PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

1. More about Autolayout and Design Optimization: ...................................................... 2

1.1.
1.2.
1.3.
1.4.

Autolayout example:................................................................................................................ 3
The Base Case: ...................................................................................................................... 4
Variation 1: Effect of Bringing the Nozzles Closer to the Vessel ............................................. 7
Variation 2: Effect of Moving Nozzles Away from the Tank And Reducing the Vertical
Separation Between Ranges ................................................................................................... 9
1.5. Variation 3: Effect of Increasing the Minimum Overlap Between Nozzles to 5 ft. .................. 11
1.6. Conclusion: ........................................................................................................................... 13

2. Air-Cooled Cooling Water System for Condensers in a Power Station .................. 13

2.1.
2.2.
2.3.
2.4.

Scenario 1 - Cooling Water Pump Trip without Valve Closure ............................................... 14

Scenario 2 - Cooling Water Pump Trip followed by Valve Closure......................................... 16
Scenario 3 - Cooling Water Pump Trips with Expansion Tank ............................................... 16
Conclusions: ......................................................................................................................... 18

3. Frequently Asked Questions ...................................................................................... 18

3.1.
3.2.
3.3.
3.4.

More about PID Controllers: .................................................................................................. 18

PIPENET Spray/Sprinkler Module: ........................................................................................ 20
PIPENET Standard Module:.................................................................................................. 22
Miscellaneous: ...................................................................................................................... 23

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

1. More about Autolayout and Design Optimization:

Let PIPENET take the strain

Autolayout is not only about saving time in generating networks. It is a new approach and a new way
of design optimization which only takes a few seconds. The rewards are enormous. For example,
when it comes to the design of tank protection systems, merely at the touch of a few buttons
optimization of number of nozzles, horizontal spacing, vertical spacing, minimum nozzle flowrates,
minimum required are all obtained. This makes accurate design easy and optimization is also
possible to a level impossible by manual calculation. In this edition of PIPENET News we show
exactly how optimal design can be carried out by evaluating several options quickly in the design of
the deluge system for a spherical tank.
There are many new developments in PIPENET. When it comes to PIPENET Spray/Sprinkler
module, we are introducing a philosophical change. Historically this module has been an
analysis/verification tool. Basically a system is designed by the user and PIPENET is used for
verification and fine tuning the design. The philosophical change we are introducing is that PIPENET
Spray/Sprinkler module is now a true design tool. In a number of cases it will design the system by
itself using a small number of system parameters. We feel that we are in the early stages of this
philosophical change. We are reliant on our customer feed-back to improve this capability further.
All the dialog boxes and diagrams in this section are from PIPENET.
This principle is best illustrated through an example. Let us consider the need to cover the lower
hemisphere of a 48ft diameter spherical tank with a minimum flow density of 0.1 gpm ft-2 at every
point. We have chosen a spherical tank because that probably poses the most difficult design
optimization challenge.
This design problem itself has an infinite number of potential solutions given the number of degrees
of freedom: There are many questions such as (i) Is it more optimal to use a higher flow rate per
nozzle or more nozzles? (ii) How far from the surface of the sphere should the nozzles be placed?
(iii) Is sufficient vertical coverage obtained etc. To do this manually would be tedious and time
consuming and the scope for true optimization is limited. Generally these choices are made by
experience. Autolayout allows the user to consider more cases and hence optimize the design in a
matter of minutes in a systematic manner.
In this article we discuss how the Autolayout facility of PIPENET be used for both piping layout
design and for optimisation. Note also that the Autolayout tool is based around the fact that NFPA15
requires 0.25 gpm ft -2 by default but initially we will work with 0.1 gpm ft -2 here. Also, note that a
nozzle library must be set up and the nozzles must have a divergence angle to use the Autolayout
facility.
We assume here that the nozzle data has already been input into the library. Please note that the
spray divergence angle must be input into the library. A dialog box for the nozzle in the library is
shown below for illustration. (Please note that this is not the data used in the Autolayout example
below. It is shown merely for illustration.)

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

Spray
divergence angle
3

In PIPENET menu bar we click on the Autolayout option as shown below:

1.1.

Autolayout example:

As mentioned above, we will consider the need to cover the lower hemisphere of a 48ft diameter
spherical tank with a minimum of 0.1 gpm ft-2 at every point. The definition of the tank and basic pipe
details are input in the dialog box below.

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

Size and shape of

vessel

1.2.

The Base Case:

The initial setup of the model is shown in the dialog box below. As we are only interested in the lower
hemisphere, we set the start point to be the equator, and set the upper hemisphere separation to be
a large distance, so that no ranges appear above the equator. We achieve this by setting the vertical
separation between the rows above the equator to 100 ft which is well above the separation
prescribed in the NFPA rules. It is set to a value higher than half the diameter of the tank. So in
effect the range above the equator is well above the top of the tank and does not form part of the
model.
The minimum spray overlap was set to 0 ft and nozzle-vessel distance was set to 3 ft in the initial
case these will be changed later in further cases. The choice of the average flow density has
initially been set to 0.05 gpm ft -2, to give a reasonable initial guess for the flow per nozzle. This is
based on the total flow impacting on the vessel, divided by the surface area of the vessel; in other
words this takes into account the whole of the sphere. As we know that the entire flow is impacting on
only the lower hemisphere, we half this value in order to obtain the total flowrate from the nozzles on
the lower half. This puts the value below the NFPA minimum for the entire sphere and turns the value
red, but we are not considering the entire sphere and so this too can be ignored.

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

Range separation
above the equator

The library had 3 nozzles in this case. Our choice of nozzle N-2 below is arbitrary just to illustrate the
dialog box. Note that the total flow for the chosen nozzle divided by the surface area of the vessel is

.




= 0.0607 gpm ft-2 which is significantly higher than 0.05. We will see why this is later on.

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

Generating the spray diagram for the model results in the following.

The same spray diagram with the sphere rotated to show the plan view results in the following. This
diagram also shows the spray cone.
The critical flow density is set to be a user defined value of 0.1 gpm ft-2, the resolution is increased
slightly, the warning colour is changed to white (in order to see the cut-off more clearly) and then
regenerated the image. In the image in the bottom pane of the window we can see that the coverage
is not good. Turning on the spray cones in the Model Menu panel, we also see that due to the wide
spray angle of the nozzle and their distance from the vessel, much of the spray from the nozzles

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

actually misses the vessel itself. That flow is being completely wasted by not hitting the vessel. This
inefficiency is the reason for the unnecessarily high value of the total flow required.

1.3.

Variation 1: Effect of Bringing the Nozzles Closer to the Vessel

Optimization at your finger tips

We return to the Autolayout dialog in PIPENET and reduce the nozzle-vessel distance to 1 ft from 3
ft. This may be considered to be too small but it is interesting to see its effect. It only costs a few
seconds in time to generate another spray diagram.

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

It is indeed interesting to note that there is a large increase in the number of nozzles but each nozzle
is required to flow less water. What is even more interesting to note is the fact that the total flowrate
required is considerably less. On the other hand, the minimum required pressure at the nozzle inlet
is also less. It is important to make sure that this is above the vendor recommended minimum
pressure. It is true that the nozzles are horizontally spaced for the circles of spray impact to touch
each other as required by the NFPA rules. The problem is not only a significant increase in the
number of nozzles and a reduction in the inlet pressure of the nozzles. It can also be clearly seen
that, as the nozzles are too close to the vessel, its coverage vertically is poor.

By generating a new spray diagram the following can be obtained:

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

The nozzle cones are shown and they are extended to full size so that they touch the sphere. We see
that, although the nozzles are now utilizing 100% of their flow, there are now large areas of
unprotected surface. So, increased efficiency and reduced flowrate have been obtained at the cost
of increased number of nozzles and poor vertical coverage. Furthermore, the nozzles may be
working below the vendor recommended minimum pressure.
1.4.
Variation 2: Effect of Moving Nozzles Away from the Tank And Reducing the Vertical
Separation Between Ranges

Effortless optimization in seconds

In variation 1 the minimum spray overlap was set to 0, so the circles of impact all (at least) touch.
Moving the nozzles away from the vessel slightly will increase the size of the circle of impact for each
nozzle, but we will also need to decrease the range separation (vertical separation between rings) to
get better vertical coverage. The adjustments are shown in the dialog box below:

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

10

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

11

In the last image I set the spray cone alpha to full so that they became completely opaque, and also
set the size to full. This gives a very clear view of the potential coverage of the model.
1.5.

Variation 3: Effect of Increasing the Minimum Overlap Between Nozzles to 5 ft.

We see that there are very small gaps in the cone coverage in the above variation 2. It is clear that
increasing the minimum overlap would be required to achieve complete coverage if this is required.
The minimum overlap is now increased to 5 ft. The effect can be seen below.
This of course leads to a large number of nozzles and reduced inlet pressure.

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

We now have total coverage of the vessel, with a generous overlap for each cone (this is a good idea
as the spray angle is wide). What is now required is an increase in flow per nozzle to achieve the
minimum 0.1 gpm ft-2. In order to do this it is necessary to switch to the nozzles tab in Autolayout
and not the flow per nozzle in the current setup. The original value was 3.9 gpm as shown below.
This was changed the flow option to flow per nozzle and changed the value to 15 gpm.
12

Change here

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

13

We now have greatly improved coverage. When the cursor is over the image in the bottom pane to
see the flow at a particular point. Further adjustment in the same manner will be required to get better
results.
1.6.

Conclusion:

PIPENET and engineer working in unison

What we have shown is this article is how to use the Autolayout facility of PIPENET for optimization
of the selection, positioning and spacing of nozzles. It is intended to show the flexibility of the
Autolayout capability rather than define a methodology. The principle is PIPENET will perform the
tedious tasks and leave the user to concentrate on what an engineer, rather than a computer
program, is good at. The best designs are obtained when PIPENET performs the tedious tasks and
the engineer uses his experience and exercises his judgement.

2. Air-Cooled Cooling Water System for Condensers in a Power Station

In this exercise the customer used PIPENET to model an indirect air cooling system in a 330MW unit
power plant in China. The model featured the following.

Carbon-steel pipes
168 air-cooled heat exchangers in the cooling system (modelled by PIPENET general
pressure loss components)
2 Cooling water pumps
A two-chamber, two-pass condenser with 22 x 0.5 mm tubes (modelled by a PIPENET
pipe bundle).

Pressure loss due to the filters before the cooling pumps is a head loss of 1.259 m fluid at the design
flowrate of 5.08 m3/s.
The objectives of the modelling exercise were the following:

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

To estimate the pressure surge caused by a pump trip, and the likelihood of the pumps being
forced into reverse rotation
To examine strategies to alleviate the problems, such as closing the pump outlet valves or
installing an expansion tank.

The network is shown below:

2.1.

Scenario 1 - Cooling Water Pump Trip without Valve Closure

In our initial scenario, the main pumps trip after 10 seconds while the outlet butterfly valves remain
open. The inlet pressure of the cooling pumps is fixed at 0 barg, as if they are linked with a low level
supply pool.

14

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

Once the pumps trip, the water level in the heat exchanger banks will drop. This has the effect of
reducing over time the pressure forcing the pumps into reverse flow. The vapour cavitation model
used for the heat exchangers and the pipes in the tower ignores this, which increases the safety
margin in our calculations.
The calculated maximum pressure is 41.25 m fluid (gauge) and the minimum pressure is -10.0376 m
fluid (gauge) which indicates cavitation occurred in the cooling tower after the pump trip. This is
because the vapour pressure is -10.0376 m. The maximum reverse rotation of the pump is 5.85
rev/sec, which is about 82.6% of the rated speed but in the reverse direction.

15

PIPENET
2.2.

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

Scenario 2 - Cooling Water Pump Trip followed by Valve Closure

In order to reduce the reverse rotational speed of the pumps, in this scenario we will close the
butterfly valves downstream of the cooling water pumps, using a cubic closure pattern over 30
seconds.
Now the calculated maximum pressure is 39.55 m fluid (gauge) and the minimum pressure is still 10.0376 m fluid (gauge), so cavitation again occurred in the cooling tower. However the maximum
reverse rotation of the pump decreased to 3.64 rev/sec, which is about 51.4% of the rated speed but
in the reverse direction.

2.3.

Scenario 3 - Cooling Water Pump Trips with Expansion Tank

16

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

In this scenario, a 5 m x 5 m expansion tank is installed at the inlet of the cooling pumps. The
butterfly valves downstream of the cooling pumps are kept open. The idea is that the water level in
the expansion tank will increase after the pumps trip at 10 seconds, which can depress the growth of
cavitation at the high parts of the network (for example, the heat exchangers in the cooling tower, or
the pipes at the outlet of the condenser). The water level in the expansion tank is at about 2.5 m
initially.
The calculated maximum pressure is 23.59 m fluid (gauge) and the minimum pressure is -10.0376 m
fluid (gauge), so again cavitation occurred in the cooling tower after the pump trip. But now the pump
speed is always positive.

17

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

18

2.4.

Conclusions:

In all of our scenarios, cavitation will happen in the network after the pumps trip. Therefore, it may be
necessary to install vacuum breakers at high positions to protect the network from a pressure surge
due to cavity collapse.
Closing the valves downstream of the cooling pumps can protect the pumps by reducing the reverse
speed after the pumps trip.
An expanding tank with a correct size and water level can significantly depress the pressure surges
and protect the pumps from high reverse speed after the pumps trip.

3. Frequently Asked Questions

3.1.

More about PID Controllers:

Basic PID controller equation is shown below

    
     
 

Controllers made simple



 !

   




"

K is the Gain.
TI is the Reset Time.
TD is the Rate Time.
u, us is the Output and output set point.
y, ys is the Input and input set point (also called reference signal).
Q: How to get initial steady state when a network has PID controllers?
A: PID controller is a time-based component, which steady state is only achievable by time.
Therefore, the run-in time should be long enough and the option of Calculate initial steady state

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

should be unselected. In addition, a reasonable output set point is helpful to get initial steady state
quickly.
Q: How the input set point affects the simulation?
A: The input set point is the control target which represents the desired value of the controlled
variable at the steady state.
Q: Does the output set point affect the simulation?
A: No. The output set point is used to set the initial output signal. However, a reasonable output set
point can save much run-in time to reach initial steady state.
Q: When the gain is negative or positive?
A: The negative gain (negative feedback) is used when the operation (setting) of the controlled
element has opposite tendency with the input signal. In the control system below, the PIDs gain is
negative because the bypass valve tends to open further when the measured flow rate is below the
set point.

If the flow sensor is moved to the main pipeline, the PIDs gain should be positive because the
bypass valve tends to open further when the measured flow rate is above the set point, see the figure
below.

Q: How to Estimate the Gain?

A: The network may need extremely long time to settle down if the gain is too small. On the contrary,
the control system may cause unstable oscillation if the gain is too big. The correct scale should be
based on the sensitivity of control signal to the operation of the control element. In the above
example, the flow rate at the main pipeline increases about 334 l/min when the bypass valve

19

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

changes the position from opening to closure. Therefore, the gain can be set as (1-0)/334 = 0.003
1/l/min.
Q: How to Set Reset Time and Rate Time?
A: Reset time is used to scale I-term (integral) whilst rate time is used to scale D-term (differential).
The output control curve tends to be smooth with big reset time but the system needs more time to
settle down. However, small reset time or big rate time may lead oscillations. Generally the rate time
should not more than 25% of the reset time.
Q: How Anti-windup Affects the PIDs Behaviour?
A: Anti-windup introduces a negative-feedback signal to offset the saturated I-term signal, which
makes PID controllers quickly respond to the input signal after the saturation is over. Tracking time is
required when the anti-windup function is activated. Ideally the tracking time should be larger than
the rate time but smaller than the reset time.

Q: How to Model a Deluge Valve in PIPENET Spray-Sprinkler Module

A: Deluge valve is a very important and common equipment in the fire protection systems.
PIPENET Spray/Sprinkler Module provides three different methods to help our users to model the
deluge valves.
3.2.

PIPENET Spray/Sprinkler Module:

Method 1: To use the deluge valve model.

The equation for the deluge valve model in PIPENET goes back to the early 1980s. Angus Fire
Armour (Thame and now part of Kidde) used to publish data for the values of K and x for their deluge
valves those days. So, some users asked for this equation to be built into PIPENET. This was done.
A few years later Angus stopped publishing this data. However, we could not remove this model
because there were users with data files which had this model. Nowadays it is there largely for
historical reasons.
The deluge valve equation in PIPENET Spray/Sprinkler Module is
#$ 

%&
'

where:
DP is the pressure drop across the valve.
Q is the (volumetric) flow rate through the valve.
K is a constant for the valve.
X is a constant for the valve (typical values are 1 or 2).
If you have a conventional valve Cv or Kv value from your valve manufacturer which has the
following formula

20

PIPENET

NEWS

VOLUME 2, ISSUE 10

(  )*+#$,

AUTUMN 2012

(  *+#$,

or

or
% 2

#$  , ./0 1

or

% 2

#$  , .'0 1

21

where:
G is the specific gravity of fluid (1 for standard water)
Then we can rewrite the above equations as
#$ 

/0 3

or

#$ 

'0 3

Comparing this equation and the equation for the deluge valve, we can get
X=2
and
  )*2 ,

or

  * 2 ,

Please note: when we use the above formula to input K for the deluge valve in PIPENET
Spray/Sprinkler Module, we should set the flowrate and pressure units in PIPENET same as the units
given with the valve Cv value in your valve data. Of course, you can always change to other units
after you input K, PIPENET will calculate the new K value according to your new units.
Furthermore, you can use the following two methods to model a deluge valve as well:
Method 2 - To use an equipment item
Some engineers, who do not wish to use the above model, use the equipment item model instead.
The equipment item model uses the equivalent length for the deluge valve.
For items such as fittings which are not found in the NFPA rules but for which you have some data,
we can use "equipment item" to model them. It is very common to model items such as deluge
valves, strainers and so on in this way. To use the "equipment item", please follow the procedure
below. I have attached some screen shots to show the steps. Suppose a strainer has an equivalent
length of 30 m, you have to go through 3 steps.
1. Pick up the equipment item using the cursor.

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

2. Please it on the pipe where the strainer is located.

22

3. Enter the name and equivalent length of the strainer.

Method 3: To use the elastomeric valve

PIPENET has the capability of modelling elastomeric valves such as Inbal valves. With this model it
is not even necessary to specify the Cv value of the valve. The user can specify the required
flowrate and PIPENET will calculate the pressure drop across the elastomeric valve which is
necessary to achieve the specified flowrate.
3.3.

PIPENET Standard Module:

Q: How to Use Fixed Pressure Drop Item in PIPENET Standard Module

A: The fixed pressure drop item
in PIPENET Standard Module enables user to specify a fixed
pressure drop in a very easy way; however, we suggest our customers to think more carefully before
you add a fixed pressure drop item to model the devices for the following reasons.
1. Those devices normally have a pressure drop as a function of velocity; however, fixed
pressure drop component means that the pressure drop is fixed for all velocities.
2. For networks with fixed pressure drop components, the given specifications might not be able
to satisfy the required pressure drop, and consequently the network cannot be solved.

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

If your network is one of the above cases, we recommend to the use of fittings to model the devices
which the fixed pressure items are used to model in your network. If you know the pressure drop
value and its corresponding flowrate, you can use the "device" fitting; or if you have enough
information from the manufacturer, you can directly use the "K-factor" fitting.
23

3.4.

Miscellaneous:

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

Q: Upwards Compatibility - How do I open older files of PIPENET in the latest version?
A: Pre PIPENET Vision versions (now known as PIPENET Classic) used .dat files. It had the
supporting files .pdf, .plb and so on. The path names for the supporting files were explicitly
embedded in the .dat files. So the first thing you have to make sure is that the path names in the .dat
files point to the actual locations of the supporting files and their correct names. If the path names or
the file names are incorrect then you have to edit them using a text editor. Once this is done you can
simply open a .dat file in PIPENET.

You must first select All Files (*.*) in the drop-down box next to the filename:

Once opened the data can be saved as a .sdf (which will also update your library file to be a .slf).
Q: Why am I getting warnings in the status of my simulation?
A: Note that the warnings status does not mean the simulation has failed. It only means that some
requirements have not been met; it may still be able perform a calculation. You may need to change
some of the parameters in your network. For example, the maximum speed along a pipe is

24

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

exceeded or a nozzle does not receive the correct flow rate. This will require you to make changes to
your network, such as change pipe diameters.
The following is also possible. PIPENET perform network flow analysis by iterative calculations. It is
possible that during an intermediate iteration a warning was raised. This warning may not be
relevant for the final solution.
Q: While using spray/sprinkler module, why is it not possible to add a pipe to the network?
A: Before you can add a pipe in the spray/sprinkler module, you must add a pipe to the pipe types:

Q: How do I change the volumetric flow units?

A: Using the user-defined units, ensure you first double-click on the Flowrate option and then
double-click on the volumetric flow rate option. The following list of volumetric flow rates shall appear:

Q: Why am I am getting the error Network is over specified in the check network dialog box?

25

PIPENET

NEWS

VOLUME 2, ISSUE 10

AUTUMN 2012

A: Note that the following rules governing specification:

Let Nio be the sum of the input and output nodes.
Phase
Design
Analysis

Pressure specifications (Np)

1
1

Flow rate specifications (Nf)

Nio-1
Nio- Np

As a rule of thumb, for the analysis phase you must have a total of Nio specifications. Ensure at least
one of these is a pressure specification. You must however set the Analysis specification pressure for
the inlet nodes of the pumps that are switched off as there is no equation relating the flow rate to the
pressure for a pump that is switched off.
Note that for the Spray/Sprinkler module, one design pressure specification is already provided
through PIPENET and hence, you need only enter Nio-1 flow rate specifications. For the analysis
phase, under the user-defined calculation, you must obey the same rules as in the table. Note,
however, for the most-remote nozzle option, the pressure and flow rate will typically be specified at
the most remote output, and the input node would be left simply as an input (but with no pressure or
flow rate specified).

CONTACT US:
Technical Support:

support@sunrise-sys.com

Sales:

pipenet@sunrise-sys.com

SUNRISE SYSTEMS LIMITED,

CAMBRIDGE CB25 9QZ,
UNITED KINGDOM
Web site: www.sunrise-sys.com
Copyright 2012 SUNRISE SYSTEMS LIMITED.
All rights reserved.

26