Modeling of hydronic radiant cooling of a thermally homeostatic building using a parametric cooling tower

Applied Energy 127 (2014) 172–181

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Modeling of hydronic radiant cooling of a thermally homeostatic building using a parametric cooling tower Peizheng Ma a,⇑, Lin-Shu Wang a, Nianhua Guo b a b

Department of Mechanical Engineering, Stony Brook University, Stony Brook, NY 11794, United States Department of Asian and Asian American Studies, Stony Brook University, Stony Brook, NY 11794, United States

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Investigated cooling of thermally

homeostatic buildings in 7 U.S. cities by modeling.  Natural energy is harnessed by cooling tower to extract heat for building cooling.  Systematically studied possibility and conditions of using cooling tower in buildings.  Diurnal ambient temperature amplitude is taken into account in cooling tower cooling.  Homeostatic building cooling is possible in locations with large ambient T amplitude.

a r t i c l e

i n f o

Article history: Received 23 October 2013 Received in revised form 24 February 2014 Accepted 10 April 2014 Available online 4 May 2014 Keywords: Heat extraction principle Building energy modeling Radiant cooling TABS Cooling tower Thermally homeostatic building

a b s t r a c t A case is made that while it is important to mitigate dissipative losses associated with heat dissipation and mechanical/electrical resistance for engineering efficiency gain, the ‘‘architect’’ of energy efficiency is the conception of best heat extraction frameworks—which determine the realm of possible efficiency. This precept is applied to building energy efficiency here. Following a proposed process assumptionbased design method, which was used for determining the required thermal qualities of building thermal autonomy, this paper continues this line of investigation and applies heat extraction approach investigating the extent of building partial homeostasis and the possibility of full homeostasis by using cooling tower in one summer in seven selected U.S. cities. Cooling tower heat extraction is applied parametrically to hydronically activated radiant-surfaces model-buildings. Instead of sizing equipment as a function of design peak hourly temperature as it is done in heat balance design-approach of selecting HVAC equipment, it is shown that the conditions of using cooling tower depend on both ‘‘design-peak’’ daily-mean temperature and the distribution of diurnal range in hourly temperature (i.e., diurnal temperature amplitude). Our study indicates that homeostatic building with natural cooling (by cooling tower alone) is possible only in locations of special meso-scale climatic condition such as Sacramento, CA. In other locations the use of cooling tower alone can only achieve homeostasis partially. Ó 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +1 6317458937; fax: +1 6316328544. E-mail addresses: [email protected] (P. Ma), lin-shu.wang@ stonybrook.edu (L.-S. Wang), [email protected] (N. Guo). http://dx.doi.org/10.1016/j.apenergy.2014.04.031 0306-2619/Ó 2014 Elsevier Ltd. All rights reserved.

P. Ma et al. / Applied Energy 127 (2014) 172–181

1. Introduction In a Foreword to Heating, Cooling, Lighting [1] the architect J.M. Fitch wrote, ‘‘The central paradox [challenge] of architecture [is] how to provide a stable, predetermined internal environment in an external environment that is in constant flux across time and space. . .’’ We want to argue that the changing external environment is both a challenge and an opportunity—a challenge in the architectural design of a building for achieving partial independence from the ambient environment and, at the same time, an opportunity in the engineering of the building for harnessing natural temporal and spatial energy gradient in its interaction with the ambient environment. One of the main challenges imposed by variable environment is the cooling and heating load-demand on a building. Heat Balance design sees a building in terms of its loads—which are functions of the specific indoor air temperature [2,3] and the weather ambient temperatures, with envelope heat transmission performance as the parameter. These loads are to be balanced with HVAC equipment. If the HVAC equipment is to guarantee thermal comfort at all time, it must be designed for peak weather ambient temperatures. Since any extreme weather condition is a statistical possibility, it would not be practical to aim for the most extreme but transitory condition (which would lead to over-sizing of equipment). The design of equipment is therefore based on fixed climatic design [peak] conditions, which for annual cooling according to the 2009 ASHRAE Handbook – Fundamentals [3] is the design condition for 0.4% or the design condition for 1% or the design condition for 2% in annual cumulative frequency of occurrence (exceeding the design condition). There are 365  24 h = 8760 h in one year. The 0.4%, 1%, and 2% design conditions are the three dry-bulb temperatures values that the instantaneous hourly temperature in the hottest months exceeded the corresponding value for a duration of 35 h (0.4% of 8760 h), 88 h (1%), or 175 h (2%) per year, respectively, for the period of record. For convenience the following discussion will be based on 1% dry-bulb temperatures or (Tout)design_1%. Notice that the climatic design conditions for conventional HVAC selection—design condition for 1%—are the peak values of hourly temperature Tout, not diurnal average temperature T out . (For simplicity, in this discussion the conventional load is said to be based on peak hourly value in the steady-state sense. Overlooked are the various refinements of the conventional load calculations by taking into consideration of the temporal thermal response of a building.) Let the peak-to-peak amplitude of the diurnal temperature variation be DTout, the peak hourly temperature Tout and the daily mean temperature T out are related as:

1 T out ¼ T out þ DT out 2

ð1Þ

The peak hourly temperature Tout, which is the determinant of the equipment capacity and the driver (causation) of the irreversible heat transfer process, depends on (and increases with) both the daily mean temperature and its amplitude: both are parts of the changing external environment: high value of T out and high value of DTout both contribute to the cooling load in the operation of a building. Rather than the heat balance design which sees a building in terms of its energy demand, a new two-step process assumptionbased design method [4–7] sees a building in terms of its autonomous and homeostatic existence. We use here the term autonomous in the narrow sense of a building’s ability of staying within a specific temperature range passively [5–7], but not at a given temperature level (homeostasis), which requires active control (the topic of this paper). By considering building energy as a building thermal processes problem, the central argument of this

173

paper is that buildings’ external environment represents both a challenge of maintaining autonomy [5,7] and a driving force for contributing to buildings’ homeostasis by the application of heat extraction principle. This paper is limited to only the cooling phase of homeostasis by heat extraction. Heat can be extracted by heat pumps, or cooling towers, or solar thermal panels. Investigation along the use of heat pumps (for cooling and heating) and solar thermal panels (for heating phase) will be conducted elsewhere. Instead, this paper addresses the use of cooling tower for extracting heat from indoor thermal mass for controlling indoor operative temperature level. A parametric cooling tower is added to the RC (resistor–capacitor) model built in Matlab and Simulink used in Refs. [4–7] to harness the natural energy gradient driven force for the buildingroom cooling. It will also give a discussion of the hypothetical design selection of cooling tower: instead of dependency on the single peak hourly temperature Tout, it will be shown that the design conditions are function of both the diurnal average temperature T out and the amplitude of the diurnal temperature variation DTout, with quite different functional relationships for the two. At last, how a building function under climatic conditions of different locations will be studied. This paper is a study following the development of Refs. [4–8], and represents the first paper on engineering for building homeostasis. A few words on thermodynamics of building systems and the importance of heat extraction for efficient operation of building systems are given below to put this paper’s content in the context of our overarching approach to building energy efficiency: This paper is one element in the goal of reducing waste heat from the cooling and heating operation (i.e., conditioning) of building systems, which is the sound scientific and engineering approach for successful building design for energy efficiency.

2. Thermodynamics of the operation of building systems Design and the refinement of heat engines and heat pumps have been based on the science of thermodynamics, which was founded by Carnot, critically developed by Thomson, and refined and made productive applications by generations of engineers including Otto and Diesel, Whittle and Von Ohain, Carrier, Keenan, etc. Heat engines and heat pumps are powered by fuel input (or heat input) for producing well-defined outputs, work or power in the former case, and, in the latter case, heat to be removed or heat to be delivered. Thermal efficiency of heat engines and coefficient of performance (COP) of heat pumps are properly defined. In contrast, whereas buildings or building systems are well defined, there is no ‘‘output’’ of a building: a building exists and we construct a building not for anything it produces but for its existence at certain indoor conditions we prefer. There is thus no yardstick for measuring the performance of a building in the same objective or scientific way we measure heat engines and heat pumps. We know how a perfect heat engine (even only as an abstraction) could perform and that is the way we measure objectively a real heat engine’s performance. We did not have an idea of how a building with perfect cooling and heating operation should perform—until today. Because we did not have a good answer to that question it is not surprising that building energy efficiency has not gained as it has been hoped for: ‘‘These 121 LEED buildings [with available energy performance data] consume more total energy per square foot than the average for the entire commercial building stock,’’ APS concluded in its 2008 study [9]; ‘‘Whereas the US has made significant progress in increasing efficiency and reducing energy use in the transportation and industrial sectors of the economy, both building sector energy use and building system energy use have shown only

174

P. Ma et al. / Applied Energy 127 (2014) 172–181

modest reductions. . .,’’ a Penn State University finding reported by ASME [10]. Below is a brief outline of a development in a few recent papers toward thermodynamics of the operation of building systems, which aims for providing a solid foundation for significant building operation energy efficiency gain. A 2013 paper [11] presented a new inference of the second law of thermodynamics that all reversible processes are equivalent to perfectly executed heat-extraction processes resulting in maximum extracted heat with zero waste heat (the definitions of maximum extracted heat, or maximum free heat, and waste heat are given there). The precise definition of waste heat provided the common denominator for objectively measuring the performance of heat engines and heat pumps as well as the performance of the operation of building systems: a perfect heat engine is hereby defined by zero waste heat and correspondingly a perfectly operated building system with a given envelope and internal-heat-gain (due to plug loads [receptacle loads and process loads], lighting, and inhabitants) can also be defined in terms of its reference waste heat, as described below. We treat both envelope heat transfer and plug loads and lighting dissipative heat as irreversibility that cannot be eliminated but can and should be minimized. The concept that there is irreversibility that cannot be eliminated and irreversibility that can be theoretically eliminated was introduced in a 2007 paper, [12] in which two fundamentally distinctive processes are defined: non-reversible processes and reversible-like processes. Whereas all processes are irreversible (a reversible process is a theoretical abstraction), non-reversible processes cannot be made anywhere close to reversibility but reversible-like processes can be made as close to reversibility as it is practically possible. Envelope heat transfer and internal-heat-gain cannot be avoided or eliminated and their irreversibility should be minimized by measures of energy conservation. In contrast, a building’s operation or conditioning can be designed for taking advantage of heat extraction processes (i.e., reversible-like processes) and their irreversibility can be reduced without limit theoretically. The limit of perfectly executed heat extraction was introduced in a recent paper, [8] in which the minimum reference waste heat resulting from a building’s existence is defined in terms of its envelope R-value and internal-heat-gain by using the thought experiment of perfectly executed heat extraction—and, today, provides the reference for perfectly conditioned buildings. What was shown in [8] is that irreversibility resulting from a conventionally heated building is one order of magnitude greater than the irreversibility of envelope heat transfer. By reinventing building conditioning, there is large room for reducing operation irreversibility beyond irreversibility minimization by energy conservation alone. That conclusion may explain why there has not been more significant building energy efficiency gain [9,10] even with widely adopted energy conservation measures by the whole built environment community. Despite the characterization by Ex-Energy Secretary Steven Chu that building energy efficiency is a lowhanging fruit, there is likely no easy solution. This paper is one element of our overarching proposed solution for building operation energy efficiency, which is, though not an easy solution, but one that promises real efficiency gain. A 2012 National Grid commissioned study on the future of heating concludes, ‘‘Electrification of heat in buildings, facilitated primarily by heat pumps, is a critical component of decarbonizing heat and meeting the 2050 [GHG] target.’’ [13] We agree with this recommendation but argue that heat pumps are only one example of heat extraction. A list of heat extraction devices comprises the following: 1. Heat pumps. 2. Hydronically activated radiant systems: – Cooling tower.

– Solar thermal panels. – Thermal energy storage (TES) tanks. – Hydronically activated radiant surfaces (ESS, TABS, and radiant panels). 3. Combined heat and power (CHP) plants. The full impact of electrically powered heat extraction paradigm will be realized only if heat extraction is understood in its general thermodynamic meaning. 3. Extraction of indoor heat from hydronically activated building with cooling tower The process assumption-based design conception of homeostatic buildings is a two-step design process. The first step is the architectural design of a building thermal system with adequate thermal mass and satisfying the requirements of minimum envelope thermal resistance and maximum window-to-wall ratio, so that the building maintain its acceptable operative temperature range with no HVAC equipment at all. The second step is the engineering selection of equipment, which can be either mechanical ones (including heat pump) or low power equipment (lowPE) ones, for maintaining acceptable operative temperature level of a building subject to given external environments. This paper focuses on the use of lowPE for cooling application only. A lowPE is by definition driven by natural energy spatial and temporal gradient—while assisted by auxiliary units of moderate power demand. Unlike mechanical equipment which are typically powered directly or indirectly by fossil fuel (i.e., until the electricity system itself becomes decarbonized [13]), lowPE are driven by natural energies (i.e., natural heat and coolness), which are low-T heat and high-(or moderate-)T coolness. Examples of natural energies are: nighttime air’s coolness in summer, an example of natural temporal gradient; solar heating in winter and sunlight in summer (as source of power as well as heat source for cooling devices), examples of natural temporal and spatial gradient; heat and coolness of water bodies and geothermal sources, examples of spatial gradient. Examples of low-power equipment are cooling tower, active solar thermal (AST) systems and solar PV systems (as well as solar powered cooling systems), water-source heat pump and ground-source heat pump, and possibly others. In this paper a hypothetical study of the design selection of cooling tower as an example of lowPE is given. A building with lowPE is driven largely by natural energy gradient and requires little mechanical power for auxiliary systems, i.e., it is largely autonomous from man-made mechanical power. The autonomous building is constructed of thermally activated building slabs (TABS), and equipped with an outdoor cooling tower for harnessing the natural energy gradient driven force in the nighttime, as shown in Fig. 1. Water pump is on when the nighttime ambient temperature Tout is lower than the water temperature at TABS outlet Trw, which is also the water temperature at cooling tower inlet. In a wet cooling tower, if the climate is not too humid, the supply water temperature Tsw can be brought to be near the dew point temperature of outdoor air. TABS was originally established in the 1990s by a Swiss engineer, Meierhans [14,15]. In TABS, water carrying (PEX) pipes are embedded in building structure (ceiling, floor or wall concrete slabs). The whole slab (with its large thermal mass), rather than the thin surface layer [16,17], is activated because of the water pipes placed in the middle of the slabs. As a result, a building’s operative temperature Top remains within the thermal comfort zone in the daytime, and in the nighttime the water is cooled by the cold ambient air. In this paper, the building system is modeled by a resistor– capacitor (RC) model in Matlab/Simulink following Refs. [4–7].

P. Ma et al. / Applied Energy 127 (2014) 172–181

175

Fig. 2. Southwest corner view of the investigated building.

Fig. 1. Thermal homeostatic building cooling: nighttime operation.

The RC model on building energy simulations is well validated in a large volume of literature [18–24]. There are several reasons of using the RC model rather than other computer-aided modeling software programs (such as TRNSYS, EnergyPlus, ESP-r, Trane TRACE, and eQuest): Weber et al. [19] noted, ‘‘It is therefore of major interest to find simplified models that are limited in their complexity but still sufficiently accurate. One such type of model is the RC-network.’’ In another research paper, Široky et al. [25] commented, ‘‘Largely used computer aided modeling tools are not considered here, as they result in complex models which cannot be readily used for control purposes. . . . Alternatively to the statistical approach, especially if there is a lack of data or some knowledge of building physics is present, the RC modeling can be used.’’ That is, whereas computer-aided modeling is useful for taking into consideration of detailed and specified design features of individual buildings when their major design decisions on the system level are made on sound building physics, building physics propositions can be advantageously generated by the RC simulation modeling—like the classical experimental tools—that can generate dependent outcomes over a range of experimental parameters (subject to constant control variables). The configuration of the investigated office building [26] is chosen according to typical characteristics of TABS-equipped office buildings: south and north orientations with normal offices along the main façades and corner offices with glazing on two sides at the front faces. It has 5 stories and each story has 12 normal offices and 4 corner offices. With a net area of 36 m2 per room, the total floor area is 2880 m2. A southwest corner view of the investigated building is shown in Fig. 2. All ceilings and floors of the building are constructed of concrete slabs with PEX pipes inside. A south-west corner room is investigated in this paper. The south- and west-facing walls are exposed to the outdoor environment in the south–west corner room. The external walls

consist of 10 cm-thick normal-weight concrete and insulation outside of the concrete. The normal-weight concrete slabs in the ceiling and floor are 25 cm thick. The internal walls are made of 20 cm-thick structural light-weight concrete. Other interior thermal mass are modeled as wood with dimensions of 6 m  6 m  0.1 m. The room has a large window-to-wall ratio (WWR) of 42% and all the windows are double glazing with lowE coating, whose thermal conductivity is 1.887 W/m2 K. All the components of the building envelope meet the requirements in the ASHRAE Standard 90.1-2010 [27]. The total air change rate of infiltration and ventilation for the room is about 0.7 ACH, which meets the requirement for indoor air quality. The heat gains of the room include internal heat gains qi00 and solar energy input qs00 . Internal heat gains include heat dissipation from lighting, persons, and equipment. The solar input is calculated from the solar geometry of Zürich, assuming that 8% of which goes through the windows because of the good shading devices. The water in the PEX pipes flows through a cooling tower in summer nighttime. ‘‘A Cooling Tower is a heat rejection device that extracts [dissipates] waste heat to the atmosphere by cooling [extracting heat from] a stream of hot water in the tower.’’ [28] There are two basic types of commonly used cooling tower: wet and dry cooling towers. In a dry cooling tower the water and air are separated, and in a wet cooling tower, the hot process water is in direct contact with the cooler air. A large volume of literature of the cooling tower application in buildings exists [29–44]. In a dry cooling tower, the heat exchange in an air–water heat exchanger is [45]:

_ w ðT w1  T w2 Þ Q water ¼ C w m

ð2Þ

  Q air ¼ a T w  T A

ð3Þ

where Qwater and Qair are the heat released at the water side and absorbed at the air side of the heat exchangers, respectively; Cw _ w (kg/s) is the water mass flow (J/kg K) is the water specific heat; m rate; Tw1 and Tw2 are the water inlet and outlet temperatures in the heat exchanger; a (W/m2 K) is the heat transfer coefficient and is related to the airflow velocity through the heat exchanger; T w is the mean water temperature in the heat exchanger; T is the outside air temperature, A (m2) is the surface area of the heat exchanger. Let us simplify the heat exchange process as follows:

Q water ¼ Q air

ð4Þ

and

Tw ¼

T w1 þ T w2 2

ð5Þ

then

 T w2 ¼ 1 

 2aA 2aA T þ T aA þ 2C w m_ w w1 aA þ 2C w m_ w

ð6Þ

176

P. Ma et al. / Applied Energy 127 (2014) 172–181

In a [wet, counter-current flow] cooling tower according to EnergyPlus [46], the cooling tower effectiveness, or the cooling tower thermal efficiency, can be defined by analogy to the effectiveness of a simple heat exchanger, if assuming that the heat _ p) for the cooling tower water is less than that capacity rate (mc for the air:

e¼

T win  T wout T win  T wbin

ð7Þ

thus

T wout ¼ ð1  eÞT win þ eT wbin

ð8Þ

where Twin is the inlet water temperature, Twout is the outlet water temperature, Twbin is the wet-bulb temperature of the inlet air, and e is the heat exchanger effectiveness:

h i 1  exp NTUð1  C_ w =C_ a Þ h i e¼ 1  ðC_ w =C_ a Þ exp NTUð1  C_ w =C_ a Þ

ð9Þ

where

_ w cpw and C_ a ¼ m _ a c pe C_ w ¼ m

ð10Þ

and NTU is the Number of Transfer Units

NTU ¼ UAe =C_ w

ð11Þ

_ w (kg/s) is the mass flow rate of water, In the equations above, m _ a (kg/s) is the mass flow rate of air, cpw (J/kg K) is the specific heat m of water, cpe is the mean specific heat of the moist air treated as an equivalent ideal gas, U (W/m2 K) is the cooling tower overall heat transfer coefficient, and Ae (m2) is the equivalent heat transfer surface area:

Ae ¼ Acpe =cp

ð12Þ

2

where A (m ) is the heat transfer surface area, and cp (J/kg K) is the specific heat of moist air. It is easy to find that Eqs. (6) and (8) have the same form, if in Eq. (6) let

e¼

2aA aA þ 2C w m_ w

ð13Þ

The cooling tower effectiveness, or thermal efficiency, e is between 0 and 1. Ref. [32] gave an example of the efficiency as a function of air and water flow rates, as shown in Fig. 3. In the TABS-equipped room, the mass flow rate of the water in the PEX pipes is 0.12 kg/s. Thus according to Fig. 3, the thermal efficiency may be higher.

Fig. 3. Tower efficiency as a function of air and water flow rate (mspray = 1.4 kg/s, Twb = 16 °C).

At the room side, according to Chapter 4 in Ref. [47], the calculation equations are:

_ w cw ðT sw  T rw Þ ðT sw  T c Þ=Rt ¼ q00room ¼ m

ð14Þ

_ w cw Þ T rw ¼ T sw  q00room =ðm

ð15Þ

where Tsw is the TABS supply-water temperature, Tc is the TABS slab’s core temperature, Rt is the TABS-resistance, q00room (W) is the heat flux from the TABS to the room (negative in the cooling case), _ w (kg/s) is the mass flow rate of water in the TABS pipes, cw m (J/kg K) is the specific heat of water and Trw is the water temperature at the TABS outlet. From the application point of view, ‘‘a wet cooling tower is more effective than a dry tower, because it relies mainly on the wet-bulb temperature of the air and not on the generally higher dry-bulb temperature, as in the case of a dry-cooling tower.’’ [48] ‘‘The ambient wet-bulb temperature of the entering air . . . is typically 10 °F [5.56 K]–30 °F [16.67 K] lower than the dry-bulb temperature, depending on the local climate.’’ [49] In EnergyPlus, cooling tower has two models. In one model, ‘‘the nominal capacity is specified for the standard conditions i.e. entering water at 35 °C (95 °F), leaving water at 29.44 °C (85 °F), entering air at 25.56 °C (78 °F) wet-bulb temperature and 35 °C (95 °F) dry-bulb temperature.’’ [46] Here the dry-bulb and wet-bulb temperature difference is 9.44 °C (17 °F). Two important parameters determine the wet cooling tower performance: range and approach, as shown in Fig. 4 [50]. Range is defined as ‘‘the difference between the cooling tower water inlet and outlet temperature.’’ [50] A high range means that the cooling tower can reduce the water temperature effectively. Approach is ‘‘the difference between the cooling tower outlet cold-water temperature and ambient wet bulb temperature.’’ [50] Lower approach means better cooling tower performance. Although both of them should be monitored, the approach is a better indicator of cooling tower performance [50]. ‘‘As a general rule, the closer the approach to the wet bulb, the more expensive the cooling tower due to increased size. Usually a 2.8 °C approach to the design wet bulb is the coldest water temperature that cooling tower manufacturers will guarantee.’’ [50] This paper only considers cooling tower as one example of lowpower equipment on the feasibility of using it for harnessing natural energy, not focuses on the design of the cooling tower itself. Therefore, a parametric cooling tower is added to the RC model of the room. In Section 5, a diurnal-mean value (design condition for 1%) of the outdoor air dry-bulb temperature T out , which is 27.10 °C, is selected in summer time to investigate the building

Fig. 4. Range and approach of cooling towers.

P. Ma et al. / Applied Energy 127 (2014) 172–181

cooling operation. As in the ‘‘Nominal Capacity’’ model adopted in EnergyPlus, the dry-bulb and wet-bulb temperature difference is assumed to be 9.44 °C, and thus the mean value of the outdoor air wet-bulb temperature is 17.66 °C. If the guaranteed approach of the cooling tower is 2.8 °C, the mean value of the TABS supply-water temperature should be 20.46 °C. Suppose that the cooling tower works from 8:00 PM to 4:00 AM the next day, and in this duration the outdoor air dry-bulb temperature is lower than its mean value. Therefore, a parametric cooling tower can contribute to the cooling of the building-room with naturally cool ambient nighttime air. We investigate in the following two sections the extent of this contribution. In the RC model, the heat flux from the TABS to the room q00room and the TABS supply-water temperature Tsw are measured; the TABS returned-water temperature Trw can be calculated with Eq. (15) as the mass flow rate of water in the TABS pipes is a known constant; Eq. (8) gives the theoretical outlet water temperature of the cooling tower Twout; a switch controls the operation of the cooling tower; between 8:00 PM and 4:00 AM the next day, the cooling tower is turned on by closing the switch and Tsw is equal to Twout; from 4:00 AM to 8:00 PM, the cooling tower is turned off by opening the switch and temperature of the water in the TABS floats freely. By varying the cooling tower effectiveness, Twout is changed and then Tsw is controlled. The model is used to investigate the cooling operations in Sections 4 and 5. It is important to point out that the comfort of the indoor environment also requires the air humidity in a certain range, which is usually controlled by the dedicated outdoor air systems (DOAS) in TABS-equipped buildings. TABS with DOAS for air-quality & moisture control is widely viewed to be a best path toward carbon neutrality in building sector [51]. Moisture controlling is not a topic in this paper, which may be a valuable research topic in the future. The radiant cooling system can only handle the sensible heat gains, and for simplicity in the RC model the air humidity and the moisture gains resulting from the ventilation and infiltration air are not taken into account, which may bring minor quantitative error into the model.   4. Hypothetical example of the maximum T out design by using cooling tower at available DTout There is considerable variation in the diurnal temperature amplitude DTout from one climatic region to another one. This diurnal temperature variation can be the driving force for building cooling conditioning: although large diurnal variation causes larger operative temperature variation, the same large diurnal temperature variation serves to maintain the proper operative temperature level at higher summer mean ambient temperature using cooling tower. The simulation condition and the parameter selection of the autonomous building-room are: (1) the mean value of the operative temperature is kept at 25.25 °C for best thermal comfort in summer [52]; (2) the approach of the cooling tower is not smaller than 2.8 K; (3) the ambient temperature amplitudes DTout are from 2 K to 34 K with a 4 K step; (4) three window-to-wall ratios (WWR): 20%, 42% and 60%; (5) five U-values (W/m2 K) of the external walls as selected in Ref. [7]: 3.293, 1.998, 0.857, 0.365 and 0.203. These five U-values cover all eight climate zones of North America. Notice that Zones 1 and 2 mean very hot and hot climates, respectively, such as Hawaii and Florida; Zones 7 and 8 mean very cold and subarctic climates, respectively; all of Alaska is in Zone 7 except for several boroughs in Zone 8. Therefore, U = 3.293 required in Zones 1 and 2 is really a high value and U = 0.203 required in Zones 7 and 8 is definitely a low value. Systematically simulation data are selectively presented in Table 1 (some data are omitted because they share the same pattern of linear variation with DTout, see below).

177

Notice that we keep the minimum cooling tower approach at 2.8 K. Under this circumstance, we can say that the cooling tower has the maximum effectiveness since higher effectiveness will cause the approach to fall below 2.8 K. In Ref. [7], the maximum window-to-wall ratio of a thermally autonomous building as a function of envelope U-value and ambient temperature amplitude was investigated under the constraint of DTop no bigger than 2 K. According to the result shown in Fig. 4 in Ref. [7], the cases meeting the constraint are highlighted in bold and the cases not meeting the constraint are highlighted in italics. Data in the table are further discussed in the following. It will be noted that the strict constraint of DTop being no bigger than 2 K will be relaxed in the general case of this paper. Now let us focus on the cases when the U-value of the external walls Uew is 0.365 W/m2 K, which is a most commonly mandated U-value in ASHRAE Standard 90.1 [27,53,54]: Maximum restriction of nonresidential mass walls (Zone 1 in ASHRAE Standard 90.12004); Maximum restriction of nonresidential steel-framed walls (Zones 7 and 8 in ASHRAE Standard 90.1-2004; Zones 4–8 in ASHRAE Standard 90.1-2007 and 2010); Maximum restriction of nonresidential wood-framed and other walls (Zone 5 in ASHRAE Standard 90.1-2007 and 2010); Maximum restriction of residential steel-framed walls (Zones 4–7 in ASHRAE Standard 90.1-2004; Zones 2–6 in ASHRAE Standard 90.1-2007 and 2010); Maximum restriction of residential wood-framed and other walls (Zone 6 in ASHRAE Standard 90.1-2004; Zone 4 in ASHRAE Standard 90.12007 and 2010). The corresponding simulation results are shown in Fig. 5(a–c). Against the ambient temperature amplitude DTout, the following variables are shown: 5(a) the maximum cooling tower effectiveness e (with minimum approach = 2.8 K); 5(b) the maximum mean ambient temperature that the cooling tower can maintain the most comfortable indoor temperature (mean DTop = 25.25 °C). These two figures should be discussed together. With the increase of DTout, both the maximum e and the maximum mean Tout increase while the WWR is constant, which means at a location with bigger DTout, a cooling tower can work more effectively: the cooling tower has a slightly higher thermal efficiency and the ambient temperature can be much higher. Taking the cases when WWR = 42% for example, if DTout increases from 2 K to 34 K, the maximum e changes from 0.463 to 0.479 and the maximum mean Tout increase from 25.67 °C to 32.03 °C. Calculating with Eq. (1), the peak hourly temperature can be raised from 26.67 °C to 49.03 °C, which is a 22.36 K difference! Obviously, locations with larger DTout are favorable for building cooling by using cooling towers. The WWR also has influence on the maximum e and the maximum mean Tout. Clearly, with the increase of the glazing area (larger WWR), the corresponding maximum mean Tout decreases even though the maximum e increases, which mean that it is harder for the cooling tower to maintain the desired Top level. The reason is obvious: larger glazing area means bigger heat exchange and more solar energy gain through the glazing, and thus the internal thermal environment is harder to be kept stable due to greater influence by the external surrounding. However, the influence of WWR is limited: when DTout is constant, the decrease of the mean Tout (and the peak hourly temperature) is only about 2 K when the WWR changes from 20% to 60%. In previous research [4,5,7] of the first architectural step, we were concerned with the indoor operative temperature variation DTop only. In this second mechanical-engineering step, we are focusing on the Top level. However, the Top variation (range) should still be monitored as shown in Fig. 5(c). In Ref. [7], we kept DTop at or below 2 K strictly. After adding the mechanical equipment in this paper, this restriction may be relaxed a little bit. In the typical case when Uew = 0.365 W/m2 K, WWR = 42% and DTout = 18 K, DTop increases from 2.0 K to 2.5 K after adding the cooling tower, which

178

P. Ma et al. / Applied Energy 127 (2014) 172–181

Table 1 Maximum mean ambient temperature, operative temperature variation, and cooling tower effectiveness, approach and range under different exterior wall U-values, ambient temperature amplitudes, and window-to-wall ratios when the cooling tower minimum approach is 2.8 K and the operative temperature level is 25.25 °C. Uew (W/m2 K)

DTout (K)

WWR = 20%

WWR = 42%

WWR = 60%

Max e

Mean Tout (°C)

DTop (K)

Max e

Mean Tout (°C)

DTop (K)

Max e

Mean Tout (°C)

DTop (K)

3.293

2 34

0.440 0.510

26.25 31.17

1.30 3.96

0.465 0.521

25.59 30.65

1.69 4.44

0.486 0.531

25.02 30.21

2.03 4.89

1.998

2 34

0.435 0.491

26.36 31.79

1.26 3.36

0.465 0.510

25.62 31.07

1.66 4.00

0.486 0.524

25.00 30.47

2.10 4.58

0.857

2 34

0.428 0.459

26.53 32.74

1.21 2.80

0.463 0.492

25.65 31.66

1.62 3.59

0.487 0.514

24.98 30.82

1.99 4.29

0.365

2 6 10 14 18 22 26 30 34

0.423 0.425 0.427 0.428 0.430 0.431 0.433 0.434 0.435

26.65 27.48 28.32 29.16 30.01 30.84 31.69 32.54 33.38

1.19 1.36 1.52 1.69 1.86 2.03 2.20 2.37 2.54

0.463 0.465 0.467 0.470 0.472 0.474 0.476 0.477 0.479

25.67 26.46 27.24 28.04 28.84 29.63 30.43 31.23 32.03

1.61 1.83 2.05 2.27 2.49 2.72 2.94 3.17 3.40

0.488 0.490 0.493 0.496 0.498 0.501 0.503 0.505 0.508

24.96 25.71 26.46 27.22 27.98 28.74 29.50 30.27 31.04

1.98 2.24 2.51 2.78 3.05 3.33 3.60 3.88 4.16

0.203

2 34

0.421 0.425

26.70 33.65

1.18 2.46

0.462 0.474

25.67 32.18

1.61 3.34

0.488 0.505

24.96 31.12

1.98 4.12

Fig. 5. Maximum cooling tower effectiveness, maximum mean ambient temperature and operative temperature variation when the cooling tower minimum approach is 2.8 K, the operative temperature level is 25.25 °C, and the external wall U-value is 0.365 W/m2 K.

is still acceptable. In fact, indoor temperature drifting of up to a rate of 4 K/h (7.2 °F/h) is found to be acceptable for most people [55–57]. Therefore, with the strict restriction in the architectural step, the designed building is guaranteed to have a comfortable Top range, except for extreme cases such as WWR greater than 60% or DTout greater than 34 K. Systematically simulation data show that when the external wall U-value changes, the variation patterns are almost the same (nearly linear with DTout), as shown in Fig. 6. With the decrease of the external wall U-value (that is, bigger thermal resistance), the corresponding mean Tout increase a little bit and the corresponding DTop decrease, which mean that for the cooling tower it is easier to maintain the desired Top level and range. The reason is also obvious: smaller U-value means greater heat exchange resistance through the external walls and cooling tower of the same performance is able to balance the heat transfer gain subject to higher Tout; higher resistance wall is also better able to isolate the ambient influence keeping DTop smaller. 5. The possibility of homeostatic building cooling by a cooling tower with minimum approach of 2.8 K 5.1. A building in locations with different micro-climate statistics In a previous paper [7], seven U.S. cities were selected: Fullerton, CA; Sacramento, CA; Wilmington, DE; Atlanta, GA; Springfield, IL; Valentine, NE; and Albuquerque, NM. The reason for their selection is that the seven cities’ design 1% daily-mean dry-bulb

temperatures may be taken to be the same 300.25 K (27.10 °C or 80.78 °F) as the design condition for 1% (Tout)design_1%—but with individual sets for each cities of their daily mean ambient temperatures and daily ambient temperature amplitudes (see [7]). The ambient temperature amplitudes in the seven cities are found to be (see Fig. 7 in Ref. [7]): (a) most amplitudes are smaller than 20 °C, except a small part in Valentine and a moderate part in Sacramento; (b) in Sacramento, majority of summer days has high amplitudes between 15 °C and 21 °C; (c) in other cities, amplitudes are mainly in the middle- or small-amplitude part. The mean ambient temperature amplitudes are between 9.5 K (Wilmington) and 18.1 K (Sacramento). The seven cities well represent a wide geographic range of the United States. The seven cities are distributed in three different climate zones: Zone 3 (Fullerton, Sacramento and Atlanta), Zone 4 (Wilmington) and Zone 5 (Springfield, Valentine and Albuquerque). Using the real-time hour-by-hour dry-bulb temperatures of the cities requested by email [7] from the DOE website [58], Fig. 7 shows the daily distribution of dry-bulb temperature mean values vs. amplitudes of the selected cities in the summer of 2007 (from June 1st to September 30th, total 122 days). In Fig. 7, the assumed maximum cooling tower effectiveness is used, that is, the minimum approach is 2.8 K. Each blue dot is one summer day and its position in the plot corresponds to its daily mean temperature vs. temperature amplitude. In each location, the red line separate days that cooling tower with the maximum effectiveness can maintain a 25.25 °C operative temperature level from days that cooling tower cannot: in days below the red line, cooling tower can maintain the

179

P. Ma et al. / Applied Energy 127 (2014) 172–181

Fig. 6. Maximum cooling tower effectiveness, maximum mean ambient temperature and operative temperature variation when the cooling tower minimum approach is 2.8 K, the operative temperature level is 25.25 °C, and the window-to-wall ratio is 42%.

Fig. 7. Daily distribution of dry-bulb temperature mean values vs. amplitudes of the selected seven US cities in the summer of 2007.

Table 2 Days that cooling tower cannot meet the cooling requirement in 122 summer days of 2007 and effective (i.e., higher) operative temperatures with zero failure day. Location

Fullerton, CA Sacramento, CA Wilmington, DE Atlanta, GA Springfield, IL Valentine, NE Albuquerque, NM

Hottest month mean DTout in ASHRAE Handbook

CT cannot meet cooling requirement

Mean Top increases 1K

2K

3K

K (°F)

Day

Percentage (%)

Day

Day

Day

°C (°F)

10.6 (19.0) 18.1 (32.6) 9.5 (17.1) 11.1 (19.9) 10.7 (19.3) 14.7 (26.4) 13.6 (24.4)

10 4 18 36 15 5 13

8.2 3.3 14.8 29.5 12.3 4.1 10.7

5 3 7 15 6 3 4

4 0 2 7 2 1 0

2 0 1 2 0 0 0

30.17 26.72 29.70 29.61 27.91 27.51 26.91

optimal operative temperature; in days above the red lines, cooling tower cannot maintain the optimal operative temperature due to high mean ambient temperature or small ambient temperature amplitude, or both. Notice that the red line is ad hoc corresponding to the investigated building with external wall U-value of 0.365 W/ m2 K and WWR of 42%. For buildings with different external wall U-value and/or WWR, the red line may be different. Notice that the red lines in Fig. 7 are the same as the red lines in Figs. 5(b) and 6(b). For buildings with different WWR or external wall Uvalue, the red lines can just be replaced with other lines in Figs. 5(b) and 6(b). In Table 2, days and percentages that cooling tower cannot maintain the optimal operative temperature are given in the third

Mean Top with zero failure day

(86.31) (80.10) (85.46) (85.30) (82.24) (81.52) (80.44)

column: in Sacramento, only 4 days (3.3%) that the cooling tower does not work well enough to maintain a comfortable indoor environment in the whole four-month summer period; but in Atlanta, there are 36 failure days (29.5%). In the fourth column, after the assumed mean operative temperature is increased by 1–3 K, the failure days drop continuously. The last column gives the ‘‘would be’’ effective mean operatives temperatures if there is zero failure day in the 122 summer days. The necessary increase in effective mean operative temperature is 2.56 K (4.61 °F) in the case of Atlanta. These effective mean operative temperatures in the column are plotted by the red squares on the 40% relative humidity line in Fig. 8. From the figure we can find that in five of the seven cities, the effective mean operative temperatures in these cities are

180

P. Ma et al. / Applied Energy 127 (2014) 172–181

outside the summer comfort zone, i.e., thermal comfort has to be compromised during extremely hot days using cooling tower alone. The most striking result is that even though Sacramento has the highest peak (design 1%) hourly temperature of 309.3 K (see Table 1 of Ref. [7]), it is the easiest location for using cooling tower to maintain its operative temperature level with 96.7% of the summer days at optimal thermal comfort. The reason is that Sacramento has the ‘‘gift’’ of large natural energy gradient (see Section 5.2) to support homeostatic buildings. In other locations either the thermal comfort has to be partially compromised (i.e., higher indoor operative temperature in extreme hot days), or equipment other than lowPE (e.g., such as heat pump) needs to be used for assisting the cooling tower for achieving full homeostasis in summer.

5.2. Most favorable locations and meso-scale climatic process It is interesting to note that the meso-scale climatic process that creates the condition in Sacramento is that the summer day time high temperature air at Sacramento rises to pull in the sea breeze through Golden Gate strait, and this breeze travels towards Sacramento and by the night time it reaches Sacramento resulting in cool nighttime air there. It is the combined result—of solar heating of air generating thermal convection, ocean’s moderating temperature effect, and the right distance of the location from ocean for the convective ‘‘cool front’’ to arrive at Sacramento in the nighttime—that creates this natural energy gradient gift. Other locations, especially prevalent in the best wine countries, with the same combination are the upper Napa valley and Paso Robles at the upstream of Salina valley. These are the most favorable locations—created by sun, air, ocean, and convective sea breeze—for experimental testing of homeostatic (radiantly cooled) buildings with minimal heat pump assistance required.

Fig. 8. Mean operative temperatures of the selected seven US cities based on zero failure days on the 40% relative humidity line in the ASHRAE comfort zone chart.

The big two of renewable energies are wind and the sun. We prefer to call renewable energies as examples of ‘‘natural energy gradient,’’ because the term includes all irreversibility resources beyond those that can be harnessed into energy stock (energy carriers) such as power harnessed from PV panel and wind turbine. The physicist and seminal thinker Amory Lovins wrote, ‘‘following the remark attributed to General Dwight Eisenhower: ‘If a problem cannot be solved, enlarge it’—until its boundaries include the options, synergies, and degrees of freedom that its solution requires.’’ [59] We must enlarge the energy problem, as Lovins has been arguing since 1976, beyond the ‘‘hard path’’ of fossil fuel to a ‘‘soft path’’ of renewable energies (along with conservation and efficiency). We must also enlarge the energy problem by thinking in terms of the irreversibility process of natural energy gradient, not just energy stock in the ground or collected from the sky. Natural energy gradient represents an even bigger world of the enlarged possibilities that Lovins refers to as the new fire [60]: natural energy gradient of the strong meso-scale climatic process of sea breeze may well be a new, the third kind of, renewable ‘‘energy’’ next to the wind and the solar energy: a renewable thermal resource in the present case, which is different from sea breeze being a dynamical resource for wind farms (e.g., San Gorgonio Pass, CA).

6. Conclusion The ‘‘architect’’ of energy efficiency is the conception of best heat extraction frameworks—which determine the realm of possible efficiency. The practice of such ‘‘architecture’’ is assisted by a toolbox: ‘‘Water is 832 times denser than air. Thermally active surfaces [for radiant cooling] are built around this basic principle.’’ [61] ‘‘Radiant cooling systems are more efficient, more comfortable, more attractive, and more healthful than [conventional HVAC] systems that circulate air.’’ [62] Hydronically activated radiant surfaces are the architectural tool in the toolbox of the ‘‘architect’’ of energy efficiency. Process assumption-based design, a two-step design method that has been initiated by the authors, is the design tool in the toolbox. Process assumption-based design, which was applied to determine the required architectural/thermal qualities of the autonomous building in Refs. [5,7], can be used for achieving building homeostasis by applying heat extraction principle: using cooling tower (the engineering tool in the toolbox) in this paper for achieving homeostasis partially. A central argument in this paper is that buildings’ external natural temporal and spatial gradient represents both a challenge to building’s indoor comfort (see Figs. 5c and 6c) and a driving ‘‘force’’ for buildings’ conditioning (see Figs. 5b and 6b). Using cooling tower, it is shown that homeostatic buildings at optimal operative temperature are possible in locations of large diurnal temperature change. We need to think in terms of how to manage irreversibility: while diurnal temperature variation is not usually viewed as a source of energy as wind power is, it is clearly a thermal resource that manifests as a temporal irreversible gradient. In locations of moderate diurnal temperature variation the use of cooling tower alone can only achieve homeostasis partially. Future studies will investigate the use of cooling tower in combination with heat pump for extracting heat from building indoor. In such case of ‘‘electrification of heat [cooling in our case] in buildings, facilitated primarily by heat pumps. . .’’ [13], electrification, cooling tower, and heat pumps are the engineering tool in the toolbox. This study shows that it is possible to reduce waste heat resulting from cooling operation of a building, in some cases almost entirely except that from the operation of hydronic pumps, and in other cases significantly resulting from reduced heat pump operation. We have suggested that waste heat is the logical

P. Ma et al. / Applied Energy 127 (2014) 172–181

common denominator for measuring the performance of building operation in the same objective way as heat engines are measured. This suggestion led to the re-examination of how buildings are conditioned. In future studies we shall give the quantified assessment of this possibility by investigating the full homeostasis of buildings in summer and in winter – and present complete assessment of the energy efficiency potential of the electrically powered heat extraction paradigm in terms of waste heat and its impact on the environment.

References [1] Lechner N. Heating, cooling, lighting. 3rd ed. Wiley; 2009. [2] Watson RD, Chapman KS. Radiant heating and cooling handbook. McGrawHill; 2002. p. 5.251. [3] ASHRAE Inc. 2009. ASHRAE Handbook – fundamentals. SI and I-P ed.; 2009. [4] Ma P, Wang L-S, Guo N. Modeling of TABS-based thermally manageable buildings in Simulink. Appl Energy 2013;104:791–800. [5] Wang L-S, Ma P, Hu E, Giza-Sisson D, Mueller G, Guo N. A study of building envelope and thermal mass requirements for achieving thermal autonomy in an office building. Energy Build 2014; http://dx.doi.org/10.1016/ j.enbuild.2014.04.015. [6] Ma P. Thermal homeostasis in buildings (THiB): radiant conditioning of hydronically activated buildings with large fenestration and adequate thermal mass using natural energy for thermal comfort. Doctoral thesis, Stony Brook University, May 21, 2013. [7] Ma P, Wang L-S, Guo N. Maximum window-to-wall ratio of a thermally autonomous building as a function of envelope U-value and ambient temperature amplitude. Appl Energy; September 2013, submitted for publication. [8] Wang L-S, Ma P. The science of low grade heat from fire. Appl Energy 2013. [9] American Physical Society. Energy Future 2008. [10] American Society of Mechanical Engineers. Integrated/Sustainable Building Equipment & Systems (ISBES) report, [4/24/2013]. [11] Wang L-S. Exergy or the entropic drive: waste heat and free heat. Int J Exergy 2013;12(4):491–521. [12] Wang L-S. The nature of spontaneity-driven processes. Int J Ecodyna 2007;2:231–44. [13] National Grid. Pathways for decarbonising heat. A National Grid commissioned study by Redpoint Energy-Baringa; 2012. [14] Meierhans RA. Slab cooling and earth coupling. ASHRAE Trans 1993;99(2):511–8 [DE-93-02-4]. [15] Meierhans RA. Room air conditioning by means of overnight cooling of the concrete ceiling. ASHRAE Trans 1996;102(1):693–7 [AT-96-08-2]. [16] Ma P, Wang L-S. Effective heat capacity of interior planar thermal mass (iPTM) subject to periodic heating and cooling. Energy Build 2012;47:44–52. [17] Ma P, Wang L-S. Effective heat capacity of exterior planar thermal mass (ePTM) subject to periodic heating and cooling. Energy Build 2012;47:394–401. [18] Ren MJ, Wright JA. A ventilated slab thermal storage system model. Build Environ 1998;33:43–52. [19] Weber T, Jóhannesson G, Koschenz M, Lehmann B, Baumgartner T. Validation of a FEM-program (frequency-domain) and a simplified RC-model (timedomain) for thermally activated building component systems (TABS) using measurement data. Energy Build 2005;37:707–24. [20] Weber T, Jóhanneson G. An optimized RC-network for thermally activated building components. Build Environ 2005;40:1–14. [21] Olesen BW, de Carli M, Scarpa M, Koschenz M. Dynamic evaluation of the cooling capacity of thermo-active building systems. ASHRAE Trans 2006;112(1):350–7. [22] Gwerder M, Tötli J, Lehmann B, Dorer V, Güntensperger W, Renggli F. Control of thermally activated building systems (TABS) in intermittent operation with pulse width modulation. Appl Energy 2009;86:1606–16. [23] Gwerder M, Lehmann B, Tötli J, Dorer V, Renggli F. Control of thermally activated building systems TABS. Appl Energy 2008;85:565–81. [24] Bueno B, Norford L, Pigeon G, Britter R. A resistance–capacitance network model for the analysis of the interactions between the energy performance of buildings and the urban climate. Build Environ 2012;54:116–25. [25] Široky J, Oldewurtel F, Cigler J, Prívara S. Experimental analysis of model predictive control for an energy efficient building heating system. Appl Energy 2011;88:3079–87. [26] Lehmann B, Dorer V, Gwerder M, Renggli F, Tötli J. Thermally activated building systems (TABS): energy efficiency as a function of control strategy, hydronic circuit topology and (cold) generation system. Appl Energy 2011;88:180–91. [27] ASHRAE Inc. ANSI/ASHRAE/IES Standard 90.1-2010: energy standard for buildings except low-rise residential buildings. S-I ed.; 2010.

181

[28] Bhatia A. Cooling towers. CED engineering course material. Course No: M07001. [29] Maiya MP. Analysis of modified counter-flow cooling towers. Heat Recov Syst CHP 1995;15(3):293–303. [30] Gan G, Riffat SB. Numerical simulation of closed wet cooling towers for chilled ceiling systems. Appl Therm Eng 1999;19:1279–96. [31] Facão J, Oliveira AC. Thermal behaviour of closed wet cooling towers for use with chilled ceilings. Appl Therm Eng 2000;20:1225–36. [32] Facão J, Oliveira AC. Heat and mass transfer correlations for the design of small indirect contact cooling towers. Appl Therm Eng 2004;24:1969–78. [33] Hasan A, Sirén K. Theoretical and computational analysis of closed wet cooling towers and its applications in cooling of buildings. Energy Build 2002;34(5):477–86. [34] Stabat P, Marchio D. Simplified model for indirect-contact evaporative coolingtower behaviour. Appl Energy 2004;78:433–51. [35] Wang SW, Tyagi SK, Sharma A, Kaushik SC. Application of solar collectors to control the visible plume from wet cooling towers of a commercial building in Hong Kong: a case study. Appl Therm Eng 2007;27:1394–404. [36] Vangtook P, Chirarattananon S. Application of radiant cooling as a passive cooling option in hot humid climate. Build Environ 2007;42:543–56. [37] Jin G-Y, Cai W-J, Lu L, Lee EL, Chiang A. A simplified modeling of mechanical cooling tower for control and optimization of HVAC systems. Energy Convers Manage 2007;48:355–65. [38] Yu FW, Chan KT. Optimization of water-cooled chiller system with load-based speed control. Appl Energy 2008;85:931–50. [39] Sarker MMA, Kim E, Moon CG, Yoon JI. Performance characteristics of the hybrid closed circuit cooling tower. Energy Build 2008;40:1529–35. [40] Tyagi SK, Wang S, Park SR, Sharma A. Economic considerations and cost comparisons between the heat pumps and solar collectors for the application of plume control from wet cooling towers of commercial buildings. Renew Sust Energy Rev 2008;12:2194–210. [41] Wu J, Zhang G, Zhang Q, Zhou J, Wang Y. Artificial neural network analysis of the performance characteristics of a reversibly used cooling tower under cross flow conditions for heat pump heating system in winter. Energy Build 2011;43:1685–93. [42] Sagia Z, Rakopoulos C, Kakaras E. Cooling dominated hybrid ground source heat pump system application. Appl Energy 2012;94:41–7. [43] Papaefthimiou VD, Rogdakis ED, Koronaki IP, Zannis TC. Thermodynamic study of the effects of ambient air conditions on the thermal performance characteristics of a closed wet cooling tower. Appl Therm Eng 2012;33– 34:199–207. [44] Liu Y, Qin XS, Chiew YM. Investigation on potential applicability of subsurface cooling in Singapore. Appl Energy 2013;103:197–206. [45] Zhai Z, Fu S. Improving cooling efficiency of dry-cooling towers under crosswind conditions by using wind-break methods. Appl Therm Eng 2006;26:1008–17. [46] EnergyPlusTM. EnergyPlus engineering reference: the reference to EnergyPlus calculations. EnergyPlus documentation. Version 7.0 documentation; October 2011. [47] Koschenz M, Lehmann B. Thermoaktive bauteilsysteme. EMPA; 2000. [48] van der Merwe J, du Toit CG. Numerical simulation of a wet cooling tower. R D J 2002;18(2). [49] Furlong JW, Morrison FT. Optimization of water-cooled chiller – cooling tower combinations. CTI J 2005;26(1):12–9. [50] United Nations Environment Programme. Electrical energy equipment: cooling towers; 2006. [51] Sullivan D, Bean R. Thermally active building design: a path toward carbon neutrality 2011. [52] Lehmann B, Dorer V, Koschenz M. Application range of thermally activated building systems tabs. Energy Build 2007;39:593–8. [53] ASHRAE Inc. ANSI/ASHRAE/IES Standard 90.1-2004: Energy Standard for Buildings Except Low-Rise Residential Buildings, S-I ed.; 2004. [54] ASHRAE Inc. ANSI/ASHRAE/IES Standard 90.1-2007: Energy Standard for Buildings Except Low-Rise Residential Buildings, S-I ed.; 2007. [55] Olesen BW. Thermo active building systems – using building mass to heat and cool. ASHRAE J 2012:44–52. [56] Kolarik J, Toftum J, Olesen BW, Jensen KL. Simulation of energy use, human thermal comfort and office work performance in buildings with moderately drifting operative temperatures. Energy Build 2011;43:2988–97. [57] Toftum J, Olesen BW, Kolarik J, Mattarolo L, Belkowska D. Occupant responses and energy use in buildings with moderately drifting temperatures. Atlanta: ASHRAE; 2008. p. 145. [58] http://apps1.eere.energy.gov/buildings/energyplus/weatherdata_download.cfm [last access: 23.02.14]. [59] Lovins A. Comment on opportunities and challenges for a sustainable energy future. RMI Outlet 2012. [60] Lovins A. Reinventing fire. Chelsea Green Publishing; 2011. [61] Moe K. Thermally active surfaces in architecture. Princeton Architectural Press; 2010. [62] Energy Design Resources. Design brief: radiant cooling 2003.