Strategies for Coupling Energy Simulation and Computational Fluid Dynamics Programs
Total Page:16
File Type:pdf, Size:1020Kb
STRATEGIES FOR COUPLING ENERGY SIMULATION AND COMPUTATIONAL FLUID DYNAMICS PROGRAMS Zhiqiang Zhai and Qingyan Chen Massachusetts Institute of Technology Cambridge, MA 02139, USA Joseph H. Klems and Philip Haves Lawrence Berkeley National Laboratory Berkeley, CA 94720, USA ABSTRACT to determine indices such as the predicted mean vote (PMV), the percentage of people dissatisfied (PPD) Energy simulation (ES) and computational fluid due to discomfort, the percentage dissatisfied (PD) dynamics (CFD) can play important roles in building due to draft, and ventilation effectiveness. With the design by providing complementary information information from both ES and CFD calculations, about the buildings’ environmental performance. designers can design environmental control systems However, separate applications of ES and CFD are for buildings that satisfy multiple criteria. usually unable to give an accurate prediction of building performance due to the assumptions However, due to the complete mixing model used in involved in the separate calculations. Integration of ES, most ES programs cannot accurately predict ES and CFD eliminates many of these assumptions energy for systems that produce non-uniform air since the information provided by the models is temperature distributions in the occupied space, such complementary. Several different approaches to as displacement ventilation systems. Moreover, the integrating ES and CFD are described. In order to spatially averaged comfort information generated by bridge the discontinuities of time-scale, spatial the single node model of ES cannot satisfy advanced resolution and computing speed between ES and design requirements. The convective heat transfer CFD programs, a staged coupling strategy for coefficients used in ES programs are usually different problems is proposed. The paper illustrates empirical and may not have general applicability, a typical dynamic coupling process by means of an either. Furthermore, most ES programs are unable to example implemented using the EnergyPlus and provide information on the airflow entering a MIT-CFD programs. building, for example, by natural ventilation, while the ventilation rate information is very important for Key words: energy simulation, computational fluid predicting room air temperature and (or) heating/ dynamics (CFD), integration, building design cooling load. CFD, on the other hand, can easily determine the INTRODUCTION temperature distribution and convective heat transfer coefficients, which ES needs. CFD is also a Energy simulation and computational fluid dynamics powerful tool for the simulation of natural ventilation programs provide complementary information about driven by wind effect, stack effect, or both. At the the performance of buildings. ES programs, such as same time, CFD also needs information from ES as EnergyPlus (Crawley et al 2000), address the inputs, such as air conditioning loads and surface performance of the building envelope, as well as the temperatures. Otherwise, CFD has to compute heating, ventilating and air conditioning (HVAC) results based on estimated boundary conditions. system, and provide the whole building energy analysis. Space-averaged indoor environmental Therefore, coupling ES with CFD is very attractive conditions, cooling/heating loads, coil loads, and and is the objective of the present investigation. energy consumption can be obtained on an hourly or Starting from the principles of ES and CFD, the sub-hourly basis for periods of time ranging from a paper describes possible approaches to ES and CFD design day to a reference year or more. CFD coupling. The current study emphasizes the explicit programs, on the other hand, make detailed coupling of individual ES and CFD programs by predictions of thermal comfort and indoor air quality exchanging the inter-coupled boundary values. (IAQ), including the distributions of air velocity, temperature, relative humidity and contaminant concentrations. The distributions can be used further 1 In order to bridge the disparities between ES and qik = radiative heat flux from surface i to surface CFD programs due to the different physical models k and numerical methods employed, the study suggests the staged coupling processes that may reduce the computing demands while keeping the advantages of coupled calculations. To demonstrate the process and benefits of coupled simulation, examples of coupled calculations for a simple office space are presented at the end of this paper. FUNDAMENTALS OF ES AND CFD THERMAL COUPLING Figure 1. Energy balance on the interior surface of a wall, Principle of ES ceiling, floor, roof or slab Energy balance equations for zone air and surface The qi can be determined by transfer functions, heat transfer are two essential equations that an weighting factors, or by solutions of the discretized energy program should solve. The energy balance heat conduction equation for the wall using finite equation for room air is differences. The radiative heat flux is N ρVCTΔ room p (1) q= h (T− T ) (3) ∑qi,c A i + Qother − Q heat _ extraction = ik ik,r i k i= 1 Δt where where h = linearized radiative heat transfer N ik,r = convective heat transfer from enclosure coefficient between surfaces i and k ∑qi,c A i i= 1 Ti = temperature of interior surface i surfaces to room air Tk = temperature of interior surface k qi,c = convective flux from surface i N = number of enclosure surfaces And Ai = area of surface i Qother = heat gains from lights, people, qi,c = hc (Ti – Troom) (4) appliances, infiltration, etc. where Qheat_extraction= heat extraction rate of the room hc = convective heat transfer coefficient ρVCTroom p Δ = room air energy change T = room air temperature Δt room ρ = air density The convective heat transfer coefficient, hc, is Vroom = room volume C = air specific heat unknown. Most energy programs estimate hc by p empirical equations or as a constant. If the room air ΔT = temperature change of room air temperature, T , is assumed to be uniform and Δt = sampling time interval, normally one room known, the interior surface temperatures, T , can be hour i determined by simultaneous solving the surface heat balance equations (2). The heat extraction rate is the same as the cooling/heating load when the room air temperature Space cooling/heating load then can be determined is maintained constant (ΔT = 0). The convective heat from the calculated convective heat transfer from fluxes are determined from the energy balance enclosure surfaces using Equation (1). Thereafter, equations for the corresponding surfaces, as shown in the coil load is determined from the heat extraction Figure 1. A similar energy balance is performed for rate and the corresponding air handling processes and each window. The surface energy balance equation HVAC system selected. With a plant model and can be written as: hour-by-hour calculation of the coil load, the energy consumption of the HVAC system for a building can N (2) be determined. qi+ q ir =∑ qik + q c,i k= 1 where Principle of CFD qi = conductive heat flux on surface i qir = radiative heat flux from internal heat CFD is the application of numerical techniques to sources and solar radiation solve the Navier-Stokes (N-S) equations for fluid 2 flow. The N-S equations are derived by applying the (supply), outlet (exhaust), enclosure surfaces, and principles of conservation of mass and momentum to internal objects. The temperature, velocity and a control volume of fluid (A thorough treatment may turbulence of the air entering from diffusers or be found in many textbooks on CFD). When windows determine the inlet conditions, while the applying CFD to the IAQ and thermal comfort interior surface temperatures and/or heat fluxes are problem, the conservation of mass for a contaminant important thermal boundary conditions for the species and energy for thermal responses also may be enclosures. applied. All of the conservative governing equations may be written in the following general form: Coupling Approaches Φ∂ 2 + (V•∇)Φ - Γφ ∇ Φ = Sφ (5) ∂t The above discussion of the principles of ES and CFD shows that the convective heat transfer from where interior surfaces of a space not only links the zone air t = time energy balance equation with the enclosure energy balance equation in ES, but also links ES with CFD. Φ = Vj for the air velocity component in the j direction The problem of model coupling is, then, focused on = 1 for mass continuity how to treat the convective heat transfer in ES and = T for temperature CFD. = C for different gas species = turbulence parameters Depending on the method used to treat the V = velocity vector convective heat transfer, two different coupling Γ =diffusion coefficient approaches are possible in practice. Since CFD φ solves the energy equation for the indoor air, a CFD S = source term φ program can be extended to solve heat transfer in solid materials, such as building enclosures, with an Multiple concentrations, C, can be used to simulate appropriate radiation model. The convective heat different species, such as water vapor and various transfer is then calculated directly in the simulation. contaminants. For buoyancy-driven flows, the This is the conjugate heat transfer method. Some Buossinesq approximation, which ignores the effect researchers have applied this method to integrated of pressure changes on density, is usually employed. calculations (e.g. Holmes et al 1990, Chen et al 1995, The buoyancy-driven force is treated as a source Moser et al 1995, Schild 1997). This approach