Method for the determination of optimal work environment in office buildings considering energy consumption and human performance

  • Published on

  • View

  • Download

Embed Size (px)


  • Mb








    Energy and Buildings 76 (2014) 278283

    Contents lists available at ScienceDirect

    Energy and Buildings

    j ourna l ho me pa g e: www.elsev ier .com/ locate /enbui ld

    ethod for the determination of optimal work environment in officeuildings considering energy consumption and human performance

    hangzhi Dai, Li Lan, Zhiwei Lian

    tate Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai 200240, China

    r t i c l e i n f o

    rticle history:eceived 29 October 2013eceived in revised form 22 February 2014ccepted 26 February 2014


    a b s t r a c t

    How to balance the contradiction between energy saving and improvement of indoor environmentalquality which consequently affects human performance has always been a problem. We put forward theeconomically optimum condition as a concept that maximized economic benefit in terms of regulationof office environment parameters. The calculation method was provided by which energy consump-tion could be reduced without compromise of human performance. A regression model predicting the

    uilt environment designuman performancenergy consumptionptimal environment

    energy consumption of a typical office building was illustrated as a function of two indoor environmentparameters, i.e., indoor air temperature and air ventilation rate. Practical factors including salary andelectric price, which have impact on the condition determination, were discussed. As a prototype, anoffice building in Shanghai achieved its economically optimal conditions at an air temperature of 25.1 Cand an outdoor ventilation rate of 17.9 L/s-person in summer.

    2014 Elsevier B.V. All rights reserved.

    . Introduction

    The global energy use has rapidly grown in the past few decades,iving rise to more and more concerns over energy security, utiliza-ion efficiency and environmental impacts. It is also believed thathe energy situation would become more and more serious in theuture, especially in emerging economies [1,2]. Currently, buildingsccount for approximately 40% of the total energy consumptionlobally [1]. Forecast made by the EIA [3] suggests that energy usen the building environment will grow by 34% in the next 20 years.nergy consumption in HVAC systems, comprising heating, out-oor ventilation and air conditioning, has proved to be the largestnergy end-use both in residential and non-residential sectors. Inhina, for example, the HVAC systems are responsible for about 65%f the energy use in building sector [4]. As there are strong rela-ionships between HVAC energy consumption and indoor climateet point [5], reasonable indoor climate parameters are essentialor energy saving today. However, the relationship between HVACnergy consumption and indoor environment quality (IEQ) factorss difficult to be theoretically deduced because of its variance with

    uilding envelope, HVAC performance, etc. Therefore, regressionnalysis and simulation methods have been widely used in thearametric studies of building energy consumption [6,7].

    Corresponding author. Tel.: +86 21 34204263.E-mail address: (Z. Lian).

    ttp:// 2014 Elsevier B.V. All rights reserved.

    In office buildings, however, energy consumption is not theonly priority for indoor climate design. Attentions are increasinglydrawn to the humanwork environment interaction [8]. A healthyand effective built environment was proposed in the domain ofgreen ergonomics [9]. As the salary of office workers is an order ofmagnitude higher than the cost of maintaining and operating thebuilding [10], even small improvements in productivity can resultin a substantial economic benefit. Fisk and Rosenfeld [11] estimatedthat improved indoor environment can bring a direct increase inproductivity, ranging between 0.5% and 5%. Proper thermal con-dition [1214] and indoor air quality [1518] have proved to beof great help for better performance. However, little is known onthe combined effects of these factors. Qualitative studies have beenconducted by Witterseh et al. [19] on the combined effects of tem-perature and recorded noise. Hygge and Knez [20] investigated howoutdoor ventilation noise, air temperature, and illuminance com-bine or interact in their effects on cognitive performance. Clausenand Wyon [21] carried out an experiment with subjects exposedto different combinations of traffic noise, lighting, access to day-light, open-plan office noise, air temperature and air quality. Dueto the lacking of quantitative studies, the combined effects ofair temperature and ventilation rate on work performance wereroughly estimated in this paper, while attention should be paid on

    the proposed methodology for determining economically optimumconditions.

    Building services engineers gradually realize that not onlyenergy consumption but also human productivity should be
  • C. Dai et al. / Energy and Buildings 76 (2014) 278283 279










    Fig. 1. A framework illustrating the core idea of the model.

    ncorporated into the economic calculations pertaining to build-ng design and operation. However, there are few methods shedight on the global economic effects caused by energy consumptionf HVAC systems and human productivity in the building. In thistudy we proposed a method to achieve such economically opti-al indoor environment by balancing the contradiction between

    uman performance and energy consumption of HVAC systems.wo important IEQ factors including indoor air temperature andutdoor ventilation rate were set as the example to illustrate howo achieve such economically optimal set points.

    . Methods

    Improved indoor environment generally brings forth higherroductivity while may cause extra investment in operating cost.o obtain economically optimum conditions, both the benefit frommproved performance and the corresponding energy cost shoulde quantitatively analyzed. A subtraction model instead of theostbenefit ratio was selected in Eq. (1) since energy charge wasnly a small part in total inputs and much less than the economiceturns.

    ax G(T, Q, E. . .) C(T, Q, E. . .) (1)here T, Q, E. . . were the IEQ factors, T was indoor air tempera-

    ure, Q was outdoor ventilation rate, E was illuminance level, G washe economic returns decided by the employee working perfor-

    ance and influenced by the IEQ factors, C was the HVAC energyonsumption which also depended on the IEQ factors.

    The core idea of the model was illustrated in Fig. 1. It workedike an open-loop control strategy. Two functions representing the

    ork output and energy cost of the certain office were processedo build the optimization model as shown in Eq. (1). And the inputata of the whole model decided what the two functions woulde. After using the optimization method and theory, the optimal

    ndoor environment conditions could be achieved.Due to the lack of quantitative relationship between human per-

    ormance and other IEQ factors, two important parameters wereiscussed in this study: air temperature T and outdoor ventilationate Q. Nuisance factors affecting either economic returns or energyonsumption were kept constant here.

    .1. The relationship between office work output and IEQ factors

    Taking air temperature as its horizontal axis and outdoor venti-ation rate as the vertical axis, a Cartesian plane could be drawn in

    Fig. 2. Method to obtain the relative performance in different conditions.

    which each point had its corresponding value of work performance,as shown in Fig. 2. The origin point (T0, Q0) was the reference pointwhere the relative performance was set to be 1. Both T0 and Q0 wereset at the values when human performance was poorest. Assumingthe magnitude of the combined effects was the sum of indepen-dent parameters, RPH represented the relative performance, butif the combined effects were replaced by the greater of the sin-gle parameters, the relative performance was then expressed asRPL. At present there was a study reporting the combined effectof outdoor ventilation and temperature on human performance.Wargocki and Seppnen [22] suggested that the magnitude of thecombined effect was at least the effect of the greater of the singleparameters, and not more than the sum of the independent param-eters. So we proposed the mean value of RPH and RPL as the finalrelative performance (Eq. (2)). 1 and 2 represented the changein relative performance when the indoor environment improvedalong with the coordinate axes.

    RPx = 12 (RPH + RPL) = 1 + 12 [1 + 2 + max(1, 2)]

    = 12 [RPt + RPq + max(RPt, RPq) 1] (2)

    Given the varying trends of the relative performance along coor-dinate axes, the relative performance interval from RPL to RPH couldbe figured out for the office environment conditions (Tx, Qx).

    As shown in Fig. 2, when the air temperature was kept to beconstant, the relative performance would be RPq. if the outdoor ven-tilation rate increased from Q0 to Qx L/s-person. Similarly, when theoutdoor ventilation rate was kept to be Qx L/s-person, the relativeperformance would be RPt by reducing the air temperature from T0to Tx C.

    The effects of air temperature or outdoor ventilation rate onhuman performance have been estimated in existing studies. Inthis paper, the quantitative relationship between productivity andthermal sensation vote (Eq. (3)) developed by Lan et al. [23] wasused. Based on this relationship, the economic optimization modelcould illustrate the optimal conditions for different seasons andcould allow for the changes of factors including clothing thermalresistance, relative humidity and so on [24].

    RP = 0.0351 tsv3 0.5294 tsv2 0.215 tsv + 99.865 (3)

    where RP was the relative performance when compared to the max-imum performance and tsv was the thermal sensation vote (3 to+3 on the ASHRAE seven-point thermal sensation scale) [23].

    Seppnen et al. [15] established a quantitative relationship(Fig. 3) between outdoor ventilation rate and productivity. The

    following equation (Eq. (4)) could be obtained from Fig. 3.

    RPq = 0.021 ln(Q ) + 0.960 (6.5 L/s-person Q 30 L/s-person)(4)

  • 280 C. Dai et al. / Energy and Buildings 76 (2014) 278283













    which is an integrated building simulation tool using base tempera-tures to couple buildings and systems, and carrying out simulationby stage for building environment and its control system [26].

    Table 1Main input data and output results of the model in case study.

    Input:Weather data:Reference year weather data of Shanghai (in practical use, real weather

    conditions are better)

    Building enclosure parameters:Open plan offices area, 454 m2; private offices areas, 21 m 6 m; floor

    size, 54 m 16.8 m 3.3 m; outer wall material, porous concrete;inner wall material, aerated concrete slab; windows, 6 mm singleglass, 145.2 m2 in area; rooms air leakage, 0.4 h1, etc.

    Heat sources:68 persons in open plan offices, 6 persons in private offices, 70 W for

    each person, 22 W/m2 for the light and facility in open plan officesand 30 W/m2 in private offices

    Staff/facility schedules:5 days a week and 8 h a day.On weekdays, 50% of the staff stayed in the office from 12:00 pm to

    1:00 pm and 5:00 pm to 6:00 pm; 10% of the staff worked overtimefrom 6:00 pm to 9:00 pm

    HVAC system parameters:All air VAV system and screw refrigerator

    Thermal sensation factors:Air temperature, T; mean radiant temperature, T; relative humidity,

    40%; air velocity, 0.15 m/s; clo value, 0.6 clo; activity level, 1.2 m

    Electric price:1.2 RMB/KWh

    Pay levels:8000 RMB/month-person in the open plan offices and

    15,000 RMB/month-person in the private offices


    Fig. 3. Relative performance in relation to outdoor ventilation rate [15].

    n which RPq was the relative performance comparing to theeference outdoor ventilation rate of 6.5 L/s-person. The outdoorentilation rate was controlled below 30 L/s-person considering theestrictions of the HVAC design and installation.

    Assuming that the staff cost was taken as surrogate of the worketurns at the referenced temperature and outdoor ventilation rate,he economic output of employees working performance could bealculated after the reference point (T0, Q0) was determined (Eq.5)):

    (T, Q ) = s RPx (5)here s was staff cost when relative performance was equal to 1

    nd RPx was given by Eqs. (2)(4).

    .2. The relationship between HVAC energy consumption and IEQactors

    Energy consumption of the building was calculated with simu-ation tool since it could vary with different building enclosures andVAC systems, especially considering the interaction of tempera-

    ure and outdoor ventilation. A typical office building in Shanghaias taken as an example in this case study. Based on the calcu-

    ated energy consumption for different temperatures and outdoorentilation rates, a relationship between energy consumption andndoor parameters (Eq. (6)) was developed by binary polynomialegression, since multiple polynomial regression models for energyse in air-conditioned office buildings have been proved valid [25].

    (T, Q ) = e Pn(T, Q ) (6)here e was the local electricity price which reflected the energyemand directly and Pn(T, Q) was the polynomial regression func-ion describing how energy consumption of HVAC system changedlong with temperature and outdoor ventilation.

    .3. The optimization problem solver

    In this case study the temperature was constrained between0 C and 28 C and the outdoor ventilation rate ranged from 6.5 L/s-erson to 30 L/s-person. Using the penalty function, the constrainedinary optimization problem as expressed in Eq. (1) was transferredo an unconstrained problem (Eq. (7)):

    in (T, Q, r) = C(T, Q ) G(T, Q ) + r


    1T 20 +

    128 T +

    1Q 6.5 +

    130 Q


    here r was the penalty factor.

    Fig. 4. Floor plan of the building.

    Then the steepest descent method was used to determine thedirection of search in Eq. (8); the step size was calculated withArmijo method. The iteration process was finished on Matlab toachieve the points of convergence, which represented the optimumwork conditions for temperature and outdoor ventilation rate.

    d = (




    3. Case study

    A prototypical office building in Shanghai was used in this casestudy. One floor of the typical office was selected for the detailedanalysis. The floor plan was shown in Fig. 4. Hourly simulation wasperformed for the summertime energy consumption with DEST-C,

    May 22September 27

    Output:Optimal air temperature T = 25.1 COptimal outdoor ventilation rate Q = 17.9 L/s-person

  • C. Dai et al. / Energy and Buildings 76 (2014) 278283 281

    Table 2the simulate...


View more >