding capacity will require a
smaller filter than a reactor with low dirt holding capacity.
Each of the mentioned parameters and its possible measurement is discussed in
subsequent sections. Some parameters may not be easily determined for various reasons
(e.g. if sampling is not possible, the feed is not available or the plant is not yet built), in
which case the best possible estimates shall be made, e.g. from experience with existing
similar units.
 PTS 31.27.21.10
February 2010
Page 12

4.2 DIRT CONCENTRATION IN THE FEED

4.2.1 General
The particle size distribution and the solids concentration of the feed need to be known in
order to predict the run length of the reactor, if defined by fouling, to determine whether an
external feed filter is needed and the rate at which solids have to be removed by the feed
filter.
Several methods can be followed to characterise the solids in the feed. The best way is to
use on-line sample filtration at the temperature of the feed. This may be difficult for high
feed temperatures (> 70 °C), in which case sample filtration may have to be performed in a
laboratory. However, cooling and storage may precipitate additional solids, which might not
completely redissolve when the liquid sample is heated to the original process temperature
and this consequently results in analysis errors. Reasons for instability are the presence of
asphaltenes or wax which may precipitate during cooling and the presence of reactive
species which may polymerise (form gum) during storage. Appropriate sampling points
should be upstream and downstream of the (potential) location of the feed filter, located
and constructed so that representative samples can be taken (see PTS 32.31.50.10).
Points of attention specific for solids are described in ISO 3170 and ISO 3171. A local
sampling procedure should be prepared to cover these points.

4.2.2 On-line sample filtration

On-line sample filtration can be performed at ambient temperatures using portable filtration
equipment. The Ssafcon sampler shown in Figure 1 has been developed in collaboration
with Shell and is safer than the “hose end” sampling method of ASTM D 2276 for
measuring the solids concentration in aviation fuel. It is easily plugged into a permanent
sampling tray via self-sealing connectors. About 5 litres are passed through a 0.8 µm
(Millipore) membrane filter as measured by an in-built volume meter. The filter is
subsequently removed for gravimetric or colorimetric analysis. It can also be used for
subsequent microscopic analysis for estimating the particle size distribution, and for
elemental analysis. The Ssafcon sampler can be used at temperatures up to 70 ºC.
The Ssafcon sampler provides only a spot measurement. A pilot type of filter should be
used to collect larger samples of dirt for more accurate determination of the time averaged
particle concentration and particle size distribution. The larger amounts of dirt collected in a
pilot filter can also be used to determine the (compacted) bulk density of the dirt in liquid.
Larger amounts of dirt can be used as test dust for laboratory filtration experiments.
On-line sample filtration devices for temperatures beyond 70 ºC are not commercially
available.

4.2.3 On-line liquid sampling off-line sample analysis

If on-line filtration of the sampled feed is not possible then a sample of the feed needs to be
tapped and analysed in the laboratory according to one or more of the various standard
procedures available. The solids can be determined using standard methods such as the
Hot Filtration Test (ISO 10307-1) and toluene insolubles (method AMS-581 MOD and/or
ASTM D-4072) for residue feedstocks. Increased amounts of liquid may be needed to
increase the accuracy of the methods. ISO prescribes the use of a glass fibre or equivalent
filter. Membrane types of filter are more suitable if the filters are to be subsequently used
for microscopic analyses to estimate the particle size distribution, but can only be used if
the viscosity of the fluid and filtration temperature allows this. It is essential that sample
bottles are clean and that the sampling line is adequately flushed prior to taking the
samples (see ISO 3170).
 PTS 31.27.21.10
February 2010
Page 13

Figure 1 Stanhope-Seta Ssafcon sampler for on-line sample filtration at ambient

temperatures
 PTS 31.27.21.10
February 2010
Page 14

4.3 PARTICLE SIZE DISTRIBUTION OF THE DIRT

There are many ways to measure the particle size distribution of a sample of particles, two
of which are:
(1) The small analytical test filter used in the methods described in (4.2) to separate
particles from the liquid can be examined with an optical microscope or a scanning
electron microscope (SEM) to count the number of particles in different size classes.
The number distribution is subsequently translated into a volume distribution assuming
a certain shape of the particles. A large number of particles must be counted to obtain
sufficient accuracy especially for the sizes larger than the volume-averaged size.
Normally the shape of a sphere can be assumed, but the microscopic analysis may
suggest other shapes. A membrane type of test filter should be used rather than the
glass fibre test filter prescribed in some procedures, since a membrane type of filter
has a better defined surface. Automatic counting may be suitable, using techniques
such as image analysis.

(2) A liquid sample may be examined with a laser diffraction instrument to yield directly the
volume-based particle size distribution. Normally, the particle concentrations in
hydroprocessing feedstocks are too low for this technique to be applied directly to a
sample of the feed. Concentrating can be carried out by first filtering the oil over a
membrane-type filter (approx. 0.8 µm), washing the solids on the membrane with a
solvent and subsequently flushing the particles from the filter into a small amount of
filtered solvent, dispersing the particles in an ultra-sound bath, possibly removing
disruptively large particles (e.g. from the scraped filter paper) and finally measuring the
particle size distribution with laser diffraction. A blank determination should be carried
out as well. Concentrating should be performed by a factor of about 50. Hence, to
obtain a liquid sample of 100 ml for liquid sample laser diffraction analysis, start with a
5 litre feed sample. The advantage of the liquid sample method is that results can be
produced quickly. The disadvantage is that no information is obtained about the
morphology of the particles or the state of agglomeration. Fibre-shaped particles will
produce significantly deviating size distributions with laser diffraction.
 PTS 31.27.21.10
February 2010
Page 15

4.4 DIRT CHARACTERISTICS SUCH AS MORPHOLOGY, STICKINESS AND ELEMENT

COMPOSITION
The morphology of the particles is important for the choice of the filter medium and the filter
rating. Slotted screens are less suitable for fibre-shaped particles unless the slot size is
smaller than the fibre diameter. The morphology can easily be determined by means of a
microscope as described in (4.3). Asphaltenes, wax and gum are known to foul up metal
filter elements, and these may have to be cleaned off by soaking the elements in hot
solvents or other cleaning methods. Their presence and concentration should be indicated
by the plant operator. The elemental composition of the dirt may help the operator to
identify the source of the dirt contamination of the feed and may help the filter designer to
classify the filtration problem and use experiences from similar filtration problems.
 PTS 31.27.21.10
February 2010
Page 16

4.5 DIRT PRODUCED BETWEEN FILTER AND REACTOR

Feed filters are preferably installed just upstream of the high pressure feed pump and heat
exchangers, as this is where the feed is still liquid without gas (hydrogen) being present.
The temperature of the feed to be filtered should be high enough for the viscosity to be
approximately as low as that of water. Dirt particles generated between the filter and the
reactor will add to the fouling of the reactor. The operator of the hydroprocessing plant
should take measures to ensure that these amounts are small compared to the dirt holding
capacity of the reactor. Small amounts are automatically taken into account by the
calibration of the dirt holding capacity of the reactor as described in (4.7).
 PTS 31.27.21.10
February 2010
Page 17

4.6 PARTICLE SEPARATION CHARACTERISTICS (GRADE EFFICIENCY) OF THE

REACTOR AND ITS INTERNALS
The reactor with its catalyst bed(s) and internals shall be considered as a series of solids
separators each with its own grade efficiency and dirt holding capacity. Settling trays and
basket filters can be installed to separate dirt from the feed before it enters the catalyst bed.
The top layers of the catalyst bed are often graded to promote an even distribution of the
dirt over the graded layers. The overall grade efficiency of the reactor can be obtained by
calculating for each particle size range the fraction of dirt particles passing through all the
subsequent solids separation stages. The overall grade efficiency of the reactor will depend
on the particle sizes in the graded layers and the catalyst bed, the thickness of these layers
and the type of reactor (vapour phase, liquid phase or trickle flow). For the purpose of filter
selection and filter rating, a constant average (over the run length) grade efficiency of the
reactor should be taken, based on actual conditions. It does not have to be accurately
known if the dirt holding capacity of the reactor can be calibrated with the actual reactor run
length (see 4.7).
The grade efficiency of the separate stages can be obtained from (model) experiments
separately for each stage. Alternatively, the grade efficiencies can be obtained by dirt
sampling each stage and calculating the grade efficiencies from the mass balances over
each separation stage for each particle size class using the cumulative amount of dirt
supplied with the feed. The dirt obtained from sampling should be used to determine the
(compacted) bulk solids density in the liquid and may also be used as actual test dust in
fouling experiments in the laboratory.
 PTS 31.27.21.10
February 2010
Page 18

4.7 DIRT HOLDING CAPACITY OF THE REACTOR AND ITS INTERNALS

The dirt holding capacity of the reactor can be calibrated with the actual reactor run length
as follows:

Alternatively, the dirt holding capacity can be determined by measuring the amount of dirt
collected in the reactor after the maximum pressure drop has been reached. This is the
actual dirt holding capacity. Part of this dirt could be due to attrition of the catalyst or
corrosion between the feed sampling point and the reactor, e.g. corrosion in the heat
exchanger. This amount shall be subtracted from the actual dirt holding capacity together
with the amount to be spared for by-passing the filter to produce the dirt holding capacity
available when feed passes through the filter in normal operation.
It should be realized that the dirt holding capacity of the reactor depends on the particle
size distribution of the dirt flowing to the reactor. The presence of a feed filter may require a
different grading scheme for uniform separation of the dirt passing through the filter and
retained by the reactor.
 PTS 31.27.21.10
February 2010
Page 19

4.8 DIRT HOLDING CAPACITY TO BE SPARED FOR BY-PASSING

A small fraction of the dirt holding capacity (say 5 %) should be spared for collecting the dirt
when the filter is by-passed. This will result theoretically in a larger size of filter, but since
filters are often offered in standard sizes, the oversize of the filter may be sufficient to allow
a certain amount of filter by-passing without the penalty of a reduced run length.
 PTS 31.27.21.10
February 2010
Page 20

4.9 PARTICLE SEPARATION CHARACTERISTICS (GRADE EFFICIENCY) OF THE FILTER

Each filter type has its own grade efficiency characteristics (e.g. see Figure 2) and therefore
the same exercise should be done for each different type of grade efficiency curve.
However this would be very time consuming and it is therefore recommended that a
standard average grade efficiency curve be used to derive the filter specification. The
prospective filter suppliers may then convert the filter specification to an equivalent one
corresponding with the grade efficiency curve for the type of filter proposed. This will lead to
approximately the correct choice of filter rating. The cost of the filter elements with the
approximately correct filter rating and the cost of the corresponding filter housing can then
be used in the subsequent cost comparison of the various types of filters proposed, and a
filter selection can then be made. The procedure outlined above may then be repeated for
the grade efficiency characteristics of the actually chosen filter types and the choice of filter
rating may then be reviewed.
 PTS 31.27.21.10
February 2010
Page 21

100
Source:
(1) Shucoscy, Chem Eng., Jan. 18, 1988, p72-77
90 (2) SLS Vol. 8 Filter Media, Figure 2.2 (1980)

80 Symbols are curve

identifiers

70
Removal efficiency [%]

60

50

40

30 Wound cartridge (1)

Pleated cartridge (1)
20 Felt (2)
Wire gauze (2)
10

0
0 5 10 15 20 25 30 35 40
Particle size [µm]

Figure 2 Grade efficiencies of various (clean) filter media

 PTS 31.27.21.10
February 2010
Page 22

5. TYPES OF FEED FILTERS AND SELECTION OF MANUFACTURERS

5.1 TYPES OF FEED FILTERS

The following five types of filters are traditionally applied in hydroprocessing units to remove
dirt particles:
• Disposable cartridge filters;
• Μanually cleanable disposable cartridge filters;
• Automatic backwash (liquid assisted) filters;
• Automatic backflush (gas assisted) filters;
• Manually cleaned candle filters;
• Filters packed with polypropylene wool/blankets/mattresses.
Granular beds (porous extrudates/pellets, activated carbon) may be useful in certain
applications, but are not generally recommended.
Other types of filters are used in similar applications such as (automatic) mechanically
cleaned filters (wiped screens) to remove fines from residual fuel oils and electric filters to
remove solids from FCC slurry oils or magnetic filters, but their application in
hydroprocessing units is not known and therefore not recommended for such service.
An example of a disposable cartridge filter is given in Figure 3 with two radial fin cartridges
stacked on top of each other within a single filter housing. Filtration is from outside to
inside. A large filtration area is achieved with radial fins of wire screen fixed to a central
cylinder. The fins are covered with a filter cloth of an appropriate material and appropriate
pore size. After the recommended maximum pressure drop has been reached, the filter
may be backwashed or the cartridges may be removed from the housing and cleaned
manually. After a number of reuses the cartridges may be disposed and replaced or may be
provided with a new filter cloth. This type of filter has been applied by Knitmesh as prefilter
for protection of coalescers in hydroprocessing units, PTS 31.22.05.12. Types with axial
fins are also commercially available (e.g. terraced filter).
Figure 4 shows an example of a filter with a bundle of filter elements each fixed to a division
sheet (also called septum plate, tube sheet). Filtration is from outside to inside. A large
filtration area is achieved with a large number of small diameter filter elements. The filter
elements may be cheap cotton wound cartridges or pleated paper, which are discarded
after the maximum pressure drop has been reached. Alternatively, the filter elements may
be expensive porous metal candles made up from spirally wound wedge wire or from one
or more layers of (pleated) wire screen, which are backwashed or manually cleaned
whenever the maximum pressure drop has been reached. Many different types of
cartridges and candles are available from a large number of Manufacturers.
Automatic backwash filters are used in residue hydroprocessing units to remove relatively
high solids loadings and are available in a few different types (wedge wire, sintered woven
wire) from only a few Manufacturers. There are also automatic types using synthetic filter
media for relative low temperatures.
The automatic backflush filters, which are sometimes referred to as gas assisted backwash
filters, can handle higher solids loading and require less backwash volume compared to the
backwash filter. These filters are normally installed in duplex configuration.
Manually-cleaned metal candle filters are mainly used in hydrocracking units operating at a
temperature too high for disposable cartridge types and these are available from many
Manufacturers.
The fifth type (polypropylene wool/mattresses) has been used as coalescer prefilters in
hydrotreater units but with mixed experiences.
The suitability of the above-mentioned types depends on various factors and can be
summarised in the Table 1.
 PTS 31.27.21.10
February 2010
Page 23

Table 1 Performance comparison of filter types

Filter type Temperature Solids loading Cleaning costs
Low high low high low high
Polypropylene wool + -- ++ + + -
packing
Disposable cartridges ++ -- ++ - + -
Manually cleanable ++ -- + + ++ --
disposable cartridges
Manually cleaned ++ ++ + + ++ --
metal candles
Automatic backwash ++ ++ + ++ + ++
filters
Automatic backwash ++ ++ + ++ + -
filters

Initial screening generally indicates that several types are possible, and the final selection
depends on technical details and costs.
The considerations in selecting a feed filter are which type of filter is the best for given
conditions, what is the best lay-out (spare vessel or not, number of parallel banks, parallel
lines with different filter ratings) and what is the economically optimum size of the filter.
These matters are dealt with in the following sections.

5.2 SELECTION OF VENDORS

There are many filter Manufacturers who can deliver a high quality filter but few or none can
deliver a guaranteed right type and size of filter tailored to the actual conditions of the
hydroprocessing unit. Many Vendors deliver only a limited range of filter types and
choosing such a Vendor means choosing a type of filter. Hence, the best approach is to
select the most likely types of filters and then choose correspondingly approved or
recommended Vendors.
 PTS 31.27.21.10
February 2010
Page 24

Vent

Sketches are
approximate

Cross section A-A

A A

Contaminated Filtered fluid

fluid inlet outlet

Drain

Figure 3 Example of a filter with two stacked radial fin or pleated cartridges
 PTS 31.27.21.10
February 2010
Page 25

Figure 4 Example of a multiple candle or cartridge filter

 PTS 31.27.21.10
February 2010
Page 26

6. DATA AND INFORMATION TO BE SUBMITTED

6.1 RESPONSIBILITIES
It is the responsibility of the Vendor to propose a filter to meet the filter specifications under
the specified process conditions. It is the responsibility of the Principal to specify the
process conditions and filter specifications.

6.2 SPECIFICATION OF PROCESS CONDITIONS

All conditions and requirements that influence the choice of the filter type, the size of the
filter, the choice of construction materials, the process control, the instrumentation and the
safety and environmental aspects should be specified.
The Principal shall define the following data:
• Normal throughput and maximum possible throughput of the filter. The filter should be
designed to operate economically and optimally at normal throughput and should still be
operable at maximum throughput although perhaps at the cost of lower efficiency.
• The operating process temperature and pressure of the filter to be taken for selection
and sizing of the filter.
• Maximum operating temperature and pressures, which may occur occasionally or as the
result of a possible but undesired event. This should include pressures that may result
from upstream pumps or from opening valves connecting pipes or vessels at high
pressure.
• The fluid properties in terms of a composition of the feed blend of distillate fractions and
the density and viscosity of the feed blend at the normal process temperature.
• For the solids loading, the nature and concentration of the solids and the particle size
distribution during normal and upset conditions (upstream of the plant) together with the
possible frequency and duration of such upsets. This will be performed with the support
of the filter vendor.
• Pressure drop available for the feed filter.
The Contractor should define:
• The design temperature and pressure according to PTS 01.00.01.30. A maximum
operating temperature of at least 150 °C and a maximum operating pressure of at least
3.7 bar (ga) should be specified to allow the filter housing to be purged with medium
pressure steam.
• The materials specifications for the filter and filter housing according to PTS 30.10.02.11
and PTS 30.10.02.13; as applicable.
• The target time-average grade efficiency characteristic of the filter together with the
corresponding amount of solids to be removed during normal operation. The grade
efficiency characteristic may be provided as a graph, but preferably as a table
containing at least the particle sizes to be separated at separation efficiencies of 99.5,
95 and 50 %. The target grade efficiency characteristic shall be replaced by the actual
grade efficiency characteristic once a choice of filter type has been made, and the filter
rating procedure shall be reviewed accordingly.

6.3 DATA AND INFORMATION TO BE SUBMITTED WITH THE TENDER

The prospective Vendors shall submit sufficient drawings and detailed information with their
tenders to enable a full evaluation of the filter, including as a minimum:
- production and engineering documents schedule;
- general arrangement sketch (lay-out drawing), showing:
 PTS 31.27.21.10
February 2010
Page 27

· number, size and location of the filter housings;

· pipework with valves and flow restrictions connecting the filter housings
together and with the pipework of the plant;
· number, size and location of the filter elements;
· facilities for removing or inspecting the filter elements;
· facilities for draining, flushing and purging each filter housing;
· instrumentation and control;
- parts lists, with references to materials specifications;
- weight of the completely assembled filter (including filter media and accessories);
- manufacturing specifications and quality checks such as those described in
ISO 2942, ISO 2943 and ISO 3723, or equivalent;
- pressure drop / flow characteristics of the filter unit without filter elements at relevant
viscosities;
- for each filter rating: pressure drop / flow characteristics of the clean separate filter
elements at relevant viscosities, e.g. determined according to ISO 3968 or
equivalent.
- the clean pressure drop over the filter unit at actual design conditions with filter
elements installed, e.g. determined according to ISO 3968 or equivalent.
- the burst / collapse pressure drop of the filter elements according to ISO 2941 and
the (terminal) pressure drop (as measured, hence across filter element plus housing)
recommended for backwashing or replacement of the filter elements;
- procedure for replacing the filter elements;
- for (manual) backwashed filters: the backwash procedure, the backwash pressure,
backwash duration, the expected time between successive cleanings (= filtration time
- time needed for backwashing parallel filter housings/banks), and the amount of
backwash liquid to be used relative to the liquid volume in the filter housing;
- for automatically backwashed filters: the characteristics of the valves determining the
initial backwash impulse and the control mechanism to maintain a high driving force
during the backwash;
- external cleaning procedure for reusable filter elements;
- the dimensions of the filter elements and the number of filter elements and the
filtration area;
- for each filter rating proposed: the dirt holding capacity at (standard) test conditions
and description of test conditions, e.g. as determined with ISO 4572 or equivalent;
- expected dirt holding capacity under actual process conditions;
- grade efficiency data (filter rating) of the filter elements at standard test conditions,
e.g. as determined using ISO 4572 or equivalent;
- expected grade efficiency of the filter elements under actual test conditions;
- list of spare parts according to PTS 70.10.90.11. including replacement parts needed
for external cleaning or for replacement of the filter elements and their costs;
- an assessment of the filter performance at maximum flow and during upset
conditions;
- prices of the filter as proposed and prices of filters of the same type, but of one step
smaller size and one step larger size (for determination of the economically optimum
size);
 PTS 31.27.21.10
February 2010
Page 28

- list of similar filter deliveries that the Vendor considers relevant for the present
proposal;
- list of references that the Vendor considers relevant to illustrate experience.

6.4 DATA AND INFORMATION TO BE SUBMITTED AFTER AWARD OF ORDER

After award of purchase order a complete set of engineering documents shall be submitted
for approval of the Purchaser, before any shop or construction work is commenced. The
number and type of documents shall be specified on the requisition for engineering
documents (PTS 40.10.01.93).
All documents shall be marked in the right-hand bottom corner with the Principal's order
and item number, together with the Manufacturer's references.
After approval, certified final documents/manuals and reproducible drawings of the
equipment shall be submitted to the Principal. The required number and type of final
documents shall be specified in the requisition for engineering documents
(PTS 40.10.01.93).
All information, including the manuals for operation and maintenance, shall be clear and not
open to misinterpretation and shall apply specifically to the equipment supplied.
For drawing presentation the "American Projection Method", according to ISO 128, should
be followed. The characteristics of the projection method used shall be shown on the
drawing.

6.5 VESSEL REQUIREMENTS

All types of feed filters considered in this PTS use pressure vessels to contain the filter
medium. These vessels shall be manufactured according to PTS 31.22.10.32, or
PTS 31.22.20.31.
Metallic materials should be selected from the standards mentioned in PTS 30.10.02.11.
If the filter is of the disposable cartridge type, then it shall be possible to block out the filter
in order to clean it and replace the cartridges. Prior to opening, the vessel shall be drained
and purged with an inert gas (preferably steam). It shall be possible to open the vessel,
replace the cartridges and close the vessel safely and quickly to minimize the time that the
filter is out of operation.
Design of Pressure Relief, Flare and Vent Systems shall be in accordance with PTS
80.45.10.10
Prevention and protection against overpressure and under-pressure shall be in accordance
with PTS 80.45.10.11.
Emergency depressuring shall be in accordance with PTS 80.45.10.12.
The filter housings of backwash and backflush filters should be provided with a conical
bottom with a slope of 60° from the horizontal to prevent solids from settling onto the
bottom of the filter housing and subsequently being re-entrained when the filter is put back
into operation.
Feed inlet and outlet nozzles shall be positioned to enhance uniform distribution of the feed
over the filter elements and to prevent distortion of the filter elements. To prevent distortion,
installation of vortex breakers such as shown in standard drawing S 10.010 should be
considered.

6.6 PIPING AND VALVES

All piping shall be in accordance with PTS 31.38.01.11
The layout of the filter shall be such that single filter housings or groups of filter housings
can be isolated for cleaning, while the remainder of the filters continue operation. Valves
 PTS 31.27.21.10
February 2010
Page 29

used in automatic backwash filters for frequent switching between operation and cleaning
mode should be of a type suitable for frequent switching without excessive wear.
The piping upstream and downstream of the filter should each have at least two sampling
point connections for isokinetic sampling of which the upstream one shall be according to
ISO 3171 for isokinetic sampling.

6.7 INSTRUMENTATION AND CONTROL

The filter should be provided with at least:
• a connection to a remote data control and processing unit. It shall be possible to set a
pressure drop triggering cleaning of the filter;
• an instrument for measuring the differential pressure between filter inlet and filter outlet.
This instrument should be capable of transmitting the signal to a remote data processing
unit. An audible alarm shall be activated and reset when the pre-set terminating
pressure drop across the filter has been reached;
• an instrument for measurement and local display of the filter inlet pressure;
• an instrument for measurement and local display of the filter outlet pressure.
Instruments and associated auxiliaries shall be in accordance with PTS 32.31.09.31. Any
control valves shall be in accordance with PTS 32.36.01.17.
Automated backwash types of filters shall have the following features:
• initiate backwashes automatically (initiated either by time or by a pre-set terminating
pressure drop across the filter);
• allow regular interchange of the first housing or the first bank to be cleaned to prevent
unequal wear;
• initiate a normal or operator-instructed backwash from the central data processing unit
independently of the pressure drop across the filter.
• exclude one or more filter housings from the automatic backwash procedure if filter
housings can be individually isolated from the process, or exclude one or more banks of
filter housings from the backwash sequence if only banks can be individually isolated
from the process;
• set a time for the duration of the backwash procedure and forward purge procedure, if
applied.
 PTS 31.27.21.10
February 2010
Page 30

7. LOCAL COST DATA TO BE SPECIFIED BY PRINCIPAL

7.1 GENERAL
The total cost of a filter depends on local conditions such as how contaminated backwash
liquid can be dealt with and waste cartridges disposed of. The Principal shall define
approximate values for the cost factors to be included in the evaluation given below.

7.2 FIXED COSTS

7.2.1 Engineering
Costs of plant change associated with installation of the filter such as preparation of the
site, modification of engineering diagrams, plot plans, hiring of engineering contractors, etc.

7.2.2 Filter disposal

The costs associated with its disposal at the end of its life such as administration,
dismantling, transport, site clearance, refund of shred value.

7.3 FILTER LIFE TIME AND DEPRECIATION

The time that the filter is expected to be in service should be the same as the expected life
time of the plant. The accounting life time is taken to be much shorter than the expected
real life time and should be defined by the Principal.
Annual depreciation is calculated as total fixed costs divided by the accounting life time of
the filter.

7.4 VARIABLE COSTS

• Cleaning the filter:
- local man-hour cost for cleaning.
• Disposal of cartridges and other filter waste material:
- cost of transport plus administration of cartridges or other waste materials to be
disposed of;
- cost of landfill, incineration, or work-up as a function of weight or volume.
• Cleaning-out costs of reactor and other equipment. Cost factors associated with
cleaning of the equipment to be protected by the feed filter such as costs for cleaning of
the reactor:
- hours to be accounted as production loss due to early terminating of the
production run, possibly corrected for hours saved in the next scheduled
shutdown;
- nominal throughput at the time of the production shutdown;
- production margin;
- man-hours needed for preparation and supervision of the shutdown;
- man-hours needed for opening/closing of the reactor and other equipment and
removal/sieving/replacing of the catalyst;
- cost per cleaning man-hour;
- cost per staff man-hour;
- utility consumption and costs;
- materials consumption and costs;
 PTS 31.27.21.10
February 2010
Page 31

- catalyst losses as a result of extra catalyst handling (separated by sieving) or due

to early replacement.
• Similar as above for cleaning of other equipment such as coalescers, heat exchangers
or other equipment.
• Maintenance costs:
- costs for (regular) inspections;
- replacement of parts (gaskets, valve seats, etc.);
• Cost of backwash liquid
• Slops processing costs.
The Principal should specify one or more possibilities for processing/disposing of the
contaminated backwash liquid and the associated costs including costs of separation of
the solids if needed.
Examples of reuse are:
- Injection into the feed of an appropriate tray of the distillation section of the plant
using the feed filter. This tray or its auxiliaries must have the capacity to deal with
the solids.
- Injection into the feed of the FCC. This may require the hydroprocessed product
to be used as backwash liquid. However, it should be realized that particles in the
feed enhance erosion of the FCC feed nozzles.
- Injection to an appropriate tray of the crude distiller. This tray or its auxiliaries
must have the capacity to deal with the solids.
- Add to the fuel pool of the refinery.
• Costs for slops disposal (if disposed of as waste)
• Spare parts:
- cost of storage and administration of spare parts.
• Utilities:
- cost of steam, nitrogen, instrument air, electricity, vacuum, draining, flaring.
 PTS 31.27.21.10
February 2010
Page 32

8. EVALUATION OF VENDOR PROPOSALS ON BASIS OF MINIMUM TOTAL COST

8.1 RECEIVED DATA

The proposals should be checked for completeness and consistency and supplementary
information sought where necessary.
 PTS 31.27.21.10
February 2010
Page 33

8.2 DESIGN PHILOSOPHY OF VENDORS

If the technology of the Vendor is considered reliable then the pressure drops and filtration
times he provides may be used to estimate the costs in (8.4).
 PTS 31.27.21.10
February 2010
Page 34

8.3 DESIGN FACTORS

8.3.1 General
Design factors such as pressure drop and filtration times shall be derived from experiments
with the same filter type. This should be done on basis of the considerations given below.

8.3.2 Pressure drop

The filter pressure drop between filter inlet and outlet is the sum of several components:
(∆P)filter = (∆P)piping + (∆P)housing + (∆P)element + (∆P)medium + (∆P)dirt
The pressure drop across the piping, housing and element will be constant for constant flow
during the filtration cycle. This is also the case for the pressure drop across the medium if
surface filtration removes all particles, in which case the pressure drop across the dirt layer
on top of the medium will increase with time. It is also possible that no cake layer is built up
on top of the medium and that some of the dirt is collected inside the pores of the filter
medium (depth filtration). In that case the pressure drop across the filter medium will
increase with time. Surface and depth filtration can also occur simultaneously.
To summarising, the time dependence of the pressure drop can be written as:
(∆P)filter(t) = (∆P)clean + (∆P)dirt(t)
The (∆P)clean of disposable cartridge filters will be constant, but the initial clean pressure
drop of automatically or manually backwashed filters may gradually increase after each
filtration cycle. Flow through the medium is normally laminar and described by Darcy’s law,
which means that the pressure drop is proportional to the viscosity of the fluid and
proportional to the velocity, but not dependent on the density of the fluid. The flow through
the housing and connecting piping is usually turbulent in commercial filters, which means
that the pressure drop is proportional to the density of the fluid and proportional to the
square of the velocity. The influence of the viscosity of the fluid is much less and only
present through the influence of the Reynolds number on the flow resistance (friction)
factor. The pressure drop across the filter elements will be determined not only by the filter
medium but also by the flow through the central core and the flow expansions/contractions
at the inlet and outlet of the filter elements. The pressure drop of the clean filter should be
specified by the Vendor as a function of the flow and the fluid viscosity both for a single
filter element and for the complete assembled filter. Such measurements should be carried
out according to ISO 3968 or equivalent. Subsequently, the results shall be translated to
the actual conditions of the filter. This can be done by interpreting the results of the
standard tests in the form of:

(∆P)clean =
The first term on the right-hand side of the equation expresses the dependence of the
pressure drop on those parts of the filter with laminar flow (filter medium) and the second
term expresses the dependence of the pressure drop on those parts of the filter with
turbulent flow (bends, valves, cross-section transitions). B and K are constants to be fitted
to the results of the tests. K is normally a function of the Reynolds number (Re):
Re = ρ v D / η
where D is a diameter characteristic of the pressure drop of the filter, e.g. the diameter
determining most of the pressure drop of the clean filter, such as the inner diameter of the
cartridges or the diameter of the feed inlet nozzle depending on the type of filter. This
dependence on the Reynolds number can be determined by variation of the viscosity of the
fluid. K can be considered as a combined friction loss factor, being the sum of the friction
loss factors of the various flow restrictions in the filter. However, the pressure drop/flow
characteristics of the filter usually allow only one friction loss factor to be correlated as a
function of the Reynolds number.
 PTS 31.27.21.10
February 2010
Page 35

With B known and K known as a function of the Reynolds number, the clean pressure drop
of the filter at actual conditions can be calculated.

8.3.3 Filtration time

The filtration time depends on the type of filter, the characteristics of the dirt and fluid, the
terminal pressure drop and the flow. Their interrelation can only be obtained confidently by
testing under representative process conditions using the same filter, fluid, dirt, dirt
concentration and process conditions. If the results of such tests are not readily available,
then the scale-up shall be carried out from less representative tests and the results from
such tests should be weighted correspondingly lower.
The filtration time should be based at least on tests such as described in ISO 4572 or
equivalent standard, using the filter type proposed and a test dust with approximately the
same particle size distribution as specified for the requested filter. Such a test provides the
grade efficiency of the filter medium, the dirt holding capacity of the filter and the pressure
increase as a function of time. If the pressure drop increases gradually, proportionally or
exponentially with time until the terminating pressure drop has been reached, then the
filtration is said to be limited by the mechanism of either depth or surface filtration. If the
pressure drop increases gradually at first, proportionally or exponentially with time, and
suddenly increases sharply before the terminating pressure drop has been reached, then
the filtration is said to be determined by the potential dirt holding capacity of the filter. In that
case nearly all space available for accumulation of dirt has been utilised and only small
diameter channels are left for the fluid to pass through the filter.
Cartridges with pleated filter media have an optimum number of pleats when the pressure
drop starts to rise suddenly shortly after the terminating pressure drop has been reached.
Otherwise, cartridges with fewer pleats should be selected.
To translate the standard tests to actual conditions the following rules can be applied as a
guide:
Potential dirt holding capacity:

The potential dirt holding capacity of the test is the cumulative amount of test dust collected
by the filter, hence the difference between the cumulative amount of test dust supplied and
the cumulative amount of test dust passing through the filter. It follows from the above
equation that the potential dirt holding capacities of highly fluffy agglomerated sub-micron
particles can be much smaller than those determined with test dusts consisting of
nonporous single particles.
It is stressed that the potential dirt holding capacity is different from the dirt holding capacity
as determined with the multipass test, being the amount of dirt accumulated when the
terminating pressure drop has been reached. This is always lower than the potential dirt
holding capacity. The potential dirt holding capacity of cartridges can be estimated from the
dimensions of the cartridges and the volume occupied by the filter medium and supports.
This is a good check on dirt holding capacity values quoted by cartridge suppliers. The
maximum filtration time can be calculated from the potential dirt holding capacity, the filter
grade efficiency and the particle size distribution of the dirt, as follows:

where e actual is the efficiency of the filter for the actual dirt, which shall be calculated from
the grade efficiency of the filter as determined with the test dust and the particle size
distribution of the actual dirt.
For filtration not limited by the potential dirt holding capacity, the filtration time is:
 PTS 31.27.21.10
February 2010
Page 36

• inversely proportional to the power n (1.5 < n < 2) of the flux, i.e. the flux based on the
actual effective filtration area (the effective filtration area might be smaller than the
installed filtration area e.g. if wedge wire filter candles are used). The value of 2 applies
for theoretical cake filtration, the value of 1.5 applies for depth filtration;
• approximately proportional to the pressure drop across the cumulative amount of dirt
collected. The compressibility of the dirt and/or medium may cause a lower than
proportional dependency. The pressure drop to be used should be calculated by
subtracting the pressure drop of the clean parts from the terminating pressure drop;
• inversely proportional to the viscosity of the fluid passing the cake, hence either the
clear fluid or the fluid containing fines passing the cake. This dependence on viscosity
means that the filter should be operated at sufficiently high temperature;
• inversely proportional to the fraction of solids concentration to be removed;
• inversely proportional to the flow resistance of the dirt and the filter medium. The flow
resistance of the filter medium plus dirt is expected to vary over a wide range according
to the solids, the particle size distribution, and the shape and stickiness of the particles
possibly present in hydrocarbon feeds. As a first approximation the flow resistance
should be taken as inversely proportional to the square of the Kozeny average size of
the retained particles.
These rules should be applied to translate an experimental filtration time measured with the
same type of filter and same filter rating to the actual filter. If parallel filter housings or
parallel filter banks are proposed, then the translation should include increased fluxes
resulting from parallel banks or parallel filter housings temporarily out of filtration service for
cleaning, maintenance or other reasons. The translated filtration time shall be called the
theoretical filtration time. The actual expected filtration time is the smaller out of the
theoretical filtration time and the limiting filtration time.
 PTS 31.27.21.10
February 2010
Page 37

8.4 OPTIMUM SIZE

Filters should be compared on the basis of minimum total annual filtration costs for removal
of the required amount of dirt particles (see Figure 5). This includes both fixed and variable
costs. Fixed costs should include all costs involved to install the filter and to get it
functioning, including acquisition, engineering, purchase, transport, installation and
commissioning. Variable costs should include replacement materials, cleaning labour,
maintenance, production losses, penalties for not meeting specifications and disposal costs
for backwash liquid and spent filter media. Some of these costs depend on local conditions
and shall be provided by the Principal (see section 7).

8.4.1 Fixed costs

The fixed costs include the following items:
• Acquisition
• Commissioning
• Engineering costs for plant change associated with installation of the filter
such as preparation of the site, modification of engineering
diagrams, plot plans, hiring of engineering contractors, as
estimated according to the complexity of the filter installation.
• Filter disposal the costs associated with disposal of the filter at the end of its
life (such as administration, dismantling, transport, site
clearance, refund of scrap value)
• Installation costs assembling;
tankage, transport lines and pumps to be installed for transport
and storage of backwash liquid and slops are charged as fixed
costs.
• Insurance cost of insurance during transport from port of departure to the
site if the purchase price is FOB
• Pre-engineering costs of studying feasibility of critical design problems prior to
opting for a particular design.
• Project supervision
• Purchase price price of equipment, auxiliaries and first set of spares, FOB or
CIF as specified by the Vendor.
• Transport cost of transport from port of departure to the site if the
purchase price is FOB.
Some of the cost factors may be insignificant for selection of a filter type. Most costs are
dependent on filter size, such as the purchase costs and the filtration time. Fixed costs
such as purchase costs will generally increase with filter size. Variable costs, such as
replacement costs for filter cartridges or work-up costs of backwash liquid, will generally
decrease with size. The size of the filter at minimum total (fixed plus variable) cost is the
optimum size of the filter. The optimum size of each type of filter for a particular application
should be determined by calculating the total costs for a number of filter sizes with the aid
of the data provided by the Principal (7) and Vendor (6) and of the best available scale
rules or those described in (8.3). For each type of filter, the Vendor's definition of filter size
should be adopted (filter size is not critically important, as filters are selected on the basis of
cost). Different types of filters will have different minimum costs, depending on process
conditions, local cost factors and rate of solids removal.

8.4.2 Depreciation
Fixed costs should be annualized over life time as specified by the Principal.

8.4.3 Variable costs

Variable costs include the following items:
 PTS 31.27.21.10
February 2010
Page 38

• filter cleaning (proportional to cleaning frequency);

• disposal of cartridges and other filter waste materials (proportional to cleaning
frequency);
• Penalty for filter by-passing: any resultant reduction of the run time of the reactor or
coalescer or other equipment to be protected shall be factored into the cost equation as
follows:
{(%reduction - %threshold)/100} * filter benefit [USD/year]
%threshold = %reduction without consequences of earlier termination of the run,
because the filter has actually been over designed and is separating more than really
required.
• Penalty for filter underperformance
If a filter type is not expected to achieve the specified performance then the reduction in
anticipated filter benefit should be forced into the variable costs.
• Penalty for production loss
If part of the filtered feed is used for backwashing the filter without increasing the feed
rate to the filter, then part of the capacity of the plant is not being utilised and should be
factored in as a penalty equal to the production loss multiplied by the margin.
If part of the plant product is used for backwashing, then the value of the product should
be factored into the equation.
• Maintenance costs
costs of (regular) inspections;
labour costs for replacement of parts (gaskets, valve seats, etc.).
• Slops processing costs:
The amount of backwash liquid needed for backwashed filters is proportional to the
cleaning frequency and the size of the filter and also dependent on the type of filter. The
costs for both obtaining the liquid and processing or disposing of the waste backwash
liquid shall be accounted for. It may be possible to offset this by using the slops in
another part of the plant or refinery.
• Spare parts: cost of replacement of disposable filter cartridges are proportional to the
filter cleaning frequency.
• Utilities: steam may be needed to purge the filters and air is needed for pneumatically
actuated valves. Most utility costs will be proportional to the filter cleaning frequency.
 PTS 31.27.21.10
February 2010
Page 39

8.5 MINIMUM COST AS FUNCTION OF RATE OF SOLIDS REMOVAL

The higher the required rate of solids removal, the greater the optimal size and minimum
costs of a filter. This increase will be different for different types of filters. Disposable
cartridge types of filters will have relative low total costs at low rates of removal, but total
costs will increase relatively steeply with increasing rates of removal. Automatic backwash
types of filters have relatively high total costs at low rates of removal, but total costs will
increase relatively slowly with increasing rates of removal. Hence the two lines of minimum
total costs versus rate of removal will intersect at a certain rate of removal. Below this
intersection point disposable types are cheaper and beyond this point automatic backwash
filters are cheaper. Manually cleanable filters may be cheaper in a region in between the
disposable cartridges and the automatic backwashed filters. The boundaries between the
regions in which a particular type of filter is cheaper depend on process conditions, plant
throughput and local cost factors.
In order to evaluate the effect of solids concentration in the feed, the whole procedure as
outlined in (4) and in (8.4) shall be redone for at least two other rates of solids removal. If
the same filter target, i.e. a specified run length of the reactor, is adhered to, then the
amount of solids in a certain size range that can be allowed to pass through to the reactors
is a constant maximum value independent of the solids supply rate. Hence, a higher solids
concentration in the feed with the same particle size distribution will require a lower micron
rating of the filter and a larger rate of removal. Generally, the particle size distribution will
also vary and the procedure for the grade efficiencies should be followed to determine the
micron rating and the rate of solids removal. It should be realised that, theoretically, the rate
of solids removal may even be lower with higher solids concentrations if the particle size
distribution in the feed has simultaneously shifted towards lower particle sizes. The
minimum costs at rates of solids removal other than those used in (8.4) can be evaluated
relatively easily if spreadsheets are used.
Plotting the minimum costs of all proposed filters as a function of the rates of removal in the
actual range facilitates the selection of the right type of filter. Figure 6 is an example of a
case study showing that, for the conditions studied, a cleanable disposable filter is cheaper
than a non-reusable cotton wound cartridge filter for the whole range of solids
concentrations. An automatic backwashed filter is cheaper than the cotton wound cartridge
filter if more than about 3.5 kg/day solids have to be removed and cheaper than the
cleanable disposable filter if more than about 14 kg/day solids have to be removed. Each
data point in Figure 6 has been determined with the procedure illustrated in Figure 5.
 PTS 31.27.21.10
February 2010
Page 40

Figure 5 Costs of a 4000 t/d gas oil feed filter

300,000
Cotton wound cartridges
Automatic backwashed filter
Minimal annual costs, USD/year

250,000
Cleanable disposable cartridges

200,000 Symbols are curve identifiers

150,000

100,000

50,000

0
0 2 4 6 8 10 12 14 16 18
Solids removed, kg/day

Figure 6 Example of minimal filtration costs of different filter systems (data are
from an actual case study)
 PTS 31.27.21.10
February 2010
Page 41

9. FILTER SELECTION

The final choice of the filter should be based not only on the minimum cost, but also on a
strength / weakness analysis considering selection criteria such as:
• reliability of the filter based on practical experiences;
• reputation and experience of the Vendor;
• short and long-term assurance of deliveries of spares and services;
• versatility of the filter;
• ease of operation of the filter with regard to factors other than those expressible as
costs;
• uncertainties in the evaluation procedure (which could be made clear by a sensitivity
analysis).
 PTS 31.27.21.10
February 2010
Page 42

10. COST / BENEFIT / RISK ANALYSIS

A cost / benefit / risk analysis and calculation of the pay-out time should make clear
whether a feed filter should indeed be purchased.
The costs of the filter are defined by the evaluations in (8).
The benefits of the filter are the costs saved by eliminating the need for cleaning the
equipment protected by the feed filter. Such costs include:
• loss of margin due to early termination of the production run, possibly corrected for
hours saved in the next scheduled shutdown;
• labour costs associated with opening/closing the reactor and other equipment and
removal/sieving/replacing of the catalyst;
• catalyst losses as a result of the extra catalyst handling (separated by sieving) or due to
early replacement.
 PTS 31.27.21.10
February 2010
Page 43

11. SITE TESTS AND COMMISSIONING

After the filter has been installed at site it shall be checked whether the installation has
been executed in accordance with the Manufacturer's instructions. The site commissioning
and performance test run shall consist of a test run under the design conditions for at least
one filtration run. The following observations shall be recorded and compared with figures
given by the Manufacturer:
- pressure drop as function of time and flow;
- solids concentrations in the feed before and after the filter at approximately the same
time and preferably repeated several times during a filtration cycle;
- filter efficiencies calculated from the solids concentrations upstream and downstream of
the filter;
- duration of a filtration run.
 PTS 31.27.21.10
February 2010
Page 44

12. REFERENCES

In this PTS reference is made to the following publications:

NOTE Unless specifically designated by date, the latest edition of each publication shall be used, together
with any amendments/supplements/revisions thereto.

PETRONAS STANDARDS
Index to PTS publications and standard PTS 00.00.05.05
specifications
Definition and determination of temperature, PTS 01.00.01.30.
pressure and toxicity levels
Non-Metallic Materials-Selection and Application PTS 30.10.02.13
Metallic materials - selected standards PTS 30.10.02.11
Liquid/liquid and gas/liquid/liquid (three-phase) PTS 31.22.05.12
separators - Type selection and design rules
Pressure vessels (amendments/supplements to PD PTS 31.22.10.32
5500)
Pressure vessels (amendments/supplements to PTS 31.22.20.31
ASME Section VIII, Division 1 and Division 2)
Piping – General requirements PTS 31.38.01.11
Instrumentation for equipment packages PTS 32.31.09.31
On-line process stream analysis - Sample take-off PTS 32.31.50.10
and transportation
Control valves - selection, sizing and specification PTS 32.36.01.17
Electrical trace heating PTS 33.68.30.32
Requisition for engineering documents PTS 40.10.01.93
(data/requisition sheet)
Note: The latest revision of data/requisition sheets are contained in PTS
binder 30.10.01.10

Spare parts PTS 70.10.90.11

Overpressure and underpressure – Prevention and PTS 80.45.10.11
protection
Emergency depressuring and sectionalizing PTS 80.45.10.12
Design of Pressure Relief, Flare and Vent systems PTS 80.45.10.10
STANDARD DRAWINGS
Vortex breakers S 10.010
METHODS SERIES
Toluene insolubles AMS-581 MOD

AMERICAN STANDARDS
Standard test method for particulate contaminant in ASTM D 2276
aviation fuel by line sampling)

Issued by:
American Society for Testing and Materials
100 Bar Harbor Drive
West Conshohocken, PA. 19428-2959
USA
 PTS 31.27.21.10
February 2010
Page 45

INTERNATIONAL STANDARDS
Technical drawings - general principles of ISO 128
presentation
Hydraulic fluid power - Filter elements – Verification ISO 2941
of collapse/burst resistance
Hydraulic fluid power - Filter elements – ISO 2942
Determination of fabrication integrity
Hydraulic fluid power - Filter elements – Verification ISO 2943
of material compatibility with fluids
Petroleum Liquids - Manual sampling ISO 3170
Petroleum Liquids - Automatic pipeline sampling ISO 3171
Hydraulic fluid power - Filter elements - Method for ISO 3723
end load tests.
Hydraulic fluid power - Filters - Evaluation of ISO 3968
pressure drop versus flow characteristics
Hydraulic fluid power - Filters - Multi-pass method ISO 4572
for evaluating filtration performance
Petroleum products - Total sediment in residual fuel ISO 10307-1
oils - Part 1: Determination by hot filtration

Issued by:
International Organisation for Standardization
1, Rue de Varembé
CH-1211 Geneva 20
Switzerland.

Copies can also be obtained from national standards

organizations.

Last page of this PTS