I
IEE REVIEW
Automatic controls in building services
M.V. Harrold
D.M. Lush
Indexing terms: Control systems, Power systems, Power utilisation
Abstract: Until recently, the term ‘controls’ was
confined, in building services, to the thermostatic
devices associated with the heating, ventilating
and air conditioning (HVAC) systems. Nowadays,
the definition is generally considered as encompassing the motor starter and control panels,
energy management and building automation
systems, lighting controls and, at least, the interfacing with fire, security and lift systems. Based on
this premise, the review covers the range and
applications of the automatic control systems now
available, with a review of the state of the art, and
some predictions as to the likely direction in
future. The aim of the review is to provide an
insight into current and future control systems.
The intention is to enable electrical engineers to
better co-ordinate their activities with their
mechanical and HVAC colleagues who commonly
specify the control systems which have to be integrated into the total services design. The topics
reviewed include the effect of correct selection on
the operation of the HVAC system, the electrical
requirements of controls, guidelines on the application of controls to particular systems, coordination of effort between all the parties
involved, the variations and scope of EMS/BAS
and direct digital control, lighting control, and
testing and commissioning.
1
Introduction
Control theory for process work and environmental purposes is, essentially, identical, but in practice, applications
are markedly different. The use of controls in industrial
processes is generally accomplished by the application of
highly-priced equipment to satisfy a very precisely
defined set of requirements, which if not satisfied may be
very costly in terms of lost production. Because reaction
times within the process may be very rapid, and instability is never far away, it is important to engineer and
check on the stability criteria for each control loop.
The controls used for environmental purposes are sold
in a highly competitive market at an order of cost lower
than their industrial counterparts, but are functionally
comparable. The controls are applied to systems which,
like process plants, have their performance requirements
Paper 5982B. first received 20th February and in revised form 6th
August 1987. Commissioned IEE Review
The authors are with Arup Research & Development, 13 Fitzroy Street,
London W1P 6BQ, United Kingdom
1EE PROCEEDINGS, Vol. 135, Pt. B, N o . 3, M A Y I988
closely specified. However, the systems are inherently
stable, the reaction times are normally much slower than
industrial processes, and less than specified performance
creates discomfort instead of financial loss. Fairly obviously, extreme or prolonged discomfort may produce less
efficient output from the people affected, with consequential financial implications.
It is often believed that expertise in process control is
readily transferred into the simpler automatic controls
systems for building services, a view possibly reinforced
by the preceding comments on stability and reaction
times. There are however major unpredictable elements
which negate this assertion - people, their use of the
building and their very personal views of what is the
most
acceptable
environment
for
themselves.
Environmental control systems are not simply technical
agglomerations of specific items to satisfy a given set of
conditions, they have to be sufficiently flexible to adapt
to a large number of options that can be set and reset
over many years with changing patterns of use. To appreciate the importance of the environmental requirements,
the topic is covered in some detail.
Costs and engineering input have already been mentioned and, while the market is competitive, it should not
be thought that engineering requirements are ignored.
While it is not necessary to carry out stability analysis for
control loops, the engineering required to design and
specify the control system for all the combinations of
usage patterns and environmental requirements is very
demanding and relatively costly when compared to the
design input for other elements of building services.
The march of technology, particularly in the electronics field, has had a marked and continuing effect on
the development of automatic controls over the last ten
to fifteen years. What used to be impossible or horrendously expensive is now commonplace and, at face value,
apparently inexpensive. This frequently creates situations
where the seductive quality of high technology obscures
the fact that a mundane solution would be simpler,
cheaper, better understood, more easily commissioned
and maintained and, most important of all, perfectly
suited to the problem it solves. It is essential to understand what technology can offer to automatic controls
systems, but it is crucial for the long term benefit of the
final user to keep the system as simple as possible to
satisfy the needs.
This review covers automatic controls for building
services from their earliest known history through their
development for environmental purposes, their theory,
systems and applications. It will concentrate on the use of
electrical/electronic controls, remembering that equivalent pneumatic systems exist for most electrical arrangements. Until the early 1970’s the more complex projects,
105
mainly air conditioned, tended to use pneumatic controls.
The advent of cheap electronics/chips has now made controls available for individual air conditioning terminal
units which are competitive with their pneumatic equivalents. The pneumatic arrangement of a temperature
sensor and small pneumatically actuated valve(s) for this
purpose were the only cost effective solution prior to this
development.
It is not the object of this review to cover the contractual and interface problems which arise and have to be
overcome on projects both large and small, simple and
complex. It is however becoming increasingly true that
automatic controls elements occur within all building services systems and that analogously they frequently form
the mortar holding the bricks together to provide a
homogeneous co-ordinated arrangement.
In this context it is advantageous to design the automatic controls as part of a single package which also
includes the motor starter and control panels, the interconnecting wiring and, whenever used, the building management system.
The term automatic controls applied to building services used to be confined almost exclusively to the
environmental controls used on heating, ventilating and
air conditioning (HVAC) systems. It now encompasses
controls for lighting, fire, security and building management and spreads into lift control and communication
systems. This review is primarily concerned with HVAC,
lighting and building management.
2
History
Controls, for most of us, have been available at whatever
level is appropriate to the HVAC systems that are in
vogue at that Particular time. Sometimes there might
have been an apparent lag between what
be ‘Onsidered ideal and what is actually manufactured, but
good engineering (design and control) has always been
able to improvise suitable control schemes. At the present
time, we all suffer from a surfeit of control devices,
systems and literature (which is not always concise or
technically relevant).
In some ways it has not always been like this, and in
other instances very little has changed. The history of
thermostatic control in the modern sense dates back to
around the middle of the 19th century, when mercury in
glass thermometers, incorporating electrical contacts, initiated alarms. However, Galileo in the 16th century produced a simple device to investigate degrees of heat and
cold. In the late 19th century, electric thermostats were
available in Europe and the USA, and pneumatics as we
know them were developed in the USA in the early 20th
century, although they were not in common use in the
HVAC industry until the 1930s in the USA, and later still
in the UK.
The first UK thermostat of note was not available
until the late 1920s when the Rheostatic Company (now
Satchwell Control Systems Ltd) produced such a device,
and the industry has always owed a lot to manufacturers
outside the UK. Until the last ten years, when a multiplicity of firms in the UK started to manufacture electronic devices, very few wholly UK companies have
operated in the commercial controls business.
Despite the advances in the design and range of suitable appliances for controls, thermostats, detectors, controllers, valves, etc., there are doubts about their use. To
quote an address to a learned society: ‘It is difficult to
understand the prejudice against automatic control. It is
106
not many years ago that the President of the Institution
‘deprecated in his own Practice the elaboration of automatic mechanisms, because in his view they were not
needed, and, in the second place they were liable to get
out of order . . . . In normal heating work, all that fancy
work was not needed‘.’
You may recognise the sentiments, but they date back
to 1941 in an address to the Newcomen Society by
Dufton. Things have not changed, as another recognisable comment is : ‘These devices are very complicated,
their cost is very high and their performance is very
uncertain, so it seems that they ought to be abandoned
completely’.
Do not despair: despite these quotes, successful
control systems can be designed which do not suffer the
implied problems, but the factors mentioned later must
be taken into account. Incidentally, the second quotation
precedes the first, having been made by Picard in 1897, in
a paper on heating & ventilating.
The tenor of comments such as those quoted should
always be in the forefront of any controls philosophy.
Good control systems always depend on designing correctly, selecting the controls to do the job as efficiently as
necessary, and ensuring that the systems will be kept in
working order by the users.
To do this, it is necessary to remember that the history
and development of controls now encompass many other
systems in addition to the simple choice of which elements will control a given loop, and whether the system
should be electronic or pneumatic.
3
Environmental and usage requirements
3.1 Genera,
The environment in which we live and work has two
basic elements: the external, Over which we have, as yet,
no control, and the internal, which can be maintained at
a specified condition according to our needs.
The three major sub,ects for which internal
environmental control is considered are:
(4 people
(b) machines
(c) processes.
While the conditions for each may differ, frequently there
are common requirements. In the case of people, the
requirements are often referred to as ‘comfort’ or
‘environmental comfort’ conditions.
Environmental control is sometimes considered only
in terms of the ability of a thermostatic control system to
maintain specific conditions, but this is a narrow concept.
It is necessary to consider the internal environment in
terms of the overall building design, as the selection and
construction of the building fabric may have a marked
effect on the environmental performance and energy consumption. Equally, the selection of suitable plant and
equipment may also affect the same parameters.
In the past, the achievement of a specified environment
tended to ignore the efficient use of the energy utilised for
the purpose. Since the fuel crisis of 1973/74, this attitude
has radically altered, and energy conservation is considered as a major design parameter in all building services systems. In the UK approximately half of the
overall energy is used in building services, overlapping
with the domestic sector, which in its own right consumes
about a quarter of the overall total. As both sectors are
directly related to the internal environment, there is considerable potential for energy conservation by suitable
integrated design and selection of equipment.
IEE PROCEEDINGS, Vol. 135, P t . E , N o . 3, M A Y 1988
The use of electrical energy is important in all forms of
environmental control, whether it is for the supply of
thermal power, circulation of air and water, or control.
Apart from thermal power, where the choice of fuel is
often governed either by its availability or the apparent
economics during the design period, electricity will virtually always be involved with the other elements. The
use of particular energy sources such as oil, gas, coal,
electricity, etc. may be governed by the specific application. The comparison of fuels in terms of economics and
costs to the client, as distinct from the primary energy
consumption for each fuel, is important and needs to be
considered in selecting building services systems, but is
an indirect element of environmental control.
the occupants. Fig. 2 illustrates the elevation in temperature required to compensate for increasing air movement.
V
’r
3.2 Internal environment
The indoor environment should be safe, appropriate for
its purpose, and pleasant to inhabit. The parameters to
be considered include the thermal, acoustic and visual
conditions.
3.3 Personal comfort
In human terms, an individual senses skin temperature,
not room temperature, although the latter affects the
former. The body loses heat by evaporation of moisture
from the skin, convection to the surrounding air, and
radiation to, or conduction with, cold surfaces. These
mechanisms, together with the degree of activity and the
type of clothing worn, tend to maintain the skin temperature constant (except for exposed extremities) over a
wide range of environmental conditions. The narrow
zone of real comfort conditions is often classified as
neutral or comfortable. This is shown in Fig. 1 in terms
of temperature, and illustrates the degree of satisfaction
for a group of people in a particular space, about the
optimum neutral temperature for the group. The specified space temperature is therefore always a compromise
and is only one of the criteria affecting comfort.
0.2
0
I
0.4
I
I
I
0.6
0.8
1.0
oir movement, mm
Fig. 2
Corrections to dry resultant temperature for air movement
3.4 Temperature and humidity
The term ‘space temperature’ has been used so far to
avoid confusion. Most people assume that space temperature, specified in terms of the dry bulb air temperatures, defines levels of warmth. The previous
comments indicate that this may not be valid, although
temperature detectors in common use are mainly calibrated for, and measure, dry bulb air temperature. Alternative temperature indices may provide better definitions
of comfort conditions, or are used for design calculations:
these include equivalent, effective, globe, dry resultant
and environmental temperatures. Resultant temperature
t,,, is considered to be a measure of comfort dependent
on internal dry bulb temperature tai, mean radiant temperature t, and speed of air movement u as defined as
follows :
(1)
tres = Ctr + taiJ(lOu)I/C1 + J(1Ou)I
Give that the recommended values of t,, are those in
Table 1, it is possible to adjust room temperature detectors or thermostats to a level suitable for comfort, for any
mean radiant temperature and air velocity.
3.5 Limits
- 0
- 5 -4
-2
0
2
4
5
departure from neutral temperature,T
Fig. 1 Comfort vote for personnel, at around the neutral temperature
for varying criteria
Comfort scale :
hot
warm
slightly warm
slightly cool
cool
cold
Other important parameters that affect comfort are the
radiant temperatures of the surfaces surrounding the
space, and the relative humidity and air speed in the
space, all of which can have a marked effect on the space
temperature necessary to provide optimum satisfaction to
IEE PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y I988
Limits need to be applied to any specified comfort conditions, particularly in the case of temperature and humidity. Generally, in terms of human comfort, limits of f 1°C
about a specified temperature and a relative humidity (rh)
of *lo%, about a mean of 50%, will be acceptable.
Limits more critical than is necessary will create additional and unnecessary costs.
There may also be statutory limits which have to be
applied in terms of energy conservation. In the UK since
1974, outside the domestic sector, space temperatures
now have an upper limit of 19°C. This is only a partial
limit, because the reference covers only the heating. To
complete the limit, the regulation would have to specify
an upper limit of 19°C for heating cycles and a lower
limit of, say, 25°C for cooling cycles. Between these two
limits there would be neither heating nor cooling input.
The level of humidity in a space can have a considerable impact on comfort, but in a heated building there is
a limited range of control over its value. Artificially
increasing the air change rate by opening doors and
windows is unlikely to be acceptable in winter, and is a
palliative in summer. Fortunately, the band of comfort
conditions in humidity terms is fairly broad for most
people, and often occurs, in practice, internally. Below
107
Table 1 : Recommended design values for dry resultant temperature
Type of building
Assembly halls, lecture halls
Canteens and dining rooms
Churches and chapels:
up to 7000 m3
>7000 m3
Dining and banqueting halls
Factories:
sedentary work
light work
heavy work
Flats, residences and hostels:
living rooms
bedrooms
bathrooms
Hospitals:
corridors
off ices
operating theatre suite
wards and patient areas
Hotels:
bedrooms (standard)
bedrooms (luxury)
Laboratories
Offices
Restaurants and tea-shops
Schools and colleges
Shops and showrooms
Swimming baths:
changing rooms
bath hall
Warehouses:
working
packing
- and .
- spaces
.
storage space
trms
("C)
18
20
18
18
21
19
16
13
21
18
22
16
20
18-21
18
22
24
20
20
18
18
18
22
26
16
13
approximately 35% rh, static electricity effects occur, and
noses and throats may be affected; above, say, 65%, the
effect of stickiness may be felt. Air conditioned systems
are normally designed to avoid these extremes.
3.6 Visual and acoustic parameters
The acoustic and visual impact on comfort conditions is
extremely important. The correct level of illuminance for
a particular task is important in its own right, but the
overall aesthetics are a combination of the lighting level,
the lighting source, the architectural finishes and their
reflecting properties, furniture and equipment. These aesthetics contribute to the comfort of the occupant.
Sound, in terms of personal comfort, is also related to
the particular task. People's reaction to sound varies
according to age and situation. Acoustics is both complex
and subjective, and only a broad outline is included for
the purpose of defining general criteria for comfort.
Because the response of the ear is non-linear and less sensitive at low frequencies, it perceives equal loudness for
various combinations of frequency and sound pressure
levels, units of loudness being defined in phons. Sound
pressure is a fluctuating air pressure sensed by the ear.
The fluctuations are minute in relation to atmospheric
pressure; sound pressure levels are specified in decibels
The levels are created by the sound power (the power
transmitted by the sound waves) which is normally considered only in reference to a sound source. Sound power
levels L are also referred to in dB:
L = 10 log (W/W,) decibels
(2)
where W is source power (measured in W) and W, is reference level (normally 1 pW).
Sound pressure is proportional to the square root of
sound power.
108
Acoustic control does not form part of the automatic
controls covered in this review. It is normally part of the
design and equipment selection process of the heating
ventilating and air conditioning (HVAC) system. The
incorrect selection or use of automatically controlled
dampers in the ductwork systems can contribute to noise
problems, as can rotating machinery which may transmit
noise through structure, pipework or ductwork.
3.7 Machines and processes
Apart from comfort criteria for personnel, there are two
other areas where environmental conditions may be
important: machine rooms and process plants. The
former covers spaces such as computer rooms and
medical machine areas and the latter covers areas such as
electronic manufacturing or food processing factories. In
many ways the criteria are similar to those for personal
comfort, but there may be requirements for closer limits
and, in particular, air filtration becomes a significant
factor, the number and size of particles being very closely
defined according to the process. In certain critical processes, the air movement patterns are also specified,
laminar flow probably being the most difficult to achieve.
While temperatures are often specified with limits of not
greater than k0.25 or fOS°C, with humidities of
- 2% rh or better, two points should be made:
(a) Limits specified may be unnecessarily stringent and
need to be questioned. As an example, computer rooms
in the past needed close limits if the machines were to
operate correctly, but nowadays this not normally necessary.
(b) The achievement of conditions throughout the
treated space. Strictly, the specified conditions can normally be achieved only at the point of detection, the
variation throughout the remainder of the space being
largely dependent on good plant design and distribution
of the heating and cooling media.
+
3.8 Safety requirements
In environmental control, the specified conditions and
their limits have to be achieved, but generally only during
normal periods of occupation. It is therefore necessary to
consider the environmental conditions necessary under
abnormal circumstances. Examples include emergency or
maintained lighting levels when the normal system
breaks down or during unoccupied hours, the low limit
temperatures to be maintained to prevent freezing or
damage to equipment and furniture and the maintenance
of humidity below a specified dewpoint condition to
prevent condensation on cold glazing.
3.9 Usage and plant
The operating patterns of plant usage vary considerably
in any particular building and from one to another. The
plant sizing, hours of operation and functional performance are interrelated as well as being dependent on
other external factors. To design a comprehensive control
system, an understanding of these matters is essential,
and a summary of the main features follows.
3.10 Plant operating periods
Periods of plant operation should be optimised so that
comfort conditions are achieved during the occupied
periods within those areas of the building being used.
With the exception of special buildings e.g. museums or
art galleries, and special areas e.g. computer rooms, the
specified conditions are not required outside occupation
periods and during the unoccupied periods only protecIEE PROCEEDINGS, Vol. 135, P t . E , N o . 3, M A Y 1988
tive conditions such as frost protection and security lighting are normally required.
For mechanical services plant, such optimisation is
achieved by variable switch-on and (switch-off) times for
the plant so that the required conditions are reached as
the occupation period commences. Optimisers may also
provide a variable stop time, set for operation some time
before the building occupation ceases.
Lighting systems will normally be switched independently of the mechanical plant, and switching may be
automatic, manual or a combination of both. Arrangements should be made to limit the lighting levels according to the task, e.g. daytime office work and night-time
office cleaning. The switching requirements may often be
programmed through a building management system,
that has been provided primarily to control other building services.
3.11 Intermittent heating and cooling
The calculation of loads for steady state and cyclic situations leads naturally to consideration of the effects of
running the plant intermittently to satisfy only the specified environmental conditions during periods of
occupation. Inherent in this are the following:
(a) the preheat period necessary to bring the building
up to temperature under varying climatic conditions; in
particular, on the coldest day for which the load is calculated
(b) the thermal response of the building during
preheat, which will depend on its construction, insulation
and ventilation
(c) the thermal response of the plant when first
switched on
( d ) the ratios between preheat, normal heating and
plant-off periods
(e) the relative running and capital costs.
3.12 Matching plant output to load
Boilers and chillers are selected to match the maximum
heating and cooling design loads with, often, a known
standby capacity. To obtain maximum efficiency, the
control system should be designed to control the plant
without the standby capacity being normally on line. In
most buildings the plant operates for the greater proportion of its life at an output well below the maximum
design figure.
Plant should be selected with the capability of being
turned down to match the minimum normal load
requirement and the control system must be designed so
that this turn down is accomplished in accordance with
the varying load. The overall efficiency of plant operation
is normally improved if the plant and controls are selected on the basis of modulating operation, rather than
start/stop cycles.
It is necessary that the degree of tolerance permitted
from the desired temperature be clearly defined and not
invalidated by incorrect selection of boilers or chillers.
3.13 Zoning
The efficient operation of any control system in terms of
energy consumption and environmental comfort is governed to a very large extent by zoning the building in
accordance with differing usage patterns and the spatial
planning for accommodation. Because a large open-plan
officewith South and West facing aspects cannot be adequately controlled from a single space temperature detector, due to the solar gains on each face varying with time,
zoning is required. There are several identifiable areas
I E E PROCEEDINGS, Vol. 135, Pt. B, No. 3, MAY 1988
which affect zoning and the adequacy, or otherwise, of
the final control system, and they can be interactive.
3.14 Internal conditions and gains
The varying usage of portions of a building will indicate
heating and cooling loads which differ from day to day,
or season to season, because of differences in occupancy,
process heat gains and specified environmental con&tions in the spaces. Each portion may constitute a zone
with separate controls to satisfy the environmental
requirements.
3.15 External load variations
Heating and cooling loads in various parts of the building vary because of solar effects, changes in direction and
intensity of wind, and changes in outdoor conditions. The
effect of these variations may suggest splitting the building into zones of aspect. This has the effect of simplifying
the control problems and reducing the overall energy
consumption. A further split into upper and lower zones
should be considered on tall buildings where the exposure at the higher levels may lead to increased heat losses
and also where one zone per aspect may create excessively large plants.
Zoning based on the proximity of other buildings,
which shelter or shadow the building being zoned, should
only be carried out after careful consideration, since the
configuration of neighbouring buildings may change after
the system is designed. Inner areas of buildings, more
than 6 m from the perimeter, are not usually affected by
external load variations.
Corner areas that are fed from two aspect zones need
special detailing to ensure that the two zone inputs do
not cause local discomfort.
3.16 Construction
Tall buildings may suffer from stack effects causing
uncontrolled heating to the upper floors. These can be
alleviated by sealing off stair wells, lift shafts, and other
vertical openings. However, with or without horizontal
zoning, the use of terminal controls will overcome this
problem. Spaces having large areas of glazing may
require special treatment as independent zones.
3.17 Economics
Piping arrangements, plant space and ductwork distribution will, in many cases, be governing factors in selecting
the number of zones. However, the segregation of those
portions of the building which are not being used, or
require different conditions from their neighbours, may
well justify the expenditure on extra zoning, both economically and in the context of energy conservation.
4
Theory
Mathematical control theory is extremely well documented in text books, and forms part of the curricula for
all electrical engineering courses, and it is not the itention
to cover it here. There are, however, some texts which
deal with the subject as it relates to HVAC plant [l].
This Section identifies those areas particularly relevant to
the subject generally [2].
4.1 Control modes
A controller can be made to operate a final control
element (FCE), e.g. a valve, damper or luminaire, in a
number of different ways in response to a signal. The way
in which the FCE acts in response to the signal is known
109
as the control mode. There are several principal control
modes, sometimes used singly or in combination, while a
number of variations also occur. The control mode does
not define the means by which control is effected, which
may be through mechanical, electro-mechanical, electronic or pneumatic systems.
Apart from the on/off mode, all other modes are continuously variable (modulating) and the means of obtaining the signal were all based on variations of the
imbalance obtained when the resistances in the legs of a
Wheatstone bridge were altered. Modern systems now
use a bridge network with voltage division capability to
provide the necessary facilities.
some of the characteristics described under Section
overshoot
a
4.I. 1 On/off control: On/off control provides only two
plant outputs, maximum (on) or zero (off). The control
sensor usually takes the form of an on/off thermostat,
pressure switch, humidistat, etc. and operates so that
when the controlled variable is below the set point, the
contacts close; this is known as direct acting. The reverse
operation i.e. contacts made when the controlled variable
is above the setting, is said to be reverse acting, and
devices with changeover contacts are common.
There will always be an interval in between the contacts making and breaking when no signal change can be
given. This range, where no control signal change occurs,
is generally referred to as the differential. Since on/off
control can normally produce only two plant outputs,
cycling is bound to occur, see Fig. 3, and such control is
'18
I
I
I
I
I
I
I
I
I
I
time
!
ZL
Fig. 3
Action of onloff controller
more suited to high capacity systems where the time
taken to reach design temperature is long. It should be
noted that the temperature swing is wider than the differential, owing to the thermal inertia of the system being
controlled. Simple room thermostats used for heating
applications often have a slow response characteristic,
resulting in wide switching differentials. This problem is
often solved by the addition of an accelerator in the form
of a very high resistance heating element which is energised only when the thermostat is calling for heat. The
effect of this is to heat the sensing element artificially,
thus anticipating the effect of the space heating on the
thermostat. This increases the speed of response, and
reduces the switching differential, as shown in Fig. 4a.
There is a complementary effect associated with this
particular arrangement. The ratio of on to off periods of
the thermostat with increasing space load varies the
internal heat input to the thermostat, and alters the controlled condition as indicated in Fig. 4b. This variation is
normally acceptable, but it does effectively produce a
proportional band (see Section 4.3), and the thermostat
110
load, %
b
Action of thermostat fitted with accelerator
Fig. 4
* Corresponds to a small differential in space temperature together with a variation in the thermostat enclosure temperature equivalent lo the differential of the
nonaccelerated thermostat
There are certain specialised forms of on/off controller
which permit multiple stages of plant capacity. These
include multistep thermostats and step controllers. The
former are uncommon except for specific applications,
but the latter are frequently used in conjunction with
detectors and conventional controllers and solid state
versions are now replacing the rotating cam types.
4.12 Proportional control: Proportional control action
refers to a control element having an output signal proportional to the change in its input signal. The proportional band is the range of input necessary to produce
the full range of control output. It can be expressed in
terms of a physical quantity (e.g. "C, Pa, '?4 humidity, lux,
etc.) or as a percentage of the controller scale range.
If the scale range of a temperature controller is from
20 to lOO"C, and the proportional band setting is such
that the controlled variable must change through 100°C
to make the valve move from full open to shut, then the
proportional band is 100/80 = 1.25 or 125%. If a change
of only 20°C is required to bring about this same 100%
control action then the proportional band is 20/80 = 0.25
or 25%. Alternatively, in the latter case, the proportional
band may be said to be 20°C. Fig. 5 illustrates the
relationship between the control action and the proportional band width, and shows that the controller will
give the desired value corresponding to the set point at
only one position of the FCE. This occurs at the 50%
open position. The diagram is drawn for a heating
system.
Under all other load conditions, within the range of
stable control, some degree of offset (a sustained
deviation) will be present depending on the proportional
band setting and the prevailing load. For a controller
with proportional only action, the output signal is given
by :
where k p is the proportional control action constant
(equal to the reciprocal of the proportional band), fii is
the input signal, and Po is the output signal.
4.7.3 Integral control: A correcting element may be
arranged to remain stationary when the controlled
medium is at the desired value, and move in a corrective
manner at an increasing speed that is proportional to the
I E E PROCEEDINGS, Vol. 135. Pt. B, No. 3, M A Y 1988
deviation from the desired value. For a controller with
integral action only, the output signal is given by :
by :
r
(7)
(4)
where k, is the integral control action constant, and 8 is
time.
6o
t
I
0
100
4.3 Availability of various modes
The stand alone proportional controller used to be the
most common in use despite its inherent offset problem,
I
I
I
I
1
J
75
50
25
0
valve p o s i t i o n (%)
open
closed
time-
7------
u c -
offset remaining if
controller h a s
proportional action only
U
-
0
C
I
c
time
t
,
!
.I!
"
oc
.Q
:or
t i m e d
,-.7 5 r
1
c
> a
25
Fig. 5
time
Action of a proportional controller
4.I .4 Derivative control: A correcting element may be
set so that its speed of operation is proportional to the
rate of change of the controlled variable. This action is
used to eliminate overshoot during a fast load change.
Derivative control is mainly applied in special process
systems. For a controller with derivative action only, the
output signal is given by :
Action of proportional plus integral ( P + I ) controller
Fig. 6
+I
I
where kd is the derivative control action constant.
4.2 Combinationsof basic modes
There are many combinations of the basic modes, but
apart from the stand alone proportional controller, the
most common combinations are:
( a ) Proportional plus integral (P I) control: This
mode combines the stability of proportional control with
the accuracy of the integral mode to eliminate offset. The
behaviour of P + I control is shown in Fig. 6, and the
output signal is given by:
+
Bo =
-k,,pi
-
k,
s
/3i d e
(6)
(b) Proportional plus integral plus derivative (PID)
control: This form of control combines the advantages of
(P + I) control with derivative, to combat sudden load
changes, while maintaining a zero offset under steady
state conditions (see Fig. 7). The output signal is given
IEE PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y 1988
Fig. 7
Action of proportional ( P + I ) plus derivative (PID)controller
111
because the cost of adding the additional mode (term) or
modes (terms) was expensive and only justified where
very close control tolerances were specified. The advent
of the microchip has effectively made it cheaper generally
for manufacturers to provide all three terms, rather than
tailored units, and depending on the application the same
unit can be used as P, P + I, I or PID.
4.4 Time lags
The full effect of corrective action is not immediately
apparent in the controlled medium, owing to various
time lags, which are functions of the plant characteristics,
involving units of quantity, potential and time. The most
common types of lag are:
4.4.1 Dead time: The time interval between a change in
a signal and the initiation of a perceptible response to
that change, see Fig. 8.
4.4.2Distance-velocity lag: The time between a signal
being sent to an element and the element starting to
respond, arising from the finite speed of propagation of
the signal, see Fig. 8.
I
y ' w
I
,I
I
valve
movement
Fig. 8
I
time
Transfer functions of a typical plant
A Calorifer coil temperature
B Calorifier secondary temperature
4.4.3 Exponential lag: An exponential lag occurs when
the change with time in the output from an element or
system (resulting from the application of a step change in
the input signal to that element or system), is of simple
exponential form. In heating and air-conditioning applications, a step change in input signal to a plant rarely
produces a pure exponential characteristic at the plant
output, partly because of distance-velocity lag and partly
because the signal is generally passed through a series of
stages. For example, if the position of a three-port valve,
feeding a calorifier is changed, in order to change the
temperature of the water flowing on the secondary side,
the following time lags are involved (relating to Fig. 8
which illustrates the general shape of the transfer functions of such a plant):
(a) distance-velocity lag, being the time taken for the
hotter water leaving the mixing valve to reach the coil
(b) time lag which occurs in raising the temperature of
the coil
(c) time lag which occurs in raising the temperature of
the water in the calorifier.
The time lag described in (b) is associated with a
temperature/time relationship which, to all intents and
112
purposes, is a pure exponential (curve A). The
temperature/time relationship in (C) is not a pure exponential (curve B) because its shape is dependent upon the
form of the preceding temperature/time relationship.
Curve A has a time constant of T . In curve B, the time
constant is not valid in the sense of the simple exponential mathematical expression, but a notional time constant of T, as illustrated, is sometimes employed if control
equipment and plant are to be analysed.
4.4.4 Capacity lag: This lag may be defined as a capacity for storing energy which may be on the demand side
of the process e.g. heated water in a tank, or on the
supply side e.g. the hot water in the primary heating coils.
Capacity lag will occur when satisfying this demand. In a
single capacity process this lag would be exponential, but
in practice most processes are multicapacity. Where heat
is transferred from one capacity to another transfer lag
also occurs.
These lags tend to slow up response, and vary considerably according to the process in hand and the magnitude and frequency of the load changes. They may also
affect the choice of control mode. An integral mode will
provide a temperature with close limits when used to
control the air temperature directly downstream of a
heater, but if the same heater battery is to be controlled
from the extract duct this mode is not suitable. The controller operates over a very narrow band of temperature
and in the first case, there is only a minimal lag between
the detector calling for a change in output and sensing
the resulting change. In the second case, the lag time
(comprising capacity, transfer and distance-velocity
elements) is large, and the change of output will continue
until the detector senses the change, which may occur
only after the general space temperature has altered very
considerably.
4.5 Open and closed loop control
Control systems are applied to control loops, each loop
normally comprising one section of an overall plant, e.g.
reheater, calorifier, smoke alarms. Loops are normally
defined as being open or closed, most in the HVAC field
being closed. When the condition of the controlled variable changes in closed loop control the sensor detects the
change and initiates correcting action to the final control
element. The effect of this is reflected in the controlled
variable and reassessed by the sensor which continuously
provides feedback to the controller. Examples of this are
space temperature control, control of secondary domestic
hot water system (DHWS) temperature, and static pressure control of supply fan volume in variable volume
systems. All modes of control may be used to provide
closed loop control.
Open loop control does not include feedback of the
type described above. The simplest form of such a loop
would be the detection of smoke, when a local and/or
remote alarm is initiated by a smoke detector without
any feedback to the detector. A more common version in
the HVAC field is the use of an external sensor to control
the output from an air curtain in a shop entrance. In this
case the output from the heater battery serving the air
curtain would be modulated in accordance with the
outside temperature so that, for example, at 0°C externally there would be full output, and at 15"C, no output.
The external sensor is not affected by any variation from
the final control element (heater battery control valve),
and therefore there is no feedback. Only on/off or proportional modes of control may be used for this type of
loop.
IEE PROCEEDINGS, Vol. 135, Pt. B,No. 3, M A Y 1988
I
4.6 Stability analysis
In the past, little attention has been paid to the stability
of heating and air conditioning systems at the design
stage. Generally speaking, systems have depended for stability upon the large thermal capacity of traditional and
heavyweight buildings with large capacity boilers, heat
exchangers, etc. However, with the trend towards lightweight and better insulated buildings, and the use of air
conditioning and low capacity boilers and heat
exchangers, it is necessary to consider system stability in
greater detail. Stability analysis makes this possible. Such
analysis is normally complex and the analytical solutions
require numerical data for a wide number of parameters
to test the stability criteria.
It should also be added that there may be difficulties
in applying these techniques to conventional HVAC
systems which frequently have nonlinearities due to actuator reactions, heat transfer characteristics, and limited
plant outputs. At present, suitable techniques for the
HVAC field are in their infancy.
4.7 Valve parameters
In defining the control modes it must be remembered
that the selected controller operates an actuator which
normally forms part of a valve assembly (or damper
assembly) which is controlling the output from a heat
exchanger such as a heater or cooler battery. The aim of
the vast majority of HVAC control loops is to achieve a
linear relationship between thermal power output and
valve spindle movement. The flow characteristics of the
valve have a marked effect on heater and cooler battery
outputs, and correct selection is an extremely important
part of control system design.
4.8 Basic valve characteristics
The relationship between valve spindle lift and area of
the valve opening is known as the valve characteristic. In
the case of shoe type (or rotary) valves a similar relationship exists between shoe rotation and area. The valves in
general use for HVAC systems are shown in Fig. 9. There
is a large range of possible valve characteristics but those
most commonly encountered in control systems are :
(a) linear, where the orifice area is directly proportional to the valve spindle movement, and the flow varies
a
b
linearly with spindle lift (this should not be confused with
power linear which is basically achieved with (6) and (c))
(6) characterised V-port, with a characteristic falling
between linear and equal percentage
(c) equal percentage, or similar modified parabolic,
where equal increments of valve spindle lift provide an
equal percentage change of the previous area. The characteristic is represented by:
s = log
(2)
where s is change in valve stroke (%), V, is volume flow
rate at zero stroke if valve was characterised to this point
(m3/s), and VI,, is volume flow rate with valve fully open
(m3/s)
(d) quick opening, where the flow increases very
rapidly from zero for small valve spindle movements,
with a fairly linear relationship between flow and spindle
movement. Beyond this initial movement the flow rate
varies more slowly with increased spindle movement.
Such valves are primarily used for on/off service.
The terminology used to describe valve characteristics
varies considerably, and it may be necessary to check
with manufacturers if there is any doubt. Fig. 10 shows in
stylised form the basic characteristic curves for valves
described in (a) to (4. The basic characteristics are
obtained by measuring flow against a constant pressure
drop across the valve for all positions of the valve. In
these circumstances, the valve is said to have an
‘authority’ of unity. In practice, these curves are modified
as the valve authority decreases. Normally, the plug
design determines the characteristics on the assumption
that the plug lift will be directly proportional to the
signal from the control loop. However, it is possible to
achieve a particular characteristic by making the plug lift
follow a particular curve not proportional to the signal
and the use of software based controllers makes this
easier.
4.9 Valve authority
Normally, a control valve is employed to regulate the
flow in a system where the total pressure drop in the
circuit concerned is derived from the circulating head of a
pump which should be reasonably constant. Figs. 11 and
C
n
d
e
Fig. 9
Some common valve types
a Three-port single seat valve
d Three-port shoe type valve
b Three-port double seat valve
e Butterfly valve
c Two-port single seat valve
IEE PROCEEDINGS, Vol. 135, Pt. E , N o . 3, M A Y 1988
valve spindle lift, %
Fig. 10 Stylised basic valve characteristics
A Linear
C Equal percentage
B Characterised V-port
D Quick opening
113
12 show examples of systems where the valve authority is
defined by :
represent the quantity of fluid which passes through a
fully open control valve at unit pressure drop.
In SI units, a flow coefficient A, based on the quantity
of fluid in m2/s for a pressure drop of 1 Pa across the
valve, has been defined in BS4740 (Part 1, 1971), to
replace the existing coefficients. In practice, manufacturers' figures are published only as C, and k, values. The
traditional imperial valve capacity coefficient C, is
defined from:
(9)
where N is valve authority, P , is pressure drop across the
valve (fully open, (Pa), P 2 is pressure drop across the
remainder of the circuit (Pa).
Suitable valve authorities in practical terms should be
kept as high as possible. To achieve this, the pressure
drop through the battery has to be known, or defined to
suit the selected valve. The pressure drop across the valve
itself is based on its flow coefficients.
4.10 Flow coefficients (capacity indices)
The pressure drop across given control valves is a function of the flow coefficients, or capacity indices, which are
published by individual manufacturers. These coefficients
v
where is volume flow rate (imperial units, gal/min), C,
is the valve coefficient (imperial units), Pi is the pressure
drop across valve (imperial units, psi), and sf is relative
density (specific gravity) of fluid.
The two indices are not compatible, but the conversion factor is given by :
A, = 28.8 x
a
b
Fig. 11
Valve authority
a Three-port mixing application
b Three-port diverting application
C,
(1 1)
4.1 1 Combination of valve parameters and battery
characteristics
The thermal output characteristics of heater and cooler
coils do not vary in a linear manner with changing water
flow rates. They tend to give a large percentage of their
rated output for a small percentage of the design flow
rate. The controls objective is to ensure that this nonlinear characteristic is modified so that thermal power
output has a linear relationship with valve spindle movement.
Fig. 13 illustrates the importance of suitable valve
authority: the upper right hand quadrant of the diagram
shows a stylised set of valve authority curves for both
linear and equal percentage valves, while the lower right
hand quadrant depicts a typical heater or cooler battery
characteristic. The relationship between valve lift and
power (heating or cooling output) is shown in the bottom
left hand quadrant and this is a function generated from
valve lift
flow
Fig. 12
Valve authority
P ,. iP ,-
P
=
(1
Basic two-port application
(typica-I)
b Two-port application,ladder arrangement
Fig. 13 Stylised valve authority and power output characteristics
equal percentage
N = Valve authority
~ _ _ _linear
~
114
I E E PROCEEDINGS, Vol. 135, Pt. B, N o . 3, M A Y 1988
a valve authority curve and the battery output characteristic. Using curve 2 (equal percentage characteristic
N = 0.7) the points on the new curve are defined by
drawing a series of lines a,b,, a , c , , a, b,, a, c , , etc. and
then drawing lines vertically through b,, b,, etc. and
horizontally through c,, c , , etc. The points of intersection d,, d,, etc. define the power output/valve lift curve.
This particular curve shows an almost linear relationship
between lift and output, but for comparison, the power/
lift curve is also shown based on curve E (linear characteristic N = 0.8). The latter illustrates why linear valves
are not recommended for this type of application. It indicates that any valve selection which provides an output
curve of this form could provide stability at or near
maximum output and be unstable at low load conditions.
Note that the latent cooling characteristic of the cooler
battery does not follow the same curve as the total
output as this may be relevant to applications where
latent cooling is paramount.
4.12 Setting out control requirements
To move from the theoretical to the practical control
system requires receipt and understanding of data on
environment and usage, plus the following conditions
and tolerances.
4.13 Conditions and tolerances
The design, selection and operation of overall systems are
related to the specified conditions and tolerances for both
the internal environment and the individual items of
plant. These parameters are of particular significance to
the control system, and care should be taken to ensure
that the tolerances are appropriate to the application.
The relationship between dry bulb temperature, wet
bulb temperature and relative humidity of air are fixed
for specified barometric pressures, and normally defined
by a psychrometric chart (Fig. 14). Moisture content on
r e l a t i v e humidity,
100
%
50
must be considered on its merits, particularly for projects
outside the UK. In special areas, the tolerances may need
to be maintained to f5% rh and f 1"C of the specified
desired values, and in critical cases these figures could be
as low as 1 2 % rh and f0.5"C.
Both the general and special cases should be quoted in
specific terms, e.g. 50% f 10% rh and 20°C f 2"C, or
24°C 4°C - 0°C. Fig. 14 indicates this problem where
there is individual temperature control of a space, with
central control of the humidity. If the specified space
requirements are 22°C f 2"C, and 50% f 5 % rh, and the
moisture content corresponds to the line A-B, then at
22"C, 50% rh will be achieved, but at 20°C and 24°C for
the same moisture content, the relative humidity will fall
outside the specified limits. Equally, at 22"C, moisture
contents designated by lines C-D and E-F will not quite
satisfy the humidity limits, but if the temperature varies
from 22°C to the permitted limits, with either of these
constant moisture contents, the humidity limits will be
decisively broken. From this example, the relevance of
conditions and tolerances to design is illustrated clearly.
Either, there should be separate control of humidity in
each space if the specified conditions are precisely what is
wanted, or the tolerances should be amended so that they
can be achieved with normal operation of the plant. In
the example this would mean specifying 22" f 2°C for
temperature control, and accepting a relative humidity of
50% + 12% - 10% for a central plant capable of
holding the moisture content between C-D and E-F.
In addition to limits for space conditions, those for
heating and cooling source flow and return temperatures,
and for particular locations within air handling plant
must be specified with suitable tolerances. Select the tolerances for the correct space requirements and then
provide the controls to satisfy these parameters. The use
of conditions or limits of less than 1 2 % rh for humidity
is not normally practical.
The stability of flow temperatures for boilers and chillers may be important to the control of individual plants
elsewhere in the system. Again, any limits specified must
be within the capability of the selected plant and control
system. Limits of f 2 " C and f0.5"C for heating and
cooling, respectively, are the lowest figures likely to be
achieved under all load conditions.
+
5
Equipment and systems
In this Section prominence is given to the electric/
electronic elements of control systems rather than the
purely mechanical components e.g. valve bodies, or pneumatic controls. Electronics encompasses microprocessorbased systems and building management systems.
Until we consider building management systems, most
control systems loops comprise a sensor, a controller
(except in the case of the on/off mode), and a final control
element (or elements) that is normally a motorised valve
or damper assembly. There are wide variations across
each of these elements and the more common types are
described with some indication of their specific features.
d r y b u l b temperature,
Fig. 14
"C
Psychrometric diagram of conditions and tolerances
the chart is indicated by horizontal lines in units of
kg/kg. Humidity in the range of 40-70% rh and temperatures of 19 to 23°C will normally be acceptable, if the
air velocity is low enough to prevent the sensation of
draught. There may be situations even for general
comfort where these tolerances are too wide. Each case
1EE PROCEEDINGS, Vol. 135. Pt. B, N o . 3, M A Y 1988
5.1 Sensors
Irrespective of the form of sensing element or the parameter being controlled, the siting of the elements is vitally
important. Pipework and ductwork applications require
that the total active portion of the element is immersed in
the fluid flow, and where large cross-sectional areas are
encountered, averaging elements serpentined across the
area should be considered. In space mounted applica115
tions, the moutning surfaces should not be affected by
solar radiation either directly or indirectly, nor should
they be part of hot or cold structural bridges. In large
spaces it may be difficult to select a representative position for a single detector, and consideration must then be
given to multiple averaging detectors, split plant or
acceptance of a reduced level of overall control. Methods
of sensing are described for the common parameters in
HVAC systems, i.e. temperature, pressure, humidity, flow
and enthalpy. Different sensor mechanisms are available
for each parameter, and manufacturers usually select particular elements for use in their own ranges of equipment.
The selected elements are suitable for general applications in building services.
Sensitivity and drift in calibration are becoming more
important with improving standards for environmental
control. Both these parameters can have a marked effect
on comfort and energy consumption.
16). Water should flow past the sensing element at not
less than 0.3 m/s to minimise the time lag, whereas air
should circulate around the sensing element at a
minimum of 2 m/s.
5.1.1 Temperature: The most commonly used temperature detectors rely on either thermal expansion or
electrical changes due to temperature variations. The
familiar bimetallic strip, and rod and tube elements both
rely on the differential thermal expansion of two dissimilar metals rigidly fastened together. Sealed bellows are
used as temperature sensitive primary elements. The
bellows are evacuated, and partially filled with a volatile
liquid such as Freon, a proportion of which is boiled off
as a vapour until an equilibrium pressure is reached. This
pressure is a direct function of the temperature of the
liquid, particularly at the liquid surface. The change in
pressure creates movement of the control mechanism.
Remote bulb sensing elements are an extension of the
vapour-filled bellows principle, a remote bulb being connected to the bellows by a capillary tube. Resistance
bulbs operate on the principle that the electrical resistance of certain kinds of wire (nickel, copper, platinum)
change with temperature. Usually a resistance bulb is
made up of a coil of wire wound round a core and
installed in a rigid capsule.
Thermocouples are formed when two dissimilar metals
are welded or twisted together at one end and this junction is heated. A voltage, proportional to the temperature
difference between the welded and free ends, is then
developed. The pairs of metals are usually copper/
constantan,
iron/constantan,
chromel/alumal
or
platinum/platinum-rhodium.
Thermistors are semi-conductor devices which have an
inverse relationship between temperature and electrical
resistance. This relationship is nonlinear, but the output
can be made approximately linear over a range of 10°C
by using linearising resistors. These are very sensitive
detectors up to 100°C.
In any practical system, the response to a change in
input will not be instantaneous. This delay in response is
called lag, and the lag coefficient is defined as the time
needed to reach 63.2% of the total change. Fig. 15 gives
the response curve for various sensors. For example, from
curve C, the lag coefficient is 1 min. It will be seen that a
protective tube or pocket causes the lag coefficient to
increase considerably, and also that sensing elements
respond much more quickly in moving liquid than in
moving air, since air has a very low thermal capacity and
conductivity.
Another factor influencing response is the type and
velocity of the fluid in which the detector is immersed. A
low fluid speed, with respect to the sensing element,
greatly increases the resistance to heat transfer (see Fig.
A Thermistor in water
B Thermistor in pocket in water
time, rnin
Fig. 15
1 I6
Typical temperature sensor response curves
C Resistance bulb in water
D Thermistor in still air
I
I
I
I
I
0.2
0.4
0.6
0.8
1.0
water velocity, ms-1
Fig. 16
Effect of water velocity on heat transfer
5.1.2 Pressure: For pressure sensors, bellows, diaphragms, Bourdon tubes, or similar devices may be used.
The medium under pressure may be transmitted directly
to the device used to operate the mechanism of a pneumatic or electric controller. Rate of flow, quantity of flow,
liquid level, and static pressure may all be controlled by
pressure sensors. The most common types of pressure
sensing devices utilise a bellows or diaphragm acting
against a spring. Elements of this type are very sensitive,
being able to detect pressure changes of about 10 Pa and
can be used over a range of absolute pressures from
0.15 Pa to 150 Pa. The Bourdon tube is made from a flattened thin-walled tube formed into a C shape and sealed
at one end. As the pressure is applied to the open end, the
tube tends to straighten out, producing movement at the
tip, proportional to the applied pressure.
5.1.3 Humidity: The primary element of a humidity
sensor can be human hair, nylon, wood or other substance that responds to humidity change. As the moisture
content of the surrounding air changes, moisture is
absorbed or released and the element expands or contracts, operating the mechanism of the controller.
A biwood element operates on a similar principle to
the bimetalic element for temperature measurement, but
in this case two woods with differing moisture absorbancies are used. Humidity sensing elements suitable for
electronic application and control may also consist of
metallic contacts fused to a piece of glass or plastic. Electrical contact is maintained between the two strips by a
hygroscopic salt, e.g. lithium chloride, painted over the
surface of the material. The resistance of the salt varies
with the amount of moisture absorbed.
IEE PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y 1988
5.1.4 Flow: Existing flow measurement instruments
operate on a variety of principles, but the most widely
used are based on the differential pressure across a
restriction. This readily lends itself to either electric or
pneumatic transmission and control. The restriction is
installed in a duct or pipe to create a pressure differential
that bears a square law relationship to the rate of flow.
This restriction may be an orifice plate, a Venturi tube or
a flow nozzle. The orifice plate is widely used because of
its convenience and economy, and reliable data are available for its calibration. Extensive use is now also made of
calibrated regulating and control valves for the same
purpose. Positive displacement and turbine instruments
are commonly used, and magnetic flux flow meters are
now available.
5.1.5 Enthalpy (total heat): A measure of enthalpy is
obtained using temperature and relative humidity sensors
(often mounted in a common housing) which feed their
signals into a controller, the output of which is characterised to give an approximately proportional signal.
Because the major application is optimising the use of
fresh and return air in air conditioning systems, some
devices include a four input controller (i.e. inputs related
to the temperature and humidity of both fresh and return
air). The enthalpies of the two air streams are compared,
and used to control mixing dampers or heat recovery
equipment.
5.2 Actuators
An actuator is a device that responds to the signals
from a controller, and enables power to be transmitted to
a final control element. A variety of types of actuator
exist, the most common of which are detailed here. The
two characteristics which most affect actuator selection
are the available torque (and its efficient transmission to
the final element) and stroke time, which may be significant in terms of overall system response. The most
common application of actuators is to the operation of
valves and dampers.
5.2.1 Electric motors: Electric motors in the 5 to 50 VA
range, operating through a reduction gear train, produce
relatively high torque at low speeds. The motors are
usually single-phase, and may operate on either line
voltage or extra-low-voltage supplies. Reversible motors
are essential for modulating systems, but either reversible
or unidirectional motors are suitable for two-position
operation. In unidirectional motors the limits of operation are determined by internal limit switches.
Where safety features are required, motors may be
fitted with a spring return mechanism that returns the
motor to a specified condition when the normal power
fails. Alternatively, an internal power pack with rechargeable batteries may be incorporated.
Some electric motor actuators are also designed to
drive a screw which is part of the actuator spindle; others
are arranged to provide a linear motion of the spindle
directly proportional to the detector signal, and some can
provide a linear motion which follows any required characteristic related to the detector signal.
5.2.2 Solenoids: Solenoid actuators are normally
electromagnets with a movable armature linked directly
to the device to be controlled. However, the use of such
actuators with dampers is unusual. The most common
solenoid actuators provide two-position operation, and
generally return to the de-energised position on power
IEE PROCEEDINGS, Vol. 135, P t . B, No. 3, M A Y 1988
failure, and are therefore suitable for failsafe operations.
To achieve the maximum force from these valves, the
armature travel is relatively short. Their application is
normally limited to small size valves, and they have a
very restricted maximum differential pressure rating.
A particular type of solenoid valve may be used for
modulating control whereby a variable voltage is applied
to the solenoid, and the valve spindle takes up a position
proportional to the applied voltage. This arrangement
normally utilises very restricted valve spindle movements.
5.2.3 Pneumatic actuators: Pneumatic actuators comprise a piston, or diaphragm, to which air pressure is
applied to provide a linear displacement. The construction of the actuator and its method of connection to the
valve or damper determines whether, with increasing air
pressure, the fluid flow is increased or decreased. Most
pneumatic actuators are of the single action type in
which the force on the diaphragm is opposed by a spring,
and the net force applied to the valve or damper is the
difference between them. When the air pressure is
removed, the spring will return the valve to the selected
extreme position, and this may be used for failsafe
requirements by specifying the necessary conditions.
In certain process applications, double acting actuators are employed in which air is applied to one or other
side of the diaphragm or piston while exhausing the
opposing side. In these cases, only the air is used to drive
the actuator, and there is no automatic failsafe feature.
Single action actuators operate from air at pressures
from 100 to 250 kPa. However, air pressures up to
1 MPa are used for double acting actuators because of
the high forces required, although this is rare in building
services applications. When pneumatic actuators are
required to deliver a large power output, or operate
against variable forces, the position of the actuator
spindle may not be proportional to the signal pressure
from the controller. To overcome this, a pneumatic relay,
known as a positioner, is fitted to the actuator which uses
main air line pressure to ensure that the actuator is correctly positioned for a given controller signal. To do this,
a positive feedback of actuator position is used.
5.2.4 Electrohydraulic actuators: Electrohydraulic actu-
ators consist of an oscillating pump, pressure cylinder
and piston for opening, and return spring and bypass
valve for closing. A variation, sometimes used for control
valves, employs wax filled actuators in which the change
in volume associated with phase change in the wax is
used to operate the valve.
5.2.5 Self operating actuators: Self-operating actuators
are based on the pressure change applied to a sealed
liquid or vapour-filled bellows system which then operates the spindle of the valve directly. When the temperature changes, the liquid or vapour in the sensing
element expands or contracts, and the consequent pressure change is transmitted, normally via a capillary, to
the bellows system which acts against a return spring and
moves the valve spindle. Radiator valves frequently
employ this form of actuator. Some wax operated actuators may be regarded as self-operating.
5.2.6 Manual actuators: Manually operated actuators
are normally added to automatic control valves to permit
the normal actuators to be overridden. Various versions
are available which allow valves to be manually positioned in one or two specific positions or in any position.
117
I
5.3 Valves
A valve which is aqtomatically controlled to regulate
fluid flow from zero tb full quantity is generally a control
valve. Manually ope ated valves which can be made to
perform the same fu ction also have a control function,
but they are more co monly referred to as regulating or
isolating valves, or y their means of construction, e.g.
gate or globe valves. he type of control valve for a specific duty depends oq the service required, conditions of
temperature and pres ure of the fluid involved, and other
circumstances, includ ng valve characteristics, described
in Section 4.7 ‘Valve arameters’.
Terminology in re pect of valves varies according to
the field of applicatign. The terms used here are those
generally associated yith the building services industry.
The use of the alterqative terms ‘two-way’ and ‘threeway’ or ‘two-port’ and ‘three-port’, respectively, is normally acceptable. A qonventional control valve consists
of four basic componeplts:
(a) actuator
(b) body (shown in Fig. 9)
(c) bonnet, containing a packing box or gland (shown
in Fig. 9)
(d) valve plug (shown in Fig. 9b).
The selection of the correct control valve for a system is
an integral part of overall design, and an essential prerequisite for a stable system which can be properly commissioned.
Whereas two position, or on/off, valves need be selected for specific throughputs under clearly defined conditions, the requirements of modulating valves need much
closer attention. It has been noted in Section 4.11 that
power output is not directly proportional to flow, and
i
4
11
l-----l
parallel
Fig. 17
A Ideal
118
U
varies with both valve and emitter characteristics. It is
important to ensure that the valve may be shut off or
opened against the prevailing pump and/or static pressures; this is particularly true of large valves, but some
terminal unit control valves have a very limited tolerance
in this respect.
5.4 Control dampers
The following details and data on dampers and their
basic application are intended to illustrate the effect of
suitable selection on the stability of air flow for control
purposes. Certain assumptions have been used to simplify
the techniques. Where detailed selection becomes necessary, a more rigorous procedure should be adopted. The
flow of air in an air handling system is regulated by
control dampers in a similar fashion to the regulation of
water by valves. Control may be either two-position, as
in the case of an isolating damper, or modulating, where
two air streams are to be mixed, as in face and bypass
control. Regulating or balancing dampers are not dealt
with here.
5.5 Characteristics
There are two arrangements of rotating blade dampers :
parallel and opposed. The relationships between damper
blade rotation and flow through the damper, for a constant pressure drop across the damper, are shown in Fig.
17. These are called the inherent, or intrinsic, characteristics of the damper. Ideally, these would be straight lines,
but in practice they take the form shown. In practice also,
the pressure drop across the damper is not constant, but
increases as the damper is closed. Consequently, the
inherent characteristics are modified to the installed characteristics (see Fig. 18).
100
80
z
6 60
c
e
3
z” 40
20
opposed
Damper types and inherent characteristics
B Parallel
C Opposed
n
“0
20
‘0
20
40
60
blade rotation, %
80
100
40
60
80
blade rotation, %
b
Fig. 18 Installed damper characteristics
a Parallel blades
b Opposed blades
100
I E E PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y I988
5.6 Time controls
There are three basic methods of time control: on/off
time switch, fixed time/boosted start, and optimum start
control. The selection for a heated or air conditioned
building requires a number of factors to be considered,
including the following:
(a) size of building and annual fuel consumption.
(b) hours of occupation and length of unoccupied
periods.
(c) thermal inertia of the building structure (e.g. heavy
or light building structure and degree of thermal insulation).
(d) boiler or chiller margin.
5.6.7 Onloff time switch: Simple time switch control is
usually limited to small installations such as small fan
convector and unit heater installations. This method is
simple, low in cost, and easy to install. The method has
the disadvantage that the start time is selected to ensure
that the space is at the desired temperature at the start of
occupation under the design temperature conditions.
This means that in less severe conditions energy is
wasted.
Some optimum start controllers include the facility for
optimum stop. The early stop time is again determined
by the comparison of inside and outside temperatures.
The application of this feature is not common at present,
owing to the uncertainties of prediction, and possible
problems associated with stopping air circulation in
ventilated and air conditioned buildings. Some of the
energy saved by this arrangement is required for boost
the following morning.
The operation of optimum start control is shown in
Fig. 19 which indicates the basic principle, coupled with
frost control. Fig. 20 illustrates how this same optimiser
copes with different situations including varying outside
air temperatures (too).
p
temperature
$
3
i-
$
F night minimum
2
g
._
cn
C
._
typically
5.6.2 Fixed time boosted start: Fixed time boosted
start is designed to reduce preheat time by allowing the
central plant to provide maximum preheat capacity.
Facilities for an earlier start time following a prolonged
shutdown (e.g. a weekend break) should also be provided
on the time switch to supply the additional heat required
to bring the building up to the correct temperature for
occupation. The system should incorporate a means of
terminating the boost by either time or temperature. This
method is now largely superseded by optimum start
control which reduces the energy used for preheat in the
intermediate seasons.
5.6.3 Optimum start and stop control: This is a time
controller for intermittently occupied buildings which has
shown substantial energy savings when compared to an
on/off time switch or fixed time boosted start. It provides
optimum plant start time by continuously monitoring
both inside and outside temperatures to provide sufficient
preheat to give the desired conditions at the beginning of
the occupancy period. Early start can also be provided
following prolonged shut-down, such as at weekends.
Early-on and day omission devices are normally provided, and a day extension timing facility with manual
setting can be added.
The controller takes into account the thermal response
of the system, structure, and duration of the off periods
over a daily cycle. The largest energy savings are generally from buildings having light structures and with a
heating system of low thermal capacity. For maximum
economy, the plant should be started at the latest possible time, allowing for the minimum preheat period at
full capacity. Termination of boost may be provided by
an internal detector in advance of occupation time if the
internal conditions are satisfactory. This detector may
also be used for frost protection.
The cost and performance of optimisers vary widely,
and care should be taken to ensure that the model selected is suitable for the particular application. Modern versions use adaptive software which simplifies initial setting
and permits the unit to monitor and correct its own performance. During unoccupied periods, control should be
based on minimum inside air temperature, to prevent
freezing or condensation and separate frost protection
override can be provided where required.
IEE PROCEEDINGS, Vol. 135, Pt. B, N o . 3, M A Y 1988
10°C
I
plant
off
Fig. 19
t i m e t
frost and
condensation
protection
optimised
on
occupation
time
Optimum start control characteristics with frost protection
)design
temperature
t i m e d
roccupation
optimum
time
intermediate
additional
boost p e r i o d
a f t e r prolonged
shutdown
Fig. 20
o p t i m u m start
for design
condition
Optimum start control characteristics for varying conditions
5.7 Controllers
There has been a marked change over the past decade in
the ‘controllers’ used in conventional HVAC control
systems. In the past, such controllers, other than pneumatic, were based on a bridge network (electromechanical,
electronic or solid state) which provided, via a balance or
motor relay, an extra-low voltage signal, energising either
the open or close terminal of the actuator. In addition,
there was, very frequently, a potentiometer on the actuator which provided feedback signals to the controller. A
version of this arrangement, for an early proportional
plus integral controller, is shown in Fig. 21.
Modern controllers and their associated actuators can
take many forms, and the overall electronics may be distributed between them. The most common ranges of controllers generally provide a ramped output of 0-10 V DC
according to load demand. This signal is fed to an elec119
logue basis using setting potentiometers for each value.
With the advent of softeware based control logic, the
setting of each parameter has been by a structured
setting-up procedure where each parameter has been
selected and set via alpha numeric key pads. The digital
nature of such software is thus replacing the traditional
analogue controllers.
tronic positioning device that may be part of, or independent from, the actuator. This device, or relay, controls the
position of the actuator via a bridge and amplifier, a
Schmitt trigger device, a triac switch and a feedback
potentiometer on the actuator. The block diagram for
such an arrangement is shown in Fig. 22 which is self
explanatory. The settings at which the actuated device
Fig. 21 Controller schematic
0Satchwell Control Systems Ltd.
bridge circuit
bridae
summer
amplifier
0
temp.
detector
p.b.
actuator
T1
-
+
-
t oc
EO
Schmitt
7
2LV
OC
- -
!
feedback signal
(motorised valve or damper) may operate can be adjusted
in terms of the voltage at which the actuator is allowed
to start opening (or closing) and the voltage span
(between 0 and 10) over which it will drive from one end
to the other. The settings may be in voltage terms or in
those of the parameters under control e.g. temperature
set point (start) and proportional band (span). The
arrangement may be set for direct acting or reverse acting
operation.
It is difficult to cover the setting variants possible in
the enormous range of proportional (P), proportional
plus integral (P + I) and proportional plus integral plus
derivative (PID) controllers and their common cousins
with different nomenclatures. Reference to manufacturers'
literature is necessary, if sometimes confusing.
Controllers for the HVAC industry will normally be
single, two or three stage (as distinct from single, two or
three term
see under Section 4). This means that a
controller may operate up to three stages of an HVAC
plant in sequence when required i.e. a heating coil valve,
dampers controlling fresh, exhaust and recirculated air,
and a cooling coil valve. Other features such as dead
zones (temperature bands between heating and damper
operations, etc.) and compensation (the variation of set
point with outside temperature) are often available.
The simplest single stage proportional controller will
require a set value (SV) adjustment and a proportional
band (PB) adjustment. A similar two stage controller
could have an additional PB setting or two: SV and two
PB settings, and in the latter case the difference between
the two SV's would constitute the dead zone (DZ) value.
As terms are added e.g. P + I, extra settings are available
for the integral action time (for each stage).
Until fairly recently, the settings have been on an ana~
120
I
+
Fig. 22 Block diagram of an AET series controller
0Satchwell Control Systems Ltd.
In effect, the use of this form of control is direct digital
control (DDC), that, in the controls field for HVAC
systems, distinguishes it from conventional analogue
control. It also leads directly to the growing use of software, resident in building management systems (BMS), to
perform the same control functions as the independently
mounted digital controllers. Whenever a BMS is incorporated into a building services system, a decision on
whether to use DDC or conventional controllers is
required. The use of DDC is not necessarily the best solution for all HVAC systems - initial cost, complexity of
project and levels of maintenance capability are all
important factors.
The emphasis on defining the meaning of DDC in the
HVAC controls field is deliberate. Jargon, particularly
that associated with computing and its high technology,
electronics/chips, etc., is a constant problem for all engineering disciplines. Nowhere is this more apparent than in
the area under discussion. Automatic controls for building services sit firmly among an enormous number of
interfaces, between such diverse elements as electrical
engineering, mechanical and building services engineering, HVAC controls suppliers, companies with a process
background who see a lucrative future in building services controls, and a multiplicity of software and computing organisations who sell building management systems.
Words of one syllable become extremely important when
specifying the requirements for a comprehensive controls
and building management system.
5.8 Building management and automation systems
A building management system (BMS) is one of a
number of terms which is used to describe an integrated
or overlaid system over and above the basic HVAC
IEE PROCEEDINGS, Vol. 135, Pt. B, N o . 3, M A Y 1988
control system. Here again a more precise definition is
required to avoid confusion. While standard definitions
for the various terms are not known to be available, a
good basis follows:
5.8.1 Building management system ( B M S ): A system
which can provide management functions e.g. billing and
programmed maintenance, in addition to building
automation and energy management.
5.8.2 Building automation system (BAS): A system
which automates the operation of building plant and services to considerably reduce the extent of manual operation which would otherwise be required. It can include
energy management functions as part, albeit important,
of the facilities it offers.
5.8.3 Energy management system (EMS): A stand
alone arrangement specifically related to the energy management functions and efficient use of energy in a building (inherently it will include some automation
functions).
The terms are loosely used and this need not be a
problem, but when a specification for a project is to be
produced the terms need be very precise, otherwise facilities which are not required may be included, and vice
versa.
5.8.4 General: Most thermostatic control manufacturers
now market building automation systems to monitor and
control other building services such as energy, security
and fire protection systems. Other manufacturers also
offer such systems, particularly in the fire and security
sector, and there is a rapidly increasing number of large
and small firms with basic processor and software capabilities who are entering the market. It is always advisable to question the level and validity of the software
offered to perform the tasks required in the context of
building services automation. Much of the software has
still not been proven in practice, and it is unwise to
accept these programs until they have been demonstrated
on a real project. There is often a deep gulf between software capability and the necessary knowledge of building
usage patterns and building services operation to
produce adequate programs.
One other item that requires emphasis is the possible
susceptibility of all processor based systems to electrical
interference, however caused. Interference phenomena
and their suppression is not fully understood, and many
building automation systems suffer intermittent faults as
a result of this [3].
The rapid advance of microelectronics and the consequential reduction in equipment costs has made the consideration and use of these systems much more
widespread. Because the technology and associated software are both still being developed, this section covers
the essential principles and requirements to obtain satisfactory systems. It also covers the facilities available, and
associated arrangements that are now being incorporated. The principles apply whether the system has one
unit with 10-20 points, or multiple processors and outstations serving 10,000 points.
The problems of communication and ambiguity lead
to one other generalised statement. It is essential that the
specification for any BMS/BAS/EMS is precisely detailed
in terms of all performance requirements, functional
sequences and clearly defined interface conditions for
every facility required. Minor facets or requirements, genIEE PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y I988
erally accepted amongst engineers as basic to any operational sequence, will not be provided by the software
unless they are identified in the specification. A detailed
specification of this nature will be more conducive to a
satisfactory conclusion than one which concentrates on
detailing the precise technology which is to be used for
each element of the system.
5.9 Basic considerations
The basis of operation of these systems is to continuously
scan the connected input data points and report unusual
occurrences whilst carrying out control functions automatically. The centralisation of the logic equipment
allows greater flexibility of operations than conventional
hard wired systems and also frees maintenance staff from
watchkeeping duties, enabling more effective use of manpower.
If such a system is considered on any project there are
essential prerequisites if it is to provide its optimum performance.
(a) It must be considered very early on in the building/
plant design concept. If it is added at a later stage it will
duplicate a number of other elements which are complete
in themselves.
(b) The client should be prepared to employ suitable
staff to operate the centre and utilise all its facilities.
(c) The cost effectiveness should be properly evaluated
and include the following:
(i) capital cost of system, including outstations, detectors, instrumentation, modified plant, standard software
and wiring, less any savings for equipment which it
replaces
(ii) interest on capital cost
(iii) value of energy savings which the system itself will
provide
(iv) cost of additional staff employed to operate the
system, less the saving on those maintenance staff who
will be replaced by the system
(v) annual cost of maintaining the building automation system
(vi) a substantial capital sum to cover the cost of collecting data from the building during the first one to
three years of operation and producing software for optimising the energy use in the building, since much of the
existing software does not cover the full range of optimisation possibilities for large buildings or complexes
(vii) costs saved by using programmed maintenance
co-ordinated by the system (which should include value
of increased plant life), less the cost of preparing such a
programme
(viii) savings in using the software capability of the
system to replace the interlocking relays, etc. normally
used in motor control centres and control panels, set
against the cost of the project oriented software.
This basic list of factors leads on to direct cost details
and interface problems that must be considered when
such a system is contemplated. Some examples are:
(a) The functions of a BMS/BAS could be achieved by
hard wired control systems and the economic justification for a sophisticated system includes the fact that the
engineering, site wiring and commissioning costs can be a
fraction of the hard wired system.
(b) If process control is to be achieved by direct digital
control (DDC) then savings will occur in the elimination
of local loop controllers and in the avoidance of duplicating sensors for control and monitoring purposes.
(c) Energy management programs will optimise
energy usage but this does not relieve the designer from
121
ular plant is off, for the period following start-up and
when they sequentially follow on from an initial alarm.
Many computer based systems allow various control and
energy management programs, as detailed later, thus dispensing with the need for conventional items of
equipment, instrumentation or controls. Ultimately, they
will reduce control panels to little more than starter
boards, where control relays are superseded by software.
Such a system permits the logic to be changed without
extensive wiring modifications, but this is not intended to
encourage delayed design decisions.
Planned and programmed maintenance schedules can
also be produced and monitored by the computer based
systems but this requires a large memory storage capability and considerable project orientated input data.
his responsibilities for the correct selection of building
services systems.
(d) Failure of power or plant may result in a direct
loss to the customer, e.g. computer rooms, specified
pathogen-free animal buildings. If it can be shown that a
building automation system minimises the number of
failures and reduces the down time, this is a strong factor
in the economic argument.
BMS/BAS with full facilities will be more easily justified
on larger projects, and offer the most savings where services are particularly complex. Options include fully
owned systems and those where third parties own the
central building automation equipment and sell a
‘remote’ supervision and control service.
It should be remembered that if a BMS/BAS is not
fully utilised or understood, the building services control
and data collection will not be adequate, and it is possible to increase rather than decrease energy used, due to
the system falling into disuse.
5.1 1 Software
Software (programs) available can be defined in two categories : operational and management. The following
describes the software currently available, but this is a
rapidly developing technology and as knowledge of
designers’ requirements increases, so will the availability
of programs. In the meantime, special requirements are
expensive and, as a general rule, best avoided. User
friendly programs, i.e. those where the operator is provided with data or instructions in plain language, are now in
common use, and should always be requested as part of
the design requirement.
5.10 Computer based systems
Transmission of information from the data collecting
outstations to the central monitoring panel is via a single
or two-wire trunk using pulse coded messages (see
Fig. 23). In some systems the scan initiation is entirely
from the central processor that addresses each individual
point in turn, and then processes the data received. The
processor may include software providing PID control of
loops, or it may be part of the outstation facilities. The
use of software for this purpose is currently described as
direct digital control (DDC), and individual conventional
controllers are made redundant. Alternatively, intelligent
outstations are provided, whereby the outstations
address the processor on alarm occurrence. A variation is
the use of microprocessors at various locations which
contain all the intelligence needed for local supervision
and control, including DDC, hosted by a central computer having overall control. This is particularly suitable
for larger projects possibly involving several buildings
where more conventional transmission systems become
overloaded. Data transmission rates are, typically, 300 to
9600 baud.
Alarms for analogue devices may be set and programmed at the central processor and/or the outstations,
for each of the separate sensors. These alarms can normally be inhibited by means of software when the partic-
5.11.1 Alarm priorities: Most systems allow several
levels of alarms to ensure that first order alarms can be
quickly identified. Low priority alarms, e.g. maintenance
alarms, can be programmed to print out only on request.
5.1 1.2 Alarm inhibition: Certain digital and analogue
points inevitably go into alarm when the associated plant
is switched off. Similarly, some alarms will not be
expected to disappear until the plant has been running
for some time, e.g. room temperature out of limits. Such
alarms may be inhibited by software.
5.1 1.3 Analogue alarms: Computer based systems
allow limits to be set by software for each of the measuring elements.
5.1 1.4 Integration: Integration of measured values
allows the calculation of other quantities, e.g. energy con-
I
II
centra I
processor
data transmission
cabling
possible data
I
selective
peripherals
flow
digital
I
1
plant
panel
keyboard
alarms
displov
[unit‘
I
IMI
t
intercom
computing
L
1
central control
Fig. 23
122
I
outstat ions i n s t a l l a t i o n
I
Block diagram of typical building automation system
1EE PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y 1988
sumption from flow rate and temperature difference. This
minimises the need for complex metering systems, and
allows alarms to be set as for other analogue devices.
5.11.5 Totalisation: The major function is the summation of motor run times from status signals which occur
as part of a control function. Alarms set for various times
may be used for maintenance reminders and duty
sharing.
5.11.6 Timeswitching: Plants can be allocated to
various daily programs and special programs account for
holidays, shut down periods, etc. Lighting control is one
specific area which utilises this feature, and programs and
hardware have been developed to provide individual
control of luminaires from the system and locally, with
the use of power cables for signalling.
5.11.7 Event initiated sequences: An external event,
normally an alarm (analogue or digital) may be arranged
to initiate a sequence of control operations.
5.11.8 Load shedding: A maximum demand meter, or
an integrated value of particular variables, may be used
for load shedding pruposes.
5.11.9 Restart after power failure: If the system detects
that power has been restored (after a failure) the program
restarts plants in planned sequence.
5.1I .IOOptimised startlstop: Summer and winter optimisation can be achieved through software, using drybulb/wet-bulb or enthalpy inputs as appropriate,
eliminating the need for local optimiser controls.
5.11 .I 1 Cycling: Where statutory regulations permit,
plant cycling may be employed to reduce energy consumption. Suitable programs permit out of limit conditions to initiate restart of plant.
5.1I. 12 Process control: Control of analogue values
has previously been achieved by local pneumatic or electronic controllers. Temperature and humidity control is
now possible through computer programs using PID
algorithms based on digital systems, instead of the traditional controllers. The software must permit overrides for
manual and emergency use. Standby systems using secondary processors or conventional controllers for priority items may be necessary. This form of control is now
commonly referred to in the thermostatic controls
industry as direct digital control (DDC). The use of terms
such as centralised control when applied to a BAS or
BMS does not normally include DDC, unless it has been
specifically requested.
5.11.13 Reheat reduction : A program allowing an
increase in the offcoil conditions of the supply plant,
determined by the zone with the greatest cooling load,
minimises energy consumption in constant volume reheat
systems.
5.1I. 14 Optimum damper control: Dampers may be
controlled through software from enthalpy sensors to
optimise the use of recirculated, and fresh, air.
5.1I .I5Security: Security patrols may be supervised by
programs arranged for a sequential or timed patrol tour
with alarm indication on default. These alarms may be
used to indicate a sequence of events.
I E E PROCEEDINGS, Vol. 135, P t . B, N o . 3, M A Y 1988
5.11.16 Fire: Air handling plant may be controlled
automatically on detection of heat or smoke. There is a
move towards the replacement of hard wired fire alarm
systems by software data transmission and control
systems, but in the UK specific agreement is required
from local fire authorities before use.
5.11.1 7 Miscellaneous: Among programs which are
commonly available is one for programmed maintenance,
and there are many other options, such as translation to
a second language.
5.12Specification and installation
In order to obtain a BMS/BAS suitable for a particular
project it is essential that the installer is provided with a
clear specification. Schedules, used in conjunction with
schematic diagrams of the systems are a convenient
method of conveying details of requirements, connections
and interfaces to the suppliers and installers. One
example of this is shown in Fig. 24.
The following are some areas to be considered:
(a) Who is to supply and install data input devices
(analogue and digital)? Are pockets provided with pipe
line mounting equipment? What special contact features
are required on the digital devices?
(b) Where are outstations to be located in order to
minimise wiring?
(c) Who is to supply and/or install input, transmission
and intercom cables? Who is to be responsible for checking and terminating these cables? In special screening or
segregation required?
(d) How is switching of 240 V start/stop circuits to be
handled, i.e. will relays be located at the outstation or the
start panel ? How will back-feeds between systems be
avoided?
(e) Who will provide the hand (test)/off/auto switches
which are essential both to enable basic commissioning
and for emergency operation of connected plant?
(f) Can power and monitoring cable terminals at the
starter panels be segregated to minimise damage to outstations due to cross connection?
(9)If sequence/control logic is to be executed through
the automation system, will safety interlocks be required
and how will they be incorporated?
(h) Can all the necessary control and energy management programs be supplied as standard software? Have
the programs been validated in actual use?
( i ) What special requirements are there for duplication
of peripherals? Are mimic diagrams and annunciator
panels necessary?
(j) Which language and units are to be used for
display on VDUs and printers, and who will bear the
cost of translation?
(k) What are the power supply requirements? Are batteries necessary for software/memory protection of
back-up facilities?
( I ) How much spare capacity is to be allowed for later
additions?
(m)Are spare parts provided? Is there a local service
organisation? What is the maximum callout time for
attendance to correct faults?
(n) Who will carry out the testing and commissioning
of the system hardware and software? What system of
servicing and maintenance is to be employed during and
after the defects liability period?
(0) How are the interfaces between all the different
systems and subcontracts to be defined in an unambiguous manner? Who will be responsible for specific
interfaces ?
123
5.13 Control panels and motor control centres
Describing such units to electrical engineers may be analogous to carrying coals to Newcastle. However, there are
certainly differences of approach between the building
services requirements for nonindustrial type buildings
when compared with their industrial counterparts. Generally, the work is carried out under contractual conditions which may differ from those for industrial process
work and the participants are often allied wholly with the
HVAC industry. While many of the requirements may
parallel those of the industrial sector and switchgear
panels, the following resume highlights the more
common areas which require particular attention for the
automation and controls systems.
Cupboard style panels with enclosed starter/isolators
behind single or double doors are more normal than the
cubicle style panel favoured by the process industry. Cupboard style panels do not suffer the problems of starters
being located very close to the floor and the age-old
dilemma of where to locate the duty selector switch for a
run/standby pair of drives.
Panels should be designed to be worked on when live,
with all components being shrouded so that general day
to day maintenance can be carried out.
5.13.1 Equipment and wiring
( a ) Electric or pneumatic controls: The choice may
affect who does what in the panel prior to, and after,
delivery.
( b ) Starters: It is essential to identify power ratings at
which changeover from direct online to star-delta, etc.
takes place. Ensure that any special features are also
stipulated, e.g. back contacts on overloads for alarm purposes, closed transition on star-delta starters. Ensure
suitability of overloads for run up times on large motors.
( c ) Fuses: Specify characteristics to suit applications,
e.g. motors, immersion heaters, SCR controllers.
( d ) Isolators: Stipulate whether suitable for on load iso-
lation.
( e ) Switches and lamps: Types and method of operation are important for control, e.g. three position switch
for selection of run and standby plant should have a
positive off position between the two operational positions.
(f)Ammeters: These are necessary only for the largest
drives.
( 9 )Relays and contactors: Suitable circuitry and performance criteria should be detailed, e.g. coil voltages,
in-rush currents and acceptable number and types of
contacts.
( h ) Controllers: The majority of environmental controllers are designed for fascia mounting and are usually
mounted on the control panel doors.
( i ) Wiring standards: Is the wiring to be loomed or
carried in trunking? Detail the methods of identifying
each connection to match drawings. Specify the type of
wiring to be used.
0’) Terminations: Detail the type of terminations to be
used both for the incoming and outgoing cables and for
interconnections in the panel.
( k ) Space and spares: Plan for additional space for
future modifications and include necessary spares, e.g.
fuses.
Sample of points for a plant associated with a building
Fig. 24
automation system
124
IEE PROCEEDINGS, Vol. 135, P t . B, No. 3, M A Y 1988
5.13.2Interlocking
( a ) Logic: Functional sequences should include all the
requirements for control ,loops as described elsewhere,
and all interfaces between items of equipment and plants
should be interlocked, to ensure that incorrect sequences
cannot occur.
( 6 ) Safety and alarms: Ensure that all safety and
limit devices to protect plant and satisfy any statutory or
other regulations are included in the logic sequences.
Include visual and audible alarms to indicate the operation of these devices at the control panel, with repeater
devices at remote locations where necessary.
( c ) Equipment for logic operations: Decide whether
the interlocking sequences are to be carried out using
conventional relays or programmable controllers
(processor software). In the latter case the software
facility of the (remote) controllers may be used to set up
the logic sequences.
5.13.3Responsibilities
( a ) Panel supplier: Who will supply the control panels?
Interface problems are minimised if panels are provided
by the controls supplier.
( b ) Wiring and pneumatic pipework: The wiring in
the panel is always carried out by the panel manufacturer. The external wiring to the panel may be by the
electrical sub-contractor, the controls supplier, or a
specialist sub-contractor. It is often advisable to make the
HVAC sub-contractor responsible for the co-ordination,
even if the work is carried out by others, thus, the wiring
is included in the HVAC sub-contract, with the exception
of main incomers to panels which are the responsibility
of the electrical sub-contractor. Connection of the incoming cables is normally carried out by the installer unless
there are exceptional reasons such as computer based
automation systems, for which the specialist equipment
supplier often makes the final connections after the cable
has been laid by the installer. The pipework is normally
carried out by the control’s supplier within, and external,
to the panel. Clarification of any alternatives which may
be permitted, e.g. pre-fabrication by the controls supplier
is necessary before fitting to the panel at works or on site.
( c ) Drawings: The information for obtaining and
co-ordinating the information required from the various
suppliers should be vested in one of the sub-contractors
alone. The HVAC sub-contractor is usually best suited to
this function. Drawings finally incorporated into panel
diagrams will include those from:
(a) Controls supplier
(b) Boiler supplier
(c) Chiller supplier
(d) Cooling tower supplier
(e) Starter supplier
( f ) Electrical sub-contractor
(9)HVAC sub-contractor.
( 6 ) On site: All connections into the panel should be
checked before power is switched on, and the functional,
operational and safety sequences should be rechecked
before the plant is operated.
Systems and applications
6
To provide a brief, but electrically orientated, description
of the application of controls to systems, this section will
be restricted to an examination of multiple boiler/chiller
controls, and the control of a dewpoint system for an air
conditioning plant. It will be seen that the selection of the
suitable combination of sensor, controller and motorised
valve(s) is relatively simple compared to the associated
interlocking required to cover all the eventualities associated with usage patterns and limit and safety features.
While the interlocking diagrams are shown in one of the
traditional forms, software based versions are becoming
much more common, and will probably become the
norm within the next decade.
6.1 Multiple source control
With multiple boiler or chiller combinations, the first
choice to be made is whether to run them for series or
parallel operation and control : chillers may occasionally
be installed for series/parallel operation. The answers to a
number of checklist questions are essential for the selection of a suitable system. The different basic arrangements are shown in Figs. 25 to 28.
constant or variable
volume primary p u r n p ( s )
Fig. 25
Parallel sources with primary pumping
For variable volume systems use connection AB
For constant volume systems use connection BC
This Figure excludes all regulating, isolating and control valves and bypasses
rm I
constant or
variable volume
primary pump(s)
secondarv
source source source
I
I
I
Fig. 26
Parallel sources with primarylsecondary pumping
This diagram excludes all regulating and isolating valves
constant or
variable
volume
primary pump(s)
5.13.4 Testing
( a ) At works: All functional, operational, and safety
sequences should be simulated and checked, and compliance with any statutory or other requirements, such as
the IEE wiring regulations, should be ensured.
IEE PROCEEDINGS, Vol. 135, Pt. E, No. 3, M A Y I988
load
Fig. 27
u
Series sources with primary pumping
This diagram excludes all regulating, isolating and control valves and bypasses
125
In the parallel mode points to be considered are:
(a) The amount of over-temperature (boilers) and subcooling (chillers) which can be tolerated during sequence
constant or
variable volume
primary pump(s)
D
loads
Fig. 28
facturers’ guarantees should not be affected but modifications to the normal single unit operating functions may
be required.
(e) Are the unit sizes selected so that operation at low
partial loads does not create additional problems? On
multiple unit installations one unit normally operates at
minimum load more frequently than the overall system
operates at maximum load. Large machines are only permitted a limited number of starts in a given period. What
is the maximum turndown ratio? At least one machine
must be selected so that its minimum output is less than
the lowest normal operating load.
(f)Do the limit devices have the span to cope with
the unit operation over the full range of loads? This is
illustrated by the simple example in Fig. 29.
Series sources with primarylsecondary pumping
This diagram excludes all regulating and isolating valves
load
operation over the total load spin. (Over temperature
situations are shown in Fig. 29). Will the return water
temperature on boiler systems drop below an acceptable
level?
(b) The choice. of primary plus secondary pumps or
primary only, in conjunction with the valving-off, of nonoperational units. This is related to both hydraulic stability and (a) above.
(c) The permitted variation in flow through the units
(flows through boilers and, particularly, chillers), should
not be reduced much below design rating.
(d) Flow balance between individual units. Short circuiting may occur for flows below maximum, unless the
system is correctly designed.
(e) The maximum number of units which can reasonably be controlled in sequence. This is related to overtemperature and sub-cooling and the availability of a
suitable control system.
In the series mode the points are:
(a) the problems of over temperature (boilers) and
sub-cooling (chillers) do not occur. What is the minimum
temperature rise (boilers) or fall (chillers) acceptable
across each unit? Is there a likelihood of flow between
the cold feed and common boiler vent?
(b) the choice of primary plus secondary pumping, or
primary only.
(c) Does the selected sequence of operation for the
sources affect their output capability? For example, for
three chillers in series on a system designed for 5°C flow
and 11°C return, any one chiller must be capable of providing its maximum 2°C drop, for inlet conditions
varying from 5°C to 9°C.
(d) Will the advantages of a series system outweigh the
additional pumping costs for units in series, particularly
chillers ?
For both series and parallel operation:
(a) The choice between modulating or stepped mode
of temperature control. Modulating control is prefereable
for system stability.
(b) Is temperature control from flow or return?
Control from the return is not acceptable except with
constant total flow through the units, and control from
the flow is not advisable with the stepped mode.
(c) Is there a suitable sequence control system available for the proposed arrangement? There may be difficulties for more than three units.
(d) What is the effect of superimposing an overall
control system over that of the individual units? Manu126
@d
”5
v,
tr
T: Boiler control detector. Isolating and regulating valves not
shown.
Case 1 (LPHW)
Load
t,,
@
=
90°C;t , ,
Case 2 (HPHW)
= 78°C
t,,
=
130°C;t,,
= 85°C
(”/.I
Return Boiler flow temp.
temp.
rd”c
t,rc
78
100
66.7 82
33.3 86
,t,
1
2
3
90
94
98
90
94
86
90
82
86
Return Boiler flow temp.
temp.
r,rc
t,rc
85
100
115
1
2
3
130 130 130
145 145 100
160 115 115
= design flow temperature
t , = design return temperature
Fig. 29
Constant volume system
It should be emphasised that the analysis of the conditions at the heating or cooling source cannot be treated
in isolation from the remainder of the hydraulic circuit.
The converse is also true.
The example shown in Fig. 29 illustrates some of the
problems listed. They are relevant to a boiler sequence
control system with continuous constant flow through all
the boilers for both low pressure hot water (LPHW) and
high pressure hot water (HPHW) systems. This clearly
shows the magnitude of the over-temperature problem
with sequence control under partial load conditions.
While the over-temperature may be acceptable for the
LPHW system, other than for roof top boiler houses, the
HPHW system reaches over-temperature conditions
which may be outside the limits of the designed system
pressure. Both systems illustrate that any boiler-mounted
thermostats must be set much higher than normal,
impossibly so in some cases, if the sequence control is to
operate correctly and efficiently. The tabulated temperatures are simply calculated from the load percentage
with a fixed flow temperature. There is also the necessary
consideration of the safety requirements and devices
under the Heatlh and Safety at Work act.
In the equivalent chiller system, similar calculations
may be carried out for the subcooling case. There is
however, little latitude for subcooling before the operating chiller is locked out by its low temperature cutout
thermostat. This must always be a primary consideration
for selecting a suitable multiple chiller configuration.
Sources in parallel are more common than those in
series and this is particularly so for boiler systems. One
I E E PROCEEDINGS, Vol. 135, Pt. B, N o . 3, M A Y I988
example of such a system is shown in Fig. 30, with the
interlocking shown in Fig. 31. The features, details and
requirements for such a system are tabulated in Table 2.
mixed flow
llmlt
detector
YT2
I
I
1
I
1
I
'
constant volume
primary pump(s)
<
3
"'
i solat Ing
control
tf
1
I
'd
-
tf
F t a n t or
vnrinhle
-
volume
secondary
pump(s)
control
detector
Basic schematic diagram for Schemes I and 2
Fig. 30
Any two sources selected for operation with the other as automatic standby
The scheme has the following parameters:
(a) sources in parallel
(b) constant volume through selected sources
(c) reciprocating chillers or high/low boilers
(d) control from the return
(e) automatic changeover to standby source.
Note that where there is variable flow through the
sources, control from the return is not a viable option,
and that if the sources can be fully modulated from low
to high, control from the flow would be the best option,
irrespective of source flow rate.
6.2 Dewpoint control
The control of air systems may be considered in two
parts, the first related to the main air plant and the
second to the terminal units. The basic controls for the
main plants are generally the same, regardless of system
type. However, some systems may require additional controls. The dewpoint control system with its various
options and special features, represents a large proportion of the problems encountered in the control of all air
systems. The psychrometrics for air conditioning and air
systems generally are covered in detail in Section B3 of
the Chartered Institution of Building Services Engineers'
(CIBSE) Guide and are an essential part of understanding the operation of the control system. Only the basic
psychrometrics necessary for particular elements of the
schemes described below are shown here.
In energy conservation terms, overall dewpoint
systems may not be the most efficient. Sequential heating
and cooling with separate humidification and a cooling
coil overrride feature for dehumidification, or dewpoint
control only for the fresh air supply, should be considered as alternatives.
6.3 Dewpoint control and auxiliary features
In the system shown in Fig. 32, the dewpoint is controlled by detector T, which sequentially modulates a
pre-heater battery control valve V,, dampers D l a , D l b
and D l c , which operate in parallel, and a sprayed cooler
battery valve V, to maintain a constant saturated temperature condition. A low-limit detector 7
' is sometimes
employed in the discharge duct to override T, and maintain the discharge temperature above a predetermined
limit. Floating, thermal feedback, or proportional plus
integral controllers should be used, because, for stability,
proportional control alone requires a wide band which
IEE PROCEEDINGS, Vol. 135, Pt. E, N o . 3, M A Y 1988
results in unacceptable deviation from dewpoint. The
reheater, which is part of many systems, is controlled by
the extract temperature detector T, , modulating the
control valve V, to maintain a constant space temperature.
The psychrometrics of the system are shown in Fig. 33,
in which the dewpoint plant is set to provide a moisture
content of H so that air is supplied to the space after
reheat in a condition S . This permits the design room
condition R to be achieved when the latent gains ( R X )
are added and the sensible losses (S,X), or gains (SsX)
are taken into account. In winter when the incoming air
is at O w ,the dampers can be modulated to achieve a
mixed air condition corresponding to condition M , with
the cooler and preheater both off (any point on the line
O,R, represents a mixed air condition with 0, defining
full fresh air and R represents total air recirculation).
Normally, there is a requirement that the fresh air quantity shall not fall below a minimum quantity, represented
by MF, which means that the dampers cannot take up a
position between M F and R . From M the air is then adiabatically saturated as it passes through the spray coil,
moving the condition point along the wet bulb line to H .
An ideal spray coil would saturate to point D, where the
wet bulb and dry bulb settings are identical, but in practice, the spray is not perfect, and the dry bulb detector
(TI) is set at F to ensure that H corresponds to the
required moisture content. When selecting the dewpoint,
allowances should be made for the inefficiency of the
spray coil and the latent gains to the space ( R X ) .
In a very cold situation, such as shown at Ow,,
the
room condition R at start up may be below the line
DM, and the mixing point M , cannot then be achieved.
Alternatively, the mixing condition may not be possible
because the minimum fresh air position MF, is below the
line D M . In either case, the dampers are modulated to
M , , (points on line M,,R not being permitted), and the
preheater control valve is then modulated from closed
toward the open position to reach the wet-bulb line at
M , , where adiabatic saturation again takes place to
move the condition point to H . In any situation where
point M (or M , , etc.) can be achieved by the use of mixed
fresh and recirculating air, with or without the preheat,
and with no output from the cooling coil, the term 'free
cooling' is used, as the dewpoint can then be maintained
without the assistance of mechanical cooling. Thus, in
very cold weather the dampers are in the minimum fresh
air position to minimise preheating and, as the external
temperature rises, the preheater valve closes and the
dampers modulate to the full fresh air condition.
In summer conditions such as O s , the preheater is off,
the dampers are in the full fresh air position, and the
cooler valve V, is modulated to achieve point H .
Although each of the previous steps have been described
as discrete operations, the three stage controller permits
continuous sequential operation of each stage according
to load conditions. When point H is achieved it is necessary to reheat the air to the desired discharge condition,
point S , or Ss for winter and summer conditions, respectively.
A reduction in mechanical cooling load can be provided by overriding the damper control in certain conditions. Fig. 34 is similar to Fig. 33, but shows specific
enthalpy (total heat) lines in addition to the wet bulb line.
For all practical pruposes, either wet bulb or specific
enthalpy can be used to measure the total heat content of
the air, and for simplification one can assume that H
coincides with D. Normally, the dampers are fully open
127
h'
l i m i t detector
in mixed flow
2,
start
signal
fleeting contacts
inst break on
energisation and
delay remake
safety
interlocks
auxiliary switch
makes when
valve I S open
_ _ _ _ _ _
detector i n
+common
return
step switches
Fig. 31
128
thermostatic controls
Interlocking wiring for system of Fig. 30
IEE PROCEEDINGS, Vol. 135, Pt. B, N o . 3, M A Y I988
to fresh air before the coder battery is required, but one
must consider the three summer conditions, Os,, Os, and
O s , . For any condition Osl between lines OR and DZ,
the minimum amount of cooling (smallest reduction in
total heat) is required when the dampers are in the full
fresh air position MI, i.e. DO, is a minimum. Any condition Os, which falls on the line 0 , R requires the same
amount of cooling DO,, irrespective of the damper
position. The cooling load for a condition Os,, above
line 0 , R can be reduced to a minimum DO, by
moving the dampers to the minimum permissible fresh
air position M , . In the past, a fixed minimum fresh air
quantity was often employed. However, energy conservation considerations now render this approach unsuitable
in most circumstances, and the system shown in Fig. 32
incorporates damper modulation to achieve the desired
conditions. In the example above, Os, represents the case
where the enthalpy of the outside air exceeds that of the
return air, and it is more economic to cool return air
than outside air. A detection device is therefore required
Table 2: System details and requirements, (Figs. 30 and 31)
Element
Reciprocating chillers
Complete units
High/low boilers
Interlaced units
System use
Constant volume through sources for all load conditions. Secondary systems may be constant or variable volume.
Two sources operating to satisfy the design load, with one as automatic standby - its isolating control valve closed.
Use where sub-cooling with only
Use where step functions in flow
Use where sub-cooling from one
At
one complete unit operating is
complete unit is not acceptable or
temperature = l,
and/or overacceptable.
when each source is composed
6
of multiple compressors.
temperature is acceptable.
The system may be extended to a greater number of steps and/or units than the example if the sub-cooling or
over-temperature is acceptable and suitable step controllers are available.
Control
wiring
As Fig. 31 but:
As Fig. 31 generally but connections
need alteration between terminal
groups 1 to 24 and 25 to 36,and:
reverse the action of the thermostats, controller etc., to suit the cooling function.
Diagram assumes that all steps are brought in by making a circuit, but there may
be instances where the circuit needs breaking to bring in steps. Check.
The single wafer sequence selector switch could be replaced by a multiple wafer switch and fewer
External
interlocks
Other
interlocks
Proof of primary pumping and fuel feed
interlocks.
On power failure and for any other form of interruption of the start signal line, the sources restart unloaded (R16).The
'safety interlock' may be a 'fireman's switch' or 'boiler house door cut-off switch' etc.
If a unit is locally isolated, restart is sequenced from sequence starting relays.
If it is necessary to prevent low
temperature return water at start up, add
necessary equipment and modify the
diagram.
The closure of the source isolating control valves should take not less than 120 s, and the actuators should be selected
and/or the wiring modified to achieve this. The valves should be of the butterfly type. The isolating control valve on the
leading source must always be held oDen.
relays.
Proof of primary pumping and proof of condenser operation
Requirements Contain high and low pressure cut-outs and low temperature limit device, all
of source
with hand reset and alarm features and integral sequence controls.
packages
Package to include switch to permit the use of the local unit package sequence
controls, or the overall sequence system.
Control
settings
For the two stages on each boiler Fig. 31
would be modified by deleting: Switch
and step controller elements 3,4,7, 8.
Relays R6, R7, R10, R1 1 , R14, R15.
Sequential relays SR1,SR2, SR3.
Contain limit, high/low and low/off
thermostats, all suitable for overtemperature setting. Package to include
switch to permit local control through
the three thermostats, or the overall
sequence control. Limit thermostat to be
oDerative for both systems.
System assumes all selected units are required for full load operation. Normally proportional band (PB) setting equals
design temperature fall (rise) across sources (ArJ but this may not be the same as the design temperature drop across the
load (At,),if the source volume (V,) exceeds the design load volume (V,) and/or the source capacity (C,)exceeds the design
load capacity ( C , ) .
The basic relationship is Ats = At, x
5 5. The Ats actually used when setting the PB, should be
x
U s
. This permits a dead zone over which the step controller continues to operate
Total no. of source switching v=
steps
after all the selected sources are allegedly operating at full capacity.
At the end of the dead zone the standby sequence is initiated. The setting of the desired value (DV) of T,should be
,t - ;At, (boilers) and ,f + ;Ats (chillers).
The differential settings on the step switches should be set so that the return temperature variation between the on and
(-)
4
off action is f x
(assuming equal sized steps).
No. of steps
The limit thermostat (T,)in the mixed flow should be set at ,t
- 2°C for cooling, and at ,t
+ 5°C for heating
Standby
facilities
Standby is automatically accomplished if the selected sources do not apparently provide sufficient capacity.
This is detected by an out of limit return water temperature - too high for cooling and too low for heating, which initiates
the changeover. The changeover moves the sequence to the next selection on the switch and closes and opens the necessary
source isolating valves. If the sequence still contains a faulty source the sequence will move on again to the next sequence
and repeat the valve changeover.
Number of
starts
Specify timers in packages or motor control centres to limit the number of starts in accordance with the requirements of
the electrical eaubment and the source unit sumliers.
Alternatives
Modular boilers may also be controlled
using the same basic concept. Sufficient
steps would be required for the number
of boilers/stages and the sequence
selection would require modification.
The over-temperature conditions could
be extreme.
If control from the flow is considered, the same general interlocking may be used as shown in Fig. 31, but a proportional
Controller is NOT suitable. Use a two-term proportional plus integral, thermal feedback, or floating controller with suitable
sensitivity. There will also be problems of hunting between steps, which will require very careful setting of the differential
stem on the steD controller and the addition of timino devices in the control svstem to limit the rate of resoonse.
IEE PROCEEDINGS, Vol. 135, P t . B, N o . 3, M A Y 1988
129
to measure the total heat (enthalpy) or wet bulb temperature. The enthalpy of the outside air can be measured
created by the plant fans can vary the surface level of the
water system.
U
T4b
Fig. 32
Dewpoint schematic
Dewpoint detector
Temperaturedetector
Low limit detector
Enthalpy detectors/comparators
T5 Boost limit thermostat
T,
T,
T,
T,
T, Secondary water control detector
T, Modulating low limit
T8 Frost protection thermostat
T9 Free cooling limit thermostat
TI, Free cooling enthalpy detector
directly T4a as being above the design room value, or by
dual detectors T4a and T4b which compare the room
and outside air total heat conditions. In either case, when
the room total heat is exceeded by the outside air, a
signal from the device drives the dampers to the
minimum fresh air position ( M 3 in Fig. 34). A wet bulb
detector may be used as T4a, provided that the wetsock
assembly remains wet. If not, the dampers will motor to
the minimum fresh air position for all conditions to the
right of D-N (see Fig. 34). The use of such detectors
therefore requires permanent water supplies to the
wetsock assemblies, and an appreciation that pressures
HI Return air humidity detector
PI Differential pressure switch
VI Preheater valve
V, Cooler valve
V, Reheater valve
V, Secondary chilled water valve
Free cooling may also be utilised to satisfy the secondary cooling requirements of terminals such as fan coil
units, that operate with water chilled to only about 12°C
(at temperatures below this, water condenses on the fan
coil units). In such circumstances, it is common to feed
the terminal circuit from the return from the main
cooling coil (points A and B in Fig. 32). It is then necessary, after the early morning boost heating (described
later and terminated by thermostat T,) to maximise the
use of fresh air to achieve cooling. In winter it is possible
to use this cool air to provide water at a suitable temperature without the use of chillers, by running the
I
'
I
-owl
E E
dry bulb temperature ,"C
Fig. 33
Psychrometric process for a dewpoint plant
0, Winter outside condition
0, Summer outside condition
S Supply condition
R Room condition
D Adiabatic saturation
H Design moisture content
M Mixed air condition
130
Fig. 34
Specific summertime psychrometric dewpoint processes
IEE PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y I988
primary chilled water pumps and spray pumps without
energising the chillers or the condenser pumps, and with
the valve on the cooling1 coil driven to the fully open
position. The amount of secondary cooling thus obtained
may vary according to the control regime adopted.
If the normal dewpoint temperature is maintained at
T, the amount of cooling obtained will be less than if the
control point of T, is lawered, or dewpoint is uncontrolled. In the latter cases the relative humidity in the
controlled space will drop, but this is often acceptable.
The case shown in Fig. 32 is for an uncontrolled dewpoint where the dampers are modulated by signals from
detector T6 with valve V, fully open (in normal operation
T6 would modulate V, to maintain the necessary 12°C).It
is necessary under these conditions to protect the preheat
and cooling coils from too low a temperature, and detector T, acts as a modulating low limit override device for
this purpose set at, say, 4°C. The absolute protection
offered by frost thermostat T8 remains in operation.
While the external dry bulb temperature remains below a
specified level the system remains in operation. Above
this limit ( 5 to 8"C),thermostat T' switches the plant controls back to normal operation. If the plant is used to
obtain secondary chilled water without mechanical
cooling the cooling coil should be sized for this purpose
so that the necessary secondary cooling capacity can be
obtained for a particular setting of T9. Detector T, can
also be used to determine when the plant should revert to
normal operation. Alternatively, an additional detector
adjacent to it may be utilised with suitable timing delays
and interlocks. Where no secondary cooling capacity is
T / S signal
required, it is possible to almost satisfy the dewpoint condition without mechanical cooling whenever the enthalpy
of the external air is below point D on Fig. 33. This is
accomplished by detector TI, which holds off the chiller
and cooling tower plants whenever the enthalpy is below
the predetermined level.
In rare cases, a humidity detector HI, mounted in the
extract duct from the conditioned space, is provided to
monitor any excessive or reduced latent gains in the
space, and resets the dewpoint detector TI accordingly.
When the plant is started up in the morning, or after a
shutdown, and the space temperature is below the setting
of the boost thermostat in the extract duct T5 a boost
heating regime is initiated. The dampers remain in the no
fresh air position (to which they run when the plant is
shut down), the chiller and spray plants remain off, the
boiler system is started, the fans are run when the water
is up to temperature, and the preheater and reheater run
fully open until T, is satisfied. When this occurs the plant
reverts to the normal operating sequence, and T5 is
latched out for the day.
A differential pressure detector P , fitted across the
plant filter is a standard control item for providing a
warning of high pressure across the filter for both maintenance and energy conservation purposes. There are
many variations to the interlocking which may be necessary, but two general features are desirable for all dewpoint plants. These are:
(a) The minimum fresh air position of the dampers is
important for ventilation and energy conservation. TO
permit ease of setting up, commissioning and adjustment,
N1
Fig. 35
EA
OA
RA
T/S
PH
RH
CIA sensor R A sensor
IEE PROCEEDINGS, Vol. 135. P t . B, No. 3, M A Y 1988
Dewpoint controls
Exhaust air
Outside air
Recirculated air
Time switch
Preheater
Reheater
Relay
-
131
the controls should include an integral minimum fresh air
position adjustment. This allows the minimum fresh air
quantity to be altered without changing the mechanical
linkages to the dampers, and also provides a simple
method of overriding the partly open fresh air and
exhaust air dampers for periods of plant shutdown and
early morning boost heating.
(b) Fire protection considerations often require that,
in the event of smoke or fire, the plant shuts down automatically and that the firemen have the facility to start
boost
A
i
I
firelsmoke
C
minimurn F A
D
time
swltch
E
F
9
I
-
I
I
PH
I
1I
I
I
FA
NIC
I
I_
I
I
I
EA
RA
N/O
NIC
dampers
n o r m a l sequence PH dampers, cooling
10
.I-+
I
---
CLG
I
--u
I
N/C&
I
I
'
I
v
I
PH
control Ler
to suit
!I
minimum F A
position
ad) ust ment
dewpoint
controller
to E P R , - ~ : ~ ~ ~ - - +
(proportional
p l u s reset)
discharge
stat
I
/
dewpoint
stat
'I
Dewpoint controls: pneumatic continuation
All necessary penuematic pipework within boundary by controls suppler
EPR Electropneumatic relays by control supplier:
5 RA open, night
1 MinimumFA
6 RA closed, smoke
2 FA closed, night and smoke
7 Cooling valve open, free cooling
3 FA closed, night
4 FA closed. smoke
132
main p l a n t
chiller E
L&
I
Fg. 36
the extract fan independently of the supply fan, with the
dampers operated out of sequence. A frequent requirement is for the exhaust damper to be open and the fresh
air and recirculation dampers to be fully closed to facilitate smoke exhaust.
As an indication of the interlocking complexity which
may occur, Figs. 35 to 37 identify many of the items
described in this section for both electronic and pneumatic control systems. The principles identified are frequently applied to other systems.
8 Dewpoint DV setdown, free cooling
9 PH open, night and boost
10 PH open, night and boost
IEE PROCEEDINGS, Vol. 135, Pt. B, No. 3, M A Y 1988
It can be seen that a knowledge of the functional
requirements of the plant for a host of extraneous circumstances is crucial if the electrical logic is to be cornprehensive.
-
minimum FA
fire/
smoke
boost
C
I
0
A
I
mainplant
chiller
D
E
I
I
I
I
F
I
I
II
I
I
time
switch
I
I
I
I
I
matic Controls and their Implications for Systems
Design’, and Butterworth & Co. (Publishers) Ltd. for permission to use extracts from ‘Electrical Engineers Reference Book’ (Chapter 28, Environmental Control). The
I
I
I
I
I
l - O ’ $ ~ ~ ~ l l e dRHvalvel
~
1
dampers(
I
I
1
I
normal
I
normal
I
flclos:.*i
I
ilve
DV
1 - O
-0
I
I
I
I
dewpoint controller
( t w o l t h r e e t e r m or floating
thermal /feedback)
I
,FA
I
I
‘.A
RA,
I
I
PH d a m p e r s , cooling
I
adjustment
damDers
I
I
also a l l r e l a y s f o r
%
v1
normal s e q u e n c e
reheater controller
t o choice
I
I
D.
I
i
I
,I
R A detector
@ R H valve
3
The use of the two examples for illustrating, in the one
case, controls for water systems and, in the other, air
systems is a suitable point to conclude this review of controls in building services. More detailed information is
avilable from manufacturers (for equipment details) and
from other sources [2] for design information.
7
I
detector
Acknowledgments
The authors wish to thank the Chartered Institution of
Building Services Enginedrs for permission to use extracts
and Figures from the CIBSE Applications manual ‘AutoIEE PROCEEDINGS, Vol. 135, P t . B, N o . 3, M A Y 1988
I
Fig. 37
I
Wiring within boundary according to particular supplier’sdetails
Dewpoint controls: electric continuation
historical section reuses some quotations from ‘Building
Services Engineering: A Review of its Development’ by
N.S. Billington and B.M. Roberts.
8
References
1 LETHERMAN, K.M.: ‘Automatic controls for heating and air conditioning’ (Pergamon Press, Oxford, 198 1)
2 CIBSE applications manual: ‘Automatic controls and their h p l i c a tions for systems design’ (Chartered Institution of Building Services
Engineers, London, 1985
3 FISK, D.J.J.: ‘Micro-electronics in building services’ (Building
Research Establishment, Garston, 1979)
133