Full-scale measurement and validated simulation of cooling load reduction due to nighttime natural ventilation of a large atrium

Highlights

• Measured cooling load reduction due to nighttime ventilation of 24% to 32%.

• Dynamic thermal simulations predict the cooling load reduction with an error of 1.7%.

• Simulated indoor air temperature and bulk airflow are well predicted, with average error of 11.2% and 14.1%, respectively.

Abstract

In buildings with exposed internal thermal mass, natural ventilation (NV) can be used for nighttime pre-cooling (NVC) of the building for the next day. In many cases, these natural cooling systems are used in combination with conventional mechanical cooling systems (an approach known as hybrid cooling). The use of these systems is still limited, in part as a result of the lack of field studies that measure actual cooling load reduction that can be obtained when using an NVC system.

This paper presents the first measured energy savings due to NVC in a non-residential building. This study had two objectives: measure daytime cooling load reduction due to NVC in the previous night; and validate thermal and airflow simulation tools that are used in design of NV and NVC systems. The results are encouraging: comparison of the cooling demand of similar days with and without NVC in the previous night shows that, on average, the NVC system reduces the daytime cooling load by 27%. The dynamic thermal simulations were able to predict the cooling load reduction, internal air temperature and bulk airflow rates with average errors of 1.7%, 11.2% and 14.1%, respectively.

Introduction

The last decades have seen an expansion in the use of building mechanical cooling systems [1]. This increase is due to several factors such as global warming, higher internal heat gains, widespread use of unshaded glazed façades [2] and more stringent thermal comfort requirements. To reverse this trend designers are encouraged to use natural cooling strategies such as natural ventilation (NV) [3]. In buildings with exposed internal mass NV can be used for nighttime precooling of the building for the next day [4]. This heat dissipation passive cooling technique, known as night cooling by NV (NVC), relies on low nighttime air temperature, buoyancy and wind to achieve convective cooling by ventilation [5]. This paper presents an experimental and simulation-based analysis of NVC performance in a large atrium.

The effectiveness of NVC strategies varies with climate and building type [6]. To quantify these variations, a recent study proposed a climate cooling potential (CCP) indicator [7]. Existing CCP calculations indicate that, in central and northern Europe, NVC could eliminate the need for daytime mechanical cooling in spaces with high internal mass. In southern Europe the mean CCP per night for July is less than a third of the value for northern Europe. As a result, in this region, in the warmer months of the year, NVC alone cannot provide sufficient cooling and must be used in combination with mechanical daytime cooling (an approach known as hybrid cooling). The main quantitative indicator of a successful NVC system is its cooling capability. When NVC is used to complement mechanical cooling there should be a measurable reduction in cooling load. Similarly, for naturally conditioned buildings, there should be a reduction in daytime indoor air temperatures whenever NVC is used in the previous night.

There is a limited number of experimental studies of the impact of NVC in NV cooling capability [8]. An early comparative study using two similar rooms, with and without NVC, showed that NVC can lead to an average daytime maximum temperature reduction of up to 2 °C [9]. Recent field monitoring work of an office building exposed to a mild climate showed that, with low internal gains, an NVC system can keep indoor temperatures within the limits of the adaptive comfort model [10]. Although in some cases, NVC systems can suffer from overcooling [11], the most commonly found limitation is overheating. A recent field study in an academic building found that NVC systems can fail to meet indoor comfort requirements during heat waves and/or periods with high occupancy [12]. An experimental study based in a smart building showed that adequate control of NVC systems is central to consistent performance [13]. There is only one published study of measured cooling energy savings due to NVC use [14]. This large-scale experimental study analyzed energy data from two hundred fourteen air-conditioned residential buildings, concluding that NVC use resulted in a 26% reduction in the cooling load. To overcome the difficulties of obtaining measured data, several authors have performed studies with annual cooling energy savings predicted using simulation models that are validated using short measurement campaigns (a few hours of data with no measured energy savings). A recent validated thermal simulation-based study [15] showed that, for an office building in southeast China, optimal use of the cool roof and night ventilation can reduce the cooling energy consumption by 28%. A validated thermal simulation-based study [16] of the impact of NVC in supermarkets with high cooling demand showed that NVC can achieve a 17% reduction in cooling annual energy use.

Atrium spaces can be particularly suited to NVC systems due to the large stack height and, in many cases, less stringent thermal comfort requirements. In most cases, atriums are transitional spaces used predominantly for circulation and access to the building. A study of an atrium used as a transitional indoor space concluded that, even with direct inflow of outdoor air at 8 °C (below the typically acceptable minimum for NV systems), the conditions can still be thermally acceptable for circulation [11]. Atriums and other transitional spaces are known to require significant amounts of cooling energy in the summer [17]. A study that measured atrium overheating concluded, using simulation, that NVC can reduce cooling energy demand in atrium spaces with high solar gains [18]. Unfortunately, in some cases, designing a functional NV stack driven system can be challenging due to wind interference [19]. Clearly, NVC systems are potentially effective for atrium spaces with exposed thermal mass and high cooling demand. The relatively low vulnerability to overcooling that characterizes these transitional spaces is a further advantage for successful application of NVC. A review of the thermal and ventilation performance of atriums with NV systems concluded that the existing knowledge on atrium passive designs is incomplete given the complexity and lack of full-scale measurements [20]. A recent review of more than 90 research studies of hybrid cooling systems from 1996 to 2016 [21] identified the need for more experimental studies with field data and full-scale measurements. The abundance of simulation-based studies focused on energy savings contrasts with the lack of studies with measured energy savings due to NV or NVC in non-residential buildings. As a consequence, building system designers lack reliable and accurate data about expected performance and operation of low energy cooling systems [22]. In this context, two significant research gaps were identified in the area of energy savings due to NVC:

• there are no measurements of energy savings due to ventilation-based hybrid cooling systems in non-residential buildings;

• and, consequently, there is not field validation of the simulation tools used to model these systems in the design phase.

To address these research gaps and with the goal of contributing to increased use of NVC systems, this paper presents an experimental and simulation study of a hybrid-ventilated large atrium of a service building, exposed to the warm summer climate of Southern Europe. The atrium is mechanically cooled during the day by a chilled floor (CF) system and naturally cooled during the night by NVC. The study will focus on measuring the cooling load reduction due to the use of NVC and validating the thermal and airflow simulation tools that are used in the design of NVC systems. The next section describes the methodology of this study (Section 2). Section 3 describes the atrium, its NVC and CF cooling systems. Section 4 presents the computational fluid dynamics (CFD) simulations performed to estimate wind-induced pressure on the atrium façades. Section 5 presents the results of the atrium NVC system performance assessment, including a thermal comfort survey. Section 6 presents the building thermal dynamic simulations. The last section presents the conclusions of this study.