The present increase in the world energy demand and depletion of fuel sources
is partly contributed by energy consumption from the built environment. The
energy consumption of buildings in developed countries comprises 20-40% of the
total energy use in the EU and USA (Perez-Lombard et
al., 2008). In Malaysia, it is reported by the Energy Information Administration
(EIA, 2010) that the country is experiencing rapid increase
in energy consumption in the last decade, mainly due to its high economic growth
and increase in the standard living of household. The Energy Information Administration
also reported that the countrys consumption of natural gas, for example,
has increased by 98% over the last decade. Not only energy is becoming more
costly but the situation is worsened by the global warming resulted from the
green house gas emission. Therefore, a more efficient energy usage such as through
reduction in the wastage of cooling energy is required.
In tropical countries with hot and humid weather pattern like Malaysia, space
cooling is required all year round in commercial buildings and some residential
buildings. Air conditioners account for 42% of total electricity energy consumption
for commercial buildings and 30% of residential buildings in Malaysia (Saidur
et al., 2009). These are relatively lower than, for instance, in
Hong Kong where air-conditioners contribute to about 60% of the total electricity
consumption for commercial buildings (Chan, 1994) during
the cooling season from late March to early November. Energy cost is likely
to increase in the future due to the anticipated decrease in the national oil
wells production rate in the next decade. Hence, reduction in the energy
consumption for cooling of built environment is a vital step to energy conservation
in Malaysia. A study on the factors affecting energy performance in buildings
is necessary for a better understanding of the design and operational strategies
related to energy conservation; these are possible with the use of energy simulation
Although development of non-residential buildings in Malaysia has increased
in the last two decades, the country is lack of legislation pertaining to building
efficiency (Aun, 2004). In many situations, little attention
has been given to the aspects of operation and maintenance of the building.
On the other hand, the architectural designs of certain buildings seems to portray
more on the aesthetics values but at the same time fail to consider the hot
and humid climate of the country. Consequently, various studies were conducted
with hopes to overcome problems related to thermal comfort and high cooling
energy in these buildings (Sulaiman and Hassan, 2011;
Bhaskoro and Gilani, 2011; Sharma
et al., 2011; Wang et al., 2008).
The extensive use of glazing for buildings perimeter walls which results
in higher cooling load due to sun radiation, is an example of poor consideration
in energy conservation since sun light is available 12 h per day all year round
in Malaysia. During the design stage, consideration for highly glazed buildings
should be studied very carefully since they have a large impact on the energy
efficiency as compared to 30 to 60% window to wall area alternatives (Poirazis
et al., 2008). A study on a typical office building in Singapore
(Hien et al., 2005) revealed that the use of
double glazing is able to maintain lower internal surfaces temperature, thus
reducing solar heat transfer to the internal spaces.
When a highly glazed building has already been constructed, little can be done
on the design in order to conserve energy. Nevertheless, there are always opportunities
for energy conservation in the operation and maintenance of the buildings
air-conditioning system. Improving operation and maintenance of energy systems
may have a major impact on energy consumption and in many cases will exceed
the potential energy savings from alternative energy and energy conservation
technologies (Drost, 1992). Overcooling, for instance,
was investigated by Lam (2000), who reported that there
could be a 3% reduction in total building electricity use for every °C rise
in the indoor design temperature. In an earlier study by Sekhar
(1995) it was revealed that the space temperatures in air-conditioned office
buildings in Singapore were around 23°C (relative humidity of 70-75%) which
were not within the recommended ASHRAE comfort zone and thus increasing the
space temperature would be able to reduce the cooling energy significantly.
Building owners can also conserve energy by installing Automatic Fault Detection
and Diagnosis (FDD) which was reported as a potentially practical method (Lee
and Yik, 2010) to detect occurrence of faults in the equipment and measuring
instruments of an air-conditioning system, although this may depend on the severity
of the energy impacts of the faults.
In this study, the effects of a few operational behaviors of centralized air-conditioning systems for glazed buildings in a university are studied by energy simulation using EnergyPlus software. A few gaps in the aspect of air-conditioning operation are identified and potential solutions are simulated using the software. One of the main gaps is the supply of cool air to rooms during the period of non-occupancy throughout the weekdays and weekends. The potential savings in energy as a result of various proposed strategies are estimated based on the computer simulation.
DESCRIPTION OF THE BUILDING
The buildings in the present study are located within an academic complex in a university campus. There are 16 buildings within the academic complex which are constructed next to each other, as illustrated in Fig. 1. The buildings which have nearly the same size and shape, are labelled as Blocks 1 to 5 and Blocks 13 to 23. Blocks 6 to 12 would be built under a future development scheme. Each of the buildings has four floors with the exception of Blocks 5 and 15 which have three floors only. The buildings aspect ratio is 3.1. The major functions of the buildings are offices, classrooms, laboratories and computer rooms. The total air-conditioned floor areas for each building are approximately 4833 m2. The building can be accessed 24 h a day by the staff but the air-conditioning is only supplied 12 h per day, typically between 7 am and 7 pm from Mondays to Fridays. On Saturdays, a few areas are air-conditioned based on requests by the occupants; this is normally the case for selected laboratories and lecturers offices.
The typical layout for the third floor of the buildings which is the most occupied floor, is shown in Fig. 2. Each of the floors comprises of two air-conditioned areas or wings which are identified as the AB side and the CD side, as depicted in Fig. 2. Each side comprises individual lecturers offices which are air-conditioned. The area in between the two sides is not air-conditioned. An Air Handling Unit (AHU) which is located in a plant room to the right of the layout (not shown in the figure), serve each side or wing. For Levels G, 1 and 2, there are roof overhangs which act as external shadings. The main control of the air conditioning system is performed in the control room which is located in the administration complex not far away from the academic complex.
The cool air from each AHU is supplied through the Variable Air Volume (VAV)
system. The installed VAV system can operate under two modes as defined by the
manufacturer; i.e., pressure dependent and pressure independent. In the pressure
dependent mode which is the default setting for the building, the VAV damper
operation is set to correspond to a minimum (non-zero) flow rate that is specified
by the manufacturer in order to assure the air motion required for the comfort
of occupants. Nevertheless, the pressure dependent mode usually leads to over
cooling since cool air is continuously supplied although at a small flow rate.
As for the pressure independent mode, the VAV damper operation corresponds to
the local temperature set point.
|| Aerial view of the academic complex
|| Typical floor layout for level 3
|| Typical envelop details of the academic building (Block 17)
In addition, with the pressure independent mode, users may manually adjust
the minimum damper opening according to their requirements, even to fully shut
the damper. In contrast to pressure dependent mode, the pressure independent
mode is capable of preventing overcooling although the consequent would be loss
in air motion. Nevertheless, air motion is not required when the room is unoccupied.
Table 1 summarizes the construction details of the building.
The external walls of the buildings, including the doors, are nearly fully glazed
with aluminium frames. Most of the walls are constructed by a single layer of
tinted glass with an overall U-value of 1.43 W m-2 K. Selected laboratories
that require noise attenuation are constructed with double glass of which the
overall U-value is 0.72 W m-2 K. External walls that are not glazed
are composed of a layer of Medium Density Fibre (MDF) board which is sandwiched
by Corian® solid panels (Dupont, 2010). The total
thickness of the composite wall is 200 mm and its overall U-value is 0.72 W
m-2 K. The indoor walls are also made of glasses of the same heat
transfer properties. The floor slab is 300 mm thick concrete. The ceiling finish
is 13 mm gypsum board.
SIMULATION OF COOLING LOAD OF AN ACADEMIC BUILDING
The present study involved simulation of cooling energy in the buildings based on a few operational conditions or patterns using computer software.
Cooling loads calculations: Hourly and annually cooling loads were calculated
by EnergyPlus version 3.0 (EnergyPlus, 2010), a building
energy simulation software developed by the US Department of Energy, based on
its previous energy analysis software; i.e., BLAST (Building Loads Analysis
and Systems Thermodynamics) and DOE-2. The EnergyPlus software was developed
to model thermal loads, lighting, ventilating and other energy related systems.
The simulation program operates based on a heat balance method which allows
for simultaneous calculation of radiant and convective heat transfer effect
at both interior and exterior surfaces during each time step. The cooling load
calculation method takes into account all the heat balances on outdoor and indoor
surfaces and transient heat conduction through building construction (Eskin
and Turkmen, 2008).
Outdoor and indoor design data: The design data was based on that for Kuala Lumpur (3°8 N, 101°42 E) in Malaysia which was available in the EnergyPlus database. The location of study, in Tronoh (4°25N, 100°59 E) is situated less than 200 km to the north of Kuala Lumpur and has nearly the same ground elevation (62 meter average) and climatic conditions. Malaysia is a small tropical country with a warm and humid climate all year round. The outdoor design conditions correspond to dry bulb and wet bulb temperatures of 32.5°C and 26.9°C, respectively. The daily temperature range of Dry Bulb (DB) temperature is 8.2°C.
The indoor dry bulb temperature was set by the building operator to be 24°C
with a relative humidity of 50% and these were within the range of values commonly
practiced in Malaysia (Leong, 2009). The indoor air temperature
was controlled by thermostats.
Existing modes of operation: The buildings within the academic complex are mainly used as lecture rooms, laboratories and lecturers offices. The official working hours is between 8.00 am and 5.00 pm and AHUs are operated approximately between 7.30 am and 5.30 pm. In some cases, the AHUs are turned off late at 7.30 pm, subject to requests by the occupants. The offices which are located on the highest floor, are generally fully occupied during the office hours. Nevertheless, a number of lecture rooms and laboratories are not always occupied for various reasons which are mainly related to the operation management of the buildings. In the present study, the pattern of rooms occupation is observed.
Figure 3 shows a typical weekly occupancy pattern for Blocks 15 to 19. The dark and light bars represent the occupied and unoccupied h, respectively. Most of the lecture sessions in the lecture rooms are scheduled to be either continuous or intermittently. The longest gap between lectures is 5 h while the shortest gap is one h. The average occupancy of lecture rooms is 25 h per week per room, throughout the 14-week semester. Such patterns of occupancy are experienced due to the nature of the study programs. As for the laboratories, the occupancy patterns are not constant throughout the semester. On average, the laboratories are unoccupied for only 26 h per week. In addition, it is observed that some laboratories are occupied only for seven weeks out of the total of 14 weeks.
During the weekends, there are also requests for usage of air-conditioning
system due to certain laboratory activities or requests by staff. An observation
is made for three weeks on the trend of operation during the weekends and the
results are presented in Table 2. It is shown that in some
places, e.g., in Block 17 on Level 1, that the air-conditioning system is operated
although the area is unoccupied. It is also observed that even if the request
is made for one side of the floor, both AHUs (AB and CD sides) are turned
|| Typical rooms occupancy pattern for Blocks 15 to 19
during the semester
|| Typical occupancy pattern during the weekends in various
buildings for three consecutive weeks
From an interview conducted with the control room personnel, such occasions
are said to occur due to insufficient information provided by the requesters;
for instance the duration of the room usage is not clearly informed. It is realised
in the present study that there is no proper mechanism to request for air-conditioning
service after regular office hours.
The trends of occupancy during the weekdays and weekends clearly show that there is unnecessary wastage of energy. For an example, at Block 17 at Level 1 on the CD-side, the AHU operates from 7.40 am until 11.30 pm even though the area is observed as unoccupied. Furthermore, the observed trends only represent the situation when the undergraduate students are on campus. Larger wastage is anticipated during the semester breaks; i.e., a total of about 16 weeks per year. In the present work, EnergyPlus is used to estimate the savings in the cooling energy if supply of cool air to these rooms is temporarily cut off. In order to reduce the wastage a suitable control strategy is required.
Experimental scheduled isolation of cooled air supply: Scheduling of
air-conditioning based on the daily and weekly occupancy schedule was proposed
in previous researches; for example by Al-Rabghi and Akyurt
(2004). In the present study, the potential saving resulted from scheduling
of cool air supply is studied. By obtaining the occupancy plan in advance from
the building administrator, the control operator may be able to schedule the
supply of cool air and therefore, rooms which are unoccupied can be totally
cut off from cool air supply by mean of VAV control. In this case the VAV is
reconfigured from pressure dependent to pressure independent mode to enable
100% damper shut at the VAV box. The original configuration of the VAV control
using the pressure dependent mode is observed to be unsuitable since it causes
unoccupied rooms to be continuously supplied with cool air although at low flow
With the pressure independent mode, the VAV box will maintain the minimum flow setting when the set point temperature is met. If the minimum flow is set to be at zero percent, after reaching the set point temperature, the cool air supply through the VAV box will be cut off. Therefore each and every unit of the VAV box can be controlled independently by varying its temperature set point. This is particularly beneficial during low occupancy periods such as in the weekends. If such change is applied to the whole academic complex, the issue of space overcooling which has been a problem to the occupants in the present, can be minimized.
The savings resulted by cutting off the supply of air to unoccupied rooms (pressure independent mode) is simulated using EnergyPlus. The scheduled isolation is experimented for Block 17 at Level 1. From the building administrators schedule, the average daily room usage is found to be only 6 h. Only three rooms are occupied at all time while the rest are unoccupied. Therefore, in the present study, the EnergyPlus simulation is performed with the consideration that cool air is supplied to the affected rooms from the start of AHU operation at 7.42 am until 2.00 pm (approximately 6 h) and thereafter the cool air supply is totally cut off (isolated) except for rooms that are occupied. The savings in terms of the electrical power of the blower motor as a result of the scheduling is measured. The actual cooling energy is not measured since the appropriate measurement devices are not available.
RESULTS AND DISCUSSION
Effect of space overcooling: The set point temperature for the buildings
in this study is 24°C which is within the indoor design temperature recommended
by ASHRAE Standard 55 (McQuiston et al., 2004).
However, it is usual that the operator in the control room receive requests
from some of the occupants to reduce their room temperatures; mainly at the
time when the sun is facing the external walls of their rooms. Although the
required change in the temperature set point should be made by the control room
operator for only one room, it is usual that the operator reduce the set point
temperature for the whole floor, usually to 20°C. As a result, other occupants
who are readily feeling comfortable start to feel colder and uncomfortable.
Through interviews with a number of the occupants, it is discovered that most
of them tend to keep quiet about the situation and thus the problem prolongs.
This is also evident when a number of them start to wear additional clothing
such as coats or sweaters in offices. Similar problems were also reported in
Hong Kong (Lam, 2000), where many commercial premises
were being over-cooled by 2-4°C.
The effect of lowering the set point temperature in the academic buildings is simulated using EnergyPlus. Shown in Table 3 is the simulated annual cooling energy for a typical building (Block 23) as a result of using different set point temperatures; i.e., 20, 22 and 24°C. The differences in cooling energy are calculated relative to the energy required at a set point temperature of 24°C. It is shown that a reduction in the air temperature from 24 to 22°C and 20°C will cause increases in the cooling energy by 36 and 68%, respectively. This shows that overcooling of a space in the highly glazed academic building involves huge amount of additional energy and thus should be avoided in the first place. Overcooling should also be avoided in order to provide a room temperature that is within the level of comfort of most occupants.
Cooling energy consumption during weekends: Operation of the air-conditioning
system during the weekends cannot be avoided, as it is the decision of the building
administrator but it is found in the study that there is a room for improvement
of the energy consumption. Table 4 shows the annual cooling
energy resulted from consumption during the weekends for various locations based
on the occupancy observed throughout the study. The cooling energy is simulated
using EnergyPlus. The weekends total cooling energy for the whole academic
complex is found to be 31,936 RTh per year. While the requests for air-conditioning
service during the weekend are made by authorised staff, it is found that postgraduate
students, who are not authorised to work in the building during the weekend,
are also supplied with cool air from the air-conditioning services. There are
also buildings that are air-conditioned but are found to be unoccupied despite
frequent inspections. The unoccupied buildings and those occupied by postgraduate
students (without authorisation) contribute to cooling energy of 24,012 RTh
per year which accounts for 75% of the weekends total cooling energy.
|| Annual cooling energy for different set point temperatures
in a typical building (Block 23)
||Simulated annual cooling energy due to consumption during
the weekends for various locations
||Simulated annual cooling load for Block 17 Level 1 with isolation
(black bars) and without isolation (white bars) of the unoccupied rooms
Such unnecessary waste of energy can be avoided in the future by imposing stricter
rules for request of air-conditioning services during the weekends.
Scheduled isolation of cool air supply: Figure 4 shows the annual cooling load estimated by EnergyPlus simulation, with and without scheduling of cool air supply for Block 17 Level 1, as represented by the dark and light bars, respectively. For simplicity, the simulation is made with the assumption that the weekly schedule is the same throughout a year. It is shown that the scheduling is capable in reducing 33% of the annual cooling energy which is equivalent to an annual saving of about 13,089 RTh. Such a significant reduction is obtained due to the fact that out of the 16 air-conditioned rooms on Level 1, only three rooms require full-time cool air supply during lectures and laboratory activities, while the other rooms are supplied with cool air for only 6 h a day. The energy saving is anticipated to be higher if further reduction in the cooling energy during the semester break is accounted for.
In addition to saving in the cooling energy, there is also saving in term of electrical power to run the AHUs blower motor. When the motorized dampers of the VAVs are partially or fully shut, it causes an increase in air pressure in the main duct. As the pressure transducer detects the built-up pressure, the system will respond to counter the pressure increase in the main duct. The control system will send a signal to the Variable Speed Drive (VSD) control of the blower motor to reduce the pressure in the main duct by reducing the motor speed which consequently reduces the electrical power consumption.
Figure 5 and 6 show the variations in the
motor loads of the AHU blowers with and without isolation of cool air supply
to the unoccupied rooms for AB and CD sides, respectively, at Level 1 of Building
17. The results are presented for three consecutive weeks. As shown in the Fig.
5 and 6, the AHUs were turned off from midnight
until 5.12 am, after which the AHU blowers were turned on for about 45 min to
purge the indoor air. The blowers were turned on again at 7.42 am when the occupants
started to arrive at the buildings. The results for the original setting; i.e.,
without isolation of the cool air supply to the unoccupied rooms, are represented
by the thick line. The ones with scheduled isolation are represented by the
continuous thin lines (for Week 1) and dashed lines (for Week 2).
||Variation of the AHU motor load (in percentage) with time
for Level 1 of Block 17 (AB-Side) with and without isolation of cool air
supply to the unoccupied rooms
||Variation of the AHU motor load (in percentage) with time
for Level 1 of Block 17 (CD-Side) with and without isolation of cool air
supply to the unoccupied rooms
The motor load for the AB-side is shown to reduce by about 15% during the operation
hours, as a result of the isolation of cool air supply.
The trend was not so clear for the CD-side, as shown in Fig. 6. During the first week of the isolation, the motor load was observed to be higher for the first five hours probably due to increase in load resulted by increase in the number of occupants. Nevertheless in the second week, a consistent saving of about 17% was observed throughout the operation period of the CD-side AHU. The blower in the CD-side AHU was shut off at midnight as opposed to 7.00 pm for the AB-side AHU (due to request by the occupants).
Table 5 shows the total daily electrical energy consumptions
for the AHU blower motors presented in Fig. 5 and 6.
Comparisons are made with the results of Week 1 (i.e., without isolation)in
order to assess the energy savings. It is shown in Table 5
that the electricity consumption for the CD-side AHU is about 2.5 times higher
than that of the AB-side AHU. This is due to the greater occupancy in the building
and also because the CD-side receives more sun radiation throughout each day
due to the buildings orientation relative to the suns position.
|| Block 17 level 1 AHU fan electricity consumptions
In addition, the CD-side AHU operates on longer hours due to the use of a computer
lab until midnight. From the comparison, it is shown in Table
5 that the isolation of cool air supply results in an average of 25% electrical
energy saving of AHU motor for the AB-side and an average of 12% for the CD-side.
The difference is resulted by variation in the cooling load of the rooms in
the AB and CD sides.
It is clearly demonstrated that the combination of scheduled isolation of cool air supply and reconfiguration of the VAV system to pressure independent mode would result in savings in the cooling load energy and the electrical energy of the blower motor. By reconfiguring the VAV system to the pressure independent mode, the VAV dampers opening is now fully responsive to the room temperature disregard of the need for a minimum air flow, as explained earlier. A greater saving is expected if the buildings operations during the semester break are taken into account since the occupancy period of these areas is further reduced. The trend of improvement is expected for other floors and buildings as they have similar trends of occupancy.
Effect of outdoor air infiltration: Outdoor air infiltration caused
by cracks or by opened doors or windows is one of the contributors of space
cooling load through increase in both sensible and latent heat gain. During
the design of an air-conditioning system, the engineers will normally incorporate
infiltration load based on the level of air tightness of the building. Nevertheless,
additional infiltration load may be imposed during the system operation if doors
are purposely left open. Through inspections in the buildings of study, it was
found that doors are usually left open, by placing objects that prevent them
from shutting, for a long period of time to enable students and visitors to
move in and out of areas within the buildings. These doors were usually locked
for security reasons and therefore students and visitors would be required to
call the staff in the area to open the door for them. In the present study,
a simulation of cooling energy was performed for Block 23 with the assumption
that each floor has one door that was purposely left open during the day time
(8 am-5 pm). The infiltration rate was set in the simulation as one cubic foot
per minute per square foot of the door area (Pita, 2001).
It was estimated from the simulation that the annual cooling energy for Block
23 would increase from 685,160 kWh to 1,026,043 kWh (higher by 50%) as a result
of infiltration load from the door openings. Figure 7 shows
the result of the simulation in term of monthly cooling energy over the period
of 12 months. The dark color bars represent the results for air-tight building
and the light color bars represent those with the specified infiltration rate.
It is probable that the occupants would be able to accept the thermal condition
caused by infiltration, as implied in one report (Kwong
et al., 2009). However, with the significant increase in the cooling
energy in the present study, it is vital for the building operator to strictly
ensure that doors are not left open for a long time to minimize infiltration
||Variation of monthly cooling energy for Block 23 with an infiltration
rate of 1.0 cfm ft-2 of door area
Simulation of the cooling energy as a result of the operational behavior for highly glazed academic buildings was performed using EnergyPlus. From the study, the followings conclusions can be drawn:
||Space overcooling by reduction of the set point temperature
of the thermostat in the glazed academic building would result in significant
increase in the cooling energy by up to 70% and this would consequently
increase the operational cost. In addition, overcooling was observed to
cause discomfort among a large number of occupants
||Occupancy of buildings during the weekends was observed to contribute
unnecessary wastage in the cooling energy if cool air is supplied to the
unoccupied areas or buildings due to failure to check for the details in
the requests made by occupants. From the present study, the wastage accounted
for 75% of the weekends total cooling energy and could be avoided
by introduction of proper mechanism for request of air-conditioning services
outside office hours
||The scheduled isolation of cool air supply and reconfiguration of the
VAV system to pressure independent mode would result in significant savings
in the cooling load energy by about 35%. Such strategy may also reduce the
electrical energy of the blower motor by up to 25% through response of the
variable speed drive of the AHU blower motor
||Outdoor air infiltration through opened doors could lead to a significant
increase in the cooling energy of the buildings of study (by about 50%).
The building operator with the cooperation of the occupants should ensure
that doors are not constantly left open during regular buildings operation
The authors would like to acknowledge the support of Universiti Teknologi PETRONAS in the present study. The assistance of M Fatimie Irzaq in providing the related design information is highly appreciated.