7 CARGO AND BALLAST PUMPS

Learning objectives
-that the suction action of a pump is really atmospheric pressurepushing the liquid into the
inlet side of the pump
-how low vapor pressure of the liquid being pumped will improve suction
-that the discharge pressure will fluctuate when the liquid boils
-the meaning of head
-the meaning of NPSH
-the meaning of typical Q-H curve and the shore installation curve
-that the actual discharge rate also depends on static and dynamicbackpressure of the shore
installation
-the meaning of pressure surge
-that the system is liable to serious pressure surges if valve-closure time is equal to or less
than the pipeline period

7.1 PUMP THEORY AND CHARACTERISTICS

7.1.1 Classification and selection of pumps
There are a number of different pump types. Each type has its own special quality and
therefore certain advantages and disadvantages. The selection of pumps is determined by a
thorough study of the capacity needs and under which operational conditions the pump will
operate. The following factors are important when you evaluate these conditions:
• Estimated back pressure
• Capacity requirement
• Capacity range
• Requirement for installation and arrangement
• Expenses for purchase, installation and maintenance
• Availability of parts and service
• Suction terms
• Characteristics for the liquid to be pumped
Selection of the right pump for a determined purpose qualifies a close co-operation between
the customer and the producer of the pump. The customer has a special responsibility to
clarify all conditions concerning the pump installation, so the producer can choose the best
pump from his product range with the best match.
When you choose a pump you must find out how much the pump needs to deliver under a
specific condition. Definition of capacity range is important. Demand for capacity or
capacity range and expected discharge pressure must be specified. The capacity requirement
is determined by the intended use of the pump. The discharge pressure is determined by
various conditions where the pump’s delivery pipeline design, the capacity of the pump and
the liquid’s characteristics, is the essential.
Alternative installation locations of the pump are limited due to special demands from Class
and Shipping Authorities and also from lack of space.
Purchase and installation cost is important. Future maintenance expenses, availability of parts
and service now and over the next years, are also important and must be included in the
evaluation of alternative pump supplies.
The liquid’s properties and which other arrangements you have to consider, often limits the
options. Density, viscosity and boiling point are important properties to consider. The liquid
temperature and corrosive properties are important factors when pump material is selected.
The pump’s suction condition is determined from where the pump is located in relation to the
liquid to be pumped. A given suction pipe creates a certain resistance that will have influence
on the pump capacity. The main principle is to minimise resistance on the suction side by
decreasing the suction pipe length, have the largest diameter possible and few as possible
restrictions in form of bends, valves and so on.
The different types of pumps are divided into two main groups, displacement and kinetic
pumps. The displacement pumps displace the liquid by reducing the volume inside the pump.
An example is a piston pump where the piston is moving up and down inside a cylinder or
when the screws revolve inside a screw pump. Kinetic pumps (kinetic energy is equal to
“movement” energy) increase the liquid’s velocity through the pump.
The diagram below gives a brief view of the different available groups and types of pumps.
The diagram would be more comprehensive if the pumps were divided in all details according
to number of rotors, design of pump inlet/outlet and flow directions.

Types of pumps

Displacement pumps Kinetic pumps

Resiprocal pumps Rotating pumps Special pumps Centrifugal pumps

Piston pumps The Ejektor pump

Single rotor Multi rotor Axial flow Diagonal flow Radial flow

Single action
Gear pumps Single suction
Double action
Screw pumps
Single stage pumps

Multi stage pumps

A kinetic pump like the centrifugal pump increases the liquid’s velocity in the pump by
means of a rotating impeller. A displacement pump, like the piston pump, mechanically
displaces the liquid in the pump, either by help of a piston or screws. Resistance on delivery
side gives a liquid pressure rise (pump delivery pressure). One should be aware of this
difference for these two pump types.
The pressure rise on a kinetic pump is restricted by the increase in velocity over the pump,
which is controlled by the pump design. All kinetic pumps therefor have a designed or built-
in limitation for maximum discharge pressure. The displacement pumps limitation depends
only on available power and the constructional strength. In contrast to a kinetic pump, such a
pump will operate against resistance with all its available power. A closed-delivery valve
after a displacement pump is damaging. The same closed valve for a kinetic pump will not
bring any immediate danger.
Piston pumps and screw pumps have good suction capacity and are used where these
characteristics are required. The weakness of these pumps is the complex construction and
the relatively low capacity.
 Centrifugal pumps are simply constructed with few parts and no valves. There are no
immediate problems if the outlet of the pump is closed. These qualities result in relative low
purchase and servicing costs. Operation at high speed makes the pump small in proportion
compared to the capacity and flexibility in relation to the pump’s location.
The most negative side of using a centrifugal pump is the lack of self-priming capacity.
This weakness is improved by constructional efforts and positioning, which consolidate the
free flow of liquid. Location of a pump, for instance below the liquid level, can reduce the
flow resistance. High viscosity liquids are therefore particularly difficult to pump due to this
condition.
A centrifugal pump’s efficiency is high only within a small range. This is the reason it is
especially important to have a clear understanding of what capacity range the pump will
operate under, in connection with the selection of a centrifugal pump.

The differential pressure over each impeller is relatively low. Using so-called multistage
pumps where several impellers are mounted in serial, increase the pump’s capacity to deliver
against higher backpressure.

A centrifugal pump will, without a non-return valve on delivery side, give complete back
flow at the time the pump stops. For all operators of centrifugal pumps, this relationship is
important to know.

Centrifugal pump

7.1.2 The ejector

The ejector design is simple and is used for stripping. This ejector has no revolving or
reciprocating parts and is thereby especially easy to maintain.
The propellant (driving water), a liquid or gas, is forced through a nozzle into a mixer tube.
The velocity of the propellant will naturally increase as it passes through the nozzle. Due to
the propellant’s velocity and direction, plus the friction force between the propellant and the
liquid, the surrounding liquid will be sucked into the ejector’s mixer tube. The mixer tube is
connected to an expanding tube, the diffusor. Here some of the kinetic energy supplied to the
liquid in the mixer tube is transformed into potential energy. The capacity depends on the
friction force between the two mediums, suction head, delivery head and the propellant’s
velocity. The ejector has the advantage that it does not lose the suction capacity even if it
sucks air or vapour.
The ejector’s efficiency is between 30% and 40%. Even if the propellant’s efficiency is up to
approximately 70%, the total efficiency for the whole ejector system is far less than compared
to a pump system, such as a centrifugal pump. Another drawback with ejectors is that the
propellant is mixed with the pumping liquid. This implies that if the ejector is to be used in
cargo transfer operation, the cargo itself must be used as propellant liquid.The ejector is
frequently used as a bilge pump in hold spaces. A common arrangement for a hold space is
as follows:
The ejector is usually submerged in a bilge sump and the propellant is normally supplied
from a seawater pump. Onboard gas carriers where the hull is the secondary barrier, the
ejector may also be used to pump cargo from hold space. In that case, the liquefied cargo
itself must be used as a propellant

.
7.1.3 Tips
• Be aware that the ejector has a limitation on the propellant’s pressure. Higher pressure
than recommended by the supplier may result in reduced suction capacity.
• Start the ejector by opening all valves on delivery side first, and then adjust the correct
propellant pressure. The ejector’s suction valves should be opened last, which will prevent
the propellant’s flow back into the tank that is to be stripped.
• Stop the ejector by using the opposite procedure.
As the drawing shows the ejector is positioned 3 meters above the liquid level. The liquid
level in the slop tank is 15 meters above the ejector and the propellant's pressure is 8 bars.
The ejector’s capacity can be found by use of the performance curve for the specific ejector.
In the performance curve the ejector capacity is set as a function of the propellant pressure.
Observe that this curve has curves for different suction lifts. The different performance
curves are marked with different suction lifts. The ejector’s suction lift in this example is 3
meters; this specific curve shall be used.
You can find the capacity of the ejector by drawing a vertical line from 8 bars on the scale for
a delivery head of 15 meters and up to the performance curve with a suction lift of 3 meters.
From this point of intersection, draw a horizontal line to the left and over to the ejector’s
capacity side. The found capacity in this case is 600 m3/h.
7.1.4 The ejector’s Performance Curves
7.1.5 Displacement pumps
Pumps are very old machines. The first types of pumps (screw pump and piston pump) are
more than two thousand years old.
A pump’s purpose is for the transport of liquid, usually from a low level to a higher level. Its
purpose can also be to press a liquid into a tank, which contains higher pressure than the
surroundings. The pump increases the liquid’s energy. The increased energy is potential
energy; the liquid is transported from a low level to a higher level. This is the kinetic energy,
the liquid’s flow has increased as pressure energy, if the liquid is pumped into a tank with a
higher pressure than its surroundings. As an example, the feed water pump to a boiler is
working using these principles.
In addition to the mentioned increase of energy, the pump also has to maintain the energy
which is lost due to streaming in the system.
As mentioned, kinetic pumps constantly have liquid streaming through the pump with
pressure increasing simultaneously. In displacement pumps, a certain volume of liquid is
branched off and moved from the pump’s delivery side. Then a pressure increase occurs.
Screw pumps and piston pumps will be viewed further in this chapter. A wide range of
displacement pumps is available, such as the lamella pump, ring pump, propeller pump, etc.

7.1.6 Piston pump

The piston pump is used for relatively small amounts of liquid with large delivery heads.
When the piston is pulled upwards, a vacuum occurs inside the pump housing. The suction
valve will then open and liquid streams into the pump. When the piston is pressed downward
the pressure will increase, the suction valve will close, the delivery valve is set open and the
liquid sent out of the pump. The liquid does not stream in a smooth flow as in a centrifugal
pump, but accelerates and slows alternately. This is of inconvenience with long pipelines.
The valve is a weak point. They are sensitive to liquid pollution and they also increase the
resistance against streaming.

Usually, the piston pump is double acting. Because of the pump’s movements, the pump
must have a relatively slow piston speed. The piston pump may sustain almost unlimited
pressure. However, the limitation is the automotive power and the material strength. The
piston pump does not have to be filled with liquid before starting. Make sure that all the
valves on the delivery side are open before starting. The efficiency of piston pumps is
higher than, for instance, centrifugal pumps. The piston pump is a well-known pump on
board an oil tanker. This is the pump, which is used to pump cargo deposits ashore at the end
of the discharging operation. These oil deposits from cargo tanks, lines and cargo pumps are
pumped ashore through a small diameter line.

7.1.7 Screw pump

The screw pump consists of two screws or more, where one of them is activated. The screws
are placed inside a pump house. A common and well-known screw pump is the Swedish
manufactured so-called IMO pump. This pump consists of one active screw placed in the
middle and two symmetrical side screws. The screws tighten to each other and to the
housing, but have no metallic contact. When the screw rotates, the threads are filled with
liquid. The liquid is displaced by axial through the pump. In this pump, the side screw
rotates in the opposite direction of the middle screw. These screws are working like an
endless piston which constantly moves forward. The liquid is not exposed to rotation. The
pump is self-priming, running almost soundless and with little exposure for wear and tear
when pumping clean liquids. The screw pumps are used a lot as a lubricating pump, but are
also used as a stripping pump on oil tankers.

Screw pumps

7.1.8 Centrifugal pump

The centrifugal pump’s mode of operation
A centrifugal pump consists of a rotating impeller inside a pump casing. The liquid inside the
impeller is affected by the “blades”, and will be lead through the “blades” due to the
centrifugal force. Energy in forms of kinetic energy (velocity energy) is added to the liquid.
New liquid is constantly lead into the impeller and put into rotation. A flow through the
pump is established.
If the delivery pipeline from the pump is open to the atmosphere and has sufficient height, the
liquid will adjust itself to a precise level given by the energy, which was added to the liquid
through the impeller. Here, all kinetic energy is transformed into potential energy.
The difference in liquid level is called net delivery head. A pump’s delivery head is
dependent on the individual pump’s construction. If the level in the tank is lowered, the
liquid level in the delivery pipeline will be correspondingly lower. Net delivery heads (H1,
H2, H3) will be equal for the same pump provided that flow disturbance does not occur on
the pump’s suction side.
However, the pump’s delivery pressure is dependent on the liquid’s density and delivery
head. In this case, the liquid is water with a density ( ) of 1000 kg/m3 and the head (H) is
100 meters, the manometer pressure (pm) after the pump will be read at:

pm = ρ x g x H = (1000 kg/m3 x 9,81 m/s2 x 100 m)

pm = 981000 Pa = 981 kPa
pm = 9,81 bars

One can see from the previous example that the delivery head of the pump is obtained from
the pump itself, and that the delivery head is independent from the pump’s position or
location. It is therefore natural that the centrifugal pump’s capacity always is given as a
function of the pump’s delivery head.
If you bend the discharge pipe from the previous example, like the illustration below, the
liquid will flow out of the pipe. Only a part of the added energy in the pump will “lift” the
liquid. The rest of the energy is still in the form of kinetic energy. From the previous taught
experiment, one can predict that the capacity of a centrifugal pump will be highest at minimal
delivery head. The capacity curve (Q-H curve) will, in practice, follow this assumption, but
the curve is not linear due to loss of energy in the pump.

If you ignore the pipe resistance, the capacity Q in this situation is determined by the delivery
head (H). The delivery head here is the static height or the static backpressure, which the
liquid has to lift.

In a real pipe system, bends and valves will create a resistance due to friction against free
liquid flow. This resistance varies with the velocity and viscosity of the liquid, and is called
the dynamic backpressure. The total pipe resistance, composed by the static and the dynamic
backpressure, is called a system characteristic curve.
The intersection point between the system characteristic curve and the capacity curve is
called the actual operation point.
It was previously mentioned that disturbances on the pump’s suction side would have
influence on the capacity.
The conditions on the inlet side are very important for the centrifugal pump’s operation. A
centrifugal pump has normally no self-priming qualities, meaning that the pump is not able to
suck liquid from a lower level. Additional vacuum equipment connected to the pump will,
however, improve the pump's self-priming qualities. When the inlet pipe and impeller is
filled with liquid, the pumping process will be able to continue without this equipment. The
liquid’s viscosity may ensure a continual flow into the pump. Too high resistance in the inlet
pipe will cause the same operational disturbance. If the flow into the pump is less than the
outlet flow, due to too high pipe resistance and/or too high viscosity, these factors will have
considerable influence on the pump’s capacity.

If you start a pump, submerged in water like the sketch indicates, the pump will have a
specific capacity at a specific delivery head. If you gradually lift the pump, the pump will, at
a specific height, have a perceptible reduction in the capacity. When this occurs, the height of
the pump above liquid level is called Net Positive Suction Head or NPSH.
The explanation of this phenomena is that when the pump is lifted up out off the water, the
pipe length and the resistance at the inlet side increases. The increased resistance creates
constant negative pressure on the inlet side of the pump. The liquid that accelerates from the
centre of the impeller and out to the periphery increases this negative pressure. When the
negative pressure reaches the liquid’s saturation pressure, the liquid starts boiling and a large
quantity of vapour is created in the pump. The output flow from the pump become irregular,
and will stop at huge vapour volumes. We say that the pump cavitates.
A centrifugal pump operates satisfactorily with approximately 2% gas in the liquid. But
cavitation will always damage for the pump. The gas bubbles created in the liquid on the
pump’s suction side will collapse when the pressure rises inside the impeller.
The consequences of cavity are:
Vibrations and noise
Reduced efficiency
Pitting or cavity erosion inside the pump
As we have observed, the cavitation is destructive and must be avoided or controlled. To
ensure limited or non-generation of vapour one must make sure that the liquid at the pump
inlet has sufficient overpressure to avoid evaporation. The resistance at the pump inlet side
should be made as low as possible. This can be done by constructing the pipeline as short as
possible, limiting the number of bends and selecting a maximum diameter on the pipeline
diameter. The pump should be positioned at the lowest possible level, and preferably below
liquid level at the suction side.
A pump’s NPSH is variable and dependent on the flow. When the flow increases, the
negative pressure generated inside the pump increases. A reduction of the flow will reduce
the negative pressure. Reducing the pump’s capacity may therefore control cavitation.
A centrifugal pump’s capacity is adjusted by throttling the delivery valve. Throttling
increases the pumps discharge pressure (backpressure) which causes reduced capacity.
The capacity may also be adjusted by changing the pump’s rotation speed. Adjustments of
the pump’s revolution move the capacity curve up or down. Reduction of the revolution
moves the curve parallel downwards, an increase in revolution, upwards. Note that these
relations are valid only if the flow conditions are unchanged.
 H
System

mlc
characteristic

Operation point

n = 1 000 rpm
Static Head

n = 800 rpm

m 3/hrs Q

7.1.9 The Pump performance diagram

All manufacturers supply a pump performance diagram with the pump delivery. The curves
in the diagram are results from practical tests in the maker’s workshop and specifies:
• Type of liquid used in the test (generally water)
• Number of revolutions
• Type and size of impeller
• The optimal operation point

The operation point is normally set at the best possible efficiency, simultaneously within the
pump’s predicted capacity range.
It is important to be aware that the pump’s diagram is made for a special liquid with specific
properties. The capacity curve will be real for all liquids, provided the free flow to the pump
inlet is not restricted due to for example too high viscosity. The power consumption curve
will of course depend on the fluid’s density.
 NPSH
H (mlc) (m)

Q-H
Effiency
(% )

Power
Consumption
(kW)

Q (m 3 /hrs)
A pump’s condition is of course vital for the curve accuracy. There are a lot of methods to
check the centrifugal pump’s condition. Monitoring the pump’s delivery head, capacity,
power consumption and development of these is obvious. Detection of many minor
operational disturbances may be difficult and not necessarily observed. Establishment of
routines ensure continuous control of vibrations. Visual inspection of the pump and regular
maintenance is important to prevent break down.

7.1.10 Draining and stripping system

The last stage during a discharging operation is the stripping, which means to empty the
deposits of cargo in the cargo tanks, lines and pumps. Stripping is a part of the operation,
which cannot be done with the cargo pumps in normal running.
Well-known stripping systems are steam driven piston pumps, vacuum stripping and ejectors.
The steam driven piston pump is an IMO demand on board crude oil tankers. This stripping
pump empties the last deposits of cargo from the tanks and lines then pumps it ashore through
the small diameter line.
This pump will also be the last pump in use during the discharging operation.
The vacuum stripping system is the most efficient method. On M/T Seagull this system is
installed. The vacuum system makes use of the main cargo pumps and the main cargo lines.

7.1.11 Vacuum strip composition

The main line, just ahead of the cargo pump, on the free flow side is connected to a vacuum
separator tank. The purpose with this separator tank is to support the cargo pump with liquid.
The separator tank has a pipe connection to a vacuum tank fitted on a higher level in the
pump room. This line is supplied with a valve.
The vacuum tank is connected to a vacuum pump. This pump’s shaft leads via a gas proof
joint to the engine room where the running unit is located.
A pipe to pick up the exhaust from the separator tank leads from the vacuum tank to the slop
tank.
7.1.12 Mode of operation
The separator tank works like a reservoir feeding the pump with liquid. The liquid level
inside the separator tank will fall when the level in the cargo tank is getting lower than the
height of the separator tank.
The void space above the liquid inside the separator tank will increase. In this stage, falling
pump pressure should be observed before the vacuum system is activated.
At a fixed limit on the separator tank, the vacuum pump will start creating a vacuum in the
void space above the liquid.
The valve between the separator tank and the vacuum tank will open and the liquid will be
sucked into the separator tank because of the vacuum.
At the same time, the delivery valve is automatically (or manually) throttled. This is done to
give time for the separator tank to refill itself.
The liquid in the separator tank supports the pump with liquid, even if the flow from the
cargo tank is poor at times.
The vacuum stripping system provides good working conditions for the cargo pumps
regarding feeding. This is achieved thanks to good co-operation between the cargo pump’s
delivery valve and the vacuum pump.
The principles for different vacuum stripping systems is more or less the same, but please
study the User’s Manual of the equipment on the specific ship.
Stripping arrangement for a deepwell pump system
 Stripping arrangement for cargo systems with a separate pump-room

7.1.13 Parallel operation of centrifugal pumps

The capacity requirement is usually higher than the performance of just one centrifugal pump.
Additional pumps are then lined up and run in a parallel operation.
The diagram below shows two equal pumps in parallel operation. We assume a symmetrical
pipeline for both pumps.

System
characteristic
mlc curve

1 1+2
Static head

m3/hrs Q

The pump’s capacity curve (1) shows the relation between the delivery head and the flow rate
for one pump. As both pumps are equal, the two pumps capacity curve is represented by the
curve marked 1+2.
When equal pumps are run in parallel, the delivery head for the system will be equal to the
delivery head for one pump. The capacity, however, will increase in proportion to the
number of pumps. If, for instance one pump has a capacity of 1,330 m3/hr. at a delivery head
of 88 meters, two pumps in parallel will deliver 2,660 m3/hr. and three pumps 3,990 m3/hr. at
the same head.
If the pumps are of different types or equal pumps are run at different speed, their individual
performance curves will be different. This is no problem provided the point of intersection is
outside the pumps individual point of intersection.
If, for instance increasing the back pressure alters the point of intersection, this may lead to
one of the pumps running without delivering any cargo. If this situation occurs and
continues, the pump may be damaged. If the system does not have any non-return valves and
the backpressure increases further, a back flow will occur through the pump and into the
cargo tank.
Even though the type of pumps is the same in a parallel operation, variation in rotational
speed, asymmetric piping, variance in opening of valves, cavitation and variance of the
pump’s condition may form the same condition.

Parallel operation
Pumps are operated in parallel when the time per unit will deliver more cargo than is possible
to do with one pump. In this example, we will view how to find the delivery amount and
delivery head by use of a pump curve. The system characteristic curve is shown on the pump
curve diagram (next page).
The pump curve “A” for one pump shows in point “D1”, a delivery of 1,400 m3/hour at a
delivery head of 88 meters.
The pump curve “B” shows two pumps in parallel operation with a delivery of 2,250 m3/hour
at a delivery head of 116 meters.
To plot in pump curve “B” add the delivery amount of the two pumps at the different delivery
heads. As shown in curve “A” the delivery at 20mlc. is 1,770 m3/hour, point 1. Plot a new
point at 20mlc. (1,770 + 1,770) = 3540 m3/hour, point 11. In the same way, we are plotting
the values according to the table above. When all the values are plotted, a new curve is
drawn through the plotted points, curve “B”. Where the new curve is crossing the system
curve, the delivery amount and delivery head for two pumps in parallel operation will be
read. The same procedure stands for 3 or 4 pumps in parallel operation.
Starting pump number 2 will not double the capacity because a higher volume of flow creates
higher dynamic resistance. The increase in capacity will then be relatively less for each
pump added.

Mlc = meter liquid column.

Mlc
240

System charcteristic curve

200

160
Pump curves
«B»
«A» 6.
D2
120 66.

5. 55

D1 4. 44.
80
3. 33.

40 2. 22.
Static head
1. 11.
Capasity - cubicm./hour
400 800 1200 1600 2000 2400 2800 3200 3600

To achieve accurate plotting of two or more pumps in parallel operation, it is recommended

to use a sheet of millimetre paper. Study the pump performance curve for the cargo pumps
(or for any other pumps) on your ship.

7.1.14 Pump calculation

All pumps are delivered with pump curves from the supplier and it is important to be able to
use these curves properly. On the example curve in this chapter, a curve is drawn for one
pump which runs with a fixed revolution.
Also drawn is the curve for the pipe, which consists of static and dynamic backpressure. The
static backpressure is caused by the difference between the shore tank’s liquid level and the
vessel’s cargo tank’s liquid level. Friction resistance in valves, bends, pipes, etc causes the
dynamic backpressure. Shore terminals usually have worked out characteristics for their
installations (system curve). If the installation’s system curve or the vessel’s pump curve is
on a transparent sheet, these can be applied to each other. For instance, the vessel’s curve is a
transparency and is placed on the top of the terminal’s curve. Then the maximum delivery (at
point of intersection) at existing conditions can be seen.
From the example’s curve, in point “A” (point of intersection), the pump delivers
1,560m3/hour at a delivery head of 58 mlc. The oil’s density in the example is 820kg/m3.
Out of this information, it is possible to find out what 58mlc. corresponds to in pumping
pressure (manometer pressure) by use of the following formula:
p= xgxh

p = pump pressure
= the liquid’s density - 820kg/ m3
g = the earth’s gravity acceleration - 9,81m/s2
h = delivery head - 58mlc.

The values used are just for this example. The denomination, which appears, is called Pascal
(Pa). 100,000 Pa is equal to 1bar.

Calculate the manometer pressure:

p= xgxh
p = 820kg/ m3 x 9,81m/s2 x 58mlc.
p = 466,563 Pa.
p = 4,7 bar. (4,66563).

With a delivery flow of 1560m3/h and a delivery head of 58mlc., the manometer shows 4,7
bar. The increased pressure above the pump is calculated, which is (on oil tankers)
approximately the same as the delivery pressure because the tank pressure is close to zero.
To calculate the correct pressure, it is necessary to allow for the manometer height above the
pump’s inlet and for the liquid height in the cargo tank above the pump’s height.
In this example the following data is known in addition what already exists:
• Tank pressure = 0,1bar(10000 Pa).
• Liquid level above pump inlet = 18 metres.
• Manometer height (manifold) above the pump inlet = 23 metres.

The delivery pressure is now calculated based on the formula:

P= x g x (h + x - y) + Po

P = delivery pressure
= the liquid density - 820kg/ m3
g = the earth’s gravity acceleration - 9,81m/s2
h = delivery head - 58mvs.
x = liquid level above the pump inlet - 18 meter
y = manometer height above the pump inlet - 23 meter
Po = tank pressure - 0,1bar (10.000 Pa)

The delivery pressure will be as follows:

P = x g x (h + x - y) + Po.
P = 820kg/ m3 x 9,81m/s2 x (58 meter + 18 meter - 23 meter) + 10,000 Pa
P = 820kg/ m3 x 9,81m/s2 x 53 meter + 10,000 Pa
P = 436342 Pa
P = 4,4bar (4,36342).

These additional values normally give very small deviations, so in the further calculations
these values are not included.
The dynamic backpressure may change, i.e. when throttling on the pump’s delivery valve. In
this example, the discharge rate will be reduced to 1000m3/h. Choose to do so by throttling
the pump’s delivery valve, and when doing so, calculate the manometer pressure. First, draw
a new curve (see the dotted curve) which crosses the pump curve at a delivery rate of
1000m3/h, which creates the new intersection point “B”.
From the point of intersection “B”, a horizontal line is drawn on the left side of the curve.
The new delivery head is 98 metres. With the same formula as before the manometer
pressure is calculated:

p= xgxh

p= xgxh
p = 820kg/ m3 x 9,81m/s2 x 98
p = 788,331 Pa
P = 7,9bar (7,88331bar)

The conclusion is, by throttling the delivery valve to the manometer pressure, we get 7,9 bar
with the delivery rate of 1,000m3/h.

Out of the same formula, it is also possible to calculate the delivery head by reading the
manometer pressure. An example using the same curve diagram, the manometer pressure is
6,3bar which compares to (6,3 x 100,000) = 630,000 Pa.

Calculate the delivery head by turning the formula p= x g x h, to:

h=p:( x g)

This will give following delivery head:

h = p : ( x g)
h = 630,000 Pa : (820kg/ m3 x 9,81m/s2)
h = 78,3 mlc.
 Mlc = meter liquid colum.

120

“B” - point of intersection

Pump curve System curve
100

“C” - point of intersection

80

60

“A” Point of intersection

40
Dynamic back pressure

Delivery amounts
20 Static back pressure

Capasity - m3/hour
200 400 600 800 1000 1200 1400 1600

7.2 PRESSURE SURGE AND LIQUID PRESSURE

When a valve on a liquid line is closed too quickly, the pressure inside the line increases to a
hazardous high level very quickly.
Quick changes to the liquid flow in a pipeline may lead to a pressure surge resulting in a
rupture in the pipeline system. This surge pressure can be recognised by a “knock” in the
pipeline. This type of pressure peak is generated very quickly in the pipeline, faster than a
common safety valve is capable to relieve. The consequence may be the breakdown of the
pipeline system and thereby high risk of pollution, fire and personal injury.
Pressure surge may appear if:
• The emergency shutdown valves are activated and closed too quickly.
(ESD/Emergency Shut Down)
• Fast closing/opening of manual or remote operated valves.
• Fast variation of the volume flow resulting that a non-return valve starts hammering.
• When a pump is started and stopped.
The pressure in a cargo transfer system has three components; the hydrostatic pressure, the
cargo tank pressure and the pressure generated by the pump. The hydrostatic pressure and
the cargo tank pressure will mainly be constant and we will refer to these as “static pressure”.
If the flow suddenly is varying due to, i.e. too fast valve closure, the moving liquid that has a
specific velocity and mass will hit a “wall” inside the valve (the valve seat or the valve flap).
The kinetic energy of the moving liquid will immediately convert into potential energy by
compression of the liquid against the valve seat. How fast the pressure peak is generated
depends on the velocity and the density of the liquid.

A pipeline of 250 meters and 150 mm in diameter is used for water transfer at a capacity of
400 m3/hrs. The total mass of the moving liquid inside the pipe is 4400 kg and moves with a
velocity of 6,3 meters/second. If a valve is closed immediately, the kinetic energy will
convert almost immediately to potential energy. The pressure surge may reach approximately
40 bars within 0,3 seconds.
If the liquid is a condensed gas or crude oil, vapour may be present. These vapour bubbles
will collapse when the pressure increases. The collapsed bubbles will generate pressure
waves that will also be transmitted through the pipeline system. In an opposite case where
the pressure is decreasing rapidly, a volatile liquid will start boiling. The above mentioned
cases illustrate why it is especially important that the valves and pumps are cautiously
operated so neither dangerous pressure peaks nor pressure drops are generated.
The enclosed diagram on the next page is from ITC Tanker Safety Guide. It shows a normal
cargo operation and pressure in the pipeline. The maximum pressure is at the pump outlet.
This pressure is the sum of the hydrostatic pressure and the pressure generated by the pump.
Due to friction in the pipeline, the pressure will gradually decrease toward the cargo
manifold. If the ESD valve is activated and the valve is closed too quickly, the liquid flow
will stop quite quickly. The liquid’s kinetic energy will convert into potential energy
immediately when the liquid hits the valve seat. A pressure peak is generated and will be
transmitted at the speed of sound (the only way possible) back towards the pump. When the
wave of pressure reaches the pump, some of the pressure will unload through the pump, but
the resistance here will also operate as a “wall”. The pressure is rebuilt and reflected back
towards the ESD valve again. A new pressure peak is generated with additional increased
pressure. A “knock” will occur each time a pressure top is generated against a “wall”. All
personnel that operate valves or pumps must be observant of these phenomena and of the
liquid pressure, which may occur consequently.
The progress and the length of the pressure surge depends on the system. If the wave of
pressure is allowed to move between two valves without pressure relief, a maximum pressure
surge will be generated.
The most vulnerable parts in the system are the shore connections and loading/discharging
arms. The operative personnel normally work nearby the manifold area. A rupture in this
area may easily lead to personnel injury. (Please note that control of cargo hoses is dealt with
on page 4 in this chapter.)
Maintenance and testing of the ESD-valves’ closing time is the most important of the above
mentioned causes. Closing time of the ESD-valves, which is too short, may lead to
generation of a dangerous pressure surge. Always consult the terminal representatives about
the required pipe line period.
7.2.1 Development of pressure surge
The closing time of ESD valves should be as short as possible to prevent overflow and
spillage. But not so fast that a risky pressure surge occurs. Necessary time for a safe closure
of valves can be calculated based on the expected maximum pressure surge when the pressure
wave has passed forward and backward through the pipeline. The speed of the sound is set to
1,320 m/s. If the pipeline is 2 km, the calculated time for maximum pressure surge at closure
of the ESD valve is:
T = (2 x L) / Speed of sound = (2 x 2,000 m) / 1320 m/s = 3 s
The maximum pressure surge will occur 3 seconds from closure of the ESD valve. This time
is called a “pipeline period”. It is assumed that the safe closing time is five times a pipeline
period, so the closing time should at minimum be:
5 x 3s = 15 seconds

7.2.2 Cargo hoses

The cargo hoses are normally the weakest part in a pipeline system transferring cargo. The
responsibility for the cargo hose condition on board lays with the ship. It is important to be
aware of this fact, in case a cargo hose is lent out to a third party or is used in transferring
cargo between ships.
Hoses for cargo should be cleaned and dried before storing. The storage area should be dry
and out of the sun, if you want to take care of the hoses as long as possible. Poor cleaning
and storage is generally the cause of damage and consequently replacement. A cargo hose
prototype is tested with the products, pressure and temperature for which the hose type is
approved. The cargo hoses should be tested yearly at a pressure that is 1,5 times the
maximum working pressure.
The normal procedure for yearly testing is:
1. The hose is laid out on deck, blinded off and filled with water.
2. The hose is pressurised with 1,5 times the working pressure.
3. The hose is checked for leakage.
4. Electric bonding is checked.
5. Linear expansion for the hose is measured (measure for strength)
6. The test result is logged.

If there is no leakage, bonding is okay and the linear expansion is less than the limitation set
by the manufacturer, the hose is approved.