Design of an ice thermal energy storage system for a building of hospitality operation

International Journal of Hospitality Management 46 (2015) 46–54

Contents lists available at ScienceDirect

International Journal of Hospitality Management journal homepage: www.elsevier.com/locate/ijhosman

Design of an ice thermal energy storage system for a building of hospitality operation Chung-Tai Wu a , Yao-Hsu Tsai b,∗ a b

Department of Tourism and Leisure Management, Lee-Ming Institute of Technology, Taiwan Department of Hospitality Management, Chung Hua University, Taiwan

a r t i c l e

i n f o

Keywords: Energy management Thermal energy storage (TES) Off-peak storage

a b s t r a c t Large chunk of cooling load is typically demanded during hospitality operation. In the pursuit of reducing its energy cost, a multi-functional commercial building incorporates ice thermal energy storage (TES) concept for storing harvested ice at off-peak hours and thawing the storage medium at peak hours. Ice built during off-peak hours is used to relief the cooling burden of hospitality operation at peak hours. The refrigeration cycle of water chilling would operate under two modes, the ice mode for thermal storage, and the chill water mode for instantaneous air cooling. The TES design uses a partial storage approach to satisfy off-peak hour cooling demands. Two screw chillers are integrated with each other for the implementation to allow feeding chill water to the air-handler coils while producing ice for storage at the same time. Not only does the TES system take advantage of the electric rate structure, super-cooled air also makes humidity levels to be lower than buildings with conventional cooling systems. Cool air is introduced to VAV (variable air volume) terminal boxes and is mixed with induced plenum air to bypass any reheat, yielding additional saving. The TES system saves electricity cost while providing thermal comfort to occupants. Indoor air quality may have been sacrificed when only 5 L/s per person of fresh ventilation is introduced into the building. © 2015 Elsevier Ltd. All rights reserved.

1. Preface Energy management is of considerable concern for officials of commercial or institutional buildings in Taiwan. The island’s subtropical climate demands large chunks of cooling loads during daylight hours across all commercial buildings and households. With limited natural energy resources, higher electricity rates during daylight peak hours are prevalent throughout Taiwan in hopes to reduce its peak demands and possible blackouts. This paper offers an alternative approach to energy management operation for commercial or institutional buildings by ways of ice thermal energy storage (TES) during nighttime off-peak hours. Load shift from expensive on-peak hours to cheaper off-peak hours has been an attractive option of peak demand management for decades (Hasnain, 1998; Parameshwaran et al., 2012; Yau and Rismanchi, 2012; Hajiah and Krarti, 2012; Sun et al., 2013). During daylight peak-hours, ice built from the previous night is released to chill water for cooling in peak-hours, thereby reducing the peak cooling load. Reduction of the peak cooling

∗ Corresponding author. Tel.: +886 910090825. E-mail address: [email protected] (Y.-H. Tsai). http://dx.doi.org/10.1016/j.ijhm.2015.01.005 0278-4319/© 2015 Elsevier Ltd. All rights reserved.

load can be met by operation strategies of full-storage, partial storage, or partial-storage demand-limiting (Dincer, 2002). In hospitality operations, it is expected that high electric demands from equipment other than air-conditioning systems would coincide with high cooling requirements in the afternoon hours. Optimal TES provides an effective method to limit the electric demand (Kintner-Meyer and Emery, 1995). The degree of savings varied across building operations, climates, and time-dependent rate structures (Fiorino, 1994; Lemort, 2006; Habeebullah, 2007; Ehyaei et al., 2010; Vetterli and Benz, 2012; Nassif et al., 2013; Rismanchi et al., 2013; Sebzali et al., 2013; Lin et al., 2014). A multi-functional commercial building, situated in the heart of Taipei City, is used for the case study. The building has a gross space of 15,110 m2 . The total conditioned space is 5950 m2 where hospitality operation is expected to demand high electricity consumption. The unconditioned spaces are mostly underground parking lots that are ventilated with outside air. Due to the fact that space is much more precious in Taipei City than in other parts of Taiwan, the challenge was to design a ventilation system that would reduce its peak demand and fit within the limited space. Because of this challenge, a unique system of air-conditioning was envisioned and subsequently installed in this building.

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

Nomenclature surface area of a wall or roof [m2 ] A bn , cn , dn conduction transfer function coefficients of a wall or roof EFT entering fluid temperature [◦ C] entering water temperature [◦ C] EWT F airflow rate [L/s] GPM gallons per minute leaving fluid temperature [◦ C] LFT LWT leaving water temperature [◦ C] NTU number of transfer units OA outside air Q cooling load of the space at calculation hour, () for the current, ( − 1) for the previous hour etc. [W] q˙ heat gain through wall or roof at calculation hour, () for the current hour, ( − 1) for the previous hour, ( − 2) for 2 h ago etc. [W/h] RA return air supply air SA SC shading coefficient SHGF solar heat gain factor by orientation, latitude, month, and hour solar-air temperature, () for the current hour, T ( − 1) for the previous hour etc. [◦ C] To , Ti outside, inside air temperature [◦ C] Troom constant indoor temperature to be maintained at a particular room [◦ C] U unit conductance or U-value [W/(m2 ◦ C)] v0 and v1 RTF weighting factors due to heat gain components w1 RTF weighting factor due to room air-circulation types

47

from the temperature differential between inside and outside of the building, the load is dependent on thermal characteristics of the construction, walls, roof, floor, fenestration, and interior generation from occupants and equipment. The driving force due to external solar radiation and internal operating schedule makes the load transient and difficult to predict. Fortunately, with the aid of computers and TMY (typical meteorological year) weather data, the task can be accomplished where a typical hourly outdoor data is shown in Table 1. A number of researchers had proposed different ways of predicting annual energy consumption in subtropical climate (Li et al., 2003; Mui and Wong, 2007). The transfer function method (TFM) is particularly well suited for use with a computer, taking into accounts of the transient effect to predict the hourly cooling load due to different types of walls, roofs, and fenestration (Stephenson and Mitalas, 1967; Al-Rabghi and Al-Johani, 1997). The method describes the heat flux at the inside of a wall or roof as a function of previous values of the heat flux and previous values of inside and outside temperatures. The TFM uses fixed, combined convection and radiation coefficients on the inside and outside surfaces, so that the conduction transfer functions are driven by solar-air temperature on the outside and by room temperature on the inside. The simulation used Songshan (Taipei City) Airport TMY data for outdoor conditions to calculate the required cooling loads. It is noted that there are no heating load for the building. The VAV (variable air volume) ventilation system used in this building would bypass any reheat in space zone’s terminal boxes. The cooling TFM assumes constant indoor and outdoor surface heat transfer coefficients (Mitalas, 1972). The heat gain through a wall or roof is calculated from ˙ = A[b0 T () + b1 T ( − 1) + b2 T ( − 2) + b3 T ( − 3) + · · ·] q() ˙ − 1) + d2 q( ˙ − 2) + d3 q( ˙ − 3) + · · ·] − [d1 q( − ATroom ()[c0 () + c1 ( − 1) + c2 ( − 2) + c3 ( − 3) + · · ·]

(1)

2. Building description The building has three underground floors, and four floors that are above ground, with a flattop-roof. All three basements are 75.6 m by 45.3 m (3425 m2 ), including unconditioned parking lots and conditioned space in the first basement (1B). The total underground conditioned space takes up 1210 m2 . Most of the cooling equipment is housed in the lowest basement floor. All underground parking spaces are ventilated with unconditioned outside air. The first floor is occupied by various catering services around the peripheral. A spacious dining area in the center of the first floor takes up approximately 336 m2 (24 m by 14 m) of the total first floor space (1350 m2 ). The ceiling of the dining area is elevated to the second floor, taking away the same amount of floor space from the second floor. The design is advantageous in buildings of hot climate because the conditioned cold air is heavier and stays in the first floor occupied space. No heating is required in winter where heated hot air would have stayed near the elevated ceiling due to air density differential. Nevertheless, the cooling load calculation still treats the elevated ceiling spaces as if there were two separate floor zones. Although the third floor has an identical gross area of 1350 m2 as the first and second floors, two terraces reduce the conditioning space down to 1070 m2 . The conditioned space in the fourth floor is identical to the third floor at 1070 m2 , but a reduced gross area without the two terraces. 3. Cooling load calculation Cooling load of the building envelope is the combination of the temperature differential between inside and outside temperatures and solar heat gains. Other than the driving force resulted

Specific conduction transfer function (CTF) coefficients for different constructions can be calculated using the software produced by Spitler et al. (1993). ASHRAE (2013) listed various groups of the walls and roofs for engineers to identify the CTF coefficients. The composition of the building envelope wall is granite, lightweight concrete, and insulation finishes. The civil/architectural engineers have indicated that the exterior stone has a similar thermal characteristic of a face brick. From ASHRAE Wall Group #17, the CTF coefficients of the building envelope walls were found, as shown in Table 2. Also included in this Table are the CTF coefficients for the building’s flattop-roof, adapted from ASHRAE Roof Group #1, steel deck with insulations. The actual cooling load depends on the magnitude and the nature of the heat gain calculated by Eq. (1). Stephenson and Mitalas (1967) related heat gains to the corresponding cooling load by a room transfer function (RTF). The heat storage characteristics of the enclosed space would be represented by three weighting factors, ˙ ˙ − 1), the correv0 , v1 , and w1 . Using the heat gains q() and q( sponding cooling load can be related to the current value of Q() and the preceding values of the cooling load and heat gain. ˙ ˙ − 1)w1 Q ( − 1) Q () = v0 q() + v1 q(

(2)

According to ASHRAE (2013), w1 is depended on the room envelope construction and the room air-circulation type. For example, a light construction mass with very high air-circulation would render w1 = −0.73. A heavy construction mass with low air-circulation would render w1 = −0.98. For this building, both the construction mass and the air-circulation may be considered as medium, then w1 = −0.94. The weighting factors v0 and v1 are related with the heat

48

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

Table 1 Example of climate design data for July 15th in Taipei, Taiwan. Hour

Dry-bulb temperature (◦ C)

Relative humidity (%)

Solar radiation (Wh/m2 )

0:01–1:00 1:01–2:00 2:01–3:00 3:01–4:00 4:01–5:00 5:01–6:00 6:01–7:00 7:01–8:00 8:01–9:00 9:01–10:00 10:01–11:00 11:01–12:00 12:01–13:00 13:01–14:00 14:01–15:00 15:01–16:00 16:01–17:00 17:01–18:00 18:01–19:00 19:01–20:00 20:01–21:00 21:01–22:00 22:01–23:00 23:01–24:00

25.3 25.1 24.9 24.8 24.7 25.0 25.7 27.0 28.4 29.4 30.1 30.5 30.6 30.1 29.4 28.4 27.9 27.4 27.0 26.4 26.1 25.9 25.7 25.6

88 88 88 89 89 87 84 78 73 68 66 65 67 68 72 73 75 79 81 82 83 84 85 86

0 0 0 0 0 0 134 258 393 412 368 367 371 389 345 279 190 76 0 0 0 0 0 0

Statistics for TWN Taipei IWEC: location – TAIPEI – TWN {25◦ 4 N} {121◦ 33 E} {GMT + 8.0 h}, elevation – 6 m above sea level, standard pressure at elevation – 101,253 Pa, data source – TWEC Data, WMO Station – 466960. Table 2 Conduction transfer function coefficients from ASHRAE. Coefficients

n=0 n=1 n=2 n=3 n=4 n=5 n=6

ASHRAE Wall Group #17

ASHRAE Roof Group #1

bn

dn

cn

bn

dn

cn

0.00000 0.00000 0.00013 0.00044 0.00030 0.00005 0.00000

1.00000 −2.00875 1.37120 −0.37897 0.03962 −0.00165 0.00002

0.00093 0.00093 0.00093 0.00093 0.00093 0.00093 0.00093

0.00487 0.03474 0.01365 0.00036 0.00000 0.00000 0.00000

1.00000 −0.35451 0.02267 −0.00005 0.00000 0.00000 0.00000

0.05362 0.05362 0.05362 0.05362 0.05362 0.05362 0.05362

gain component and the envelop construction, as shown in Table 3. Then, the cooling load due to conduction heat gain through a wall or roof is ˙ ˙ − 1) + 0.94Q ( − 1). Qcond. () = 0.681q() − 0.621q(

(2a)

(2b)

The cooling load due to radiation heat gain is ˙ ˙ − 1) + 0.94Q ( − 1). Qrad. () = 0.197q() − 0.137q(

q˙ = UA(To − Ti )

(3)

The radiation heat gain by glass is

The cooling load due to convection heat gain is ˙ ˙ Qconv. () = q() − 0.94q() + 0.94Q ().

calculation procedure bypasses Eq. (1). For example, the convection heat gains by glass is

(2c)

Calculations for the convection and radiation heat gains are much simpler. The heat gain at each hour is instantaneous, thus the

q˙ = A(SC)(SHGF)

(4)

The convection and radiation heat gains at the current hour are not affected by the previous hour heat gains. The heat gains at each hour are independent from each other. However, the actual cooling loads calculation, Eqs. (2b) and (2c), takes into account of the transient effect within the space zones.

Table 3 Room transfer function coefficient, v0 and v1 (weighting factors). Heat gain component

Room envelop construction

v0

v1

Solar heat gain through glass with no interior shade; radiant heat from people and equipment

Light Medium Heavy

0.224 0.197 0.187

1 + w1 – v0 1 + w1 – v0 1 + w1 – v0

Conduction heat gain through exterior walls, roofs, partitions, doors, windows with blinds or drapes

Light Medium Heavy

0.703 0.681 0.676

1 + w1 – v0 1 + w1 – v0 1 + w1 – v0

Convective heat generated by people and equipment, and from ventilation and infiltration air

Light Medium Heavy

1.000 1.000 1.000

0.0 0.0 0.0

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

49

Fig. 1. Highest and lowest peak cooling loads by hour during annum.

4. Designed cooling system The simulation showed minimum cooling loads during January and maximum cooling loads during July or August, depending on the hour of the day. Hour-by-hour required cooling loads for the building are shown in Fig. 1. In the middle of the day, hours 12 through 16 (noon–5:00 pm), the maximum cooling loads were derived from August. For the rest of the day, from hours 7 to 11 (7:00 am–noon) and from 17 to 22 (5:00–11:00 pm), the maximum cooling loads were derived from July. The maximum cooling load is 244.1 tons, which is at hour 16 (4:00–5:00 pm). It does not mean that every room and every floor of the building would reach their peak loads simultaneously at hour 16. It only means that the sum reaches its maximum at hour 16. For example, a breakdown

of the maximum cooling load required for each floor is shown in Table 4. Table 4 showed that the first basement reaches its highest peak cooling load at hour 17 (August), with 17.2 tons. By comparison, the first floor would need the most cooling at hour 16 (August), with 99.1 tons. The second floor requires less cooling, 25.3 tons when it reaches its maximum peak at hour 16 (August). The maximum peak cooling loads for the third and fourth floors are 57.5 tons and 46.2 tons, at hour 16 (July) and hour 15 (July) respectively. If the maximum peak cooling loads from all five floors were added without regard to the occurred time, the chiller system would have demanded 245.2 tons of refrigeration. The design would have been oversized and economically impractical because each space zone reaches its maximum peak cooling load at different times.

Table 4 Composition of the peak cooling load by floor and time. Occurrence

August (hour 17)

August (hour 16)

August (hour 16)

July (hour 16)

Floor levels

First basement

1st floor

2nd floor

3rd floor

Load types

Sensible

Latent

Sensible

0 0 0 0 7.72 0.78 8.70 8.66 2.47 0.19 3.92

0 0 0 0 0 0 9.82 17.17 0 0 0.98

0.99 38.64 0 1.23 42.19 0.57 37.21 38.92 17.71 0 34.93

Total

32.43

27.98

212.39

Chiller load (sen. + lat.)

60.4 kW (17.2 tons)

Solar Wall Roof Fenestration Lighting Equipment Occupancy Outside air Supply fan AHU ducts Miscellaneous

Latent

July (hour 15) 4th floor

Sensible

Latent

Sensible

Latent

Sensible

Latent

0 0 0 0 0 0 41.9 85.94 0 0 8.38

3.04 19.18 0 3.62 9.23 1.15 7.17 8.46 5.30 0.46 3.69

0 0 0 0 0 0 8.11 19.24 0 0 0.41

5.75 36.59 0 5.01 24.63 1.09 19.71 19.84 11.33 0.97 9.42

0 0 0 0 0 0 22.16 44.54 0 0 1.11

4.30 31.90 33.02 5.44 19.28 0.21 7.52 8.97 12.10 1.07 7.21

0 0 0 0 0 0 8.86 22.06 0 0 0.44

136.22

61.30

27.76

134.33

67.81

131.04

31.36

348.6 kW (99.1 tons)

89.1 kW (25.3 tons)

202.1 kW (57.5 tons)

162.4 kW (46.2 tons)

50

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

4.1. Air handling unit The cooling system can be divided into two parts, the primary system (water-side) and the secondary system (air-side). The primary system is the chill water plant that supplies chilled water to the secondary system. The secondary system receives chilled water and converts cooling from water to air. The AHU (air handling unit) is the air and liquid handling equipment, including coils, mixing boxes, pipes, and ducts. Then, cool air is distributed to each space zone’s VAV terminal devices, air diffusers, and fans. The ventilation design does not use just one large AHU because it would have taken up a large chunk of precious space, as well as longer path of ducts thereby extended duct space. Instead, there are nine smaller AHUs scattered within the building, at cramp corners. The specification of each AHU is roughly summarized in Table 5. From the top floor, one of the two AHUs serves the 4th floor stores. The other AHU would condition the open space shared with the 3rd floor. That gave way to only one AHU needed for cooling of the 3rd floor stores because the elevated open space is conditioned by the 4th floor AHU. There are three AHUs located in the 2nd floor, two corner AHUs provide ventilation to the peripheral stores. The third AHU on the 2nd floor ventilates the open space in the middle, which is shared with the elevated 1st floor zone. There is no AHU in the 1st floor. The 1st floor zones are air-conditioned by one of two AHUs located in the 1st basement. One basement AHU serves the 1st floor peripheral stores, excluding the open zone in the middle. The other basement AHU serves its own basement rooms. The last of the nine AHUs is located underneath the 3rd basement, where the chilled water system plant locates. This last AHU provides cooling to the open dining area in the middle of the 1st floor where cooling is demanded the most. The proximity of the chillers with ice thermal storage tanks in the 3rd basement to the AHU below (at the 4th level underground) shortens the piping, reduces its heat gain, and saves energy cost. 4.2. Chill water system The AHU provides cooling air via heat transfer between chilled water and air for conditioning at the cooling coils. The chill water system provides refrigerated water feeding to the cooling coils. Ice thermal energy storage (TES) system was implemented to store harvested ice and thaw the storage medium later. Energy efficient control is accomplished by the hybrid ice TES system of water phase transition, solid–melting–liquid and liquid–freezing–sold (Ezan et al., 2010; Zhai et al., 2013; Beghi et al., 2014). The intend was to build ice during off-peak hours when the electricity is cheaper and use the harvest to relief the cooling burdens at peak hours. A sophisticated routine to optimize the use of storage was realized with a partial storage operation (Zhou et al., 2005; Braun, 1990; Chaichana et al., 2001; Sanaye and Shirazi, 2013). During offpeak hours, the capacity that is not needed to meet the load is applied to the charging storage. Two identical screw chillers are integrated with each other to allow two modes of refrigeration cycle, the ice mode, and the chill water mode, as shown in Table 6. The chillers use environmental friendly refrigerant R-134a, which is the replacement of R-22. R-22 has been phased out by the refrigeration industry due to its harmful CFCs (chlorinated and fluorinated hydrocarbons) to the earth’s stratospheric ozone. During the chill water mode, the refrigerant (R-134a) provides cooling to the working fluid at the evaporator, as shown in Table 6 from 12 ◦ C to 6 ◦ C. Recall that the AHU cooling coil water absorbs heat from the air, the water temperature rises from 6 ◦ C to 12 ◦ C in Table 5. The air–water cycle completes when water returns to the chiller from cooling coil for the completion of the water loop. At the evaporator side of the chiller, the working fluid contains 30% ethylene glycol (EG) to allow below-freeze cooling (Rogers and Stefl, 1993).

The working fluid (30% EG) dissipates the absorbed heat to the condensing water (from 32 ◦ C to 38 ◦ C). During the ice mode, the refrigerant (R-134a) chills the working fluid (30% EG) from −3 ◦ C to −7 ◦ C and feeds the freezing ice-water (−7 ◦ C) to three identical ice TES tanks, each having a capacity of 760 ton h. With either mode of the refrigeration cycle, the condensing water flow rate is 1050 GPM, and the working fluid (30% EG) flow rate is 880 GPM at the evaporator. Because the utility company’s electrical rate structure strongly favors the use of cooling produced during off-peak hours, the ice storage tanks reserve the ice harvest at night and release the medium during peak hours when the cooling demand is at its highest. It was recognized that there may be some nighttime cooling load that had to be addressed. Rather than adding a separate chiller for this duty, the two operating chillers are integrated with each other during ice-building cycles through a back pressure valve that controls the refrigerant temperature in the chiller for the water-chilling duty. During peak hours in the day, the ice harvest from the previous night is used to handle a portion of the cooling demand, and the remaining portion of the cooling responsibility would come from the chillers operating under the chill water mode. The compressors are programmed to operate at a higher suction temperature. Thus, the maximum compressor motor load would reduce from 302 kW (ice mode) to 257 kW (chill water mode). Consequently, energy cost savings are achieved with balances of cooling from nighttime cheaper hours. In addition to electric cost savings by the ice TES system, rejected heat from the condensers can be used for further savings. A heat exchanger (HX) is used for the energy recovery of rejected heat from the condensers. Domestic water warmed by the rejected heat via HX is used for baptizing. As shown in Table 7A, the cold side of heat exchanger plate is actually the hot water from the condensers. The hot side of the HX plate is actually the cold domestic water to be heated. From the cold side, leaving water (32 ◦ C) would be fed to two cooling towers along with portions of the condensing water (38 ◦ C) that had bypassed the HX. The cooling towers located at the top of the building roof dissipate heat from the water to the outside air. Another heat exchanger is used for a different function, as shown in Table 7B. Melted ice liquid mixture (25% EG) at 3 ◦ C would enter the cold side of the HX. Through heat transfer, the hot side liquid (cooling coil water) would enter at 13 ◦ C and leave at 6 ◦ C. Then, the cold side leaving liquid (at 12 ◦ C) is fed to the evaporator for the next refrigeration cycle at the chiller. One omitted heat exchanger (No. HX-2) is reserved for future addition of the chill water system when the building capacity expands. The future addition includes one more chiller, one more cooling tower, and three more ice thermal storage tanks. The omitted HX will have identical characteristics, with 418 effectiveness number of transfer units (NTU).

5. Analysis One of the more difficulties of the designing process is accurately predicting the peak demand loads. Underestimating the required peak demand load will lead to a high frequency of thermal uncomfortableness by the occupants. Although an undersized cooling system may save the initial installation cost, it will lead to future penalties when the off-peak hour ice thermal storage does not sufficiently supply peak hour demand loads. On the other hand, if the cooling system is oversized, then obviously a higher and unnecessary initial cost is spent. Over-installed chillers, ice storage tanks, pipes, and ducks will also waste precious building space. Hopefully, the loads simulation results had derived an optimum sizing of the chilling system that can adequately handle its cooling responsibility and ensure occupants’ thermal comfort.

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

51

Table 5 Air and liquid handing equipment specifications at the AHU. Air handling units

Fan (air-side)

Coils (water-side)

No.

Location

Servers

FSA (L/s)

FOA (L/s)

TRA (◦ C)

TSA (◦ C)

GPM

EWT (◦ C)

LWT (◦ C)

AH-4-1 AH-4-2 AH-3-1 AH-2-1 AH-2-3 AH-2-2 AH-B1-1 PAH-B1-1 AH-B1-2

4th floor 4th floor 3rd floor 2nd floor 2nd floor 2nd floor 1st basement 1st basement 4th basement

4F stores 3F/4F open zone 3F stores 2F stores 2F stores 1F/2F open zone 1F stores 1B rooms 1F/2F open zone

5100 3000 5550 2100 2350 4200 5125 1700 4350

1020 600 1110 425 470 900 1025 1700 1250

26 26 26 26 26 26 26 35 26

13 13 13 13 13 13 13 20 13

65 38.2 70.7 26.8 30 53.5 65.2 48.4 55.4

6 6 6 6 6 6 6 6 6

12 12 12 12 12 12 12 12 12

Table 6 Two modes of operation for the screw chillers. Chill water mode

Ice mode

Condenser ◦

Evaporator ◦

◦

Compressor motor ◦

Condenser ◦

Evaporator ◦

◦

Compressor motor ◦

EWT ( C)

LWT ( C)

EFT ( C)

LFT ( C)

Maximum load (kW)

EWT ( C)

LWT ( C)

EFT ( C)

LFT ( C)

Maximum load (kW)

32

38

12

6

257

32

38

−3

−7

302

Table 7A Plate heat exchanger specifications (baptismal water). No.

HX-3

Service

Baptismal

Plate

Cold side (water)

Hot side (water)

NTU

Thickness (mm)

GPM

EWT (◦ C)

LWT (◦ C)

GPM

EWT (◦ C)

LWT (◦ C)

10

0.5

23.6

38

32

6

15

37

5.1. Load analysis Standard deviation of the peak cooling demand was calculated to statistically measure the range of the mean peak load, and was reported as an annual number at each hour. As shown in Fig. 2, the highest standard deviation was found to be 21.8 tons at hour 18 (6:00–7:00 pm), and the lowest standard deviation was found to be 12.5 tons at hour 12 (noon–1:00 pm). When a peak cooling load has a high standard deviation, its range of cooling demands has been very wide, indicating that there is a greater potential for volatility (from 6:00 pm to 7:00 pm). Conversely, a low standard deviation indicates that the cooling demands are very steady (from noon to 1:00 pm). Since the peak cooling demands vary greatly by the hour of the day, a high standard deviation of the loads implies high volatility. A secondary y-axis was inserted into Fig. 2 relating the standard deviation by percentage of the mean peak load. It was found that at the first hour of the building operation (hour 7: from 7:00 pm to 8:00 pm), its associated standard deviation percentage-wise was found to be the highest, at 12.9%, representing high volatility of its peak demand. Conversely, the lowest standard deviation was found to be 12.5 tons, at hour 12 (noon–1:00 pm), representing low volatility of its peak demand for the hour. Statistical analysis of the peak cooling demand also showed high volatility of the loads, 21.8 tons, at hour 18 (6:00–7:00 pm). This is the period when the transient nature of the building envelope’s thermal mass block takes its greatest effects on the total cooling load. In other words, the thermal energy stored within the envelope mass is releasing its heat to the exterior direction

and to the interior direction. The building envelope is a composite of three primary layers, granite (thermal characteristics of face bricks), concrete blocks, and insulation. By breaking down the wall composition, the thermal mass storage phenomenon is analyzed from each homogenous layer. As an example, the heat flux through the exterior surface of the face brick with time is shown in Fig. 3. Similarly, the heat flux through the interior surface of the face brick is also shown in the figure. The example applied a triangular pulse of T = 10 ◦ C (outdoor air minus indoor air). Initially T = 0 ◦ C, it rises linearly to 10 ◦ C in 1 h, and then decreases linearly to 0 ◦ C in the next hour. Positive heat flux means that the heat flow direction is inward, from outside to inside. Conversely, negative heat flux means that the heat flow direction reverses itself and shifts outward. Interior brick surface heat flux peaks at 1.75 h in Fig. 3, it implies that the face brick thermal mass takes 45 min (1.75−1 = 0.75 h). Therefore, face brick of the envelope walls did not cause the transient effect to delay by 2 h (hour 16–18). The middle lay of the wall composite is filled with concrete. If the concrete blocks were to stand alone as the envelope walls, applying the same triangular pulse of T = 10 ◦ C as the previous example, Fig. 4 is resulted. This figure showed that the thermal mass effect of the concrete would store and delay heat flows by approximately 2 h (3 − 1 = 2 h). Thus, concrete layer of the envelope wall is identified as the main cause of thermal delay and the greatest transient effect on the cooling loads, from hour 16 (highest peak cooling demand) to hour 18 (volatility of the cooling loads). In addition, immediately after the triangular pulse decreases, the inner surface heat flow direction briefly changes its direction outward for about 15 min

Table 7B Plate heat exchanger specifications (cooling coil water). No.

HX-1

Service

Chill water

Plate

Cold side (25% EG)

Hot side (water)

NTU

Thickness (mm)

GPM

EWT (◦ C)

LWT (◦ C)

GPM

EWT (◦ C)

LWT (◦ C)

418

15

852.4

3

12

925

13

6

52

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

Fig. 2. Standard deviation of the mean peak cooling load by the hour.

(1.25 − 1 = 0.25 h). The shifts of the heat flow direction contributed to the cooling loads volatility. If the interior insulation finishes of the envelope walls were to stand alone as an independent wall, by applying the same triangular pulse of T = 10 ◦ C, Fig. 5 is derived. Figs. 3–5 showed that the exterior layer (face bricks) of the envelope wall would reject the most heat (i.e. 83 W/m2 ). The middle layer (concrete) of the envelope wall would store and delay heat flow for the longest time period (i.e. 2 h). The interior layer (insulation finishes) of the envelope wall would transmit the least heat flow (i.e. 1 W/m2 ) into building spaces. The transmission would delay by 90 min (2.5 − 1 = 1.5 h). These three figures demonstrated the transient effect of the conduction heat flow through building envelope.

5.2. Cost analysis The cost of the HVAC system was approximately one million US dollars, or $168 per m2 ($16 per ft2 ) of conditioned building space. Although the ice TES approach represents a premium over the conventional centrifugal chillers, there are significant savings of downsizing (i.e. chiller, piping, ductwork, fans) due to reduced peak load. The cost redistribution is what makes the integrated ice TES design concept viable. The savings of system downsizing offset the additional investments in the ice storage, giving a net impact of instant payback. Specifically, a conventional system requires 245.2 tons (862 kW) of refrigeration capacity where as the ice TES system requires only 172 tons (604 kW) of chiller capacity.

Fig. 3. Surface heat flux of a face brick due to a triangular pulse of T = 10 ◦ C.

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

53

Fig. 4. Surface heat flux of concrete due to a triangular pulse of T = 10 ◦ C.

Fig. 5. Surface heat flux of wall insulation due to unit triangular pulse.

Electricity tariff schemes in Taiwan are very complicated. A simplified estimation takes into consideration of only time-dependent differential (NT$3.89 and NT$1.99 per kW h for on- and off-peak rates). Simulation showed an estimated annual cooling consumption of 740,648 kW h for the building. With disregard to savings of peak load penalty, as well as other tariff schemes, the competitive first cost in combination with an estimated annual savings of over NT$1.4 million (or roughly 46,000 US dollars) is easily achieved. 6. Conclusions and recommendations The primary reason for the use of ice thermal energy storage (TES) is economic. In no particular order, some of the key benefits are listed as follows: • Downsized equipment meets the peak cooling load when portions of the demand are met by cooling from ice TES. Thus, capital cost (first cost) savings are achieved.

• Energy cost savings are achieved with significant reductions of time-dependent energy costs, such as demand charges from the electric company, and the higher peak-hour rate structures. • Although ice TES is designed to shift energy usage rather than reduce its total energy use, energy conservation may have been achieved with off-peak hour chilling. The TES system permits chillers to operate at night when lower condensing temperatures improve equipment efficiency. • Aside from energy or cost savings, the TES system provides increased flexibility to the total cooling capacity. • Super-cooled air is provided to each space zone’s VAV mixing boxes via thawed ice from the TES tanks. In a very humid climate (Taipei), super-cooled air dehumidifies and improves occupant thermal comfort. In addition, super-cooled air allows the VAV terminal to reduce its air velocity. Therefore, the fan speed reduces, the fan energy consumption reduces, and also the fan noise level. The space environment is quieter as a result.

54

C.-T. Wu, Y.-H. Tsai / International Journal of Hospitality Management 46 (2015) 46–54

For the most part, this paper has focused on the primary side of the cooling system. From the ventilation standpoint, the delivery and diffusion of conditioned air into space zone has not been studied. The effectiveness of air mixing in space zones may be studied in the future. The cooling load calculation followed ASHRAE Standard 62.1 (ASHRAE, 2010) guidelines to estimate maximum occupancy per floor area of each space application. The authors are very interested to find out if the indoor air quality is up to par in a future study of this multi-functional building. References Al-Rabghi, O.M.A., Al-Johani, K.M., 1997. Utilizing transfer function method for hourly cooling load calculations. Energy Conserv. Manag. 38 (4), 319–332. ASHRAE, 2010. Ventilation for Acceptable Indoor Air Quality. ANSI/ASHRAE Standard 62.1-2010. American Society of Heating, Refrigerating, and Air-Conditionings Engineers (ASHRAE) Inc., Atlanta, GA. ASHRAE, (Chapter 18) 2013. Nonresidential cooling and heating load calculations. In: ASHRAE Fundamentals Handbook. ASHRAE, Inc., Atlanta, GA. Beghi, A., Cecchinato, L., Rampazzo, M., Simmini, F., 2014. Energy efficient control of HVAC systems with ice cold thermal energy storage. J. Process Control 24, 773–781. Braun, J.E., 1990. Reducing energy costs and peak electrical demand through optimal control of building thermal storage. ASHRAE Trans. 96 (2), 876–888. Chaichana, C., Charters, W.W.S., Aye, L., 2001. An ice thermal storage computer model. Appl. Therm. Eng. 21, 1769–1778. Dincer, I., 2002. On thermal energy storage systems and applications in buildings. Energy Build. 34, 377–388. Ehyaei, M.A., Mozafari, A., Ahmadi, A., Esmaili, P., Shayesteh, M., Sarkhosh, M., Dincer, I., 2010. Potential use of cold thermal energy storage systems for better efficiency and cost effectiveness. Energy Build. 42, 2296–2303. Ezan, M.A., Ozdogan, M., Gunerhan, H., Erek, A., Hephasli, A., 2010. Energetic and exergetic analysis and assessment of a thermal energy storage (TES) unit for building applications. Energy Build. 42, 1896–1901. Fiorino, D.P., 1994. Energy conservation with thermally stratified chilled-water storage. ASHRAE Trans. 100 (1), 1754–1766. Habeebullah, B.A., 2007. Economic feasibility of thermal energy storage systems. Energy Build. 39, 355–363. Hasnain, S.M., 1998. Review on sustainable thermal energy storage technologies, part II: cool thermal storage. Energy Convers. Manag. 39 (11), 1139–1153. Hajiah, A., Krarti, M., 2012. Optimal controls of building storage systems using both ice storage and thermal mass – part II: parametric analysis. Energy Convers. Manag. 64, 509–515.

Kintner-Meyer, M., Emery, A.F., 1995. Optimal control of an HVAC system using cold storage and building thermal capacitance. Energy Build. 23, 19–31. Lemort, V., 2006. A numerical comparison of control strategies applied to an existing ice storage system. Energy Convers. Manag. 47, 3619–3631. Li, D.H.W., Wong, S.L., Lam, J.C., 2003. Climatic effects on cooling load determination in subtropical regions. Energy Conserv. Manag. 44, 1831–1843. Lin, H., Li, X.H., Cheng, P.S., Xu, B.G., 2014. Study on chilled energy storage of airconditioning system with energy saving. Energy Build. 79, 41–46. Mitalas, G.P., 1972. Transfer function method of calculating cooling loads, heat extraction rate, and space temperature. ASHRAE Trans. 14 (12), 52. Mui, K.W., Wong, L.T., 2007. Cooling load calculations in subtropical climate. Build. Environ. 42, 2498–2504. Nassif, N., Tesiero, R.C., Singh, H., 2013. Impact of ice thermal storage on cooling energy cost for commercial HVAC systems. ASHRAE Trans. 119 (1), 1–7. Parameshwaran, R., Kalaiselvam, S., Harikrishnan, S., Elayaperumal, A., 2012. Sustainable thermal energy storage technologies for buildings: a review. Renew. Sustain. Energy Rev. 16, 2394–2433. Rismanchi, B., Saidur, R., Masjuki, H.H., Mahlia, T.M.I., 2013. Modeling and simulation to determine the potential energy savings by implementing cold thermal energy storage system in office buildings. Energy Convers. Manag. 75, 152–161. Rogers, E.C., Stefl, B.A., 1993. Ethylene glycol: its use in thermal storage and its impact on the environment. ASHRAE Trans. 99 (1), 941–949. Sanaye, S., Shirazi, A., 2013. Thermo-economic optimization of an ice thermal energy storage system for air-conditioning applications. Energy Build. 60, 100–109. Sebzali, M.J., et al., 2013. Comparison of energy performance and economics of chilled water thermal storage and conventional air-conditioning systems. Energy Build. 69, 237–250. Spitler, J.D., McQuiston, F.C., Lindsey, K., 1993. Development of a revised cooling and heating load calculation manual. ASHRAE Trans. 99 (1), 175–182. Stephenson, D.G., Mitalas, G.P., 1967. Cooling load calculations by thermal response factor method. ASHRAE Trans. 73 (2), 1. 1-1.7. Sun, Y., Wang, S., Xiao, F., Gao, D., 2013. Peak load shifting control using different cold thermal energy storage facilities in commercial buildings: a review. Energy Convers. Manag. 71, 101–114. Vetterli, J., Benz, M., 2012. Cost-optimal design on an ice-storage cooling system using mixed-integer linear programming techniques under various electricity tariff schemes. Energy Build. 49, 226–234. Yau, Y.H., Rismanchi, B., 2012. A review on cool thermal storage technologies and operating strategies. Renew. Sustain. Energy Rev. 16, 787–797. Zhai, X.Q., Wang, X.L., Wang, T., Wang, R.Z., 2013. A review on phase change cold storage in air-conditioning systems: materials and applications. Renew. Sustain. Energy Rev. 22, 108–120. Zhou, J., Guanghua, W., Deng, S., Turner, D., Claridge, D., Contreras, O., 2005. Control optimization for chilled water thermal storage system under complicated timeof-use electricity rate schedule. ASHRAE Trans. 111 (1), 184–197.