Integration of Reciprocating Metering Pumps Into A Process

GBH Enterprises, Ltd.

Engineering Design Guide:
GBHE-EDG-MAC-1013

Integration of Reciprocating Metering
Pumps into a Process

Information contained in this publication or as otherwise supplied to Users is
believed to be accurate and correct at time of going to press, and is given in
good faith, but it is for the User to satisfy itself of the suitability of the information
for its own particular purpose. GBHE gives no warranty as to the fitness of this
information for any particular purpose and any implied warranty or condition
(statutory or otherwise) is excluded except to the extent that exclusion is
prevented by law. GBHE accepts no liability resulting from reliance on this
information. Freedom under Patent, Copyright and Designs cannot be assumed.

Refinery Process Stream Purification Refinery Process Catalysts Troubleshooting Refinery Process Catalyst Start-Up / Shutdown
Activation Reduction In-situ Ex-situ Sulfiding Specializing in Refinery Process Catalyst Performance Evaluation Heat & Mass
Balance Analysis Catalyst Remaining Life Determination Catalyst Deactivation Assessment Catalyst Performance
Characterization Refining & Gas Processing & Petrochemical Industries Catalysts / Process Technology - Hydrogen Catalysts /
Process Technology – Ammonia Catalyst Process Technology - Methanol Catalysts / process Technology – Petrochemicals
Specializing in the Development & Commercialization of New Technology in the Refining & Petrochemical Industries
Web Site: www.GBHEnterprises.com

Engineering Design Guide:

Integration of
Reciprocating Metering
Pumps into a Process

CONTENTS

1

SCOPE

2

PRELIMINARY CHOICE OF PUMP

SECTION A - TYPE/FLOW/PRESSURE/SPEED RATING
Al

Pumping Pressure

A2

Pump Flowrate and Capacity

A3

Guide to Pump Speed & Type

A4

Metering Criteria

A5

Pressure Pulsation

A6

Over Delivery

SECTION B - INLET CONDITIONS
B1

Calculation of Basic NPSH

B2

Correction for Frictional Head

B3

Correction for Acceleration Head

B4

Calculation of Available NPSH

B5

Corrections to NPSH for Fluid Properties

B6

Estimation of NPSH Required


B7

Priming

SECTION C - POWER RATING
C1

Pump Efficiency

C2

Calculation of Absorbed Power

C3

Determination of Driver Power Rating

SECTION D - CASING PRESSURE RATING
Dl

Calculation of Maximum Discharge Pressure

D2

Discharge Pressure Relief Rating

D3

Calculation of Pump Head Outlet Losses

D4

Casing Hydrostatic Test Pressure

APPENDICES
A

RELIABILITY CLASSIFICATION

FIGURES
A3.1
A3.3
B5.1
B6
C1.1

ESTIMATE OF CRANK SPEED
SELECTION OF PUMPING HEAD TYPE
ESTIMATE OF VISCOSITY OF FINE SUSPENSIONS
ESTIMATE OF NPSH REQUIRED
GRAPH - VOLUMETRIC EFFICIENCY VS MEAN DIFFERENTIAL
PRESSURE

DOCUMENTS REFERRED TO IN THIS ENGINEERING DESIGN GUIDE


1

SCOPE

This Engineering Design Guide deals with:
(a)

The specification of the pump duty for enquiries to be sent to pump
vendors,

(b)

The estimation of the characteristics and requirements of the pumps in
order to provide preliminary information for design work by others.

It applies to pumps in Group 3 as defined in GBHE-EDP-MAC-33 Series but may
be used for general purpose pumps in Group 4.
It applies to metering duties where metering accuracy is important. Metering
pumps are often used where metering accuracy is unimportant but where this
type of pump is readily available for low flows and high heads. This Engineering
Design Guide may be used for such duties with the special metering
requirements discounted.


2


2.1

Preliminary Choice of Number of Pumps

Reciprocating metering pumps fall into reliability classes 4, 5 and 6 as defined in
Appendix A. The standard arrangement is one running pump rated at 100% duty,
with one identical pump as the spare either installed or available as a
replacement.
Other considerations are:
(a)

Systems that cannot tolerate an interval of no-flow upon changeover to a
standby pump require two running pumps each rated at 50% duty together
with a third identical pump as standby. Check whether the average flow
obtained from an integration of the flow over a time period including the
Change over should be kept constant.

(b)

Systems requiring approximately constant pressure should include an
accumulator, but then accurate short-term metering is impossible.

2.2

Estimate of Pump Flow

Normal flow is taken as the largest process flow required for operation at the
rated plant daily output.
Metering duties are seldom known accurately. Unless the process duty
specifically limits the maximum flow, take 110% of this normal flow through one
pump or divided between pumps as considered in Clause 2.1, to obtain the pump
flow rate Qo in 1/s.
2.3

Estimate of Pump Mean Differential Pressure
Estimate the maximum mean differential pressure Po as Section a Clause
A1 for the normal flow.


2.4

Consideration of Flow Metering Requirements

2.4.1 Small Flows

(a)

Qo x ρ
Po

< 0.8 and Q0 < 0.8

Where

If

ρ

=

liquid density kg/l

Accurate flow metering
Use a reciprocating variable stroke metering pump.
Take the flowrange as 0 - 100% pump flow, unless otherwise specified by
process considerations.

(b)

Rough flow metering
If metering accuracy worse than + 20% pump flowrate is acceptable:
(1)

Use a reciprocating variable stroke metering pump
Or

(2)

A rotary pump with variable speed drive if

Ѵ > 16 Po
Qo

Where Ѵ = kinematic viscosity of liquid, cSt

See GBHE-EDG-MAC-1012.
or


(3)

A variable capacity sliding vane pump if liquid is clean and

Po < 8
See GBHE-EDG-MAC-1012

(c)

Batch metering
Use a reciprocating metering pump with stroke counting or time controlling
device.

2.4.2 Large Flows

(a) If Q o x ρ

> 0.8

Po
Use a centrifugal pump with a standard flowmetering system.
See GBHE-EDG-MAC-1014.

(b) If Q o x ρ

< 0.8 and Q0 > 0.8 l/s

Po
Use a multiplex plunger reciprocating pump or use a rotodynamic pump
with a flowmetering system.

2.5

Multiple Stream Applications

The combination of variable stroke metering pump heads for different streams on
to a common drive unit should be considered for all multiple stream applications.
Proportioning should be effected by stroke adjustment of individual heads with
total flow controlled by a variable speed drive.


SECTION A
TYPE/FLOW/PRESSURE/SPEED RATING


A1

Pumping Pressure

The pulsating flow from each stroke of a metering pump results in pulsating
pressures at pump inlet and discharge which fluctuate with plunger position.
Consequently the differential head and pressure have not the same significance
as for centrifugal pumps.
A1.1 Calculation of Static Pressures
All systems can be reduced to a transfer of liquid between two reservoirs or to
circulation to and from one reservoir.
The inlet static pressure (Psi) is given by:-

Psi =

PI +

ρ x hsi

bar g

10.2

The discharge static pressure (Psd ) is given by:

Psd =

PD +

ρ x hsd

bar g

10.2

The differential static pressure (P s ) is given by:

Ps =

Psd - Psi

bar

Where

hsd

=

static liquid head from pump centre to delivery vessel
free liquid surface or constant pressure point

hsd

=

m

static liquid head from pump centre to inlet vessel
Free liquid surface

m

(Note that h sd and h si are negative when the appropriate liquid
surface lies below the pump centre)


PD

=

gas pressure at free surface of the liquid in the delivery
Vessel, or the pressure at the constant pressure point in the
delivery system
bar g

PI

=

gas pressure at the free surface of the liquid in the inlet
Vessel

bar g

(Note that PD and PI are negative for vessels under vacuum)
Ρ

=

liquid density

kg/l

Check the proposed layout to see if the piping system rises above the delivery
vessel to form a syphon. If it does then calculate the maximum discharge static
pressure Psd with the static head hsd taken to the point of highest elevation of
the piping system. There should be no syphons on the inlet piping system.

A1.2

Calculation of Duty Pressures
The duty pressures should be taken as:

PI

=

Psi

-

ρ x hri

bar g

10.2

Pd

=

Psd

-

ρ x hrd

bar g

10.2

Po =

Pd - Pi

bar

Where:
PI
Pd
hri
hrd
Po

=
=
=
=
=

mean inlet pressure
mean discharge pressure
frictional head loss in inlet piping
system at mean flow
frictional head loss in delivery piping
system at mean flow
mean differential pressure

bar g
bar g
m
m
bar


Check the maximum and minimum values of PI, Pd and Po which the pump
should still operate. Self consistent conditions are given on the Process Data
Sheet.
A.2

Pump Flowrate and Capacity

Take the pump flow, Q o from Clause 2.2. Note that as metering pumps are
manufactured in a limited number of plunger diameters, stroke lengths and
speeds, the actual pump capacity, Qp is likely to exceed the pump flow, Q o.
A3

Guide to Pump Speed and Type

A3.l

Rotational speed

The rotational speed of metering pumps is limited by the following
Considerations:
(a)

Suppression of gas release and NPSH availability

(b)

Unidirectional loading of motion work

(c)

Avoidance of pressure wave enhancement

(d)

Viscous losses affecting volumetric efficiency

As a first estimate take crank rotational speed, N, as 1 rev/s for low viscosity
liquids and the lower speeds shown in Fig.A3.l for pumps handling viscous
liquids.
Enter section A3.3 with Q e where:

Qe

=

Qo
N


FIGURE A3. 1

ESTIMATE OF CRANK SPEED


A.3.2

Pump Arrangement

The preliminary choice is a simplex pump without pulsation dampers.

A.3.3

Pumping Head Type

Pump heads of the plunger type should be used for all duties except where:
(a)

Slight gland leakage cannot be tolerated - sterility, toxicity, corrosion,
expense of loss.

(b)

Process fluid contains abrasive particles or precipitates in gland and an
alternative flushing medium compatible with the pumped fluid and
acceptable to the process is not available. Note that the gland flush
reduces metering accuracy.

(c)

Q e > 0.25 l/s and a lubricating medium compatible with the pumped fluid
and acceptable to the process is not available.

(d)

For high pressure and low flows where loss of metering accuracy due to
gland leakage and hold-up is unacceptable. As a guide take loss of
flowrate repeatability due to glands as:

N.Z
10 x Q o

Where

2/3

log Pd

%

Z = number of plungers (take as 1 for first estimate)


A3.3.2

Diaphragm Pump
Diaphragm pump heads should be used for other duties.

(a)

Mechanically actuated metal diaphragm
Use if Q e < 0.02 l/s
And

Pd < 5 bar g

And leakage on diaphragm failure is acceptable.
Note:
PTFE or elastomer diaphragms and bellows are available for flows to 0.8 l/s but
are only suitable where short diaphragm lives (typically 1000 hours) and low
accuracy ( + 10%) are acceptable.
(b)

Double diaphragm
Use if cross contamination of process and hydraulic fluids on diaphragm
failure is unacceptable.

(c)

Hydraulically actuated single diaphragm
Use on duties other than the above. If the process fluid contains large
solids or abrasive particles a design without a support plate in the process
fluid should be specified. This may be either a single or double diaphragm
head.

A3.4 Confirm Selection
Confirm selection of pump type and speed by considering:
(a)

Metering criteria (Clause A4)

(b)

Pressure pulsation (Clause A5)

(c)

Over-delivery (Clause A6)

(d)

Inlet conditions (Section B)


(e)

Engineering Design Guide GBHE-EDG.MAC.16 Series for the liquid
concerned.

(f)

Catalogues of preferred vendors

(g)

Plant experience from similar installations from Engineering Design Guide
GBHE-EDG.MAC.52 Series


Fig A3.3

SELECTION OF PUMPING HEAD TYPE.


A4

Metering Criteria

The accuracy of metering is not solely determined by the accuracy of
pump manufacture but also by factors connected with the installation such
as fluid, pressures, temperatures, contamination and electrical supply
frequency and voltage. Although extremely high accuracies (errors of the
order of ~ 0.05%) can be achieved under favorable conditions, accuracy
in practical installations should be considered no better than + 2.5%. To
achieve this accuracy the fluid has to be clean and operating conditions
stable and constant. Slurries or widely ranging conditions, particularly with
high pressures/low flows, preclude accurate metering.

A4.1 Flowrate Repeatability
Flowrate repeatability is defined as the ability of a metering pump to consistently
deliver a measured volume after repeated resetting of the capacity adjustment. It
should be considered no better than + 2% of the pump capacity Q at any flow
setting and to achieve this accuracy, operating conditions have to be stable and
constant.
If good flowrate repeatability is required over the 0-20% range of pump capacity
the stroke frequency should be varied as well as the stroke length.
Wide ranges of operating pressures, temperatures and fluid properties will result
in a considerably reduced repeatability. See Section C1.1 for the change in
volumetric efficiency with discharge pressure.

A4.2 Accuracy of Mean Flow Measurement
Although the flowrate corresponding to a given setting value may be repeatable,
the absolute value of this flowrate is subject to a further tolerance of the order of
0.5% of the pump capacity.

A4.3 Flowrate Continuity
The instantaneous flowrate from a metering pump may fluctuate markedly over
each crankshaft revolution. See Section B2 for prediction of the peak
instantaneous flow: mean flow. If duties require continuous delivery flow, install a
discharge pulsation damper or specify a multiplex pump.

For smoothing delivery flow with a damper, the delivery system should be
damped by resistance. Typically for a tolerance on flow of the same order as the
flowrate repeatability, ensure for low viscosity liquids:

hrd

>

5.1 x Pr 0.7
ρ

and hrd

>

1.3 Prd
ρ

The flow of viscous liquids is more readily damped by using multiplex pumps.
Dampers necessarily have a long time constant which limits the rate at which the
desired flow setting under automatic control can be changed and requires a
substantially constant discharge pressure. For a tolerance on flow continuity of
the same order as the flowrate repeatability this limit on rate of flow setting gives
a minimum of;

500
Z
times the period of one crankshaft revolution for full travel of the flow setting
(from 0 - 100%), where Z is the number of plungers.

A5

Pressure Pulsation

To avoid pulsation enhancement, the length of piping between the pump ports
and a large discontinuity such as a vessel should be less than 4% of the
fundamental wavelength (λ ) where:

Λ

=

C
N.Z

Where
C
N
Z

= Velocity of sound in fluid
= crank rotational speed
= number of plungers

m/s
rev/s


Then calculate the pressure pulsation at both inlet and discharge
of pump from:

δP

=

700 ρ N Q
Z

Ʃ

1
d2

Where
δP = Pressure pulsation (peak to peak)
L=
length of line to large discontinuity
D = diameter of line bore
D=
diameter of a section of line of length

bar
m
mm
l

Alternatively, if the system presents no discontinuity in the piping close to the
pump, then calculate the pressure pulsation at inlet and discharge of pump from:

δP

=

kρCQ
D2

Where
K

=

40 for simplex pump
20 for duplex pump
3 for triplex pump
1 for quintuplex pump

For lines of varying diameter, take D as the minimum line diameter excluding
short lengths (< 0.01 λ).
If

δP > 0.01 P

and

δP > 0.04 P0.7

where

P = mean pressure at pump inlet & discharge

bar a

then the pressure pulsations are unacceptable and piping vibration is likely.


Reduce the level by:
(a)

Increasing line diameter

(b)

Specifying pulsation damper

(c)

Specifying multi-headed pump

A6

Over Delivery

The pump discharge pressure should always exceed the inlet pressure to avoid
flow through the pump which is not controlled by the plunger.
For simplex pumps without pulsation dampers ensure that:

Ps > δP i
2
And

Ps > δP d
2
For other arrangements ensure that:

Ps

>

δP I + δP d
2
2

where

δP i =

pressure pulsation at pump inlet (from Clause A5)

bar

δP d =

pressure pulsation at pump discharge (from Clause A5 )

bar

Pressure retaining valves or spring loaded pump valves to develop an artificial
pressure difference should be avoided.

On batch systems, drips from the delivery line should be prevented by installing
valves at the outlet or by arranging for the outlet to be at the highest part of the
pipeline and of small bore.
SECTION B INLET CONDITIONS


BI

Calculation of Basic NPSH

Basic NPSH =

hsi

+ 10.2
ρ

B
1000

+ P1 – Pv

m

where

hsi = static liquid head over pump inlet

m

This is measured from the lowest liquid level to the pump centre and is
negative when the liquid level is below the pump centre. The lowest level
is not necessarily the extreme value but is the one at which the pump is
still operational.
B = minimum barometric pressure at pump location

mbar

Use 0.94 of the mean barometric pressure from local meteorological data
to allow for weather changes.

P1 = minimum working value of the gas pressure on the free liquid surface in
the inlet vessel. This is negative for vessels under vacuum.

bar g

Pv = vapor pressure of the liquid at the maximum operating temperature bar a
ρ=

B2

liquid density at the maximum operating temperature.
For slurries the density of the MIXTURE should be used.

kg/I

Correction for Frictional Head

Calculate the maximum frictional head loss in system, Hri at peak instantaneous
flow rather than at mean flow and include the maximum operating loss in the
the inlet filter when this is dirty.


Obtain the peak instantaneous flow from the following table:
Configuration
Quintuplex
Triplex
Duplex
Simplex

B3

Peak flow / mean flow
1.03
1.1
1.6
3.2

Correction for Acceleration Head

Calculate Hai the head required to accelerate the fluid in the inlet piping during
each pulsation cycle from:

Hai =

δP1
ρ

m

where

δP1 = pressure pulsation at pump inlet (from Clause A5)
B4

bar

Calculation of Available NPSH

Calculate the available NPSH at the pump suction flange from:
(a)

For simplex and duplex pumps
Available NPSH = Basic NPSH - (Hai2 + Hri2) ½ - 1

(b)

m

For triplex and quintuplex pumps
Available NPSH = Basic NPSH - (Hai + Hri) - 1

m


B5

Correction to NPSH for Fluid Properties

The following corrections should NOT be quoted to the pump supplier. Use only
for discussions on layout and system design.

B5.1 Suspended Solids Content
The behavior with slurries is notoriously difficult to predict; the only reliable
test is to run the pump on the plant. Pastes, gels and other non-Newtonian
fluids should always be the subject of experimental tests. For suspensions
of hard crystalline solids, use the following empirical rules as a guide:
(a)

Very fine suspensions where all the particles are less than 3 µm are
best treated as viscous liquids - see Fig B5.1.

(c) For slurries having a range of particle sizes take:

NPSH slurry =

1
1 +X

x NPSH carrier

where
x = fractional content by volume of solid particles and the NPSH for the
carrier liquid has already been corrected for viscosity.


FIG.B5.1

ESTIMATE OF VISCOSITY OF FINE SUSPENSIONS

B5.2 Viscosity
An increase in viscosity increases the NPSH required. For high values of
viscosity the losses across valves are approximately proportional to the liquid
velocity. Consequently the NPSH required is directly proportional to viscosity at
constant speed and flow.
However, it is normal practice to reduce pump speed as viscosity increases (see
Clause A).I) and so, in effect, limit the variation of NPSH required with viscosity
change.


B5.3 Gas Content
For accurate metering, the liquid should contain no free gas.
As pressure reduces across the pump inlet valves on the inlet stroke the
dissolved gas may be released from solution if the available NPSH is too small. It
is essential to avoid gas release for metering duties.

B6

Estimation of NPSH Required

The NPSH required by a pump is dictated by its detail design and its operating
speed.
As a guide for layout system and pump specification, a preliminary estimate of
the NPSH required by plunger & mechanically actuated diaphragm pumps may
be obtained from Fig B6 where:
For single valve:

A=

24 Ѵ Q
Z dv3

+ 5x105

ρ Q2
Z2 dv4

For double valves:

A=

Where

80 Ѵ Q
Z dv3

+ 15x105 ρ Q2
Z2 dv4

d v nominal valve size.

mm

Take as next size up from (5000 Q ) ½
Z


Select spring loading for duties where metering accuracy is important. Take the
value of spring pressure to be approximately 0.2 Ph without spring loading.
For hydraulically actuated diaphragm pumps, estimate as above and take
additional diaphragm and support plate losses as for a single valve without spring
loading.
The actual NPSH required should be checked after selection of the pump
supplier.
FIGURE B6 ESTIMATE OF NPSH REQUIRED


B7

Priming

For accurate metering, pumps should have a flooded inlet and the gland subject
to a pressure above atmospheric pressure to avoid inward air leakage.
The pressure at the pump inlet should be greater than the pressure differential
required to open both inlet and discharge valves against their spring forces.
When the pump does not auto-prime through the process piping but is primed by
operation of a vent valve, check the line diagram to ensure that:
(a)

The vent is shown

(b)

The vent is piped to a safe disposal point when the pump is handling a
hazardous liquid

(c)

When the inlet vessel is working under vacuum, the vent is piped to the
gas space of the inlet vessel

(d)

When the vent is piped, then a visible indicator is fitted in the vent line
adjacent to the pump.

With liquefied gases at sub-ambient temperature, the discharge pressure at startup should be less than the pressure reached if a cylinder fills with vapor and the
plunger acts as a compressor. As the clearance volume is likely to be of the
same order as the swept volume, the maximum absolute pressure developed in
the cylinder is unlikely to exceed twice the absolute inlet pressure. The discharge
pressure should be less than this value to clear the cylinder of vapor. Note that
vapor compression does not result in condensation because it is nearly adiabatic
and hence yields superheated vapor.
When the pump stops, it eventually loses prime. Consequently inlet lines should
be lagged.
This condition normally requires a special bypass provision. Check the line
diagram for inclusion.


SECTION C POWER RATING


C1

Pump Efficiency

C1.1 Volumetric Efficiency

Volumetric efficiency,

ᵑv

= delivered volume. 100

%

swept volume

For preliminary estimation of the volumetric efficiency refer to Fig C1.1. The
upper line at each band refers to large pumps (Q/N ~ 0.8), and the lower line to
small pumps (Q/N ~ 10- 3). This figure is based on a liquid such as water with a
bulk compressibility of 2.1 x 103 MN/m2.


C1.2 Mechanical Efficiency
Mechanical efficiency,

ᵑm

=

theoretical hydraulic power. 100 %
power input to pump

For preliminary estimation of mechanical efficiency take:

ᵑm

=

100
1.3 + 5
Qo Po

%

Note that as reduction gears are usually integral with the pump, their losses are
included in the above estimate.

C1.3 Overall Efficiency
The overall pump efficiency

ᵑ

is defined as

ᵑ = ᵑv ᵑm 10-2
C2

%

Calculation of Absorbed Power

E=

10 Q o P o

kw

ᵑ


C3

Determination of Driver Power Rating

Multiply the mean power requirement for the pump by the following additional
factors as appropriate:
belt drive

1.07

simplex pump

1.3

duplex pump

1.1

equipment and data tolerance

1.1

For multiple pump heads driven by one electric motor drive, sum the power
requirements for the individual feeds.
For an electric motor drive select the next highest power rating in kW from:

Consult Power Group for appropriate motor rating and type for variable speed
drives.
Note:
This power rating should be used only for a PRELIMINARY estimate of the
electrical distribution load. Check the required power rating and frame size after
selection of pump using the manufacturer’s data.
Standard motors have a starting torque of about 150% full load torque which is
adequate for reciprocating metering pumps.


SECTION D CASING PRESSURE RATING


D1

Calculation of Maximum Discharge Pressure

Calculate maximum discharge pressure, Pdmax' from:

Pdmax = PR + ρ m ( hsd + hrd )

bar g

10.2
Where

PR
ρm

=

set pressure of relief valve on delivery vessel

bar g

=

maximum liquid density

kg/l

maximum static head from pump to the free liquid surface
in the delivery vessel

m

maximum frictional head loss in the delivery pipe system
at the peak instantaneous flow (see Clause B2 for
peak flow : mean flow)

m

H sd =
H rd

=

Note: pressure pulsations due to acceleration reads are assumed sufficiently
small to be accommodated within relief valve setting.

D2

Discharge Pressure Relief Rating

A discharge pressure limiting device should be installed unless there is no
possibility of a restriction occurring in the discharge system between pump and
vessel. Protection by the motor overloads is not sufficient.
The conventional pressure limiting device is a relief valve in the discharge
pipework or in the hydraulic system of diaphragm pump heads. However, relief
valves tend to leak after they have operated and detection of a small leak is
difficult. Consequently for metering duties specify a bursting disc in the discharge
Pipework or, for a diaphragm pump, the hydraulic relief valve setting at least
125% of the maximum discharge pressure P dmax.

The relief valve should be sized for at least 110% of the pump displacement
capacity at the maximum pump speed. If long term operation against the relief
valve could induce a high temperature, a temperature trip should be fitted.

D3

CALCULATION OF PUMP HEAD OUTLET LOSSES

As a preliminary estimate take the head outlet losses, PH as twice the head inlet
loss Ph from Clause B6. Check after selection of pump using the manufacturers
data.

D4

Casing Hydrostatic Test Pressure

The hydrostatic test pressure for the pump read should exceed 150% of the
maximum attainable pressure, taken as the sum of the pressure setting of the
discharge pressure limiting device and the pump head outlet losses, PH.
Pumps handling hot liquid (above 175°C) require individual consideration taking
into account piping loads and reduction in material strength with increase in
temperature.
The rating of the inlet connection should be identical to that of the casing and
discharge connection.


APPENDIX A
RELIABILITY CLASSIFICATION
Installations having high availability are conveniently classified as follows:
Class 1
A Class 1 installation achieves high availability by having units of high intrinsic
reliability, and is characterized by:
(a)

The machine being a single unspared unit upon which the process stream
is wholly dependent.

(b)

The plant section having a single process stream with a long process
recovery time after a shutdown so that the loss of product owing to a
machine stoppage is large even though the shutdown is for a short time.

(c)

A capability of continuous operation within given process performance
tolerances over a period of more than three years, without enforced halts
for inspection or adjustment.

(d)

Component life expectancies exceeding 100,000 hours operation.

Class 2
As for Class 1 but where infrequent plant shutdowns of short duration are
acceptable because the process recovery time is short. Consequently the period
of continuous operation capability can be reduced and is taken as 4,000 hours
for this classification.

Class 3
As for Class 1 or 2 but with a machine performance deterioration during the
operating period accepted, or countered by adjustment of process conditions
or by other action on the part of the plant operators.


Class 4
A Class 4 installation achieves high availability by redundancy and is
characterized by having:
(a)

One or more machines operating with one or more standby machines at
instant readiness at all times to take over automatically upon malfunction
of a running machine.

(b)

Operating and standby machines designed specifically for their functions
so that they are not necessarily identical.

(c)

Component life expectancies exceeding 25,000 hours operation.

Class 5
A Class 5 installation follows the Class 4 redundancy concept and is
characterized by:
(a)

One or more machines operating with one or more identical machines
installed as spares to take over the process duty at the discretion of the
plant operators.

(b)

One or more machines operating in plant sections where product storage
is sufficient to give the plant operators adequate time to assess the
malfunction and take remedial action. Alternatively, plant sections where a
single machine stoppage does not cause a disproportionately large
process upset. Spare machines are not installed.

Class 6
Machines intended for batch or intermittent duty.
Where high demand availability is essential, such machines lie within Class 4.


DOCUMENTS REFERRED TO IN THIS ENGINEERING DESIGN GUIDE
This Engineering Design Guide makes reference to the following documents:
ENGINEERING DESIGN GUIDES
GBHE-EDG-MAC-1012

Integration of Rotary Positive Displacement
Pumps into a Process (referred to in Clause
2.4.1).

GBHE-EDG-MAC-1014

Integration of Special Purpose Centrifugal
Pumps into a Process (referred to in Clause
2.4.2).

GBHE-EDG.MAC. 52.Series

Machine Inventories (referred to in Clause A3.4
and block diagram page 7).


Integration of Reciprocating Metering Pumps Into A Process

More Related Content

What's hot (20)

Viewers also liked (17)

Similar to Integration of Reciprocating Metering Pumps Into A Process (20)

More from Gerard B. Hawkins (20)

Recently uploaded (20)

Integration of Reciprocating Metering Pumps Into A Process