Printer Friendly

Mechanically Ventilated Double Skin Facade with Semi-Transparent Photovoltaics, Implementing Electrical Storage and Heat Pumps to Reduce Peak Demand.

INTRODUCTION

In both the US and the European Union, the building sector has the highest electricity consumption among all the other sectors (Argonne et al., 2017; European Comission 2019). Buildings consume more than 55% of the total electricity consumption in Canada, while Quebec is the province with the highest electricity consumption among other provinces, being responsible for 35% of the total electrical consumption (NRCAN 2019). To reduce the power demand during winter peak times, Hydro-Quebec has initiated a demand response program and a dynamic electricity pricing (Hydro Quebec 2019).

The double skin facade (DSF) consists of an exterior and an interior skin that are separated by an air cavity that creates a buffer space around the building (Poirazis 2005). Photovoltaic panels can be integrated in the exterior skin of the DSF, and increase the area of the building which can generate electricity (Ioannidis et al. 2017). Figure 1 depicts a schematic of a DSF integrating STPV (DSF-STPV) and its thermal network. The STPV is integrated in the exterior layer of the DSF, while the interior layer consists of an insulating glazing unit (IGU). STPV, allow light to pass through and generate electricity at the same time, increasing the energy performance of the building (Kapsis and Athienitis 2015).

A mechanically ventilated DSF, can take advantage of the preheated air from the cavity and introduce it into the building or into the HVAC system to enhance its energy performance. A DSF can help regulate the temperature of the envelope components that exchange radiation with the occupants, allowing in this way the occupants to operate closer to the window area while maintaining their thermal comfort.

Electric utilities are considering the adoption of battery storage to shift the peak demand and avoid the upgrade of the electrical grid in order to accommodate the peak demand. In cold climates, such as the location considered in this study (Montreal, latitude 45 N), during the winter months, the peak demand hours are between 7am and 10am and 4pm and 8pm. For a typical winter day in the US, there is an increase of 20% and 10% in the electricity demand is the morning and the evening peak respectively (U.S. Energy Information Administration 2019).

The objective of this study is to shift the peak demand of the building out of the grid peak hours by utilizing electric storage and different heating strategies and reduce the electricity demand during the grid peak periods.

MODELING

In this paper, a DSF integrating STPV (DSF-STPV) is modeled based on a transient finite difference mathematical thermal network (Figure 1). The model also implements an electric battery that gives the possibility to offset the peak demand and model the grid interaction with the building. The fan driven flow inside the cavity of the DSF-STPV is modeled and the wind and buoyancy effects that assist the flow are taken into consideration. The numerical model for the assessment of the energy performance of a multi-story DSF-STPV has been implemented in MatLab (Mathworks) and models the heat-pump (HP) based on available in the market models (Ioannidis 2016; Ioannidis et al. 2017).

DSF-STPV thermal and flow network

The simulation model is based on a lumped parameter thermal network approach and takes into consideration all of the heat transfer mechanisms that happen within the cavity of the DSF-STPV. A set of explicit finite difference equations are obtained for each node of the thermal network. The cavity of the DSF-STPV is discretized into 46 control volumes. The multi-story building on which the DSF-STPV is integrated on is also divided in thermal zones, equal to the number of the floors. A more detailed description of the model is provided in previous publications of the authors (Ioannidis 2016; Ioannidis et al. 2017). The flow of the air within the cavity is mechanically assisted and the pressure drop due to the mechanical system is calculated by equation (1) where the used orifice equation employs a discharge coefficient of [C.sub.d] = 0.62. The pressure drop due to natural effects ([[DELTA]P.sub.nat]) is the summation of the stack and wind effects. The pressure difference due to stack effect is given by equation (2), while the pressure drop due to wind effect is given by equation (3). The pressure coefficients, determined experimentally by Lou et al. (2012) are used to calculate the difference between the pressure coefficients at the exterior and the interior skin of the DSF-STPV ([DELTA][C.sub.p]).

[mathematical expression not reproducible] (1)

[mathematical expression not reproducible] (2)

[mathematical expression not reproducible] (3)

DSF-STPV flow coefficients

Local Nusselt number coefficients (Nu) are used in order to represent the convective heat transfer coefficients along the channel height. The Nu correlations developed by the authors are used and for the brevity of this study, the procedure followed for the development of the Nu correlations is briefly presented here and is presented in more details in another study. A set of energy balance equation is solved to calculate the convective heat transfer coefficient towards the side of the IGU ([h.sub.cb]). Then, this calculated convective heat transfer coefficient towards the side of the IGU ([h.sub.cb]) is used to calculate the convective heat transfer coefficient towards the side of the STPV ([h.sub.ca]). These convective heat transfer coefficients are later used to calculate the Nu correlations. The authors developed different Nu correlations for the two boundary sides of the flow, the side towards the STPV ([Nu.sub.STPV]) and the side towards the interior layer, which in most of the DSF cases are insulating glazing units ([Nu.sub.IGU]), as it can be seen in Figure 1.

For the side of the cavity towards the STPV, the average Nu correlation was developed as a function of the wind heat transfer coefficient ([h.sub.c,out]), the Reynolds number, the efficiency of the STPV, the incident solar radiation and a factor that depends on the absorbed solar radiation by the glass section of the STPV (1-0.339 [[tau].sub.STPV]). Similarly, the Nu correlation for the interior side of the DSF-STPV, includes the thermal transmittance of the IGU ([U.sub.glazing]), the Reynolds number and the overall solar transmittance of the STpv. In this way, the losses from the building to the DSF-STpv over the gains from incident solar radiation are multiplied by the transmittance of the STPV. Later, by using the definitions of Nusselt and Reynolds number, the local convective heat transfer coefficients were calculated.

In most wind-driven flows around buildings, turbulent forced convection is observed and the convective heat transfer coefficient correlations are function of the wind velocity, the wind direction and the surroundings of the concerned area (Vasan and Stathopoulos 2014). The exterior convective heat transfer coefficients (hout) are calculated based on the correlation developed by Sharples and Charlesworth (1998), where Vwmd is the wind velocity.

[mathematical expression not reproducible] (4)

[mathematical expression not reproducible] (5)

[mathematical expression not reproducible] (6)

STPV and battery modeling

To model the STPV integrated on a DSF, it is assumed that the PVs operate at the maximum power point condition in each moment and the STPV efficiency is assumed to be linearly dependent on the operating temperature of the PV cells, on the PV cell efficiency at standard test conditions ([n.sub.STC]), on the PV temperature coefficient ([[beta].sub.PV]) and the PV cell temperature under standard test conditions ([T.sub.STC]).

[n.sub.PV] = [n.sub.STC] x [1 - [[beta].sub.PV] x ([T.sub.PV] - [T.sub.STC])] (7)

The STPV is modeled under the assumption that it consists of two different components; the opaque and the transparent one, with the opaque being the PV cells and the transparent being the glass between the PV cells of the STPV. In addition, the PV cells are arranged in strips thus, opaque and transparent strips are modeled.

The electric battery charging and discharging modes are modelled by implementing their characteristic curves of a Li-ion battery (Andwari et al. 2017). Its charging and discharging behavior is modelled by means of an equivalent circuit by considering a scaled model associating the battery internal charge curve with the operating power(Tremblay et al 2007). Specifically, the behavior of a battery cell is modelled with respect to the discharge/charge power and state of charge (by considering a constant current for both the charging and discharging modes).

Model validation

An outdoor test-room with a custom-built DSF-STPV module is installed in Montreal, Canada (45[degrees] 30' N / 73[degrees] 35' W). The STPV module has a packing factor (or covering factor--ratio of area covered by solar cells--i.e. opaque) of 63.4%. The cavity of the DSF-STPV is 17.5cm wide and 1.95m high. The insulating glazing unit that comprises the interior surface of the DSF has a hard low-emissivity coating and has a 13 mm wide air-gap that is filled with Argon. It has a SHGC of 0.624 and its U-value ranges between 1.64 W/[m.sup.2]K and 1.56 W/[m.sup.2]K for winter and summer conditions.

The STPV is the exterior layer of the DSF and integrates 48 monocrystalline cells with a nominal efficiency of 17.80% and the temperature is measured with thermocouples, mounted on the different layers of the DSF and in the air space along the z-axis direction at 15 equally distributed heights. The wind velocity and direction is measured at the weather station close to the test location and the data are averaged per 15 minutes. The experimental facility that is used for the development of the Nu correlations, is also used for the validation of the model. For the experimental validation, the ambient temperature ranges between -5[degrees]C and 30[degrees]C. The Reynolds number ranges between 10,000 and 26,000 and the incident solar radiation between 250 W/[m.sup.2] and 1200 W/[m.sup.2]. A validation is performed for different Reynolds numbers, ambient temperatures and incident solar radiation with time intervals of 15 minutes.

Using the Nu correlations presented in a previous section, a model validation is conducted for the temperatures of the air at the outlet of the DSF-STPV. The measured temperatures of the STPV and the IGU are used as boundary conditions and the measured airflow is used as the mass flow rate within the cavity to predict the outlet temperature of the DSF-STPV. The measured temperature of the air at the outlet of the DSF-STPV when compared with the predicted outlet temperature of the DSF-STPV presents a difference that does not exceed 1[degrees]C. This difference appears negligible for warmer ambient temperature conditions and lower incident solar radiation.

CASE STUDIES

A double skin facade integrating semi-transparent photovoltaics (DSF-STPV), is simulated for a 3-story south facing DSF system located in Montreal, Canada. This case study falls under the 6th climate zone of ASHRAE (ASHRAE 2019). The DSF-STPV has a height of 3.28 m per floor and consists of 14 rows of PV cells (0.16 m height each) and 13 rows of glass (0.08 m height each). This results in a packing factor of 68% and an effective transparency of around 29%. The PV cell efficiency is assumed to be 20%. The DSF-STPV has a width of 3.6 m and the cavity gap is 0.4 m. Because of the size of the cavity of the DSF-STPV and the smoothness of the materials of the layers, which are glasses, the pressure drop inside the cavity is small. With less than 5 Pa pressure drop along the cavity, the impact on the total fan power will not be significant but is taken into consideration in the simulations. Each thermal zone has the same width with the DSF-STPV (3.6 m) and is 8.2 m deep, following the dimensions of a typical reference office (Reinhart et al. 2013). The interior side of the facade is a curtainwall and consists of an IGU with U-value equal to 2 W/[m.sup.2]K.

The DSF-STPV is coupled with an electrical storage and the battery that is selected for this case study has a capacity of 9.9 kWh (3.3 kWh per floor). The maximum charging power is 5 kW while the maximum and the minimum total capacity is set to 95% and 10% respectively. The DSF-STPV is also coupled with a heat pump that has a maximum capacity of 2.5 kW for each zone, meaning 7.5 kW for three floors. The heat pump performance data are provided by manufactures and are in accordance with the standard EN14511 and it includes defrost cycles losses. Each indoor space is assumed to have a plug load factor of 3.55 W/[m.sup.2], and each office/workstation is 12.5 [m.sup.2], while all the occupants use notebooks and one printer is assigned for every ten occupants (Sarfraz 2018).

Strategies

A series of strategies are adopted to reduce the electricity consumption for heating and to achieve the shift in the peak demand from the grid. The dampers that allow air to enter the DSF-STPV are set to open when the incident solar radiation is greater than 200 W/[m.sup.2]. This threshold is selected to minimize the losses from the building and to enhance the heat recovery from the STPV. For the supply of the heat pump with preheated air by the DSF-STPV and not mix it with the colder ambient air, the velocity set-point of the air flowing within the cavity of the DSF-STPV must be selected appropriately. The velocity set-point for the air flowing within the cavity was selected at 0.4 m/s.

To reduce the electricity consumption for heating, during the grid peak hours, the setpoint of the temperature of the thermal zone varies depending on the time. A new adjusted temperature schedule is introduced where the temperature set-point is at 22[degrees]C from 5 am to 7 am and at 26[degrees]C from 11 am to 5 pm with a temperature set-back at 18[degrees]C. The typical schedule for heating a commercial building has a set-point at 22[degrees]C from 7 am to 7 pm with a temperature set-back at 18[degrees]C. The aim of the new schedule is to consume electricity and preheat the building prior the peak in the grid and use the stored electricity in the battery for the grid peak hours. At the same time, the temperature set-point of the adjusted schedule is selected so that it does not create discomfort to the occupants.

RESULTS AND DISCUSSIONS

Figure 2 depicts, the ambient temperature, the incident solar radiation, the average temperature of the air within the cavity of the DSF-STPV and the temperature of the air at the outlet of the DSF-STPV. The time period investigated in this study (January 12th to 15th), presents a series of sunny and semi-overcast days, followed by an overcast day. The ambient temperature fluctuates between -30[degrees]C and -15[degrees]C. The temperature rise between the ambient temperature and the temperature at the outlet of the DSF-STPV reaches more than 10[degrees]C for the colder days and during the peak of the incident solar radiation, while the average temperature of the cavity is always 2.5[degrees]C warmer than the ambient air, creating in this way a buffer zone and decreasing the heating losses of the building. The incident solar radiation on the DSF-STPV reaches 800 W/[m.sup.2], for the sunny days, while for the overcast days, it barely reaches 200 W/[m.sup.2]. The mechanical system that assists the flow within the DSF-STPV operates when the incident solar radiation is greater than 200 W/[m.sup.2].

In Figure 3 the interaction of the system with the grid is shown. The STPV electricity production, the total building consumption and the consumption of the electricity from the grid is depicted for the days chosen for this study. For the hours prior to the morning peak and after the evening peak period (before 7 am and after 8 pm) the consumption of the building coincides with the consumption of electricity by the grid. During the peak periods, the consumption of electricity by the grid is decreased and the battery is used to drive the heat pump. The heating demand during the grid peak hours is also decreased as the strategy to heat up the building prior the peak demand hours is followed. The battery is mostly discharged on the overcast day, while a full discharge cycle takes around three days. The temperature of the thermal zone increases before the grid peak demand hours and is let free to float during them. This pre-heating prior the grid peak hours can maintain the temperature of the zone above the set point for two to three hours. For the< remaining grid peak hours, the stored electricity can run the heat pump to keep the zone over the set-back.

In Figure 4 the electricity consumption of the building and the coefficient of performance (COP) of the heat pump is shown. The advantage that the DSF-STPV provides to the system by preheating the air that passes through the cavity can be seen in this figure as the increase of the COP at the time around noon is significant. The COP can reach values more the 3, thus the electricity consumption for heating decreases a lot, allowing at the same time the electricity produced by the STPV to recharge the battery. In Figure 4, the velocity of the air within the cavity is also plotted and signifies the moment of the time that the preheated air from the DSF-STPV is used to increase the COP of the HP.

This increase in the COP is due to the higher ambient temperature and to the increased temperature at the outlet of the DSF-STPV. The temperature of the air at the outlet of the DSF-STPV is mixed with ambient air prior supplying the heat pump. The selection of the velocity set-point within the cavity of the DSF-STPV is selected to decrease the mixing with the colder ambient air and therefore increase the COP. The percentage of the air supplied to the heat pump by the ambient air and the DSF-STPV is shown in Figure 5.The DSF-STPV can fully supply the heat pump for certain periods, while depending on the need and the available heat, the DSF-STPV can supply the heat pump with preheated air at percentages that exceed 65%. There is an increase that ranges from 4[degrees]C to 10[degrees]C between the ambient air, that would normally be used, and the supplied air that is preheated by the DSF-STPV depending on the incident solar radiation. During the night-time, the heat pump is supplied solely by the ambient air, because the DSF-STPV is closed and acts as a buffer zone.

GRID INTERACTION

A comparison between a sunny day and the overcast day is shown in Table 1. In this table, the difference between these two schedules is presented with respect to the reduction of the peak demand, for the specific geometry simulated but also as a percentage difference between the two schedules. The savings in the electricity consumption during the grid peak demand hours ranges from 10% to 90%, with saving of more than 80% for many hours.

CONCLUSION

A validated numerical model of a double skin facade integrating semi-transparent photovoltaics has been utilized to assess the impact on the grid of the coupling of such system with battery storage and a HP. A set of consecutive winter days, for the month of January have been investigated. The set of days is simulated for a three-story commercial building located in Montreal, Canada which is in the 6th zone of the ASHRAE climatic zones.

The mismatch between the electricity production by the STPV and the electricity needed for heating and lighting by the building zones is investigated and an adjusted heating schedule has been proposed in order to reduce the heating demand of the building. To take full advantage of the pre-heated air that flows within the cavity of the DSF-STPV, the heating set-point during the solar peak hours is increased. At this time, the temperature of the air flowing within the cavity, which later is introduced to the heat pump, is highest and thus the COP of the heat pump is increased.

With the use of such a predictive heating strategy, the peak demand of the building can coincide with the peak of the electricity production resulting in substantial reduction in the electricity consumption by the grid during the peak hours of the grid. This reduction can reach more than 80% even for cold and overcast winter days.

ACKNOWLEDGMENTS

The authors acknowledge the support of the Natural Sciences and Engineering Research Council of Canada (NSERC) through a NSERC/Hydro Quebec Industrial Chair.

NOMENCLATURE
hc:   convective heat transfer coefficient (W/[m.sup.2]K)
I:    Incident solar radiation (W/[m.sup.2])
Nu:   Nusselt number (-)
Re:   Reynolds number (-)
T:    Temperature ([degrees]C or K)
a:    exterior side of the DSF cavity
abs:  absorbance

b:    interior side of the DSF cavity
cv:   control volume
gl:   glass/glazing
gli:  glass inside
gk:   glass outside
p:    peak
rad:  radiation


REFERENCES

Andwari, Amin Mahmoudzadeh, Apostolos Pesiridis, Srithar Rajoo, and Ricardo Martinez-. 2017. "A Review of Battery Electric Vehicle Technology and Readiness Levels." Renewable and Sustainable Energy Reviews. Vol. 78. Elsevier Ltd.

Argonne, Zhou, Steve Capanna, Ookie Ma, Paul Spitsen, Ian Hamos, Dave Rench-, Patrick Gilman, et al. 2017. "2017 Renewable Energy Data Book."

ASHRAE. 2019. "ANSI/ASHRAE/IES Standard 90.1-2019 Energy Standard for Buildings Except Low-Rise Residential Buildings."

European Comission. 2019. "The Energy Performance of Buildings Directive."

Hydro Quebec. 2019. "GUIDE DU PARTICIPANT HIVER 2019-2020 Marches Commercial et Institutionnel Ainsi Que Petites et Moyennes Entreprises Industrielles."

Ioannidis, Z., A. Buonomano, A. K. Athienitis, and T. Stathopoulos. 2017. "Modeling of Double Skin Facades Integrating Photovoltaic Panels and Automated Roller Shades: Analysis of the Thermal and Electrical Performance." Energy and Buildings 154: 618-32.

Ioannidis, Zissis. 2016. "Double Skin Facades Integrating Photovoltaic Panels, Motorized Shades and Controlled Air Flow." Kapsis, Konstantinos, and Andreas K. Athienitis. 2015. "A Study of the Potential Benefits of Semi-Transparent Photovoltaics in Commercial Buildings." Solar Energy 115: 120-32.

Lou, Wenjuan, Mingfeng Huang, Min Zhang, and Ning Lin. 2012. "Experimental and Zonal Modeling for Wind Pressures on Double-Skin Facades of a Tall Building." Energy and Buildings 54: 179-91.

NRCAN. 2019. "Electricity Facts | Natural Resources Canada." 2019. https://www.nrcan.gc.ca/science-data/data-analysis/energy-data-analysis/energy-facts/electricity-facts/20068.

Poirazis, Harris. 2005. Single Skin Glazed Office Buildings Energy Use and Indoor Climate Simulations.

Reinhart, Christoph F, J Alstan Jakubiec, and Diego Ibarra. 2013. "Definition Of A Reference Office For Standardized Evaluations Of Dynamic Facade And Lighting Technologies." 13th Conference of International Building Performance Simulation Association, 3645-52.

Sarfraz, Omer, Christian K. Bach, and Christopher K. Wilkins. 2018. "Plug Load Design Factors." ASHRAE Journal 60 (1): 14-19.

Sharples, S, and P S Charlesworth. 1998. Full-Scale Measurements of Wind-Induced: Convective Heat Transferfrom a Roof Mounted Flat Plate Solar Collector. Solar Energy. Vol. 62.

Tremblay, O, L Dessaint, and A Dekkiche. 2007. "A Generic Battery Model for the Dynamic Simulation of Hybrid Electric Vehicles." In 2007 IEEE Vehicle Power and Propulsion Conference, 284-89.

U.S. Energy Information Administration. 2019. "Real-Time Operating Grid." 2019. https://www.eia.gov/beta/electricity/gridmonitor/dashboard/electric_overview/US48/US48. Vasan, Neetha, and Theodore Stathopoulos. 2014. "Experimental Study of Wind Effects on Unglazed Transpired Collectors." Solar Energy 101: 138-19.

Zisis Ioannidis, MASc

Student Member ASHRAE

Ted Stathopoulos, PhD

Member ASHRAE

Andreas K. Athienitis, PhD, PE

Fellow Member ASHRAE

Anamaria Buonomano, PhD

Zisis Ioannidis is a PhD student in the Department of Building Civil and Environmental Engineering, Concordia University, Montreal, Canada. Andreas K. Athienitis is a Professor of Building Civil and Environmental Engineering, Concordia University, Montreal, Canada. Ted Stathopoulos is a Professor of Building Civil and Environmental Engineering, Concordia University, Montreal, Canada. Anamaria Buonomano is an Assistant Professor at the Department of Industrial Engineering, University of Naples, Federico II, Naples, Italy.
Table 1 Difference Between the Typical and the Adjusted Schedule for a
Sunny and an Overcast Day.

                       Morning Peak
Hour            7        8            9        10
Sunny day      -5.4kW   -2.5kW         -2.4kW   -0.5kW
Overcast day   -4.1kW   -2.1kW         -2.0kW    0.4kW
                        Morning Peak
Hour            7        8              9       10
Sunny day     -89.1%   -76.9%         -88.5%   -20.6%
Overcast day  -91.2%   -66.8%         -68.2%    12.1%

              Evening Peak
Hour           16      17       18       19       20
Sunny day      -0.4kW   -0.5kW   -3.3kW   -3.4kW   -1.9kW
Overcast day   -0.2kW   -0.3kW   -2.8kW   -2.9kW   -2.5kW
               Evening Peak
Hour           16       17       18       19       20
Sunny day     -12.2%   -13.4%   -91.3%   -91.6%   -47.6%
Overcast day   -6.4%    -8.6%   -90.0%   -90.1%   -76.8%
COPYRIGHT 2020 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. (ASHRAE)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2020 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:VC-20-C015
Author:Ioannidis, Zisis; Stathopoulos, Ted; Athienitis, Andreas K.; Buonomano, Anamaria
Publication:ASHRAE Transactions
Geographic Code:1CANA
Date:Jul 1, 2020
Words:4152
Previous Article:Smart and Efficient Building Envelopes: Thermal Switches and Thermal Storage for Energy Savings and Load Flexibility.
Next Article:Improving the Interoperability of gbXML Data Model through Redefining Data Mapping Rules of HVAC Systems.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |