SIMULATING COOLING WATER

CIRCUIT DYNAMICS

MENTOR GRAPHICS, MECHANICAL ANALYSIS DIVISION

W H I T E P A P E R

M E C H A N I C A L A N A L Y S I S

w w w . m e n t o r . c o m
Simulating Cooling Water Circuit Dynamics

Public concern of the nature and extent of our power generation network has increased at a time when global
consumption of energy has escalated dramatically. Concerns over issues such as security of fuel supply, plant safety
and emissions, and climate change means that energy policy is a subject of intense public debate. In such an
environment, it is more important than ever that those involved in the design, build and operation of current and
future generations of power plants have at their hands all the tools required to demonstrate that they are meeting
and exceeding the efficiency, safety and emissions targets expected of them. As designers modernize, retrofit, add
additional safety measures, meet environmental regulations or change fuel sources, they can no longer afford to
depend on rules of thumb or expensive and time consuming prototyping. As well, to accurately analyse transient
operation, the use of approaches such as spread sheets is lacking.

Flowmaster® has provided designers and operators of power stations crucial information on proposed or existing
installations for decades. Underpinned by a raft of empirical data (much of which was documented by DS Miller in
Internal Flow Systems), Flowmaster was originally conceived as a tool for evaluating the nature and extent of water-
hammer events in liquid systems. While still acknowledged as the class leading tool in this area, Flowmaster has
since diversified into compressible gas and two-phase steam simulations.

This paper will focus on Flowmaster being applied to a cooling water network, in order to demonstrate how the
condenser component can be used to help generate a detailed picture of the response of such systems to
transient events.

THE COOLING WATER CONDENSER CIRCUIT

Cooling water networks are critical to the operation of all thermal power plants. The Rankine cycle upon which all
fossil fuel and nuclear power plants depend requires that the steam that is expanded across the turbines is
condensed back to water before being returned to the furnace. This requires vast quantities of water be passed
through the steam on one side of a large heat exchanger (see below) before being returned to the source; usually
the sea or an inland waterway of sufficient capacity.

Figure 1 - A Surface Condenser being loaded for delivery

(Image courtesy of Bronswerk Heat Transfer)

w w w. m e nto r. co m /m e ch a n i c a l
2 [ 8]
Simulating Cooling Water Circuit Dynamics

Condensers are critical and expensive components, so it’s vital that their performance be understood in the
context of the system around them. As well, optimising the cooling water system for its design point, it’s essential
to understand how it will respond in the event of unplanned shut downs arising from pump trips or safety valve
closures.

Consider the following illustrative example; we will assume that Flowmaster has already been used to size the major
components at the steady-state ‘design point’. The remaining questions relate to shut downs and ensuring that the
plant is able to do so in a safe and controlled manner, even if the control procedure itself fails.

Figure 2 - Example Cooling Water Network

In the network shown above, two condensers are fed by two centrifugal pumps each. After passing through the
condensers, the flow is discharged back to the water source, in this case simply modelled by a pair of reservoir
components. Although not shown in this simple illustrative example, Flowmaster allows users to model weirs of
different types at the end of such lines. The standard component library contains models for broad-crested, sharp-
edged rectangular and V-notched weirs, each of which can be added to a simulation should the sizing or transient
response of such components be a requirement.

In the event of a pair of pumps tripping, an intermediate control valve is designed to open in order to allow the
remaining two pumps to drive both condensers. Simulating such a set-up in Flowmaster allows the user to:

■■ Verify that the designed procedure is adequate for safe operation (with reference both to the pressures during
the transient and the cooling water flow rates following it)
■■ Judge the impact of this procedure on the performance of the circuits
■■ Establish the consequences of an unmanaged shut down (i.e. when the switch over valve fails to operate)
■■ Judge what needs to be put in place to ensure that even in this extreme case equipment integrity and plant
safety aren’t put at risk
In order to allow such scenarios to be considered, Flowmaster contains detailed models of cooling water
condensers which allow for factors such as water box geometry, installation of gas injection valves and the
presence of entrained gas cavities in the water boxes to be considered.

w w w. m e nto r. co m /m e ch a n i c a l
3 [ 8]
Simulating Cooling Water Circuit Dynamics

In order to check the validity of the controlled shut down procedure described above, time histories of pressure at
key points in the circuit and water box liquid levels can be reviewed. The perturbation on the system is as follows:

■■ The two pumps powering the uppermost (in the schematic) of the condenser lines trip. Note how the rate of
change of speed of the pumps is a function of the torque placed on the shaft by the impeller and is inversely
proportional to the pump and motor inertias.

■■ Two seconds after the pump trips, valves immediately downstream of the pumps begin to close in order to
prevent any damage arising from reversed flow (red line).

■■ At the same time, the valve connecting the two circuits begins to open allowing the remaining two operating
pumps to supply both networks.

w w w. m e nto r. co m /m e ch a n i c a l
4 [ 8]
Simulating Cooling Water Circuit Dynamics

In order to judge whether this procedure is enough for a controlled shut-down, we might choose to observe
whether any pressure peaks or vapour cavities are formed in the pipe network. In addition, it may be prudent to
check if the levels in the water boxes are affected.

In terms of checking whether unduly high or low pressures have occurred, the most convenient way is to annotate
the schematic with the maximum and minimum values attained at any given point over the course of the
simulation. A review of this output (see below), indicates that the designed procedure doesn’t cause any pressure
excursions of a magnitude that may cause concern.

It is of course possible to gain more detail at any node or component in the network, but this method offers a
convenient means by which potential problems can be quickly identified.

Having considered the pressures along the entire circuit, a more detailed examination of the condenser itself is in
order.

Figure 3 - Condenser Symbol & model schematic

The condenser model itself is unique to Flowmaster and has been developed over many years to allow users to
accurately predict the transient response of a shell and tube type condenser and its connecting circuit. It allows
users to consider changes in water box level, flow rate and pressure as well as the effect of installing gas injection.
This particular model isn’t intended to cover heat transfer, but this is possible using other Flowmaster heat
exchanger components discussed later in this document.

w w w. m e nto r. co m /m e ch a n i c a l
5 [ 8]
Simulating Cooling Water Circuit Dynamics

The geometry of the condensers in this model is

that the water boxes are mounted 3.7m above the
datum level, are 2.6m in height and are shaped as
per the schematic shown on the right of Figure 3.

Note that it is possible to specify any required water

box geometry via an area vs. level plot. The tube
bundle extends vertically to 0.35m below the top of
the water boxes. One assumption that will be made
is that the liquid level must not drop below this
point. Were it to do so, water hammer and plant
shut down are more than likely going to happen.
Examining the levels in both water boxes (right)
shows that the inlet level remains unaffected,
whereas the outlet water box sees a small, oscillating drop in level. However, this doesn’t extend beyond the top of
the tube bundle and ultimately refills once the supply from the two remaining operating pumps is in place.

In broad terms, the designed procedure is adequate. It would be entirely possible to use Flowmaster to gain more
detail on any individual aspects of the network, but for the purposes of this example it will be assumed that this is
adequate and the case where the planned changeover doesn’t happen will now be considered.

In this case, the two-pumps trip as before, but the

valve linking the two circuits fails to open and as a
result the uppermost condenser completely shuts
down. Predictably, this is a much more dramatic
case. As the pumps spool down and their shut
down valves close, the pressure in the system
immediately downstream of the valve drops off and
a vapour cavity forms which grows and ultimately
collapses around 22 seconds into the simulation
(see the red line in the plot right). The resulting
pressure spike reaches 11bar which is significantly
above the normal operating pressure of the line.
This combination of cavity growth, collapse and the
pressure spike which follows are classic water
hammer phenomena. The impact on the
condenser is equally dramatic. Both water boxes
drain far below desired levels, with the outlet box
only restoring its level through flow reversal: water
which has previously been through the tube bundle
is ‘sucked’ back in to the condenser and fed back
through the tubes. The inlet water box doesn’t
recover and the negative level reported in the trace
opposite represents that the water level has
dropped below the bottom of the box and now
only extends along a portion of the pipe feeding
the condenser. Another consequence is a pressure
surge within the inlet water box of over 8bar, which
occurs as a result of an initial re-filling of the water
box; around 29 seconds.

w w w. m e nto r. co m /m e ch a n i c a l
6 [ 8]
Simulating Cooling Water Circuit Dynamics

Under these circumstances, it’s extremely likely that

there would be considerable damage to equipment
if no attempt were made to ameliorate these
effects.

Flowmaster allows the user to virtually prototype a

range of such measures, ranging from gas injection
in to the condenser itself, to surge tanks and
bladders elsewhere in the system.

The simulation has indicated that there are two

problems to be addresssed:

■■ Cavity formation and collapse immediately downstream of the pumps. Depending upon the pressure rating
of the pipework selected, this may mean that damage is possible.
■■ A reduction in water box levels such that significant vapour cavities form within the component. Damage to
this piece of equipment appears extremely likely.
The time histories reprinted above show that the pipeline cavity begins to form long before the perturbation
reaches the condenser, so it may be sensible to address this issue first; ideally, this may even resolve the other issues
within the condenser itself and so remove the need for any complex gas injection devices.

It is possible to use Flowmaster’s standard library to simulate a number of surge alleviation tactics, ranging from
bladders containing compressed gas through to surge tanks and gas admission and release valves. In this particular
case, two surge towers will be considered (see Figure 4) and

Figure 4 - Two Surge Towers are Placed Immediately downstream of the Pumps

the network re-run with these devices in place. If appropriately sized, they may be able to prevent the formation of
the vapour cavity and the resulting pressure spike.

w w w. m e nto r. co m /m e ch a n i c a l
7 [ 8]
Simulating Cooling Water Circuit Dynamics

Taking the pressure at this location shows that this

does indeed prove to be the case. Compare the
plot shown opposite with the unprotected case at
the same location where the pressure spiked to over
10bar following the formation of a significant
vapour cavity. Although the pressure does drop
marginally below atmospheric for a brief period, no
cavity forms and thus no water hammer event
occurs. Flowmaster can be used in order to further
optimise the sizing of such devices to ensure that
the correct device is specified and excessive
installation costs are avoided. Significantly,
Flowmaster also allows the user to gauge the
impact of such an installation elsewhere in the
network, as can be seen when considering the
response of the condenser.

Examination of the condenser waterbox levels

shows that the installed surge tanks are sufficient to
maintain levels within each. It can therefore be
seen that the installation of the upstream towers
protects the condenser as well. The potential is
therefore to insure against an uncontrolled pump
trip event through the use of a single measure
upstream of the condenser.

Although this example network has been simplified to some degree, it serves to demonstrate how the use of
Flowmaster’s condenser component as part of a transient simulation can help engineers understand the nature of
controlled and uncontrolled transient events. The requirement for potentially expensive surge aleviation devices
can be analysed and their nature and extent properly sized.

For the latest product information, call us or visit: w w w . m e n t o r . c o m

©2013 Mentor Graphics Corporation, all rights reserved. This document contains information that is proprietary to Mentor Graphics Corporation and may
be duplicated in whole or in part by the original recipient for internal business purposes only, provided that this entire notice appears in all copies.
In accepting this document, the recipient agrees to make every reasonable effort to prevent unauthorized use of this information. All trademarks
mentioned in this document are the trademarks of their respective owners.

MGC 07-13 TECH11230