A Thermal Comfort based Controller for a Direct Expansion Air Conditioning System

Available Available online online at at www.sciencedirect.com www.sciencedirect.com

ScienceDirect ScienceDirect

Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com Available Energy Procedia 00 (2017) 000–000 Energy Procedia 00 (2017) 000–000

ScienceDirect ScienceDirect

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

Energy Procedia 142 Energy Procedia 00(2017) (2017)1817–1822 000–000 www.elsevier.com/locate/procedia

9th 9th International International Conference Conference on on Applied Applied Energy, Energy, ICAE2017, ICAE2017, 21-24 21-24 August August 2017, 2017, Cardiff, Cardiff, UK UK

A Comfort A Thermal Thermal Comfort based based Controller Controller for for aa Direct Direct Expansion Expansion Air Air The 15th International Symposium on District Heating and Cooling Conditioning Conditioning System System Assessing the feasibility of using the heat demand-outdoor a b b Huaxia Huaxia YAN YANa,, Jiajia Jiajia NIU NIUb,, Shiming Shiming DENG DENGb * * temperature function for a long-term district heat demand forecast Faculty of Science and Technology, Technological and Higher Education Institute of Hong Kong, Hong Kong Special Administrative Region

a a Faculty of Science and Technology, Technological and Higher Education Institute of Hong Kong, Hong Kong Special Administrative Region b bDepartment of Building Services Engineering, The Hong Kong Polytechnic University, Kowloon, Hong Kong Special Administrative Region

Department of Buildinga,b,c Services Engineering, a The Hong Kong aPolytechnic University, b Kowloon, Hong Kong cSpecial AdministrativecRegion

I. Andrić

*, A. Pina , P. Ferrão , J. Fournier ., B. Lacarrière , O. Le Corre

a Abstract IN+ Center for Innovation, Technology and Policy Research - Instituto Superior Técnico, Av. Rovisco Pais 1, 1049-001 Lisbon, Portugal Abstract b

Veolia Recherche & Innovation, 291 Avenue Dreyfous Daniel, 78520 Limay, France Département Systèmes et Environnement - IMT control Atlantique, 4 rue Alfredenergy Kastler,efficiency 44300 Nantes, France For building HVAC systems,Énergétiques better indoor thermal comfort and higher could be achieved c

For building HVAC systems, better indoor thermal comfort control and higher energy efficiency could be achieved using using aa thermal thermal comfort comfort index index as as the the controlled controlled variable variable instead instead of of air air temperature. temperature. Although Although there there can can be be various various simplification approaches, indoor air humidity has been commonly assumed constant or left uncontrolled. Neglecting simplification approaches, indoor air humidity has been commonly assumed constant or left uncontrolled. Neglecting the importance Abstract the importance of of indoor indoor humidity humidity control control would would adversely adversely affect affect the the control control performance performance of of an an air air conditioning conditioning system. In addition, it is difficult to control local air velocity in the vicinity of a user. Therefore, either system. In addition, it is difficult to control local air velocity in the vicinity of a user. Therefore, either aa constant constant local local air velocity or air varying aa preset was in aa thermal comfort model. In paper, aa District heating are commonly the literature as one of most effective decreasing air velocity or an annetworks air velocity velocity varying to toaddressed presetinpattern pattern was assumed assumed in the thermal comfortsolutions model. for In this this paper, the thermal comfort based controller is developed and reported. Controllability tests were carried out which showed that greenhouse gas emissions from the building sector. These systems require high investments which are returned through the heat thermal comfort based controller is developed and reported. Controllability tests were carried out which showed that sales. Due toefficiency the changed the climate conditions and building renovation policies, heat demand in the future could decrease, better better energy energy efficiency of of the DX DX A/C A/C system system could could be be achieved achieved when when aa higher higher local local fan fan speed speed was was used. used.

prolonging the investment return period. Ltd. © © 2017 2017 The The Authors. Authors. Published Published by by Elsevier Elsevier Ltd. ©The 2017 Thescope Authors. Published Ltd. main ofresponsibility this paper isby to Elsevier assess the feasibility of using demand –Conference outdoor temperature function for heat demand Peer-review under of the scientific committee of thethe 9thheat International Energy. Peer-review under responsibility of scientific committee of International Conference on on Applied Applied Energy. Peer-review under responsibility ofthe the scientific committee ofthe the9th 9th International Applied Energy. forecast. The district of Alvalade, located in Lisbon (Portugal), was used as a Conference case study.on The district is consisted of 665 buildingsThermal that vary in both construction period and typology. Three weather scenarios (low, medium, high) and three district Keywords: comfort; direct expansion; air conditioning; control; air velocity; Keywords: Thermal comfort; direct expansion; air conditioning; control; air velocity; renovation scenarios were developed (shallow, intermediate, deep). To estimate the error, obtained heat demand values were compared with results from a dynamic heat demand model, previously developed and validated by the authors. The results showed that when only weather change is considered, the margin of error could be acceptable for some applications 1. 1. Introduction Introduction (the error in annual demand was lower than 20% for all weather scenarios considered). However, after introducing renovation scenarios, the errorcomfort value increased up to 59.5% (depending on the weather and renovation scenarios combination dependent considered). Indoor thermal derives from aa complicated non-linear interaction between the four Indoor thermal comfort derives from on complicated non-linear interaction between the decade, four environment environment dependent The value of slope coefficient increased average within the range of 3.8% up to 8% per that corresponds to the variables (air temperature, velocity, humidity and mean temperature) and person dependent variables variables temperature, velocity, humidity and during mean radiant radiant temperature) and the the two two person dependent variables decrease (air in the number of heating hours of 22-139h the heating season (depending on the combination of weather and (activity level and [1]. large number of systems, however, only regulate air controlled (activity level and clothing) clothing) [1]. A A number of HVAC HVAC only7.8-12.7% regulate per air temperature, temperature, controlled renovation scenarios considered). Onlarge the other hand, function systems, intercept however, increased for decade (depending on the by air regulators (ATRs). If of the other parameters varies significantly, aa user may thermally by air temperature temperature If one one ofbe theused othertofive five parameters varies significantly, user may feel feel thermally coupled scenarios).regulators The values(ATRs). suggested could modify the function parameters for the scenarios considered, and uncomfortable. Over the years, a large number of comfort index regulators (CIRs) have been developed and improve the accuracy demand estimations. uncomfortable. Over of theheat years, a large number of comfort index regulators (CIRs) have been developed and used used for for © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and * Corresponding author. Tel.: +852 27665859; fax: +852 27657198. Cooling. * Corresponding author. Tel.: +852 27665859; fax: +852 27657198.

E-mail address: [email protected]. E-mail address: [email protected]. Keywords: Heat demand; Forecast; Climate change 1876-6102 © 2017 The Authors. Published by Elsevier Ltd. 1876-6102 © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the scientific committee of the 9th International Conference on Applied Energy. Peer-review under responsibility of the scientific committee of the 9th International Conference on Applied Energy.

1876-6102 © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and Cooling.

1876-6102 © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the scientific committee of the 9th International Conference on Applied Energy . 10.1016/j.egypro.2017.12.569

Huaxia Yan et al. / Energy Procedia 142 (2017) 1817–1822 Author name / Energy Procedia 00 (2017) 000–000

1818 2

guiding the design of HVAC systems [1-4]. These include the PMV (predicted mean vote) index, proposed and developed by Fanger [1]. Nomenclature A/C CIR DX PMV Tdb Twb

Air conditioning Comfort index regulator Direct expansion Predicted mean vote Dry-bulb temperature Wet-bulb temperature

A simplified thermal comfort model is normally necessary when designing a CIR. Although there can be various simplification approaches, variations in indoor air humidity and air velocity around occupants have always been regarded as influencing thermal comfort to only a small extent. Consequently, indoor air humidity is usually assumed constant or left uncontrolled [5-6]. However, although ASHARE Standards [7] recommend a wide comfort range of relative humidity from 30% to 60%, humidity levels above 60% do adversely affect indoor thermal comfort and indoor air quality significantly [8]. A practical reason for this is however the difficulty for building A/C systems to simultaneously control indoor air temperature and humidity. On the other hand, the increase in the air velocity around occupants would help improve comfort level, although the averaged room temperature increased. Remarkable effects could be obtained from people who had moderate activity [9]. Accordingly, the personalized ventilation resulted in an energy saving without causing occupant dissatisfaction [10]. Energy efficiency is also important in judging the performance of an HVAC system. Previous research work has shown that it is possible to reduce HVAC energy use without sacrificing indoor thermal comfort while using CIR technology [11], by increasing PMV setting but still keeping it within the comfort range. However, without an appropriate humidity control strategy, increasing PMV setting can only be accomplished by increasing air temperature. Very often, the scope for increasing air temperature is limited, particularly in hot and humid climates. If indoor humidity can be controlled to a low level and a higher air velocity was achieved at working places, however, at the same PMV setting, Tdb could be set a little higher, leading to a smaller indoor sensible cooling load. When applied to HVAC system control, fuzzy logic controller attracted much research interests because of its characteristics of capturing the approximate and inexact nature of a controlled process, such as human thermal comfort. [5, 12]. In this paper, an experimental study on a direct expansion (DX) A/C system with an emphasis on indoor thermal comfort is reported. Firstly, the development of a fuzzy logic controller is presented. This is followed by presenting a simplification approaches to predicted mean vote (PMV), the commonly used CIR. Finally, examples of controllability tests results are given. 2. Development of a fuzzy logic controller

Fig. 1 Flow chart of the fuzzy logic controller for a DX A/C system Fig. 1 shows the complete flow chart of the fuzzy logic controller which was developed based on a previous work [13]. In [13], a novel control principle was developed to decouple the temperature and humidity control loops by introducing two interim variables, i.e., the sensible and latent output cooling capacity of the DX A/C system. Then, a defuzzifier was applied to convert the calculated sensible/ latent load obtained from the fuzzy logic controller to the speeds of compressor and supply fan, as shown in Fig. 1. An artificial neural network model for a variable speed DX A/C system was trained and used as a defuzzifier [13]. The training progress for the artificial neural network model is



Huaxia Yan et al. / Energy Procedia 142 (2017) 1817–1822 Author name / Energy Procedia 00 (2017) 000–000

1819 3

complicated. Therefore, in this paper, the defuzzifier was replaced by an experimental figure, i.e., Fig. 2, obtained through an experimental DX A/C system. Fig. 2 was obtained under different speed combinations of compressor and supply air fan at a constant indoor air condition of 25 ºC and 50% relative humidity. It was plotted as a trapezoid shape with the total output cooling capacity and the equipment sensible heat ratio, i.e., the ratio of the sensible cooling capacity to the total cooling capacity, directly showing the relation between cooling capacity and the speeds of compressor and fan. In order to simplify the use of the trapezoid, it was divided into nine zones, each with a representative point. The nearest representative point to the calculated point of total cooling capacity and sensible heat ratio would be chosen to determine the speeds operation of compressor and fan in the experimental DX A/C system. 0.80

Compressor speed (%)

Equipment sensible heat ratio

Fs: 100 0.75

Fs: 90 Fs: 80 Fs: 70

0.70

Fs: 60 0.65

0.60

C

Fs: 50

B

80 60 40

A I D

IV II

VII

V III

VI

70 50

G

E VIII

F IX

H

Y

J

Fs: Fan speed (%) 4.5

5.0

5.5

6.0

6.5

7.0

7.5

8.0

8.5

Total output cooling capacity (kW)

Fig. 2 Proposed zoning with nine representative points in the trapezoid 3. Simplification to the comfort index regulator The CIR have two functions, the set-point establishment and thermal comfort evaluation. The set-point establishment module will generate an appropriate PMV setting, where the actual required settings for indoor dry-bulb temperature ( Tdb ) and wet-bulb temperature ( Twb ) will be determined. These settings will be input to the fuzzy logic controller to simultaneously control Tdb & Twb by varying compressor and supply fan speeds in the DX A/C system. 4. Experimental rig and controllability test results 4.1 Experimental rig

Fig. 3 The schematic diagram of the experimental rig As shown in Fig. 3, an experimental rig was established to resemble a typical DX A/C system. It mainly consisted

Huaxia Yan et al. / Energy Procedia 142 (2017) 1817–1822 Author name / Energy Procedia 00 (2017) 000–000

1820 4

of two parts, a DX refrigeration plant and an air distribution sub-system. The nominal output cooling capacity from the DX refrigeration plant 7.5 kW. The refrigerant was R410a, with a total charge of 5.8 kg. A load generating unit was placed in the indoor space to simulate both sensible and latent loads. A two-speed fan was placed 0.8 cm away from an occupant, which could be manually controlled. When the local fan was switched off, the measured air velocity around working place was measured at about 0.15 m/s. The averaged measured air velocity were 0.21 m/s and 0.34 m/s when the fan was low and high speed operated, respectively. 4.2 Experimental test conditions Using the developed fuzzy logic controller, two examples of experiments were conducted to test its control performances. During the experiments, PMV was used as regulator. Each test lasted for 3 hours with local fan disabled in the first hour, low speed operation in the second hour and high speed in the last hour. Detailed experimental conditions were listed in Table 1. During each test, sensible and latent loads were kept constant through adjusting the inputs to the load generating unit. Table 1 Experimental test conditions Test Fixed setting Altered local fan speed/ air velocity No. PMV Relative Sensible load Latent load --/ m/s humidity (%) (kW) (kW) 1 0 50 4.40 2.18 Low speed/ High speed/ Off/ 0.15 0.21 0.34 2 -0.3 50 4.53 2.09

PMV

4.3 Experimental test results 0.6 0.4 0.2 0.0 -0.2 -0.4 -0.6

0

1800

3600

Temperature (ºC)

32

5400

Test duration (s) (a)

7200

9000

10800

Tdb Twb

28 24 20

Compressor speed (%)

16 0

1800

3600

0

1800

3600

0

1800

3600

100

5400 7200 Test duration (s) (b)

9000

10800

80 60 40 5400

7200

9000

10800

5400

7200

9000

10800

Test duration (s)

Fan speed (%)

100 80 60 40

Test duration (s) (c)

Fig. 4 Variation profiles of indoor air status and compressor/ supply fan speed in Test 1 In the first test, PMV was set as 0, which is neutral for most of people’s sensation. When the local fan speed was



Huaxia Yan et al. / Energy Procedia 142 (2017) 1817–1822 Author name / Energy Procedia 00 (2017) 000–000

1821 5

increased from 0 to low speed between 3600 s and 7200 s, and further high speed in the rest of experiments, the air velocity around occupants was increased form 0.15 m/s to 0.21 m/s and further 0.34 m/s in the last operation hour. The PMV setting was kept constant, but that for Tdb / Twb were increased, from 25.36 °C/ 18.19 °C to 25.69 °C/ 18.47 °C in the second hour and 26.16 °C/ 18.85 °C in the last hour, respectively. As seen in Fig. 4 (a), the calculated PMV in the indoor space of the DX A/C system were oscillated around 0, within acceptable range of [-0.10, 0.03]. When the local fan was switched on or turned to a higher speed, a sudden drop in PMV value would be observed, due to the increase in air velocity. In Fig. 4 (b), the variation profiles of Tdb / Twb were plotted. Before the local fan was enabled, Tdb and Twb were varied

PMV

with small ranges. In 3600 s, a low speed for the local fan was introduced. Settings for Tdb and Twb were increased. It took about 400 s for Tdb and Twb reaching their new settings. During the experiments, the space sensible and latent cooling loads were kept at fixed values through using the indoor load-generating unit. Therefore, when the variation rages and cycles were similar in the three periods. As seen from Fig. 4 (c), speeds of compressor and the supply air fan in the DX A/C system were continuously adjusted according to the output of the fuzzy logic controller. This could also explain the variation trends of Tdb / Twb . It is assumed that the variation in supply air speed had small impact on the air velocity near an occupant. Therefore, the value of air velocity for calculating PMV would be changed when there was adjustment to the local fan speed. 0.0 -0.2 -0.4 -0.6 -0.8 -1.0

0

1800

3600

7200

9000

10800

Tdb Twb

Temperature (ºC)

32

5400

Test duration (s) (a)

28 24 20

Compressor speed (%)

16 0

1800

3600

0

1800

3600

0

1800

3600

100

9000

10800

80 60 40

100

Fan speed (%)

5400 7200 Test duration (s) (b)

5400

7200

9000

10800

5400

7200

9000

10800

Test duration (s)

80 60 40

Test duration (s) (c)

Fig. 5 Variation profiles of indoor air status and compressor/ supply fan speed in Test 2 In the second test, a lower PMV value was set and similar results as that in test 1 were observed. Energy efficiency is also important in judging the performance of an A/C system. The energy performance of the experimental DX A/C system in each test was also monitored as detailed in Table 2. The power input to the supply fan and the local fan were also taken into account in comparing the energy performance. Under a constant indoor thermal loads and control strategies, the total power input to the DX A/C system in the last two hours are 6.2% and 7.1%, respectively, lower than that during the first hour, where the local fan was disabled. Similar, 4.0% and 7.6%

1822 6

Huaxia Yan/ Energy et al. / Procedia Energy Procedia 142000–000 (2017) 1817–1822 Author name 00 (2017)

energy would be saved when a low-speed and a high-speed local fan were adopted in Test 2. Table 2 Performance data during the three hours’ operation in each test Energy consumption (kWh) Test Period (Hour) Air velocity (m/s) Compressor Fan Local fan Total 0-1 0 2.108 1.460 0 3.568 1 1-2 0.21 1.891 1.451 0.015 3.548 2-3 0.34 1.824 1.458 0.035 3.489 0-1 0 2.205 1.526 0 3.731 2 1-2 0.21 2.074 1.492 0.015 3.581 2-3 0.34 1.936 1.475 0.035 3.446

Saving (%) -6.2% 7.1% -4.0% 7.6%

5. Conclusion A fuzzy logic controller was developed for a DX A/C system based on previous work. PMV was used as a comfort regular to determine the settings for indoor air temperature and humidity. A higher air velocity around working places can be achieved through using a local fan. Therefore, increasing air dry-bulb temperature setting was achieved at a constant PMV value. Experiments have been conducted, and the experimental results have demonstrated that the fuzzy logic controller can result in smooth control over PMV and 4.0% to 7.6% energy consumption of the DX A/C system could be saved when a higher local fan speed was used. Acknowledgements The authors would like to acknowledge the financial supports from The Hong Kong Polytechnic University (PolyU G-YBDB) References [1] Fanger P.O., Thermal comfort, analysis and application in environmental engineering. McGraw Hill New York 1972. [2] Gagge A.P., Fobelets A.P., Berglund L.G., A standard predictive index of human response to the thermal environment. ASHARE Transactions, 1986, 92: 709-731 [3] Sherman M., A simplified model of the thermal comfort. Energy and Buildings, 1985, 8: 37-50. Hamdi M., Lachiver G., Michaud F., A new predictive thermal sensation index of human response. Energy and Buildings, 1999, 29: 167-178. [4] Hamdi M., Lachiver G., Michaud F., A new predictive thermal sensation index of human response. Energy and Buildings, 1999, 29: 167-178. [5] Calvino F., Gennusa M.L., Rizzo G., Scaccianoce G., The control of indoor thermal conditions: introducing a fuzzy adaptive controller. Energy and Buildings, 2004, 36: 97-102 [6] Castilla M., Alvarez J.D., Berenguel M., Rodriguez F., Guzman J.L., Perez M., A comparison of thermal comfort predictive control strategies. Energy and Buildings, 2011, 43: 2737-2746 [7] ANSI/ASHRAE Standard 55-2004, Thermal Environment Conditions for Human Occupancy [8] Toftum J., and Fanger P.O., Air Humidity Requirements for Human Comfort. ASHRAE Transactions, 1999, 105(1): 641-647 [9] Jones, B. W., Hsieh, K., & Hashinaga, M. (1986). The effect of air velocity on thermal comfort at moderate activity levels. ASHRAE Trans.;(United States), 92(CONF-8606125-). [10] Schiavon, S., & Melikov, A. K. (2008). Energy saving and improved comfort by increased air movement. Energy and buildings, 40(10), 1954-1960. [11] Liang J., Du R., Design of intelligent comfort control system with human learning and minimum power control strategies. Energy Conversion and Management, 2008, 49 (4) : 517-528. [12] Yang, K.T., Artificial neural networks: A new paradigm for thermal science and engineering. Journal of Heat Transfer, 2008, 130: 1-19 [13] Li Z., Xu X. G., Deng S. M. and Pan D. M. (2015). A novel neural network aided fuzzy logic controller for a variable speed (VS) direct expansion (DX) air conditioning (A/C) system. Applied Thermal Engineering, 78, 9-23.