70 40

‘Delta T’ Design Guide

Low Carbon System Design - a whole system approach

“70ºC flow / 40ºC return”

www.sav-systems.com
Rev: 2.0 10/2014
References
• Danfoss Technical Paper: Results and experiences - Lystrup 12/2013
• CIBSE Knowledge Series Guide KS7: Variable flow pipework systems 2004
• CIBSE Commissioning Code W: Water Distribution Systems 2010
• CIBSE AM12 2013
• BSRIA Commissioning Water Systems guide BG2/2010
• BSRIA Guide BG12/2011: Energy Efficient Pumping Systems 2011
• SAV FlatStation Design Guide 2014
• SAV LoadTracker CHP Design Guide 2014
• GLA’s District Heating Manual for London 2013

This document is based on the best knowledge available at the time of publication. However no responsibility of any
kind for any injury, death, loss, damage or delay however caused resulting from the use of these recommendations
can be accepted by SAV Systems, the authors or others involved in its publication.
 ‘Delta T’ - DESIGN GUIDE
70 40

SYMBOLS

ISOLATING VALVE
PRESSURE INDEPENDENT
CONTROL VALVE
DRAIN OFF COCK

PUMP
LOCKSHIELD VALVE

SAFETY RELIEF VALVE

DOUBLE REGULATING VALVE

FIXED ORIFICE FLOW MEASUREMENT

PRESSURE GAUGE
DEVICE (ORIFICE PLATE)

FIXED ORIFICE DOUBLE REGULATING TEMPERATURE GAUGE

VALVE (COMMISSIONING SET)

FLEXIBLE HOSE
CONSTANT FLOW REGULATOR
FLEXIBLE COUPLING
DIFFERENTIAL PRESSURE
CONTROL VALVE STRAINER

2 PORT CONTROL VALVE BLANKED FLANGE PIPE END

  PRESSURE TEST POINT

3 PORT CONTROL VALVE

NON-RETURN VALVE

4 PORT CONTROL VALVE

AUTOMATIC AIR VENT

THERMOSTATIC RADIATOR VALVE

Guide’s technical author, Chris Parsloe of Parsloe Consulting Ltd

Chris is the author of numerous industry guides on variable flow heating

and chilled water systems including:
• CIBSE Knowledge Series Guide KS7: Variable flow pipework systems 2004
• CIBSE Commissioning Code W: Water Distribution Systems 2010
• BSRIA Guide BG12/2011: Energy Efficient Pumping Systems 2011

3
 ‘Delta T’ - DESIGN GUIDE
70 40

1. INTRODUCTION
In order to realise the predicted energy saving benefits of commercial heating systems and heating networks, it is
essential that pipework distribution systems be designed with proper regard to system operating temperatures.

This design guide makes frequent reference to ‘delta T’, which is the difference in temperature between the flow
and return water in a piped heating distribution system. The selection and maintenance of system flow and return
temperatures requires close attention if the energy and cost saving potential of a system are to be realised.

In those cases where the anticipated energy savings of heating distribution systems do not materialise, these failings
can frequently be traced to the inability to maintain ‘delta T’.

There are three key objectives of successful ‘delta T’ designs:

• The operating ‘delta T’ across heat sources should be as close as possible to their optimal
design values. This will enable them to operate at close to their peak efficiencies.
• The heat dissipation from, and hence ‘delta T’ across secondary distribution circuits should
be maximised. This will give rise to lower system flow rates, smaller pipes and smaller pumps that
consume less energy. Furthermore, buffer vessels or thermal stores can be reduced in size relative
to the size of the system.
• Secondary flow and return water temperatures must be kept at as low as possible. This will
help to minimise pipe distribution heat losses.
Some designers take the view that to influence system operating temperatures is impossible and that other
considerations should take priority. This guide challenges this view by explaining how good ‘delta T’ design
can be undertaken in a way that delivers significant energy and cost saving advantages.

4
 ‘Delta T’ - DESIGN GUIDE
70 40

2. PRIMARY CIRCUIT DESIGN

Proper design of the primary circuit is essential. Most low carbon heat sources have specific flow and return water
temperature requirements that enable them to perform at their optimal efficiencies. The aim should be to ensure
that these conditions are maintained for the majority of the operating period.

Table 1 shows typical recommended operating temperatures for different types of low carbon heat source.

Heat source Optimal primary circuit Reasons

‘delta T’

Flow (°C) Return (°C)

Gas condensing 55 30 Condensing of flue gases is only possible at

boiler temperatures of less than 55ºC and full
condensation only occurs at less than 35ºC.

Biomass boiler 80 60 High grade heat is available from biomass

boilers but return temperatures must be
maintained at greater than 60ºC to avoid
condensation of flue gases which might lead
to corrosion within the boiler.

CHP 80 -90 70 High grade heat is available from CHP units but
low flows and excessive delta T values may
cause uneven cooling of the CHP unit.

Heat pump 40 35 Heat pumps produce low grade heat.

The efficiency and output of heat pumps is
significantly reduced at higher flow
temperatures.

NB: These values apply to localised primary flows only. Secondary circuit design should be with
the lowest possible return temperatures and the largest possible ‘delta T’ values for the reasons
explained under the Importance of ‘delta T’ (see page 12).

Table 1: Typical primary circuit operating temperatures

It is likely that different types of heat source will be used in the same system (such as gas boilers providing back-up
for a biomass boiler or CHP unit). Where this is the case the lead heat source should be the one that has the lowest
carbon emissions, as recommended by Part L of the Building Regulations. It then follows that the primary circuit flow
temperature should be dictated by the lead heat source e.g. if water at 80°C is available from a CHP unit, then this
should be the design flow temperature for the primary circuit. However, for reasons that will become clear, secondary
circuit temperatures at terminals should be as low as possible.

5
 ‘Delta T’ - DESIGN GUIDE
70 40

Figure 1 shows a primary / secondary circuit layout that enables this objective to be achieved. The following notes
apply to each of the labelled features.

WEATHER
COMPENSATION

DELTA ‘T’ D
CONTROLLER

AUXILIARY AND BUFFER/THERMAL

BACK-UP HEAT STORE
SOURCES
F
H
B
G
A

B
PRIMARY INJECTION
A CIRCUIT
C
CIRCUIT

B
E
A AUXILIARY HEAT
LEAD HEAT SOURCE
SOURCE

PRESSURISATION
UNIT

Figure 1: Primary circuit schematic showing low carbon heat source options

6
 ‘Delta T’ - DESIGN GUIDE
70 40

A. All heat sources connected to the primary circuit should be fitted with a circulating pump and mixing valve
arrangement. This is for one or both of the following purposes
a. T o ensure that the flow temperatures from each heat source are held constant and at the same set point value
- this being dictated by what best suits the lead heat source. The maintenance of a constant combined flow
temperature from heat sources is important for proper functioning of buffer vessels and thermal stores. If the
flow temperature from a heat source is ever less than the set value (as might be the case during system start-up
conditions) flow water can be mixed in with return water until the required flow temperature is achieved.
b. T o ensure that the return temperature does not drop below a minimum value as specified by the heat
source manufacturer. This is particularly relevant in the case of biomass boilers that require a minimum return
temperature of 60°C to prevent internal corrosion. As for (a) above, if the return water temperature is less than
the required value, flow water can be mixed in with return water until the required return temperature is achieved.
B. Each of the pumps serving the various heat sources should be sized such that it is capable of delivering the design
flow rate through the heat source it serves against the design pressure loss in the entire circuit between the heat
source and the thermal store. These pumps should be variable speed but, during commissioning, they should each
be set at a fixed speed sufficient to achieve the required design flow rate through the heat source when circulating
through the thermal store. The duties of pumps supplied as integral to the heat source should be checked to ensure
that they have sufficient capacity to deliver flow to the thermal store.
C. A “de-coupler” should be included that includes a buffer tank or thermal store. The direction of flow through the
de-coupler and storage vessel may be in either direction, depending on the system operating condition.
D. The secondary pumps should be variable speed pumps sized to deliver the anticipated maximum design flow rate
against the secondary circuit design pressure loss. The flow temperatures for each secondary circuit should be at
as low a value as possible and each circuit should have the means to vary the flow temperature to suit the particular
operating conditions. Constant temperature secondary circuits should never be used since they risk delivering high
temperature return water back into the primary circuit.
E. Where operating temperatures are compatible, the lead heat source can be supplemented by a heat source
located in the main system return that pre-heats the water before it reaches the lead heat source. For example,
this opportunity exists in a system where the lead heat source is a CHP unit that has been selected to heat water
from 70°C to 80°C whilst secondary circuits are designed to deliver a return temperature of 40°C. In this case a
heat pump might be used to pre-heat the returning water from 40°C to 45°C or, alternatively a gas condensing
boiler might be used to pre-heat the water from 40°C to 55°C. It can be seen that this enables the auxiliary heat
source to operate at temperatures that achieve optimal efficiency. The heat source must include its own pump
capable of injecting heated water back into the return pipe. This means that the heat output available is limited
by the design flow rate in the primary circuit - since the flow rate through the heat source cannot exceed the
design flow rate in the primary circuit.
F. Additional auxiliary or back-up heat sources can be connected in parallel with the lead heat source. An auxiliary
boiler will supplement the heat output of the lead heat source whereas a back-up boiler will only run when other
heat sources are not available. Since each of these heat sources must achieve the same flow temperature as the
lead heat source, it may be necessary to diverge from optimal operating temperatures. For example a condensing
boiler operating as an auxiliary or back-up to a CHP unit would have to operate at a flow temperature that suited
the CHP unit (typically 80°C) thereby reducing the potential to operate in condensing mode. For this reason,
auxiliary or back-up heat sources in these locations should be the last to operate under sequencing controls.

7
 ‘Delta T’ - DESIGN GUIDE
70 40

G. An injection pump is required to push water from the primary circuit into the header from which secondary
circuits are attached. This will ensure that the secondary circuits cannot affect flows in the primary circuit.
A two port control valve ensures that the flow of water into the header is sufficient to maintain the required
main secondary flow temperature. If necessary, this temperature can be further mixed to lower temperatures
at individual secondary circuits. An additional sensor in the low loss header can be used to detect any rise in
temperature suggesting a reversal of flow through the header. The two port valve should then throttle to
reduce flow if this should happen.
H. Buffer stores and thermal stores are generally of benefit to all heating systems and are a critical aspect of system
design. The following subsection explains their role in detail:

Buffer vessels / thermal stores

Buffer vessels / thermal stores are located in the de-coupler of the primary circuit. Although a single vessel often
serves a shared function, the separate roles of buffer vessels and thermal stores can be considered as follows:

• Buffer vessels. A buffer vessel enables the heat source to run at its optimal return temperature for as
long as required. For example, if there were no buffer vessel in Figure 1, a proportion of the heated
water would inevitably short circuit through the de-coupler back to the heat source thereby raising the
temperature of the return water. For condensing boilers, this might force them into a non-condensing
condition reducing their efficiency. By installing a buffer vessel, each heat source can operate for a fixed
period, as determined by the size of the buffer vessel, without the risk of elevated return temperatures.
In the case of biomass boilers this is particularly important since elevated return temperatures may result
in over-heating of the water inside the boiler. Furthermore, biomass boilers continue to generate heat
for a significant period after being switched off. This residual heat needs to be accommodated by the
system, so an appropriately sized buffer vessel is essential.
• Thermal stores. A buffer vessel that is designed to store heated water generated during periods of
relatively low heating demand, in preparation for an anticipated high demand, is referred to as a
“thermal store”. For CHP units, it may be the case that the unit runs during periods when there is a
demand for electricity but little demand for heat. An appropriately sized thermal store enables the CHP
unit to run during these periods, creating a store of heated water which can then be used later on when
a demand for heating arises. Alternatively, a thermal store may be required in order to deal with high
short term heat loads in the system. This might typically be the case in a system serving FlatStation heat
interface units (HIUs) that provide instantaneous heating of hot water. These units cause high short term
energy demands during periods of peak hot water draw-off. Rather than sizing the main heat sources to
cope with these short term loads, a thermal store can be used to create a store of heated water capable
of dealing with a high short term demand. This benefits the sizing of the heat sources.
Buffer vessels and thermal stores can be used to control the on / off sequencing of each heat source.
Figure 2 shows a typical buffer vessel or thermal store, where the position of the horizontal separation layer
between flow and return water is used for on / off sequencing. Sensors located in the side of the vessel determine
the movement of the separation layer enabling the various heat sources to be sequenced on and off.
For a buffer vessel, the distance between the sensors will be dictated by the required run time of each heat source.
For a thermal store, the volume of the vessel may be increased to enable the lead heat source to operate for as
long as possible.

8
 ‘Delta T’ - DESIGN GUIDE
70 40

If temp < set-point, auxiliary heat source ON (location F)

If temp < set-point, auxiliary heat source ON (location E)

If temp = set-point, auxiliary heat source location F OFF
BUFFER/THERMAL
STORE If temp < set point, lead heat source ON
If temp = set-point, auxiliary heat source location E OFF

If temp = set point, lead heat source OFF

Volume available for

residual heat if required.

(red – separation layer rising = thermal store emptying)

(blue – separation layer falling = thermal store filling)
Locations E & F shown in Figure 1

Figure 2: Buffer vessel / thermal store used for heat source sequencing

In all circumstances, the key to satisfactory control is to ensure that there is good stratification of temperature
within the vessel. This means ensuring that:

• The pipes entering and leaving the vessel are from the side (rather than top and bottom) in order to
minimise any turbulence that might disturb temperature stratification. Furthermore, the water should
enter and exit the vessel through diffusers, i.e. at reduced velocity, and be directed upwards at the top
connection and downwards at the bottom connection.
• The temperature of the water entering the top of the vessel must be at a consistent fixed value. This
can be achieved by including mixing arrangements on each of the main heat sources as explained in the
preceding section.
• The temperature of the water entering the bottom of the vessel (when the vessel is discharging) must
be maintained at a value that is at least 25°C lower than the heated water at the top of the vessel.
This ensures a good separation layer between flow and return water enabling accurate measurement
of the heated water level in the vessel.

The approach indicated in Figure 2 is simple to implement and provides an effective method of sequencing heat
sources. Figures 3, 4, 5 & 6 demonstrate how EC Power LoadTracker CHP controls various heat sources based on
thermal store separation layers.
EC Power LoadTracker CHP thermal stores take this concept further by incorporating a controller that analyses how the
temperatures in the store have changed over the preceding seven days. This information is then used to calculate the
optimum start time for the CHP unit thereby prolonging CHP run times.

9
 ‘Delta T’ - DESIGN GUIDE
70 40

EC Power LoadTracker under

varying load conditions

Figure 3: Low Load (20%) - Heat Source 1 Only

Figure 4: Medium Load (40%) - Heat Source 1 & 2 Only

10
 ‘Delta T’ - DESIGN GUIDE
70 40

Figure 5: Medium Load (70%) - Heat Source 1, 2, & 3 Only

Figure 6: High Load (100%) - Heat Source 1, 2, 3 & 4

11
 ‘Delta T’ - DESIGN GUIDE
70 40

Importance of ‘delta T’
The selection of secondary flow and return temperatures (which produce ‘delta T’) have a major impact on the installed
cost and energy efficiency of a system such as that illustrated in Figures 3-6.
The implications are best explained by comparing two possible alternatives. These are as illustrated in Figures 7 and 8.
For each pair of flow and return pipes the flow rate can be estimated using the equation

qm = P / (cp DT)

where,

qm = mass flow rate (kg/s) (≈ volume flow rate in litres / second)

P = Heat load (kW)
cp = specific heat capacity of water (= 4.18kJ / kgK)

12
 ‘Delta T’ - DESIGN GUIDE
70 40

SECONDARY PUMPS

80°C
1.2 litres/sec
100kW

80°C
BUFFER/THERMAL 1.2 litres/sec
100kW
STORE
2.4 litres/sec
BACK-UP BOILER

2880 litres

AUXILIARY BOILER 100kW

80°C 60°C
PRIMARY
CIRCUIT
60°C
60°C
1.2litres/sec
CHP UNIT 100kW
80°C

2.4 litres/sec 60°C

60°C
1.2litres/sec

Figure 7: System operating at ‘delta T’ of 20 (=80 / 60) in primary and secondary circuits
 

Figure 7 shows a primary circuit into which a CHP unit and an auxiliary gas condensing boiler, each rated at 100kW,
are supplying water at 80°C. A single constant temperature secondary circuit distributes the water at 80°C to terminal
units that are sized to generate a 20°C ‘delta T’ giving a return temperature of 60°C. A buffer vessel is included to store
water at the 80°C flow temperature.
For this system (ignoring the intervention of heat source sequencing controls):

• The peak flow rate in the primary circuit is 2.4 l/s - i.e. 200 / (4.18 x (80-60))
• The peak flow rate in the combined secondary circuits is also 2.4 l/s (since load and ‘delta T’ are
the same as for the primary)
• A buffer vessel sized to provide a notional 20 minutes of flow at 80°C would need to have
a volume of 2880 litres i.e. 2.4 l/s for 20 minutes.

13
 ‘Delta T’ - DESIGN GUIDE
70 40

WEATHER
COMPENSATION
SECONDARY PUMPS

70°C
DELTA ‘T’ 0.8 litres/sec
CONTROLLER 100kW

70°C
BUFFER/THERMAL 0.8 litres/sec
100kW
STORE
1.2 litres/sec
BACK-UP BOILER

1440 litres

AUXILIARY BOILER 100kW

80°C 40°C
PRIMARY
CIRCUIT
INJECTION
40°C
CIRCUIT
40°C
0.6 litres/sec
CHP UNIT 100kW
80°C

1.2 litres/sec 40°C

40°C AUXILIARY HEAT
0.6 litres/sec
SOURCE

PRESSURISATION
UNIT

Figure 8: System operating ‘delta T’ of 20 (=80 / 60) primary and 30 (=70 / 40) secondary circuits

Figure 8 shows the same primary arrangement but with a variable temperature secondary circuit where the flow
temperature is 70°C and terminal units are sized to generate a 30°C ‘delta T’ giving a return temperature of 40°C.
For this system:

• The peak flow rate in the primary circuit is 1.2 l/s - i.e. 200 / (4.18 x (80-40))
• The peak flow rate in the combined secondary circuits is 1.6 l/s - i.e. 200 / (4.18 x (70-40))
• A buffer vessel sized to provide a notional 20 minutes of flow at 80°C would need to have
a volume of 1440 litres i.e. 1.2 l/s for 20 minutes.

14
 ‘Delta T’ - DESIGN GUIDE
70 40

There are significant system cost and energy efficiency benefits of the Figure 8 system compared to the Figure 7 system.
These are summarised as follows:

• The primary flow rate is halved resulting in smaller primary pipes and pumps
• The primary pump power is halved
• The required buffer vessel / thermal store volume is halved
• The secondary flow rate is reduced by a third resulting in smaller secondary pipes and pumps.
• The secondary pump power is reduced by a third
• Heat losses from secondary distribution pipes are reduced (due to smaller pipes operating
at lower temperatures)
• There is the potential to add an auxiliary heat source such as a gas condensing boiler in the secondary
return (i.e. feature E in Figure 1) where it can operate in full condensing mode by pre-heating the
water from 40°C to 55°C.

These benefits will have a major impact on life cycle costing calculations for the alternative systems. In general, the
energy and cost saving benefits increase if secondary flow temperatures are reduced and / or secondary ‘delta T’ values
are increased. The only cost penalty involved is the fact that terminal units may need to be larger in size in order to
compensate for operating at lower flow temperature and larger ‘delta T’. However, in modern buildings where localised
heating loads are relatively small, the implications of having to install marginally larger terminal units is far outweighed
in cost and energy terms by the benefits listed above.

3. SECONDARY CIRCUIT DESIGN

It is essential that secondary circuits dissipate heat effectively from the circulating water. This will enable the designer
to maximise the design ‘delta T’ value and help to ensure that this is maintained (or exceeded) under all operating
condition. If the return water should begin to increase in temperature under certain operating conditions, this may
severely disrupt the operation of the system by:

• Preventing any auxiliary heat sources located in the secondary return from operating;
• Enabling high temperature water to enter the bottom of the buffer vessel / thermal store thereby
disrupting the stratification layers within the vessel.

If the secondary distribution systems are to maintain consistently low return temperatures under all operating
conditions, then careful consideration needs to be given to the elimination of any unwanted routes by which flow
water can find its way directly back into return pipes. The following sub-sections explain the main secondary circuit
design considerations for achieving and maintaining optimal ‘delta T’ conditions.

Variable flow systems

In order to gain any control over system ‘delta T’, the heating system must have a variable flow with 2 port control
valves. This is in contrast to a constant flow system utilising constant speed pumps and 3 or 4 port control valves.
Constant flow systems operate at a fixed pump speed and flow rate throughout the year, regardless of system load.
When zones are satisfied, heat output is reduced by diverting flow water through a by-pass which sends it straight

15
 ‘Delta T’ - DESIGN GUIDE
70 40

back to the heat source. This inevitably causes return temperatures to increase rapidly towards the flow temperature
thereby reducing the ‘delta T’.
By comparison, variable flow systems respond to part load conditions by throttling and reducing flows through terminal
devices. At the same time, pump speed reduces simultaneously to suit the reduced flow requirement. The benefits of
variable flow systems over constant flow systems are therefore:

• Reduced pump energy consumption – pump laws predict that each percentage reduction in pump
speed produces a pump energy reduction of the same percentage cubed;
• Larger ‘delta T’ and lower return temperatures - when the flow rate through a terminal device is
reduced, the water has more time to cool meaning that it returns to the heat source at a lower value.

Variable speed pumping is therefore critical to any ‘delta T’ design.

Pump speed control

Proper control of secondary pump speed will reduce the temperature increase across each pump and minimise pump
energy consumption.
There are three common ways of controlling pump speed as described in the following bullet points.

• Constant pressure method - pump speed is controlled such that the pressure differential across the
pump is maintained at a constant value equivalent to the pressure differential around the system at the
maximum flow rate.
• Proportional method - pump speed is controlled such that the pressure differential across the pump
reduces in proportion to flow rate towards a pre-selected value, typically equal to approximately 50%
of the pressure differential around the system at the maximum flow rate.
• Remote sensor method - pump speed is controlled such that the pressure differential across the pump
reduces towards the pressure differential across the most remote differential pressure controlled sub-
branches, (or alternatively, at a point in the system where the differential pressure is approximately two
thirds of the full load pump differential pressure).

These methods of control are illustrated by the pump and system curves illustrated in Figure 9 - 10 and Figure 11.
For each method of control, the diagrams show the typical change between maximum load and the 50% part
load condition.
It can be seen that the remote sensor option produces the biggest reduction between full load and part load operating
speeds (i.e. between the red and red dotted lines). Since the percentage pump energy saving is equivalent to the
percentage reduction in speed cubed, the remote sensor option provides by far the biggest saving in pump energy
consumption.
This means that for larger systems, the aim should always be to control pump speed based on the signal from a
remote differential pressure sensor located in the system. As a rule of thumb, in systems with a uniform load pattern,
the sensor can be located at a point where the differential pressure is approximately two thirds of the maximum pump
pressure differential. In systems where the load pattern is not uniform (i.e. the load is likely to shift between different
system extremities) multiple sensors should be installed at each extremity to ensure that sufficient pressure and hence
flow is available under all conditions.
Table 2 summarises the alternative methods of pump speed control and the applications for which they are best suited.

16
 ‘Delta T’ - DESIGN GUIDE
70 40

Figure 9: Pump speed controlled to maintain constant pressure

Figure 10: Proportional control of pump speed

17
 ‘Delta T’ - DESIGN GUIDE
70 40

Figure 11: Remote sensor control of pump speed.

Pump speed control method Suitable applications

Constant pressure control Primary circuits feeding multiple heat sources.

Proportional control Secondary circuits with total system pressure

losses of less than 100kPa at maximum load.

Remote sensor control Secondary circuits with total system pressure

losses of more than 100kPa at maximum load.

Table 2: Pump speed control options and applications

Terminal device selection / sizing

Terminal devices emit heat into occupied spaces. In general, the terminals best suited to ‘delta T’ designs are those
that can work with low flow temperatures and generate large temperature differentials. The ability to dissipate heat
is crucial to the performance of the system - an ideal terminal would be one that returns water at ambient
air temperature having lost all of its heat to the surrounding space.
Of the various options, under-floor heating comes closest to meeting these requirements. With it’s large heat emission
areas, flow temperatures can be maintained at 30-40°C so that low return temperatures can be easily achieved.
Other types of passive heat emitters such as radiant panels and radiators can also be made to conform to ‘delta T’
designs. For these types of emitters, design flow temperatures in the range 50-70°C can be selected enabling return
temperatures in the range 20-40°C. Furthermore, circuits can be provided with weather compensation whereby the
flow temperature is reduced as the external ambient temperature increases (this is discussed below under “Variable
temperature circuits”).
Forced convection terminal units such as air handling units or fan coil units are more sensitive to flow variations, and
the control of flow and return temperatures therefore requires more care. For these types of terminals, manufacturers
may recommend that flow temperatures are maintained constant and design ‘delta T’ values do not exceed 20°C.
Nevertheless, during selection, the aim should be to reduce the design flow temperature to its lowest possible
value (typically in the range 55-70°C) in order to generate return temperatures in the range 35-50°C.
Outputs from forced convection terminal devices in variable flow systems are typically modulated by two port control
valves. In such cases it is essential that these two port control valves are able to achieve accurate modulating control
over heating water flow rates, otherwise return temperatures may not respond as intended. For example, if a 2 port
control valve is ineffective at reducing flow until it is nearly closed, then the return water temperature will increase
18
 ‘Delta T’ - DESIGN GUIDE
70 40

above its intended design value simply because the flow is not being slowed sufficiently by the control valve.
This highlights the importance of selecting control valves with the correct “control characteristic” and with
adequate “authority”. These terms are explained in the following sub-sections.
Underfloor Heating (UFH)
The ideal terminal unit is one that returns heating water at the ambient air temperature, having lost all of its heat to
the surrounding space. Underfloor heating (UFH) comes closest to achieving this due to the large surface area of the
heat emitter. Flow temperatures can be maintained at relatively low values resulting in return temperatures of around
30°C and less.
The ideal UFH heating system should achieve a high rate of heat dissipation enabling it to operate at a high flow rate
with a smaller ‘delta T’ as opposed to low dissipation UFH which operates with a low flow rate and higher ‘delta T’.
High rates of heat dissipation enable systems to be more reactive and operate at lower temperatures.

Figure 12: Thermal image comparison between Low and High Dissipation Underfloor Heating Systems

Figure 13: Performance data comparing Low and High Dissipation Underfloor Heating Systems

UFH Floor Temperature Distribution

Another challenge encountered with low dissipation UFH is uneven temperature distribution across the floor surface.
A circuit with a low flow rate and high ‘delta T’ dissipates a greater proportion of the heat energy in the region of
floor surrounding the start of the UFH circuit, leaving the area furthest from the manifold distinctly cooler to the touch.
This temperature differential will create regions of hot and cold over the surface, which is not ideal from a comfort or
control perspective. The design of the pipe layout can help to mitigate this problem with the counter-flow spiral being
the most effective.
High dissipation UFH further reduces this problem by flowing at a higher rate and a lower ‘delta T’. The temperature
differential between the start and end of the circuit is minimised resulting in a floor surface which radiates thermal
energy evenly.
The normal objective of 70 / 40 is to increase the ‘delta T’ as much as possible to reduce the energy losses in the
distribution pipework and achieve return temperatures which are as low as possible. However, in the case of UFH
it is necessary to achieve the opposite and reduce the ‘delta T’ by increasing the flow rate. UFH pipework must be
considered as a terminal unit, where “distribution losses” (or the output of the unit) need to be maximised.
A reduction in ‘delta T’ in this case has the effect of increased reactivity and improved comfort for the user.
The importance of having high dissipation UFH is highlighted as this is only achievable if the heat taken from the
pipes is dissipated rapidly, otherwise return temperatures will be elevated and the circuit becomes detrimental to
‘delta T’ design.
19
 ‘Delta T’ - DESIGN GUIDE
70 40

UFH - Infra-red Floor Sensor

One method of improving underfloor responsiveness is to monitor the temperature of the floor surface directly by
infra-red floor surface temperature controllers, ensuring it remains inside the limits imposed by the BS EN1264-3
(29°C in occupied areas, 33°C in bath and shower rooms and 35°C in peripheral areas).

Figure 14: Infra-Red Floor Temperature Sensor

UFH - Heat Interface Units

In the multi-occupancy residential sector, including weather compensation
within each HIU is usually cost prohibitive. Hence the UFH flow temperature
is typically fixed for design conditions.
For such applications, high dissipation UFH systems offer value in terms
of better reactivity and lower flow and return temperatures.

Figure 15: Circuit Diagram of HIU with DHW and UFH

UFH - Difutec
SAV offers a high dissipation underfloor heating solution, for more information please refer to the Difutec
UFH Design Guide.

20
 ‘Delta T’ - DESIGN GUIDE
70 40

Control valve characteristic

In order to achieve effective modulating control of heat transfer across a forced convection (i.e. fan driven) heating or
cooling device, the control valve must be selected to suit both the emitter and the circuit in which it is located. The aim
is to achieve an acceptable flow control characteristic.
A valve’s control characteristic is the relationship between the flow through the valve and its degree of opening.
The characteristic is a feature of the design of either the valve itself, or the valve and actuator combination.

100%

ff
O
n/
O
Percentage Flow Rate

ar
50% Li
ne

ge
en l
rc ua
ta
Pe Eq

0% 50% 100%
Percentage Valve Opening

Figure 16: Typical valve control characteristics

Standard valve control characteristics include on / off, linear and equal percentage. These are illustrated graphically in
Figure 16 and can be described as follows:

• An on / off characteristic causes a large increase in flow over the initial 10-20% of its opening, and
then a small increase in flow over the rest of its travel. For this reason, it is sometimes referred to as
a “fast opening” valve.
• A linear characteristic causes an increase in flow that is directly proportional to the degree of
valve opening.
• With an equal percentage characteristic, flow capacity increases exponentially with valve travel.
Equal increments of valve travel produce equal percentage changes to flow.

On / off characteristics are suitable for applications where heat transfer is insensitive to flow variations e.g. passive
heating or cooling devices. Linear characteristics are suitable for valves with a regulating function such as double
regulating valves.
For active forced convection heating and cooling terminal devices, the best solution is an equal percentage
characteristic. This is because for these types of terminal device, the heating or cooling output stabilises as water flow
increases until a point is reached where the output becomes relatively unresponsive to further increases in flow.

21
 ‘Delta T’ - DESIGN GUIDE
70 40

Therefore, in order to achieve good modulating control over the heating or cooling output from forced convection
terminal devices, the control valve needs to be most effective over the first part of its travel. This is best achieved by
the equal percentage characteristic.
This characteristic can be produced by the design of the valve’s internals. Alternatively, it can also be generated by the
valve actuator’s controlling software, by providing sufficient control points (i.e. discrete valve settings) to give an equal
percentage characteristic. Where this type of “characterised actuator” is adopted, it must be used in conjunction with
an appropriate and compatible control valve. Random mixing and matching of valves and actuators may result
in ineffective equal percentage control.

Valve authority
If a valve used for controlling flow has a low resistance relative to the circuit in which it is located then, regardless of its
intended flow control characteristic, it would need to close a long way before it starts to have an impact on flow rate.
Such a valve would struggle to achieve control and would be described as having “poor authority”.
Therefore, in order to achieve effective modulating control, a valve needs to be sized such that its pressure loss, when
fully open, is a significant proportion of the total pressure loss of the circuit containing the valve. This ensures that
flow rate responds as soon as the valve begins to close.
The ratio of the control valve pressure loss when fully open to the pressure loss in the controlled circuit is referred
to as “valve authority”. If the pressure loss across the fully open valve = p1 and the pressure loss across all other
components in the controlled circuit = p2, valve authority (N) can be defined as,
N = p1 / (p1 + p2)
The maximum possible authority is 1. This occurs when p2 is zero and the only pressure loss in the controlled circuit
is across the control valve itself.
Figure 17 shows how an equal percentage flow characteristic can be distorted by installing the valve in a circuit
within which it has poor authority.

Figure 17: Effect of valve authority on an equal percentage flow characteristic

This shows that a valve authority of 1 provides a perfect equal percentage characteristic, whereas at 0.5 the control is
generally accepted as adequate. Two-port control valves have relatively low pressure loss and typically it may be difficult
to achieve an authority of 0.5 or above. In such cases, authorities of 0.25 to 0.3 may have to be accepted on some
valves depending on the criticality of the application. At an authority of less than 0.2 the flow characteristic becomes
so distorted that it more closely resembles a linear characteristic.
22
 ‘Delta T’ - DESIGN GUIDE
70 40

It follows from the formula for N, that in order to give control valves sufficient authority, the pressure losses in the
circuit in which they control flow must be limited. This will make it possible to select a control valve with sufficient
authority for modulating control. The best way to limit pressure differentials across pipe circuits is by the use of
differential pressure control valves, as described in the next sub-section.

Differential pressure control valves (DPCVs)

Differential pressure control valves (DPCVs) are used to control and limit pressure variations. A DPCV contains a
diaphragm that separates the upper / lower chambers of the valve, which is connected to a spring-loaded piston.
The valve opens when differential pressure rises and closes as the differential pressure falls, remaining steady during
periods of equilibrium. In variable flow circuits, DPCVs are typically used to maintain a constant pressure differential
across a sub-branch, protecting downstream control valves from excessive pressures and neutralising the effects of
pressure variation caused by other control valves in the circuit.
A typical situation is illustrated in Figure 18 which shows a heating circuit with a radiator thermostatic radiator
valves (TRV). Without DPCV protection, then as an individual TRV begins to close, the pressure differentials across
the remaining fully open radiator circuits could be expected to increase.

Pressure Pressure
gradient at gradient at
full load part load

Pressure Variation in
held pressure
constant across end
by DPCV Min terminal

Max

Figure 18: Potential for excess pressures and hence flows through terminal devices

This situation would lead to excess flow through the radiator units still with fully open TRVs and a consequent
reduction in the system ‘delta T’ values. The action of room temperature controls cannot be relied on to correct this
problem. Excess flows through radiators do not cause significant increases in heat output, and, it may therefore be
quite some time before the excess flow causes an increase in room temperature sufficient to get the TRVs to throttle.
In order to maintain the system design ‘delta T’ value at part load, it is important to avoid large increases in differential
pressure across terminal branches. This can be achieved simply and effectively by placing a DPCV across the flow /
return connections local to the first radiator TRV, as shown in figure 18.
Typically, DPCVs on branches serving groups of radiators with TRVs should be sized to maintain a differential pressure
of not more than 10kPa. Similarly, DPCVs on branches serving groups of terminal units with two port valves should
be sized to limit pressure differential to no more than 1.5 times the design pressure loss across the highest resistance
terminal unit branch (inclusive of its two port valve). This will limit the maximum flows through the terminals under
part load conditions, thereby helping to maintain control of the system ‘delta T’.
23
 ‘Delta T’ - DESIGN GUIDE
70 40

DPCV Proportional Band

As DPCVs are not electronically compensated, deviations from their set point are inevitable. The proportional effect
depends on the correlation between the controller valve’s degree of opening and the deviation between the controlled
and set differential pressure. Furthermore, the deviation depends on the actual differential pressure across the control
valve and the actual control setting.
Reference figure 19 it can be seen that the Danfoss AVPL DPCV has a narrow proportional band of ± 1…3 kPa,
depending on the actual differential pressure and setting. In this example a 3 kPa “P” band means the valves controlled
differential pressure can fluctuate between 15 and 18 kPa across its flow and differential pressure range.
NB: Correct DPCV KVS selection is vital for achieving a narrow “P” band.
Settings

Figure 19: Danfoss AVPL DPCV “P” Band

Alternative solutions to the problem of excess pressures across terminal branches are described in the BSRIA Guide
BG12/2011 Energy Efficient Pumping. This guide recommends the following options:
• The use of DPCVs as part of valve modules (see SAV FloCon range). The action of DPCVs to maintain a
narrow differential pressure range constant across circuits fed from a manifold arrangement will help to
ensure that flows through individual terminal units remain within acceptable limits regardless of closures
in adjacent circuits.

Pump minimum flow rate

When in service, a pump requires a minimum flow rate. Without such flow, most of the power generated by the pump
motor is converted to heat, with undesirable consequences.
In variable flow systems, by-passes should be provided to ensure that the minimum pump flow requirement is always
satisfied. To help maintain design ‘delta T’ values, the flow in the system at minimum load must be kept as low as
possible. An assessment should therefore be made of the minimum acceptable flow through the pumps and this value
should be used as the target to be achieved under minimum load conditions.
The pump speed turndown ratio, and hence the minimum flow rate that can be tolerated, is dependent on the type
of pump. As a general rule, the pump speed turndown ratio should not drop below 10%.
With glandless motor pumps, some water flows through the pump casing in order to cool the motor bearings.
However, if the pump flow rate is too low, this would lead to overheating of the bearings, producing an alarm
condition and shut down. As a rule of thumb, glandless pumps should not be operated at less than 7.5% of their
nominal flow rate where the nominal flow rate can be estimated as approximately 75% of the manufacturer’s
maximum flow rate value. See Figure 20.

24
 ‘Delta T’ - DESIGN GUIDE
70 40

Figure 20: Minimum flow rate for canned rotor pumps

For pumps where the motor is separated by a drive shaft, the minimum permissible flow rate is dictated by what
is considered to be an acceptable temperature increase across the pump. At near zero flow, the majority of power
generated by the motor is converted to heat, causing the water in the pump to increase in temperature. The maximum
possible increase in temperature can be calculated from the equation:
ΔT = Pzero / (cp qm)

Where: ΔT = temperature increase of the liquid as it passes through the pump (K)
Pzero = pump power at zero flow (W)
cp = specific heat capacity of fluid (which for water is 4187 J/kgK)
qm = mass flow rate of water (kg/s).

The pump power available to heat the liquid at zero or near zero flow can be determined from the pump
manufacturer’s published data. A typical manufacturer’s pump power curves are shown in Figure 21. The value of
Pzero is the pump power at zero flow; it is assumed that when the pump is operating at its minimum load condition
i.e. at minimum speed, its power consumption will be approximately equal to this value.
Hence, using the above equation, a value for the maximum temperature increase across the pump can be calculated.
The designer must then decide what level of temperature increase is acceptable.
For heating systems, the temperature increase must be limited to avoid increasing water temperature to a point
where a high temperature alarm condition might be triggered. A temperature increase of 2-3°C is usually acceptable
in heating systems, enabling minimum flow rates in the range 5 - 10%.

It can be seen from Figure 21 that the lower the pump pressure at minimum load, the lower the value of Pzero and
hence the smaller the temperature increase across the pump. It is therefore essential that secondary pump speed is
controlled in such a way that pump pressure can drop to the lowest possible value. This is best achieved by remote
sensor control of pump speed.

25
 ‘Delta T’ - DESIGN GUIDE
70 40

Figure 21: Pump power consumption at different operating speeds

System by-passes
System by-passes can be thought of as any feature in the system which allows flow water to return to the heat
source, without having been cooled by passing through some form of heat exchanger. Since by-passed water is
uncooled water, it will have the effect of increasing return water temperatures. If the optimal design ‘delta T’ value
is to be maintained, it is essential that secondary system by-passes should be minimised.
However, some by-passes are essential for satisfactory system operation. By-passes are commonly included in
secondary distribution systems in order to:

• Ensure that system flow does not drop below the pump minimum flow value (as explained
in the preceding sub-section);
• Make allowance for a future heating load to be installed at a later date;
• Maintain pipes in a “live” condition ready for heating demand;
• Maintain the circulation of water treatment chemicals to system extremities.

To best satisfy the last two bullet points on this list, by-passes should always be installed at system extremities
i.e. at the ends of branches. The three best by-pass options are as follows:

• Variable flow. A variable flow by-pass is shown in Figure 22. The by-pass includes a fixed orifice
double regulating valve for flow regulation and flow measurement. Because there is no control of the
differential pressure across the by-pass, the flow through the by-pass will vary under different load
conditions. The by-passes should be sized such that their summated flows are equal to the target
minimum flow rate value.
• Constant flow. A constant flow by-pass is shown in Figure 23. This type of by-pass is particularly suited
to systems feeding terminal devices fitted with PICVs. The by-passes should be sized such that their
summated flows are equal to the target minimum flow rate value. Constant flow regulators should be
included in each by-pass to ensure that under all operation conditions the target flow rate is maintained.
• Integral. An integral by-pass is shown in Figure 24. Integral by-passes are those included within the
control valves themselves. Examples are 3 or 4 port valves that divert flow through a by-pass when
the load is satisfied.
26
 ‘Delta T’ - DESIGN GUIDE
70 40

To keep pipes ‘live’ as far as possible, multiple low flow by-passes should be installed at system extremities. The total
amount of by-passing flow should be in the range of 5-10% of the total design flow (having checked that this is not
less than the required pump minimum flow rate). If this leaves some parts of the system without regular flow, the
circulation of water treatment chemicals can still be achieved by exercising the valves periodically (e.g. cycling all two
port valves fully open for say, 10 minutes in every 48 hour period during the non-heating season).

Variable flow
by-pass

Figure 22: Typical circuit with variable flow by-pass at end of run

Constant flow
by-pass

Figure 23: typical circuit with constant flow by-pass at end of run

27
 ‘Delta T’ - DESIGN GUIDE
70 40

Integral
by-pass
 
(4PV)

Figure 24: Typical circuit with integral (4PV) by-pass at end of run

Variable temperature circuits

In order to achieve good ‘delta T’ design, terminal device flow and return temperatures must be as low as possible.
The use of variable temperature circuits (as opposed to constant temperature circuits) is critical to achieving this objective.
Figure 25 shows constant and variable temperature secondary circuits. The key difference is that in the case of variable
temperature circuits, a three port valve is used to mix some of the returning water back into the flow water thereby
reducing its temperature. The degree of mixing will determine the flow temperature. Constant temperature circuits
have no such provision to control flow temperature.

CONSTANT
VARIABLE TEMPERATURE
TEMPERATURE

PRIMARY
HEADER
Figure 25: Constant and variable temperature secondary circuit layouts

The choice of either constant or variable temperature circuits has traditionally taken into account of the type of
terminal devices to be fed from the system and how best to control their heat output. Natural convectors such
as radiators or under-floor heating tend to be fed from variable temperature circuits because they are relatively
insensitive to flow variations. Heat output from these emitters is therefore best controlled by varying the flow
temperature. On the other hand, forced convection heat emitters such as fan coil units have traditionally been
fed from constant temperature circuits. This is because their heat output is sensitive to flow, which can be varied
in turn by modulating control valves.
However, regardless of the type of terminal device, in ‘delta T’ designs variable temperature circuits are essential
if optimal flow and return temperature values are to be maintained.
This is because a variable temperature circuit enables the secondary flow and return temperature to be reduced to their
lowest possible values, thereby maximising system efficiency. For example, the benefits accruing from the system shown
in Figure 8 of this guide, relative to that shown in Figure 7 are only possible because the flow temperatures in the
secondary circuits can be reduced to values that are lower than those required in the primary flow pipes.
28
 ‘Delta T’ - DESIGN GUIDE
70 40

‘Delta T’ controllers - Weather compensation and return temperature regulator

It is important that every opportunity be taken to reduce the heating system flow temperature whenever possible.
Intelligent ‘delta T’ controllers can achieve this through weather compensation alone or a combination of weather
compensation and return temperature monitoring. This will have the following benefits:
• Heat emissions from distribution pipes will be reduced thereby avoiding energy wastage. This is
particularly relevant for large district heating systems where pipes are run externally and are exposed
to low surrounding air temperatures.
• The potential for high return temperatures under low load operating conditions will be mitigated. Return
temperatures tend to increase towards the flow temperature under low load conditions because the
proportion of flow circulating through by-passes increases relative to the overall flow through the system.
• Heat outputs from terminals will become easier to control, resulting in less risk of over-heating
occupied spaces.
To some extent the heating system flow temperature will be determined by the application. For example, a system
serving under-floor heating can be run at a relatively low flow temperature (35°-40°C). However, heating systems
serving hot water heaters may require flow temperatures in the range of 50°C, in order to generate hot water at
40°C. Where it makes sense, pipe systems should be split so that terminal devices with different temperature
requirements are fed from separate secondary circuits.

‘Delta T’ controllers can be provided as stand-alone units by SAV (UK) Ltd.

Self standing ‘delta T’ controllers

In situations where different types of terminal device are to be fed from the same heating system, it may be more
cost effective and user friendly to provide localised self-standing ‘delta T’ controllers to manage variable temperature
circuits. A typical configuration is shown in Figure 26.

WEATHER
COMPENSATION Delta ‘T’ Controller

Variable
RECIRCULATION
LINE temperature
circuit

DPCV MODULATING
CV

Heating
mains

Figure 26: Localised weather compensated variable temperature UFH sub-circuit feeding off a higher temperature radiator circuit.

It can be seen that this solution requires a separate pump, sized to cope with the flow through the variable
temperature circuit.
If the flow temperature sensor goes above set point (e.g. 35ºC), the ‘delta T’ controller instructs the 2-port valve to
close, thus promoting movement through the recirculation line. Similarly, if the return temperature exceeds set point
(e.g. 30ºC), the 2-port valve also closes. Only 1 temperature excess signal is required to actuate the 2-port valve,
thus providing control of flow and return temperatures.
29
 ‘Delta T’ - DESIGN GUIDE
70 40

‘Delta T’ Monitoring - Continuous digital monitoring of HVAC systems

SAV’s Flocon Watchman commissioning modules enable continuous digital monitoring of flow rate and return
temperature for individual zones and terminals within building HVAC systems.
The combination of a flow meter on the main supply pipe with two port control valves on each of the terminal
branches means that precise flow rate information can readily be obtained. By closing the two port control valve
on any particular terminal branch, even ultra low flow rates can be derived from the drop in total flow through
the meter. This is known as the subtraction method.
Digitally monitoring the return water temperature from individual terminals enables a full soft landings approach
whereby the theoretical return temperature predictions are tested under operating conditions. If return temperatures
are not as predicted producing a shortfall in ‘delta T’, this will be highlighted by the BMS. This provides the information
to enable fine tuning of the system.
Each module also includes an energy meter that measures and records the energy consumptions of up to 7 terminal
units at a time. This means that if a building is sub-divided into multiple tenancies, each lettable area can be metered
individually and the tenants billed accordingly.

‘DELTA T’ ENERGY METER CALCULATION METHODOLOGY

The following example shows the methodology for calculating the average ‘delta T’ between two points in time using
energy meter data.

Reading 1 Reading 1 The 860 is a constant, and is defined as the quantity of m3

870 m3 00101 MWh of water that will be heated by 1ºC by 1 MWh

Reading 2 Reading 2 So in this example the ‘delta T’ calculation is as follows:

01340 m 3
00120 MWh
19 x 860 = 29°C
560
Cubic Meter Consumption in period: 1340 – 780 = 560 m3

Energy Consumption in period: 120 – 101 = 19 MWh

‘Delta T’ calculated as follows:

The average ‘delta T’ in this
example is therefore 29ºC
MWh x 860 = heating in °C
m3

C1 C2 C3 C4 Chilled H1 H2 H3 H4 Heating
Water Water

B C
FloCon BMS
Interface

A = FloCon BMS Interface

B = MID Certified Digital Flow and Energy Meter - Chilled Water
C = MID Certified Digital Flow and Energy Meter - Heating Water
C1, C2, C3, C4 = Return Temperature - Chilled Terminal Unit
H1, H2, H3, H4 = Return Temperature - Heating Terminal Unit

Figure 27: Schematic of FloCon module for heating and cooling duties
30
 ‘Delta T’ - DESIGN GUIDE
70 40

Hot water heating and Heat Interface Units

Secondary circuits feeding direct heated hot water storage cylinders are not always compatible in systems that require
low return temperatures and high ‘delta T’ values. Hot water cylinders are usually designed to heat water to 60°C, the
minimum temperature required to kill legionella bacteria. To heat the entire contents of a cylinder to this temperature
within a reasonable time span requires a high flow temperature and low ‘delta T’, resulting in return temperatures
equal to or greater than 60°C.
As an alternative to storage cylinders, hot water can be generated by the use of plate heat exchangers that can
provide much higher heat dissipation than traditional cylinders. This solution is available from SAV Systems in the
form of packaged heat interface units (such as Danfoss FlatStations).
This type of unit provides instantaneous heating of the incoming domestic cold water by means of a plate heat
exchanger. Heating water circulates on one side of the plate heat exchanger transferring heat to mains cold water
passing through on the other side. The heated water can then be fed straight to the hot water taps, or to a hot
water cylinder.
A typical heat interface unit (or FlatStation HIU) is shown in Figure 28. This identifies three valves (A, B & C) which are
critical to the function of these units.

A. A DPCV must be installed in the circuit feeding the radiators. This enables the pressure differential across the
radiator circuit to be set to be maintained at its optimum value - typically less than 10kPa for apartments and
standard sized houses. This enables the radiator TRVs to control more effectively and will help to maintain a
consistent balance of flows through the radiators. During commissioning, the DPCV should be set to maintain the
required design flow rate through the radiator circuit with radiators sized at the optimal ‘delta T’. Because the
DPCV is a self-acting valve it will modulate automatically to maintain a constant pressure differential across the
circuit regardless of changes in pump speed or valve closures elsewhere in the system.
B. A fast acting temperature control valve is required which simultaneously regulates both the heating water flow and
the flow of cold water into the plate heat exchanger. This is to maintain constant domestic hot water temperature,
regardless of variations in heating or cold water conditions. This type of valve maximises the ‘delta T’ across the unit
during hot water production.
 ote: This type of valve is critical to ‘delta T’ design. Heat interface units that rely on slow acting control valves will
N
inevitably produce variations in domestic hot water temperatures and result in unwanted high return temperatures.
C. A DPCV must be located in the heating circuit feeding the plate heat exchanger, to maintain to a constant pressure
across the heating side of the exchanger at times of domestic hot water demand. Because the DPCV is self-acting, it
modulates automatically to maintain a constant pressure differential, regardless of changes in pump speed or valve
closures elsewhere in the system.

Figure 28: Danfoss FlatStation Heat Interface Unit for instantaneous indirect
hot water heating and direct space heating
31
 ‘Delta T’ - DESIGN GUIDE
70 40

Since hot water is heated instantaneously in a FlatStation HIU, there is no necessity for hot water storage. Due to the
high rate of heat transfer into incoming cold water, heating water entering at 60-70°C can be consistently cooled to
25°C during draw-off, making these units ideal for ‘delta T’ designs.
It is important that low primary return temperatures be maintained at part load, regardless of fluctuations in demand
or primary pressures.
The relationship between hot water demand and system operating temperatures using FlatStation HIUs is illustrated in
Figure 29. As can be seen from the performance chart, the primary return temperature is consistently kept below 25°C
as recommended by CIBSE AM12 2013 and as required by Greater London Authority’s District Heating Manual
for London 2013.
When there is no draw-off of hot water, the self-acting control valves inside FlatStation HIUs throttle the heating flow
to a minimum, thereby causing the pump to reduce its speed and power consumption. Instead of being closed off
completely, the flow through the unit is finely controlled to an amount just sufficient to maintain a heating return
temperature 8°C below the set hot water tapping temperature. This feature prevents the pipes feeding the unit
from becoming dead legs and ensures that hot water is available as soon as taps are opened. It also keeps return
temperatures as low as possible.

1m 2m 3m 4m 5m 6m

Figure 29: Domestic hot water demand and operating temperatures using Danfoss FlatStation HIU

Where storage of hot water is a requirement (as might be appropriate in a centralised hot water system for a school
or similar building) a plate heat exchanger can be used to control system ‘delta T’ whilst providing heat to a storage
cylinder. Figure 30 shows a typical configuration.
In order to generate domestic hot water at 60°C (the recommended temperature for eliminating legionella), the
heating water flow must be at a temperature higher than this. Hot water is pumped to a storage cylinder which fills
from the top down. A specialised fast acting pressure independent control valve (DHW-PICV) is used to control the
flow of heating water through the plate heat exchanger. This valve closes when the hot water recirculation increases
above its set point value (indicating that the cylinder is full). As hot water is drawn from the system, it is replaced by
cold water resulting in the opening again of the DHW-PICV to start generating more hot water.
Hence, heat input is only provided when hot water generation is required, enabling the heating system temperature
to remain low at 40°C or less.

32
 ‘Delta T’ - DESIGN GUIDE
70 40

Figure 30: Plate heat exchanger for heating of hot water in cylinders

For systems with multiple distributed outlets, hot water circulation can be provided with either instantaneous or storage
type hot water solutions, as indicated in Figures 26 and 29. The design of systems incorporating these units is explained
in the SAV FlatStation Design Guide 2014.

33
 ‘Delta T’ - DESIGN GUIDE
70 40

4. ‘DELTA T’ COMMISSIONING AND SOFT LANDING

Commissioning
Proper commissioning of the system is essential if the ‘delta T’ design is to be effective. It is not sufficient to simply
check that all design flow rates are achievable. There must also be a check of system operating temperatures.
Proper commissioning of ‘delta T’ designs should comply with the requirements of CIBSE Commissioning Code W:
Water Distribution Systems 2010 and BSRIA Guide BG2 / 2010 Commissioning Water Systems. However, particular
attention should also be paid to the following issues:

• Design flow rates should be tested under both maximum and minimum load conditions. It is particularly
important to prove that the specified minimum flow rate is achieved when the system is at minimum
load (i.e. with all control valves fully closed). This should normally be a value in the region of 5-10%
of the maximum.
• Pump pressure differentials should be measured at both maximum and minimum load conditions. This
will demonstrate whether the specified method of pump speed control has been properly implemented.
• Variable flow by-passes should be commissioned at minimum load (all control valves fully closed). If
the specified by-pass flow rate were set with the system operating at full load, the flows through these
by-passes would inevitably increase in an uncontrolled way as control valves close and the differential
pressure across the by-pass increases.
• The differential pressures controlled by DPCVs should be measured and recorded to ensure that the
values are in line with design and that there is no risk of excessive flows through terminal branches.

Having ensured that commissioned design flow rates at maximum and minimum load conditions are correct,
the system return temperature should be monitored to ensure that its specified design value is being maintained.
If the ‘delta T’ design is successful, the returning water temperature should remain at or below the target return
temperature under all operating conditions.

Soft Landings
The refinement of building systems operation to optimise system temperatures should continue after construction
hand-over. The proper functioning of the system is strongly reliant on maintaining ‘delta T’ values. There is always the
potential for the system ‘delta T’ to be squeezed by a poorly functioning control valve or an open by-pass. For this
reason, systems should be monitored during their operation and any reductions in ‘delta T’ investigated and corrected.

Please contact SAV Systems for more detailed information on

energy metering systems or any other aspects of this design guide.

34
MAKING MODERN LIVING POSSIBLE

‘Delta T’ - DESIGN GUIDE

70 40

MAKING MODERN LIVING POSSIBLE

Extracts - Danfoss Lystrup Technical Paper
Technical paper

Results and experiences

from a 2-year study
Technical paperwith measurements on

Results DH
a low-temperature andsystem
experiences
for low energy buildings
from a 2-year study with Danish
Christian Holm Christiansen1,
measurements on
Technological Institute, 1

a low-temperature DH system
Alessandro Dalla Rosa2, Marek Brand2, for1, low
Gregersensvej DK-2630energy buildings
Taastrup, Denmark
Peter Kaarup Olsen3, Jan Eric Thorsen4 2
Technical University of Denmark, Dept. of Civil Engineering,
TECHNICAL
DHC13, PAPERHolmSymposium
Christian
the 13th International Christiansen1, 1
Danish
Brovej, DK- 2800Technological
Kgs. Lyngby,Institute,
Denmark
Alessandro Dalla Rosa2, Marek Brand2, Gregersensvej 1, DK-2630 Taastrup, Denmark

Results and experiences

on District Heating and Cooling
Peter Kaarup Olsen3, Jan Eric Thorsen4
September 3 to DHC13,
September
3
COWI A/S,
2 Parallelvej 2, 2800 DK-Kgs. Lyngby, Denmark
Technical University of Denmark, Dept. of Civil Engineering,
the 13th4 International
, 2012,
rd th
Symposium 4
Danfoss District Energy,
Brovej, DK- Nordborg,
2800 Kgs. Denmark,
Lyngby, Denmark
from a 2 - year
Copenhagen, Denmarkstudy with
on District Heating measurements
and Cooling on +45 74883 COWI
4494,A/S,
[email protected]
Parallelvej 2, 2800 DK-Kgs. Lyngby, Denmark
a low - temperature DH
September 3 system for4 low
to September
rd - energy buildings
, 2012, th
Danfoss District Energy, Nordborg, Denmark,
4

Copenhagen, Denmark +45 7488 4494, [email protected]

A new low-temperature district heating system for low-energy buildings Author(s)

that operates with supply temperatures slightly above 50 °C was presented
at the 11th International Symposium of District Heating and Cooling in 2008;
the design includes newly developed substations and efficient distribution
pipes, resulting in reduction of heat losses up to 75 % compared to traditional
layouts. Since then, the first area using the new system has successfully been
put in operation. This paper presents the results of a 2 -year study with
detailed measurements of a low heat density area with 40 low - energy
terraced houses in Denmark. The investigations include the determination Jan Eric Thorsen,
of the heat losses from the distribution network, the pumping electricity Danfoss District Energy, Nordborg, Denmark,
+45 7488 4494 · [email protected]
consumption, the user behavior in terms of indoor temperature and domestic
Christian Holm Christiansen,
hot water consumption as well as detailed simultaneity factors to be used Danish Technological Institute,
for network design. Moreover, the paper presents solutions for using the Gregersensvej 1, DK-2630 Taastrup, Denmark
return water of existing networks to supply district heating to newly built Alessandro Dalla Rosa,
Marek Brand,
areas and summarizes in general on how to integrate low - energy houses Technical University of Denmark,
and district heating systems. Finally, it points to the potential of integrating Dept. of Civil Engineering, Brovej,
DK- 2800 Kgs. Lyngby, Denmark
low - temperature district heating systems in existing buildings as an effective
Peter Kaarup Olsen,
solution towards energy - sustainability in the heating sector. COWI A/S, Parallelvej 2,
2800 DK-Kgs. Lyngby, Denmark

Project information
Introduction
owner Housing association BF Ringgården
An innovative low-temperature District
year of construction 2008– 2010
Heating ( DH ) system for low - energy
site area
buildings [haoperates
that ] with supply tem- 1.7information
Project
peratures slightly above 50 °C has been
building units (residential )
successfully put in operation in 2010. This
40 terraced houses
owner Housing association BF Ringgården
paper presents the results of a  2 - year
residents year Seniors, young families
of construction 2008 – 2010
study with detailed measurements of site area [ ha ] 1.7
number
a low of residents
heat density area with 40 low - energy 92 (estimated ) districtenergy.danfoss.com
terraced houses building units ( residential ) 40 terraced houses
building unitsand a communal
(teritary ) build- 1 communal building
ing in Lystrup, Denmark, see table 1 and residents Seniors, young families
heated
figure area
1. The [m2] dealt with the inte-
project 4115
number of residents 92 ( estimated )
gration of sustainable
plot ratio 1 solutions both for 0.24
the end-user side and the energy supply building units ( teritary ) districtenergy.danfoss.com
1 communal building
sidebuilt
1
and floor
aimedarea
to: / site area heated area [ m ] 2
4115 50 30
• Demonstrate the operation and energy plot ratio 1
0.24
TABLE 1: Basic information on the project 35
demand of DH applied to low-energy 1
built floor area / site area
Technical Paper Results and experiences…
‘Delta T’ - DESIGN GUIDE
perature DHW systems described in [ 1 ],
70 40
number of dwellings
[ 2 ]: the low - temperature Instantaneous block number total size [ m2 ]
Heat Exchanger Unit ( IHEU ) and the low- type c1 type c2
temperature- Danfoss
Extracts District Heating Storage
Lystrup 1
Technical Paper 771 5 3
Unit ( DHSU ), see figure 2. 2 727 2 5
Heat demand
The layouts of the DHW distribution pipes 3 594 3 3
The building plan
and the floor of the dwellings
installations, in terms wereof 4 528 1 4
carefully designed, so that there
heating system, consist of a combination is a sep- DH supply DHSU principle
arate
of pipe supplying
radiators – based on each DHWsupply/
design fixture 5* 479 1 2
and the length of the pipe is minimized,
return / room temperature of 55/ 25 / 20 °C 6 484 3 2
see figure 3. Consequently, the water con-
tent inis each DHWbysupply 7 532 6 DHW 0
DHW prepared one ofline, including
the low - tem-
the volume in the secondary side of [the 120l A = 170 m
* including the communal building, 2
perature DHW systems described in 1 ],
[DHW
2 ]: theheat
lowexchanger,
- temperature is kept to a mini-
Instantaneous
mumExchanger
Heat and it is below
Unit3(liter:
IHEUthis
) andis the
the max-
low- TABLE 3: Type and floor area of the buildings
imum allowable
temperature Districtwater content
Heating Storagethat
assures
Unit ( DHSUsafety in relation to
2. the Legionella
DH return DCW
risk, even without any treatments ( ther-
The
mal,layouts of or
UV -rays thechemical
DHW distribution
), according pipesto
the German guidelines for DHW systems
carefully
( DVGW, W551 designed,
). so that there is a sep- IHEU principle
DH supply
and the length of the pipe is minimized,
Heat distribution network
3. Consequently, the water con-
A sketch
tent of the
in each DH network
DHW supply line,withincluding
the loca- thermostatic DHW
tionvolume
the of flow meters for monitoring
in the secondary side isofseen
the bypass
in figure
DHW heat4. exchanger,
Besides the isnormal
kept toend - user
a mini-
heat meters
mum and it isand
below the3 liter:
mainthismeter placed
is the max-
at the pumping station
imum allowable water content that two additional
meters safety
assures are placed at thetoend
in relation the of two dif-
Legionella
ferent
risk, streets.
even withoutOneanymeter is measuring
treatments ( ther- DH return substation DHW pipe DCW
a partUV
mal, of-the
raysnetwork with),11according
or chemical DHSU's, the to
other
the is measuring
German on a network
guidelines for DHWpart with
systems
(11 IHEU's.
DVGW, W551). Figure 2: Sketch of the DHSU and IHEU principles of DHW production

Network dimensioning FIGURE 3: Sketch of the floor plans with the layout of the DHW distribution
pipelines. Type C1 ( left ), type C2 ( right )
The network consists of flexible plastic
twin pipes for dimensions up to DN32 and Pumping station
flow rate meter:
of steel twin pipes for larger dimensions. 0.8 – 12 m 3 / h
Heat loss coefficients are calculated Pumping station
according to [ 2 ] and pipe manufacturer Instantaneous flow rate meter:
Exchanger Unit 0.8 – 12 m3 / h
data, figure 4. (IHEU)
The other assumptions for the design flow rate meter:
Instantaneous
0.01 – 0.38 mUnit
3
/h District Heating
Exchanger Storage Unit
were: (IHEU) (DHSU)
• Maximum pressure level: 10 bar. It is flow rate meter: flow rateHeating
meter:
0.01 – 0.38 m3 / h District
reasonable to design the network 0.01 – 0.09
Storage m3/h
Unit
according to the maximum hydraulic (DHSU)
flow rate meter:
load that can be withstood by the 0.01 – 0.09 m3 / h
distribution pipeline; in this case the
limit is drawn by the plastic service
pipes, which requires pressure levels flow rate meter (street):
below 10 bar. In fact the pipeline 0.26 – 3.4 m 3 / h
systems must by regulations withstand
pressures 1.2 – 1.5 times the nominal flow rate meter (street):
0.26 – 3.4 m3 / h
value. Figure 4: Sketch of
- the low temperature network with the location of the meters.
• Thermostatic by-pass valves of IHEU's DH is delivered from the utility Lystrup Fjernvarmeto the pumping station
set to 40 °C, in the customer's FIGURE 4: Sketch of the low-temperature network with the location of the meters.
substation at the end of each street line DH is delivered from the utility Lystrup Fjernvarmeto the pumping station
and set to 35 °C, in all the other
customers' substations.
•4 Design supply temperature from the Danfoss District Energy
mixing shunt: 55 °C; design return
temperature: 25 °C.
• Maximum water velocity: 2.0 m/s; also
in branch pipes.
36
 together with the shunted supply tem- 2 substations, where the control valves
0.088 0.053 1300.02 78 211 were defected2.2and bar allowed65a°Crelative
perature. The maximum monitored heat
0.085 0.056 0.1 90load is 161.3 kW82compared to maximum 240 large amount of 1.6 bar to flow uncooled
water 25 °C
32weekly average 88 of 87.4 kW. The 246 figure to the return pipe. The best performing
0.099 0.053 0.1 ‘Delta T’ - DESIGN GUIDE
shows how the shunt has been adjusted IHEU registered a return temperature of
0.096 0.06
70 40 0.1 16 122 268 26 °C in summer. Observing the same elev
during the period in order to get the low
mixed supply temperature of just slightly weeks, a year 0 later,50showed 100 improved100 150
x: Aluflex twin pipes; Tws: Steel twin pipes, series 2, diffusion abovebarrier at the outer
50 °C. Further the resultcasing. of trouble- results as seen in figure 9 for the 11 IHEU's,
Extracts - Danfoss Lystrup Technical shooting Paperin individual building installa- even though theTosupply the maintemperature
pipe had
Ringgården
tions has secured a low return tempera- been reduced in the meantime.
(Lystrup district pumping station lo
Monitoring building
ture. This design is expected values. to beFurther,
even lower the In general, the return temperature in the
heating)
indoor
after the temperature,
local „boiler as man”
measured beginningin the heating season ( week 65 °C39 – 47 ) was lower
An extensive monitoring program and living
2012 room,
has was
been 2 – 4 °C
provided above a the
tool design
that than during the summer, which confirms FIGURE
the
data acquisition system was established;
Technical Paper Shunt Result Th
n indoor
updates temperature
him every week of 20
with°C during
return the
tem- that the radiators delivers low return tem-
the measurements presented here are heating
peratureseason, and other see relevant
figure 7. In data[ 1 ] of
it was
the peratures ( 28 – 33 °C ). This Return occurred
mainly from the first monitoring period valve
shown
individual that 1 °C higher
houses. Basedthan on the expected
tool, he although the indoor temperatures (closeduring
/ half open) at the fa
; also conducted during the weeks 26–47, 2010 room
can guide temperature
the users towards can lead better to oper-
20 % operation were some degrees higher tisticall
with the main meters and individual higher
ation ofSH thedemand
substations in low - energy
which hashouses,
already 6
than the design conditions ( 20 °C ), which the sug
med to meters in 22 dwellings in place; in these so
beenthebeneficial.
room temperature alone, almost increased the minimum achievable return side. Th
e low dwellings both DH meter, meter for DHW explains
In the firstthe high consumption.
monitoring period of 2010 the temperature from the radiators. Overall, consum
and a room temperature sensor were
DHW consumption
two different substation was typesmeasured were to com- be the demonstration project has shown for the
ctor installed. Late 2010, DH meters in the
65 literspecifically.
pared / ( day × house) In the in 11 average.
homes with It is that the concept works, and that is further consum
remaining 19 dwellings were connected
a low value, which is
DHSU, the average return temperature partly related to the confirmed by the fact that there were no the dia
DH
used and the monitoring
Technical Paper has continued ever number Results and experiences… FIGURE 5: Simplified pressure line / temperature
was 39.4of °C occupants
in the weeks and26their– 47; compo-
in sum- complaints from residents about the lack expecte
DHSU: since and will go on until end of 2013. during typical operating conditions
sition. Based on
mer – weeks 26 –an 38estimate
– the average of thereturnnum- and quality of SH / DHW. The ex
ber of residents
temperature was 43.6 in the °C. dwellings,
The high return it is setting
Operating
Space heating temeratures
and domestic hot water assessed
temperature that wasthe DHW primarilyuse was dueequiva-
to the Simultaneity factors valve, w
tation: Lystrup Fjernvarme that supplies heat to lent to
malfunctionapprox. of a28 liter
single / ( day
unit. ×
The person
best ).
per- It
Based on the measurements in the first
the new low - temperature should
formingbeDHSU noticed that theaaverage
registered return tem- cold In order to define design Week loads 38 in
– 47areas
(y = −85.0 x + 1168 [W
monitoring period, a heat loadarea vs out- is
a medium - temperature DHwas system. DH water
perature temperature
of 29 °C was
in approx.
summer. 15 °C
The and11 with low - energy buildings, simultaneity
door temperature curve estab- Week 26 – 38 (108 [W])
is supplied with up the
homes average
with IHEU, DHW temperature
the average return tem- was curves were developed based on moni-
lished, see figure 6. to
For80 °C during
IHEU winter
the standby
and down 40 – 45 °C,was
perature giving 34.7 an°Caverage
in the weeks temperature 266.5
– 47; toring data sampled1500 every 4 minutes dur-
s case heat loss istoabout60 °C during summer.
25 W; for DHSU In 80fig-
W. bar
difference
in summer of 25 – 3026
– weeks °C–in 38the– the first moni-
average ing the summer weeks 24 – 38 in the year
Iture 8, thethat average weekly supply and

installation heat losses per house [W]

is seen the average heat load dur- 1400
a low- 2010. Totally 38,000 data sets, or time

Heat load for space heating incl.

return
ing summertemperatures
( week 26 –and heat
38 ) is higher loadthan
are toring
return period.
temperature According was 40.3 to [ 8°C.
], DHW
The high use 55 °C
e inte- of 30 – 40 liter /(day ×was person) and adue tem- stamps, were recorded1200 for each of 10
seenstandby
the for the losses2 year which
monitoring means period
that return temperature primarily to dP=0.3bar
at has perature difference of approx. 40 °C are IHEU's and 10 DHSU's. The methodology
together
there withbeen
has also the shunted
a SH demand supplyintem- this 2 substations, where the control valves 1000
ere are typical values 2.2 for Denmark. used to develop the curves is: Data is
perature.
period The
for somemaximum
houses monitored
e.g. for heat
floor were defected bar
and allowedIn 65a°Cthe case
relative
heat is study this would give an sorted
25 °Cin the way that 800 the combination
load is 161.3
heating of the kWbathroom.
compared to maximum large amount of
1.6 water
bar to expected
flow uncooled heat
m - tem- demand for DHW ofThe
12 – 16 kWh / ( m × yr ).
2 of the highest group load E(N) is calcu-
weekly average of 87.4 kW. The figure to the return pipe. best performing
varme. Based on the curve, the SH demand per In average the DHW consumption mea- lated for each time 600
stamp. To avoid too
shows how the shunt has been adjusted IHEU registered a return temperature of
shunt dwelling was estimated to 5.1 MWh for sured was 8 kWh/(Observing
m × yr ), which high simultaneity
elevation = 53 m curves,
AMSL it is accepted
during the period in order to get the low 26 °C in summer. 2
the issame less 400
se. The the Danish reference year corresponding than the design value ( table 2 ). When to exceed the suggested design load pr.
mixed supply temperature of just slightly weeks, a year 0 later,50showed 100 improved100 150 200 250 300
essure to about 51 kWh / ( m2 × yr ), 70 % higher added to the SH demand a total annual consumer e(N) in 1 % of
200 the
distance tapping
[m] time.
above 50 °C. Further the result of trouble- results as seen in figure 9 for the 11 IHEU's,
critical SH demand than building regulation DH consumption of 5.8 MWh per house This is equal to shortcomings for a time
shooting in individual building installa- even though the supply
To the main pipe temperature had
Ringgården Ringgården 0
hunt is design value ( see table 3 ). It was not the was periodlow of-15 minutes if assumed the 0 tap- 5
tions has secured a low return tempera- beenfound.
reducedMonitoringin(Lystrup districtalso pumping
the meantime. demon- station energy network - 5 10
a tem- purpose of the project group to look at strated that DHW can be produced at ping is occurring all 24h / day. In practice Outdoor temperatu
ture. This is expected to be even lower heating)
In general, the return temperature in the
y pipe the building „as - built” vs. designed. temperatures the period will be quite shorter than 15
he sys-
after the local „boiler man” beginning
Analysis of the measurements also rather heating seasonof( week just653°C39°C–below
47 ) was the pri-
lower FIGURE
minutes.
FIGURE 6:On the load
Heat other hand the analysis
ith the
2012 has been provided a tool that
indicates that the reason should be
mary supplythe
than during temperature,
summer, which e.g.confirms
47 °C at Thermostatic by-passvs. outdoor temperature curve
Consumer
a  DHthe supply temperature Shunt assumesatthat the consumersthe
monitored
street end duringunit
withfirstthemonitoring pe
mping
updates him every week with return tem-
found in user behavior. that radiators delivers lowof return50 °C tem- as each
Results of a 2 year study
perature and other relevant data of the in Lystrup, Denmark
expected on a
with lowthe temperature
used substations. highest DHW weekload 38 by
– 47default
( heating are placed
season )
peratures ( 28 – 33 °C ). This Return occurred valve
Floor heating
district during
heating summer
(DH) system
individual houses. Based on the tool, he is a param-
for low energy buildings (close / half open)
eter that isthe
can guide notusers
takentowards
into account
betterin the
oper- 50 30
6 ‘Delta T’ for low temp DH zone
ation of the substations which has already
80 240
temperature, weekly average [ °C ]

been beneficial.
heat load, weekly average [ kW ]

70
In the first monitoring period of 2010 the 5 210
80
two different substation types were com-
60 180
temperature, weekly average [ °C ]

70
pared specifically.
50 In the 11 homes with 150
DHSU, the average return temperature 60
40 FIGURE 5: Simplified pressure line / temperature diagram of the mixing shunt 120
was 39.4 °C in the weeks 26 – 47; in sum- during typical operating conditions 50
30 90
mer – weeks 26 – 38 – the average return 40
20 60
temperature was 43.6 °C. The high return 30
10
temperature was primarily due to the 30
20
malfunction of a0single unit. The best per- 0
10
forming DHSU registered a return tem- Week 38 – 47 (y = −85.0 x + 1168 [W], R2 = 0.95)
26, 2010 2011 25, 2012
perature of 29 °C in summer. The 11 Week 26 – 38 (108 [W]) 0
Week, year
homes with IHEU, the average return tem- 26, 2010
1500
perature was 34.7 °C in the weeks 26 – 47;
T supply to low temperature area T supply DH T return DH Heat load
in summer – weeks 26 – 38 – the average
heat losses per house [W]

1400
for space heating incl.

return temperature was 40.3 °C. The high T supply shunt, low temperature area
Average weeklywas
supply, return,due
shunted 1200
return temperature primarily to supply temperatures and heat load for the 2 year monitoring period
2 substations, where the control valves 1000
were defected and allowed a relative 37
large amount of water to flow uncooled 800
FIGURE 8: Average weekly supply, return, shunted s
to the return pipe. The best performing
design assumption of e(n) = 1 is leading pipe of the main district heating network. In the case considered, yearthe distribution
to an oversize of the DH net. Anyhow, This solution can be installed in an existing heat loss for the area with DHSU's are
‘Delta T’ - DESIGN GUIDE
since the smallest available DH branch district heating network at a location hav- slightly lower DKthan ref in the area with 2011IHEU's.
pipe dimensions
70 40 are installed, this has no ing a sufficient flow in the return pipe. In ( calc. )
The sum of the distribution heat loss ( meas. ) and
practical impact on the branch pipes. addition
total heat thedelivered
monitoring continues
to LTDH network in Lys- MWh the standby 287.2 heat loss from the273.9 substation
Additional information's for figure 10 can trup with further analysisheat of substations
demand MWhis on the other238.1 hand larger in the DHSU
219.4
be found in [ 10 ].
Extracts - Danfoss Lystrup Technical Paper and distribution network. case than in the case with IHEUs, because
MWhthe 49.1 54.5
Technical Paperdistribution heat loss Results and experiences… stor-
additional heat loss due to the
Distribution heat losses and pump %age tanks more 17.1 than counteracts 19.9 the
Conclusions
electricity consumption heat power, yearly avg. kWreduction of the
- distribution heat loss.
31.3
pipedemonstration
The of the main district project heating
of a low network.
- tem- However,
In the caseinconsidered,
areas with hydraulicthe distribution limita-
For a Danish reference year DH demand, This solution
perature DHcan supply temperature,
be installed
network for in low DH
an-existing
energy °Ctions,
heat loss suchfor as the-outer urban
area withareas, 67.4DHSUs
DHSU's are
heat loss in the distribution network and district heating
buildings has network
shown
supply at a the
that
temperature, locationconcept
LTDH hav- °Cofferslightlyin turn
lowersome than advantages,
55 in the area with thanks
52.7 IHEU's. to
the annual electricity use of the pump ing a sufficient
works. The results flow in
show thethatreturn
it is pipe.
possible In the
The lower
sum of peak
the pressure
distribution / load
heat require-
loss and
return temperature DH °C 30 34.1
were calculated based on duration addition
to supplythe monitoring
the customers continues
with a supply in Lys- ments.
the standbyMoreover,
heat loss the smallest
from themedia substation pipe
curves divided in 8 representative inter- trup with
temperature electricity
further use, pumping
analysis
of approx. 50of station
°Csubstations
and satisfy kWh diameters
is on the other of2600
thehandhouse larger in2566
connection the DHSUpipes
vals combined with load vs. temperature and distribution
both network. and the safe
the SH requirements in
casethethan
market
in the have
casestill
with a valuable
IHEUs, because water
curves derived for the first monitoring TABLE 4: of Key
provision DHW. dataThisoffactnetwork operation flow
is confirmed overcapacity
the additional heatand lossthis
duesuggestto the stor- that
period of 2010 ( as figure 6 ). In addition, by the fact that there were no complaints smaller
age tanks volume
moreofthan the storage
counteracts tank can the
a full year ( 2011 ) of measurements is Conclusions
from residents about the lack of SH or be chosen,ofinthe
reduction case of DHSU,heat
distribution and loss.this
available, see table 4. DHW. The energy efficiency
The demonstration project of target a low - tem- was would
However, reduce
in areas the with
substation
hydraulic heat loss,
limita-
The heat loss of the reference year for the met,
perature being DHthe distribution
network for low heat
- energyloss space
tions, suchoccupation
as outer costs
and ( 2010
urbancosts ) somehow.
areas, DHSUs
entire network is in line with the expected equal
buildings to 17has % ofshown
the totalthat heatthe production
concept The
offerconclusion
in turn some is that withintotal
advantages, the tested
thanks to
heat loss calculated in the design phase for the Danish
works. The resultsitemshow year.
reference that itEven better
is possible [ € / m ]substations,
the lower [peak € /the
unit ] is a better
IHEU
pressure / load[ €solution
require-
]
and comparable with the present share of real - life performance
to supply the customers is expected
with a supply when in regards
ments. to energy
Moreover, theperformance,
smallest media instal-
pipe
pipes* 120 lation costs and space requirements. 65,000 Any-
the heat loss in the existing city-wide dis- unintended
temperaturenon - insulated
of approx. 50 pipes
°C andare get-
satisfy diameters of the house connection pipes
tribution networks in Denmark. However, ting
bothinsulated.
the SH requirements pipe fittings* and the safe 32 how in thethere
market is nohave superior
still a valuable17,000water
substation con-
the measured heat loss for the full year 2011 provision of
In DH networks of this DHW. This fact
pipekind, is
laying** confirmed
serving low131 cept
flow for all
overcapacity purposes, and but
this the
suggest system
100,500 that
is about 11 % higher, which can be by the fact
heat density areas that there were no complaints should
smaller be chosen
volume oftaking
the into
storage account
tank can the
DHSUwith no possibilities
substation* 3,700 41,000
explained by a distance of unintended from residents about
for future expansion, the design should the lack of SH or specific
be chosen, characteristics
in case of of the
DHSU, siteand and of
this
non-insulated pipes before the DH meter DHW. The energy IHEU substation*
efficiency target was the
would demand.
reduce 2,600 the substation 78,000
heat loss,
envisage the exploitation of the maxi-
in each dwelling with IHEU. Considering met, being
mum the
substation
pressure that distribution
installation**
can be withstood heat loss by space occupation 1,000 and costs41,000 somehow.
these pipes insulated, the distribution heat equal to 17 %pipes.
of theThe totalnetwork
heat production The conclusion is that within the tested
the media
pump + frequency controller* design Aknowledgement
2,400 + 2,000 4,400
loss for a network with 11 IHEU's is only for the Danish
method can thusreference year. Evenso
be optimized, better
that substations, the IHEU is a better solution
slightly above the corresponding distribu- real -distribution
life performance total cost when
is expected in regards
The projects tohave
energy performance,
received 346,900
grants frominstal-
the
the heat loss can decrease
tion heat losses for the 11 DHSU's. The total unintended non
even further, at expenses - insulated
cost perofhouse pipes are
an additional, get- lation
Danish costs
EUDP and- space
program requirements.
making it
8,460 Any-
possible
distribution heat loss in the low-tempera- tingless
but insulated.
significant, pumping demand. how
to there and
develop is nodemonstrate
superior substation the low-tem- con-
ture network are approx. ¼ of the esti- InTABLE
DHanalysis
networks of this cept
perature for all purposes,
concept. but the system
The 5: investment
points at kind,
costs
the serving
fact that low the
mated heat loss in the case of conventional heat density areas should be chosen taking into account the
dimensioning of DHwith systems no possibilities
need a bet-
medium-temperature network (single for future expansion, the design should specific characteristics of the site and of
ter basis for simultaneity factors and that
pipes, series 1, 80/40 °C, 6 bar system, 1 m/s the demand.
aenvisage the exploitation
greater consideration mustofbe thegivenmaxi-to
flow velocity). The electricity use for pump- mum pressureof that
Costs
the operation thecanSH be andwithstood
DHW instal- by the end-users and their installations for
ing was estimated to be 2,600 kWh/yr, the media pipes. The network design Aknowledgement
low-temperature district heating; strate-
lations, for the calculation of the optimal
equivalent to 9 kWhel/MWhth. This is com- The totalcan
method investment
thus of the system
be optimized, so thathas gies than can be used in a nearby area with
size of the heat distribution system.
parable with the electricity demand for been
the estimated heat
distribution to 346,900
loss can € ordecrease
approx. The
1000projects have received
single family houses. The grants secondfromarea the
pumping purposes in existing well-estab- The
8,460 results
€ perdemonstrate
house, see that table it is5 possible
[ 4 ]. It is Danish EUDP - program making it possible
even further, at expenses of an additional, is in Høje Taastrup near Copenhagen,
to guarantee antheenergy -efficient theopera- to develop
lished systems [4]. According to the design also
but
tion,
seen
less
but 30
from
significant,
it is
table,
pumping thatdemand. DHSU where a lowand demonstrate
-temperature DHthe low-tem-
system sup-
method, it was expected to measure is about % very
moreimportant
expensivetothan obtainthe perature
plying 75concept.
houses with floor heating has
The
proper analysis points at the fact that the
a larger pumping demand; the lower elec- IHEU. functioning of each substation, been built. This area is supplied with an
dimensioning
otherwise of DH systems
unacceptable return need a bet-
tempera-
tricity use for the pump is explained in alternative mixing system as described in
ter basis
tures for simultaneity factors and that
result.
practice by the fact that the pressure levels
aOngoing R & D project
greater consideration must be given to [9]: a 3-pipe shunt arrangement is con-
in the network were still well below the lim- nected to the pumping station supplying
the operation of the SH and DHW instal-
its set. This points that there is room for The investigations continueofinthe an ongoing mainly DH return water to the low-tem-
lations, for the calculation optimal
optimizing the network design method R&Dofproject where low-temperature
size the heat distribution system. dis- perature area. When the return tempera-
even more, so that the heat loss can be sig- trict heating is demonstrated in two exist- ture is not sufficient, a portion of water
nificantly decreased, at expenses of an The results-family
ing single demonstrate that it is possible
house neighbourhoods.
to guarantee an energy - efficient opera- from the supply pipe can be added into
additional, but less significant from the The first area is in Tilst near Århus, where the mixing shunt. In this case the low-tem-
overall primary energy point of view, tion,
the low but-temperature
it is very important concepttoisobtain being
proper perature network is supplied by water
pumping demand. Hydraulic limitations tested infunctioning
a street withof each substation,
8 houses with radia- mixed from the supply pipe and the return
and noise must not be forgotten, though. otherwise
tors. The focus unacceptable
is on strategies returnto tempera-
prepare
tures result.

Danfoss District Energy 10 9

38

10
 SAV has been operating in the UK since 1988.
SAV Systems is a leading provider of innovative building services
solutions designed to boost energy efficiency and cut carbon emissions.

Products vary from sophisticated mini-CHP systems to simple water

meters, but all are purpose-developed to serve a better internal
environment and a greener world.
Used as individual products, or integrated into custom-designed
systems, the SAV range makes a perfect partner for low energy
technologies from renewables to central plant and even district
heating systems.

FloCon It’s a full family of choice sourced from some of the world’s leading
specialist companies - our Partners in Technology. Product groups
Modules
cover the full spectrum of modern building services.

www.sav-systems.com

UK CUSTOMER SUPPORT CENTRE

SAV Systems, Scandia House, Boundary Road, Woking, Surrey GU21 5BX
Telephone: +44 (0)1483 771910 EMAIL: [email protected]
GENERAL INFORMATION: Office Hours: 9.00am - 5.00pm Monday to Friday.