Energy benefit of a dedicated outdoor air system over a desiccant-enhanced evaporative air conditioner

Applied Thermal Engineering 108 (2016) 804–815

Contents lists available at ScienceDirect

Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng

Research Paper

Energy benefit of a dedicated outdoor air system over a desiccant-enhanced evaporative air conditioner Hui-Jeong Kim, Sung-Joon Lee, Sang-Hyeon Cho, Jae-Weon Jeong ⇑ Department of Architectural Engineering, College of Engineering, Hanyang University, 222 Wangsimni-Ro, Seungdong-Gu, Seoul 04763, Republic of Korea

h i g h l i g h t s
The energy performances both DOAS and DEVap systems were compared.
The sensible and latent cooling performances were analyzed using energy simulation.
The DOAS showed 22% less annual primary energy consumption than the DEVap system.

a r t i c l e

i n f o

Article history: Received 3 June 2016 Revised 26 July 2016 Accepted 28 July 2016 Available online 29 July 2016 Keywords: Dedicated outdoor air system Decoupled system Desiccant evaporative cooling Desiccant-enhanced evaporative air conditioner

a b s t r a c t The purpose of this study is to comparatively evaluate the energy performances of a dedicated outdoor air system (DOAS) and desiccant-enhanced evaporative air conditioner (DEVap) in building applications. The DOAS effectively accommodates latent cooling loads and some of the sensible cooling loads of the space by introducing cooled and dehumidified ventilation air into a building while integrating a parallel system aimed at reducing the remaining sensible load. The DEVap enhances the energy performance of a variable air volume system by reducing cooling coil loads through preconditioning of the supply air before it reaches a coil. The preconditioning is accomplished by using a liquid desiccant system and dew-point indirect evaporative cooler. In this paper, the operating and annual primary energy consumptions of both the DOAS and DEVap systems are compared based on detailed energy simulations. The results indicated the energy saving potential of DOAS to be greater than that of the DEVap. Specifically, a DOAS with ceiling radiant cooling panels experienced 20% less primary energy consumption compared to a DEVap. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction An air conditioning system should provide ventilation as well as sensible and latent cooling functions to maintain acceptable indoor air quality, temperature, and humidity set points for a conditioning zone. However, attempting to satisfy multiple air conditioning requirements by using one air conditioning system may result in problems due to inefficient control [1]. The concept of a decoupled system [1–3,5], which involves decoupling the ventilation function from the air conditioning function, or decoupling the sensible cooling from latent cooling, has been proposed for effective control and energy conservation of air conditioning systems. A dedicated outdoor air system (DOAS) is a decoupled system solution that independently controls the latent and sensible air conditioning loads. The DOAS accommodates overall latent load

⇑ Corresponding author. E-mail address: [email protected] (J.-W. Jeong). http://dx.doi.org/10.1016/j.applthermaleng.2016.07.185 1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.

and a certain amount of sensible load by treating outdoor air ventilation flow supplied to the conditioned zone with a total energy recovery component assisted by a cooling coil and sensible heat exchanger. Various configurations of DOAS, including desiccant systems for the allocation of humidification load of the cooling coil, have been produced. However, a typical dual wheel type DOAS, which is comprised of an enthalpy heat exchanger, cooling coil, and sensible heat exchanger, provided the highest energy conservation effect among various configurations [4,5]. Additional sensible cooling is conducted by a separate air conditioning system operated in parallel with the DOAS to maintain building space conditions [6]. Regarding the energy conservation of a DOAS with parallel cooling systems, Zakula et al. [7] investigated a low-lift cooling system that decouples thermally activated building surfaces (TABS) and a DOAS. Hallenbech [8] examined the DOAS with a fan-powered induction unit (FPIU) system. Both systems showed significant operating energy savings (i.e., approximately 50%) as compared to conventional variable air volume systems (VAV). A DOAS with

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

805

Nomenclature A C CAPFT cpa DFR EIRFPLR EIRFT h k K LG Lc _ m NTU P PLR Q QCRCP QR spd T U V V_ W

heat transfer area (m2) heat capacity rate (kJ/kg K) capacity of a chiller specific heat of moist air (kJ/kg K) driving force ratio part-load efficiency of a chiller full-load efficiency of a chiller enthalpy (kJ/kg) thermal conductivity (W/m K) enthalpy change to wet-bulb temperature change ratio on the wet side of the IEC liquid to gas ratio liquid desiccant solution concentration (%) mass flow rate (kg/s) number of heat transfer coefficients power (kW) part-load ratio load sensible load accommodated by ceiling radiant cooling panel exhaust air to outdoor air flow rate rotation speed dry bulb temperature (°C) overall heat transfer coefficient (W/m2 K) face velocity (m/s) volume flow rate (m3/s) humidity ratio (kg/kg)

Greek symbols a; b; F p ; F s ; n model coefficients Dp pressure drop in pump and fan (kPa) e effectiveness g efficiency £ relative humidity (%) Subscripts air air stream c cold fluid

ceiling radiant cooling panels (CRCP), as suggested in the literature [9–12], has also shown a relatively high operating energy conservation (i.e., over 40%) over conventional air conditioning systems. The decoupled systems that integrate desiccant dehumidifier and evaporative cooling system for separate latent and sensible load control are also proposed in several open literatures [2,3,13–16]. Dai et al. [13] introduced a hybrid air conditioning system that preconditions the process air through a liquid desiccant system (LD) combined with an evaporative cooling system. This hybrid system reduces the operating energy consumption of a conventional vapor compression-based air conditioning system. Experimental research on the energy conservation benefits of a desiccant rotor-integrated vapor compression air conditioning system are described in the literature [14]. An application of evaporative cooling and liquid desiccant in an air conditioning system has also been suggested in some experimental research. Kim et al. [2,15] suggested an application of LD, indirect evaporative cooling system (IEC), and direct evaporative cooling system supplied with 100% outdoor air. They experimentally showed their proposed system to use 68% less energy than a conventional VAV. Ham et al. [16] also suggested a non-vapor compression air conditioning system that operates an LD and

cws chilled water supply EA exhaust air eq equilibrium fan, design design value of the fan operation h hot fluid lat latent in inlet OA outdoor air out outlet max maximum p primary air r ratio RA room air ref reference s secondary air SA supply air SA, set supply air set point sen sensible t target w water film Superscript wb Wet bulb Abbreviations CC cooling coil CRCP ceiling radiant cooling panel DEVap desiccant-enhanced evaporative air conditioner DP-IEC dew-point indirect evaporative cooler DOAS dedicated outdoor air system DPT dew point temperature EW enthalpy wheel HC heating coil IEC indirect evaporative cooler LD liquid desiccant SW sensible wheel VAV variable air volume system

dew-point evaporative cooling system in parallel. The results indicated that the system conserved primary energy over 12% compared with the conventional VAV system. Woods and Kozubal [3] also suggested a desiccant-enhanced evaporative air conditioner (DEVap), which initially dehumidifies process air using a liquid desiccant dehumidifier and then cools the dry process air through a dew-point evaporative cooler close to its dew-point temperature [17]. Consequently, the cool and dry supply air meets the sensible and latent loads of the conditioned zone. The mass flow rate of the supply air in the DEVap can be modulated in a manner similar to that for a conventional VAV to meet the air conditioning load of the building. Both the DOAS and DEVap are well known for their energy saving potentials. However, the performances of the systems have not been systematically compared. The two decoupled systems require different air conditioning strategies as the DOAS and VAV systems have different system configurations. Therefore, in this research, the energy performances of a DOAS with a parallel system and DEVap are estimated using a detailed energy simulation that simulates supplying conditioned air to a model office building. The potential energy savings of the two systems are then compared.

806

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

2. System overview 2.1. Dedicated outdoor air system The DOAS considered in this research (Fig. 1) takes the form of a DOAS configuration that consists of an enthalpy wheel (EW), sensible wheel (SW), and cooling coil (CC) [5,6]. CRCPs are used as a parallel cooling system to accommodate the remaining sensible load after the DOAS conditions the entire latent load and a percentage of the sensible load [9,11,12]. In the DOAS unit, hot and humid ambient air during the summer season primarily undergoes cooling and dehumidification by the EW before passing through the CC. The EW contributes to reducing the load of the CC by transferring heat and moisture of the process air to the exhaust air stream. An EW rotating at its maximum capacity (e.g., 18–20 rpm) provides a latent and sensible effectiveness as high as 80%. The effectiveness of the EW decreases as the rotating speed decreases [18]. The process air preconditioned by the EW is additionally cooled and dehumidified by the CC, until its dew point temperature (DPT) meets the predefined setpoint. In a DOAS, the supply air (SA) passing through the CC should be sufficiently dry to accommodate the overall latent load of the space. However, when the temperature of SA is too low for air conditioning, the SW should be activated to reheat the SA with the waste heat transferred from the exhaust air stream. The effectiveness of the SW varies along with its rotation speed. In this research, the maximum efficiency of the SW at full rotating speed is set to 85%, which is a common upper limit of commercial sensible heat exchangers [19]. The CRCP system in this research is employed as a parallel cooling system used to remove the remaining sensible load of the conditioned space. When the DOAS fails to condition the whole

sensible load of the conditioned zone, the CRCP system is activated to provide additional sensible cooling needed for maintaining the room temperature setpoint. The EW moderates the introduced cold and dry outdoor air conditions during winter by regaining waste heat and moisture from the exhaust air that then reduces the required heating and humidification loads of the system. The sensible wheel reheats the SA up to a neutral temperature (e.g., 20 °C) by reclaiming waste heat from the exhaust air. A parallel heating system is required to provide additional heat into the space to meet the sensible heating load of the model space. As a parallel heating system, an electric heater is used in this research. Additional information on DOAS operation strategies, including intermediate season operations, is provided in detail in the literature [11,12,18,20]. 2.2. Desiccant-enhanced evaporative air conditioner As shown in Fig. 2, the DEVap initially dehumidifies the process air, which is a mixture of the introduced outdoor air and recirculation air with LD. The DEVap then cools the dehumidified process air using a dew-point indirect evaporative cooler (DP-IEC) [3,21]. Dehumidification is the first process stage performed by the LD to produce dry SA to meet the total latent load required by the conditioning zone. Dehumidification also enhances the evaporative cooling in the second stage process, which is completed by using the DP-IEC (Fig. 3(a)). In the first stage (Fig. 3(b)), liquid desiccant films on the flat-sheet microporous membranes within the primary channels of the LD unit absorb the moisture from the mixed air. The heat generated by this exothermic reaction of the liquid desiccant solution is removed by direct evaporative cooling in the secondary channels of the LD unit. The water and scavenger air are

Fig. 1. Schematic of a DOAS with a CRCP system.

Fig. 2. DEVap schematic.

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

807

Fig. 3. Schematic of DEVap air and water flow.

supplied to the secondary channels to cool the scavenger air through the direct evaporative cooling process. The heat transferred from the primary channels is absorbed by the secondary air during the dehumidification process. In the second stage, the DP-IEC shown in Fig. 3(b) cools dry process air conditioned by the LD unit to approximately its dew point temperature. The primary air that passes through the dry channels (i.e., primary channels) of the DP-IEC loses heat to the secondary air passing through the wet channels (i.e., secondary channels) where the vaporization occurs. The water is distributed into the secondary channels of the DP-IEC, and a portion (e.g., 30%) of the primary air leaving the dry channels is directed to the wet channels [3,22,23]. The cool and dry outlet primary air directed to the wet channels enhances both the evaporative cooling in the secondary channels and sensible cooling in the primary channels of the DPIEC. In the heating season, the DEVap operates as a conventional VAV system by deactivating the LD and DP-IEC. The electric heater is activated to provide heating for the conditioned zone [24]. 3. Simulation overview To compare the energy consumption and performance of a DOAS operating in parallel with CRCP to those of a DEVap, detailed energy simulations were performed with the assumption that each system serves a small open-plan office space, as defined in this research. 3.1. Model space The thermal loads of the design space were attained using the TRNSYS 17 program. Fig. 4 shows the dimensions of the model space. The floor area of the space was 100 m2, and the floor height was 3 m. Two 10 m2 windows were located in the exterior walls facing south and west. The window-to-wall ratio of the model space is 0.17. Based on ASHRAE Standard 90.1 [25], sensible and

latent heat generation rates of an occupant were set to 75 W and 45 W, respectively. Typical schedules of occupancy and system operation for an office building recommended by ASHRAE standard 90.1 were also adopted for the load calculation. The indoor temperature and relative humidity setpoints of the summer season were 24 °C and 40%, respectively, while those of the winter season were 20 °C and 50%. In the intermediate seasons, the space was conditioned with the setpoint of 22 °C and 50%. Table 1 summarizes the physical conditions of the model space. 3.2. DOAS with CRCP system simulation 3.2.1. Enthalpy wheel The effectiveness values (i.e., esen ; e;at ) of the enthalpy wheel used in the DOAS were predicted by the established polynomial models (Eqs. (1) and (2)) found in the open literature [18,20]. The sensible effectiveness is predicted by four variable parameters: incoming supply air face velocity (VSA;in ); temperature (TSA;in Þ; relative humidity of the supply air (£SA;in ), and the ratio of exhaust air to outdoor air flow rate (Q R ). Additionally, the latent effectiveness can be predicted by the temperature and relative humidity parameters of the exhausted air (TEA;in ; £EA;in ). The operation mode of the enthalpy wheel in the DOAS was described in previous research [12,20,26]. Fig. 5 describes operation strategies of the enthalpy wheel according to outdoor air conditions. In region 1, when the outdoor air enthalpy (hOA ) is higher than the enthalpy of room air (hRA ) that returns to the DOAS unit, the enthalpy wheel is activated at the maximum speed (e.g., 20 rpm) to obtain the maximum values of sensible and latent effectiveness. The enthalpy of the process air can be reduced as much as possible by exchanging the maximum amount of heat and moisture with the room air. This results in decreasing the cooling coil load of the DOAS. The enthalpy wheel should be inactivated to avoid an undesirable enthalpy increase within the process air as this could lead to

808

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

Fig. 4. Model space.

Table 1 Physical conditions of the model office space. Site Weather data Building type U-values [W/m2 K]

Internal heat gain Occupants Room set point

Seoul, South Korea TMY2 weather data Open-plan office (10 m  10 m  3 m) Roof 0.630 Wall 0.468 Floor 0.952 Windows (0.17 window/wall ratio) 5.6 Equipment 140 W per person (based on a PC with a monitor) Load 75 W (sensible), 75 W (latent) per person Summer Temperature 24 °C Relative humidity 50% Intermediate Temperature 22 °C Relative humidity 50% Winter Temperature 20 °C Relative humidity 50%

an increase in the system load of the cooling coil. Therefore, the enthalpy wheel operation should be halted when hRA is greater than hOA where the outdoor air humidity ratio (W OA ) exceeds the SA humidity ratio setpoint (W SA:set ) (i.e., in region 2). In region 3, when W oa is lower than W SA:set , the speed of the EW (Spd) is controlled to meet the target SA humidity ratio (i.e., W SA:set ) by recovering moisture from the exhaust air. The sensible and latent effectiveness affected by varying rotation speed of the EW are expressed by Eqs. (3)–(6) [18].

esen;max ¼ fðTSA;in ; £SA;in ; Q R Þ elat;max ¼ fðTSA;in ; TEA;in ; £SA;in ; £EA;in ; Q R ; VSA;in Þ Q 1 ðW RA  W OA Þ=ðT RA  T OA Þ DFR ¼ lat  2 ¼ 2430:6  Q sen £OA £2OA

Spd ¼ f ðDFR; elat Þ W  W OA elat ¼ SA:set  100 W RA  W OA esen ¼ 13:844  lnðSpdÞ þ 38:469

ð1Þ ð2Þ ð3Þ ð4Þ ð5Þ ð6Þ

3.2.2. Sensible wheel The sensible wheel effectiveness (Eq. (7)) varies with rotational speed. In this research, the rotational speed dependence of the sen-

sible wheel is assumed to be expressed by Eq. (6), and the maximum effectiveness is assumed 85% at full speed operation [19]. With a DOAS, the speed of the sensible wheel is modulated according to the amount of heat the SA needs to reclaim from the exhaust air to maintain the SA setpoint.

eSW ¼

T SA:set  T SW;in  100 T RA  T SW;in

ð7Þ

3.2.3. Ceiling radiant cooling panel (CRCP) The CRCP provides additional sensible cooling for the space to maintain the room temperature setpoint [12]. The CRCP is assumed to accommodate the remaining sensible load of the space (Eq. (8)).

Q CRCP ¼ Q sen;total  Q sen;DOAS

ð8Þ

3.2.4. Chiller The chiller model used in this simulation was the air-cooled chiller model of the DOE-2 (Eqs. (9)–(13)). The discharge temperature of the chilled water was set to 15 °C [27,28]. The chiller model consists of three performance curves that represent the capacity (CAPFT), full load efficiency (EIRFT), and partial load efficiency of a chiller (EIRFPLR). CAPFT and EIRFT are functions of outdoor air temperature (TOA) and chilled water supply temperature (Tcws). The EIRFPLR is determined by the part load ratio of the system. The chiller power can be calculated by using the following equations for the above three factors.

CAPFT ¼ f ðTcws ; TOA Þ

ð9Þ

EIRFT ¼ f ðTcws ; TOA Þ

ð10Þ

EIRFPLR ¼ f ðPLRÞ Q PLR ¼ Q ref  CAPFT

ð11Þ

P ¼ Pref  CAPFT  EIRFT  EIRFPLR

ð12Þ ð13Þ

3.3. Desiccant-enhanced evaporative air conditioner (DEVap) 3.3.1. Liquid-desiccant (LD) Based on the linear regression function drawn from the LD test data provided by Kozubal et al. [3], the outlet humidity ratio of the primary air was predicted using Eq. (14). The coefficients of the liq-

809

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

Fig. 5. Operation strategies for the enthalpy wheel on psychrometric chart.

Table 2 Coefficients of the liquid desiccant model. a

b

c

d

e

7.027490E3

1.302521E1

1.256766E2

5.393109E4

1.208599E2

Table 3 Variable parameters range of the liquid desiccant model.

Minimum Maximum

LG

LC [%]

TLD;in [°C]

WLD;in [kg/kg]

0.054091 0.223065

35.1 43.9

24.5 35

0.0128 0.0182

uid desiccant model (i.e., Eq. (14)) is given in Table 2. The variable parameters were inlet temperature (TLD;in ), humidity ratio (WLD;in ) of LD, liquid to gas ratio (LG), and liquid desiccant solution concentration (LC ). The valid range of each parameter is described in Table 3. For the primary air conditions not within the range of the regression model, the humidity ratio of the dehumidified process air was predicted by assuming a dehumidification effectiveness of 70% as can be determined by Eq. (15). The dehumidification effectiveness provides the ratio between the actual and potential moisture removal rates of the liquid desiccant system. The maximum dehumidification rate is calculated as the difference between the inlet primary air and equilibrium humidity ratio.

WLD;out¼ a þ bðLGÞ þ cðLC Þ þ dðTLD;in Þ þ eðWLD;in Þ WLD;out¼ W LD;in  ðW LD;in  Weq Þ  eLD

ð14Þ ð15Þ

In the simulation, the concentration of the LD solution is assumed to be maintained at 38% by the regenerator, and the regeneration rate is equal to the dehumidification rate. The regeneration heat is supplied by a gas-fired water heater. Additionally, the counter flow heat exchanger (SHX) transfers heat to the regenerator to increase the concentration of the weak liquid desiccant solution. In this case, the effectiveness of the water source heat exchanger is expressed by the e-NTU method [29]. The operation of the LD system was determined according to the humidity state of the mixed air. As shown in Fig. 6, the LD operates in regions 1 and 4, where the humidity ratio of the mixed air exceeds that of the supply air setpoint. The LD is deactivated in regions 2 and 3, where the latent cooling is not needed. 3.3.2. Indirect evaporative cooler The process air condition after the DP-IEC was predicted using the modified e-NTU method (Eqs. (16)–(25)) [16,30,31]. A simplified model of the DP-IEC was developed based on the experimental datasets of Lee [23]. The K value was adopted to simulate the wetside of the DP-IEC. As shown in Eq. (18), K, which represents the slope of the saturation line, is expressed as the ratio of the enthalpy difference to the wet-bulb temperature difference of the inlet and outlet secondary air. K is determined through iterative simulation using this model. To solve this iterative loop, the heat transfer coef-

810

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

Fig. 6. Regional operation modes of DEVap.

ficient of the primary air stream should be determined. However, the determination is complicated when attempting to express the heat transfer coefficient of partially turbulent fluids. To solve this problem, Liu [32,33] defined the overall heat transfer coefficient using the primary and secondary air mass flow rate parameters based on the findings of Pescod, as can be calculated using Eq. (20). Pescod demonstrated the heat transfer coefficient between a moving fluid and interfering wall can be expressed as n-th power of velocity [34]. The coefficients of n, F p , and F S are determined as 0.9262, 63.8777, and 3.4965 by regressing the experimental data as shown in Eq. (19). Another iteration loop was applied for predicting the outlet primary air temperature of the IEC. To predict the outlet primary air temperature with the modified e-NTU method, the wetbulb temperature of the inlet primary air was used as the initial value of the inlet secondary air temperature. Then, the equations were iteratively solved until the calculated supply air temperature converged to the value of the inlet secondary air temperature. The discrepancies between the value of the predicted outlet primary air temperature and the experimental data were less than 5%.

_ s K; Cc ¼ m

_ p cpa Ch ¼ m

C min ¼ minðC c ; C h Þ; K¼

hs;out  hs;in wb T wb s;out  T s;in

UA NTU ¼ C min

C max ¼ maxðC c ; C h Þ;

ð16Þ C min Cr ¼ C max

ð17Þ ð18Þ ð19Þ

1 l l ¼ þ UA F p mnp F S KmnS

eIEC ¼

1  expðNTUð1  C r ÞÞ 1  C r expðNTUð1  Cr ÞÞ

ð20Þ ð21Þ

Q max ¼ Cmin ðTp;in  Twb s;in Þ

ð22Þ

Q IEC ¼ eIEC Q max Q Tp;out ¼ Tp;in  IEC Ch Q IEC wb wb Ts;out ¼ Ts;in þ Cc

ð23Þ ð24Þ ð25Þ

The operation of the DP-IEC was determined according to the temperature state of the mixed air. For sensible cooling, the DPIEC activates in regions 1 and 2 of Fig. 6, where the mixed air temperature is higher than that of the supply air (15 °C). In this case, approximately 30% of the process air is utilized as secondary air to encourage evaporative cooling while the remainder of the process air is supplied to the space for air conditioning (i.e., the SA flow). The process air directed to the secondary channel of the IEC should be comprised of additional outdoor air flow introduced to the minimum outdoor air ventilation rate. Alternately, the DPIEC deactivates in regions 3 and 4 of Fig. 6, where the mixed air needs heating instead of cooling. 3.4. Fan and pump The design fan power for both the DEVap and DOAS was calculated using Eq. (26) with a fan efficiency of 0.6. The air-side pres-

811

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

sure drop in each component of the DOAS and the DEVap applied to the energy simulation are summarized in Table 4. In the simulation, both the DOAS and DEVap were assumed to be applied with variable volume supply and return fans. Although the DOAS is supplied with constant air volume in many field experiments, it was operated as a variable air flow unit to supply the required amount of ventilation air to meet changes in the occupancy schedule. The fan power at the variable flow rate was derived by Eq. (26) based on the generic fan power curve suggested by ASHRAE 90.1 [25].

Pfan ¼ ð0:0013 þ 0:1470  PLRfan þ 0:9506  PLR2fan  0:0998  PLR3fan ÞPfan;design

The pump power for both the DOAS and DEVap is estimated using Eq. (27). In a DOAS with a CRCP system, the constantspeed pump supplies the chilled water to the cooling coil of the DOAS unit and the CRCP. In a DEVap, a pump is used to supply chilled water required for the cooling coil in addition to pumps for the LD and IEC. The energy consumptions of the chilled water supply pumps used in the DOAS and the DEVap are determined according to the mass flow rate of the chilled water supplied to the cooling coils. The pump efficiency (gpump ) is simulated as 60%, and the water-side head loss is assumed as 20 m in both systems [27].

ð26Þ

Table 4 Fan pressure drop. DOAS/CRCP

DEVap

System component

Design DP

Component

Design DP

SA fan

Enthalpy wheel Sensible wheel Cooling coil Balance of system

120 Pa [35] 100 Pa [36] 100 Pa [37] 100 Pa

SA fan

LD + IEC Cooling coil Heating coil Balance of system

268 Pa [17] 100 Pa [37] 200 Pa [37] 200 Pa

RA fan

Enthalpy wheel Sensible wheel Balance of system

120 Pa [35] 100 Pa [36] 200 Pa

RA fan

Balance of system

200 Pa

LD fan IEC fan

LD IEC

115 Pa [17] 98 Pa [17]

Fig. 7. Daily performance of DEVap and DOAS during a peak summer day.

812

Ppump ¼

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

V_ fluid Dppump

gpump

ð27Þ

4. Simulation results The annual operating energy consumptions of the DOAS with CRCP and the DEVap systems were estimated by developing detailed hourly energy simulations. Energy performances of both systems for a peak summer day were also determined and compared with each other. 4.1. Cooling performance Fig. 7 shows the hourly SA temperature profile of the DEVap and DOAS on a summer peak day. In the DEVap system, the outdoor air initially mixed with the recirculated air is cooled by the IEC (Fig. 7a). The CC or HC at the terminal unit are activated to meet the SA temperature setpoint (TT ) after the IEC. During a peak summer day, additional cooling is required after the process air passes through DP-IEC to meet the target SA temperature (TT ). This is additional cooling is necessary since the sensible cooling of the DP-IEC is greatly affected by the humidity ratio of the inlet primary air. The inlet primary air of the DP-IEC, which indicated relatively a high wet-bulb temperature (i.e., 17 °C), is responsible for the insufficient cooling effect of the DP-IEC.

Alternately, Fig. 7b shows the SA temperature profile of the DOAS operating on a peak summer day. One can see that the introduced outdoor air is initially conditioned by the enthalpy wheel (TEW ), and then cooled more deeply by the CC (TCC ) for dehumidification. Unlike the CC in DEVap that was applied to provide additional cooling for meeting the target SA temperature TT , the CC in the DOAS was mainly operated for removing the moisture from the process air through deep cooling. When deep dehumidification was required in the DOAS, the temperature of the process air after the CC was lower than its setpoint temperature (TSA ). Consequently, the sensible wheel reheated the SA up to its setpoint temperature (TSA ) of 13 °C by reclaiming waste heat from the exhaust air. 4.2. Dehumidification performance Fig. 8 shows the dehumidification performances of the DOAS and DEVap systems on a summer peak day. For the DEVap, the LD is solely responsible for controlling the latent loads of the mixed air, which is a mixture of the recirculation air and the outdoor air. As can be seen in Fig. 8a, the LD contributed to dehumidifying the mixed air to the target supply air humidity ratio (WT Þ for most hours during the peak summer day. Alternately, for the DOAS, deep dehumidification of the introduced outdoor air should be performed using the enthalpy wheel and cooling coil (Fig. 8b) to accommodate the entire latent cooling

Fig. 8. Latent cooling loads of DOAS and DEVap.

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

813

load of the conditioned zone. Minimum ventilation outdoor air (i.e., process air) introduced into the DOAS unit exchanged moisture with the returned room air (Wew ). Additional dehumidification for the process air was completed using the cooling coil until the humidity ratio of the process air met the target setpoint (WT ).

the DEVap system due to its relatively low air flow rates compared to that of the DEVap. Consequently, a DOAS with a CRCP system consumed 26% less primary energy than the DEVap system during the cooling season. The primary energy consumptions are calculated by applying local primary energy conversion factors of 2.67 and 1.0 for electricity and natural gas [38,39].

4.3. Seasonal and annual energy consumption

4.3.2. Intermediate season (March, April, May, September, October, November) Fig. 10 compares the energy of the DOAS with CRCP and DEVap systems for an intermediate season. One can see that the DEVap consumed 38% more electrical energy than the DOAS with a CRCP system due to the electric heater required for reheating the SA and providing parallel heating for space. The DOAS with a CRCP system consumed 28% less electric energy than the DEVap system. The deactivation and bypass of the DP-IEC caused by relatively dry mixed air reduced fan energy consumption in the DEVap system. The DEVap system also showed less chiller energy consumption than the DOAS with a CRCP system. However, this advantage of the DEVap system was offset by heater and regeneration energy

4.3.1. Cooling season (June, July, and August) Energy consumption comparison of both systems for a cooling season is presented in Fig 9. The left-hand side scale of Fig. 9 represents the electric energy consumptions while the right-hand side scale represents the primary energy consumption. One can notice that the chiller consumes 77% of the total electrical energy required by the DOAS with a CRCP system. The DEVap consumed 71% less electricity when compared to the DOAS with CRCP, while the required gas consumption for regeneration of liquid desiccant solution offsets the advantage in electric energy consumption. Moreover, DOAS consumes approximately 32% less fan energy than

Fig. 9. Energy consumption of summer season.

Fig. 10. Energy consumption of intermediate season.

814

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

Fig. 11. Energy consumption of winter season.

Fig. 12. Seasonal primary energy consumption of the DOAS and DEVap.

consumption. Consequently, the DOAS with a CRCP system indicated 29% less primary energy consumption than the DEVap system during the intermediate season. 4.3.3. Heating season (December, January, February) The energy consumption comparison of the DOAS and DEVap systems during winter are presented in Fig. 11. The LD and DPIEC of the DEVap system are deactivated and bypassed, therefore the DEVap system operates as a conventional VAV system during the winter. Without the demand for dehumidification, electrical energy consumption is the main energy usage. In the DOAS, the enthalpy wheel and sensible wheel are operated to recover the latent and sensible heat from the exhaust air stream. One may notice that the heat recovery strategy of the DOAS unit contributed an 18% reduction in primary energy consumption compared with the DEVap system.

5. Conclusion In this research, seasonal energy performances of a DOAS with a CRCP system and DEVap system were analyzed based on results from detailed energy simulations. Despite significant differences in operation strategies, both systems properly utilized the system components to reduce the operating energy consumption. The DOAS took advantage of heat and moisture recovery to provide a significant energy saving potential for the 100% outdoor air system. The DEVap system was able to enhance the evaporative cooling performance by adopting LD and DP-IEC. The DEVap system, which is based on VAV application, was also able to improve the performance of sensible cooling by substituting conventional IEC that sends outdoor air into the secondary channels with DP-IEC. However, the annual energy simulation results disclosed that the DOAS operating in parallel with a CRCP system consumed

H.-J. Kim et al. / Applied Thermal Engineering 108 (2016) 804–815

approximately 22% less primary energy than the DEVap system. Specifically, the DOAS with a CRCP system was able to save 30%, 29%, and 23% of primary energy during the summer, intermediate and winter seasons, respectively, compared with the DEVap system, as shown in Fig. 12. The regeneration energy consumption during LD operation was the critical factor offsetting the energy savings potential of the DEVap system during summer season. However, the heater energy consumption of the DEVap system, which was approximately 90% and 95% of overall primary energy consumption during the intermediate and winter seasons respectively, was responsible for the increase in the amount of energy usage while that of regeneration was negligible owing to dry outdoor air conditions. These results clearly show that the DOAS, which introduced outdoor air with a minimum ventilation rate, was able to condition the space more effectively than the DEVap system. However, through the application of renewable energy sources such as solar thermal energy or geothermal energy to supply the regeneration heat of the LD system, the energy cost performance of the DEVap during the summer seasons can be improved significantly.

[14]

[15]

[16]

[17]

[18] [19] [20]

[21] [22]

Acknowledgements

[23]

This work was supported by a National Research Foundation (NRF) grant (No. 2015R1A2A1A05001726) and by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) and the Ministry of Trade, Industry & Energy (MOTIE) of the Republic of Korea (No. 20164010200860).

[24]

[25]

[26]

References [1] J. Ling, Y. Hwang, R. Radermacher, Desiccant-Assisted Cooling, 2014. [2] M.H. Kim, J.Y. Park, M.K. Sung, A.S. Choi, J.W. Jeong, Annual operating energy savings of liquid desiccant and evaporative-cooling-assisted 100% outdoor air system, Energy Build. 76 (2014) 538–550, http://dx.doi.org/10.1016/j. enbuild.2014.03.006. [3] J. Woods, E. Kozubal, A desiccant-enhanced evaporative air conditioner: Numerical model and experiments, Energy Convers. Manage. 65 (2013) 208– 220, http://dx.doi.org/10.1016/j.enconman.2012.08.007. [4] H.J. Kim, D.S. Yoon, J.W. Jeong, Regional optimum outdoor air system and energy consumption characteristics, Archit. Inst. Korea 31 (10) (2015) 211– 218. [5] S.A. Mumma, K.M. Shank, Achieving dry outside air in an energy-efficient manner, ASHRAE Trans. 107 (1) (2001) 553–561. [6] S.A. Mumma, Designing dedicated outdoor air systems, ASHRAE J. 43 (2001) 28–31. [7] T. Zakula, P.R. Armstrong, L. Norford, Advanced cooling technology with thermally activated building surfaces and model predictive control, Energy Build. 86 (2015) 640–650, http://dx.doi.org/10.1016/j.enbuild.2014.10.054. [8] M. Hallenbeck, An innovative dedicated outdoor air application for a government office tower, ASHRAE Trans. 118 (1) (2012) 285–293. [9] S.A. Mumma, Overview of integrating dedicated outdoor air systems with parallel terminal systems, ASHRAE Trans. 107 (1) (2001) 545–552. [10] C. Stetiu, Energy and peak power savings potential of radiant cooling systems in US commercial buildings, Energy Build. 30 (1999) 127–138. [11] J.W. Jeong, S.A. Mumma, Ceiling radiant cooling panels, ASHRAE J. 48 (2006). [12] C.L. Conroy, S.A. Mumma, Ceiling radiant cooling panels as a viable distributed parallel sensible cooling technology integrated with dedicated outdoor air systems, ASHRAE Trans. 107 (1) (2001) 578–585. [13] Y.J. Dai, R.Z. Wang, H.F. Zhang, J.D. Yu, Use of liquid desiccant cooling to improve the performance of vapor compression air conditioning, Appl. Therm.

[27] [28] [29] [30]

[31]

[32]

[33] [34] [35]

[36]

[37]

[38] [39]

815

Eng. 21 (2001) 1185–1202, http://dx.doi.org/10.1016/S1359-4311(01)000023. C.X. Jia, Y.J. Dai, J.Y. Wu, R.Z. Wang, Analysis on a hybrid desiccant airconditioning system, Appl. Therm. Eng. 26 (2006) 2393–2400, http://dx.doi. org/10.1016/j.applthermaleng.2006.02.016. M.H. Kim, J.Y. Park, J.S. Park, J.W. Jeong, Application of desiccant systems for improving the performance of an evaporative cooling-assisted 100% outdoor air system in hot and humid climates, J. Build. Perform. Simul. 8 (2014) 1–18, http://dx.doi.org/10.1080/19401493.2014.899395. S.W. Ham, S.J. Lee, J.W. Jeong, Operating energy savings in a liquid desiccant and dew point evaporative cooling-assisted 100% outdoor air system, Energy Build. 116 (2016) 535–552, http://dx.doi.org/10.1016/j.enbuild.2016.02.011. E. Kozubal, J. Woods, R. Judkoff, Development and analysis of desiccant enhanced evaporative air conditioner prototype development and analysis of desiccant enhanced evaporative air conditioner prototype, NREL Technical Report (NREL/TP-5500-54755) (2012). J.W. Jeong, S.A. Mumma, W.P. Bahnfleth, Energy conservation benefits of a dedicated outdoor air system, ASHRAE Trans. 109 (2) (2003) 627–636. M. Rabbia, G. Dowse, Energy recover ventilation-understanding energy wheels and energy recovery ventilation technology, 2000, pp. 15. J.W. Jeong, S.A. Mumma, Practical thermal performance correlations for molecular sieve and silica gel loaded enthalpy wheels, Appl. Therm. Eng. 25 (2005) 719–740, http://dx.doi.org/10.1016/j.applthermaleng.2004.07.018. A. Boranian, An investigation of optimal control of desiccant-enhanced evaporative air conditioning, Thesis, 2012, pp. 267. B. Riangvilaikul, S. Kumar, An experimental study of a novel dew point evaporative cooling system, Energy Build. 42 (2010) 637–644, http://dx.doi. org/10.1016/j.enbuild.2009.10.034. J. Lee, D.Y. Lee, Experimental study of a counter flow regenerative evaporative cooler with finned channels, Int. J. Heat Mass Transf. 65 (2013) 173–179, http://dx.doi.org/10.1016/j.ijheatmasstransfer.2013.05.069. J.W. Jeong, S.J. Lee, H.W. Dong, Energy performance of a desiccant-enhanced evaporative (DEVap) system, in: 2015 APCBE Conf., Dover, 2015, pp. 335–350, http://dx.doi.org/10.1017/CBO9781107415324.004. ASHRAE, ANSI/ASHRAE Standard 90.1-2013, Energy Standard for Buildings Except Low-Rise Residential Buildings, Refrigerating and Air-Conditioning Engineers, Inc., Atlanta, 2013. J.W. Jeong, S.a. Mumma, Binary enthalpy wheel humidification control in dedicated outdoor air systems, ASHRAE Trans. 113 (2) (2007) 220–226. U.D.O. Energy, EnergyPlus engineering reference: the reference to energyplus calculations, US Dep. Energy (2010) 1051. doi: citeulike-article-id:10579266. M. Hydeman, K.L. Gillespie, A.L. Dexter, Tools and techniques to calibrate electric chiller component models, ASHRAE Trans. 108 (1) (2002) 733–741. Y.A. Cengel, A.J. Ghajar, Heat and mass transfers: fundamentals and applications, 2007. Z. Liu, W. Allen, M. Modera, Simplified thermal modeling of indirect evaporative heat exchangers, HVAC&R Res. 19 (2013) 37–41, http://dx.doi. org/10.1080/10789669.2013.763653. A. Hasan, Going below the wet-bulb temperature by indirect evaporative cooling: analysis using a modified e-NTU method, Appl. Energy 89 (2012) 237– 245, http://dx.doi.org/10.1016/j.apenergy.2011.07.005. P. Taylor, Z. Liu, W. Allen, M. Modera, Simplified thermal modeling of indirect evaporative heat exchangers, HVAC&R Res. 19 (3) (2013) 37–41, http://dx.doi. org/10.1080/10789669.2013.763653. Z. Liu, Analysis of indirect evaporative heat exchanger modeling and experimentation, 2014. http://dx.doi.org/10.1007/s13398-014-0173-7.2. D. Pescod, Heat exchanger for energy saving in an air-conditioning plant, ASHRAE Trans. 85 (1979) 238–251. S. De Antonellis, M. Intini, C.M. Joppolo, F. Pedranzini, Experimental analysis and practical effectiveness correlations of enthalpy wheels, Energy Build. 84 (2014) 316–323, http://dx.doi.org/10.1016/j.enbuild.2014.08.001. S. De Antonellis, M. Intini, C.M. Joppolo, C. Leone, Design optimization of heat wheels for energy recovery in HVAC systems, Energies 7 (2014) 7348–7367, http://dx.doi.org/10.3390/en7117348. M. Hydeman, S. Taylor, J. Stein, E. Kolderup, T. Hong, Advanced Variable Air Volume System Design Guide Technical Report: P500-03-082-A-11, Taylor Engineering, Alameda, CA, 2003. Energy Star, Energy star performance ratings methodology for incorporating source energy use, Energy 12 (2009) 9. Ministry of Trade Industry and Energy. Energy caloric value conversion standard, Energy Act Article 5, 2011.