Printer Friendly

Net-zero Energy Impact Building Clusters Emulator for Operation Strategy Development.


Buildings are responsible for over 40% of primary energy consumption in the U.S. Among that consumption around 30% of the energy used in buildings is consumed by the heating, ventilating and air conditioning (HVAC) systems (DOE 2011). It is also estimated that 4% to 20% of the energy used in HVAC and lighting systems is wasted due to equipment and operation problems (Katipamula and Brambley, 2005). Research has shown that with optimal control strategies, building energy consumption can be reduced by 10% to 30% and energy cost can also be saved by 7% to 15% by utilizing thermal storage (Henze et al., 2005). Therefore, there is a great need to develop better building control and operation strategies to improve building energy efficiency and occupant comfort. Because it is expensive to develop and validate building control and operation strategies in real building, it is necessary to develop a testbed that can simulate real-world building system behaviors under different control and operation strategies.

Currently, with the development of smart grid, the power infrastructure in U.S. is experiencing a revolutionary transformation, from a centralized one-way communication to a decentralized network with two-way communication. With the trend of moving from centralized building operation decision to decentralized operation control, it is envisioned that neighboring buildings will have the tendency to form a building cluster, within which smart grids, distributed power generation, and storage devices, can freely share energy resource locally and globally and the entire cluster will achieve maximum energy efficiency. It is anticipated that this building cluster concept will fundamentally transform the energy industry by shifting expensive on-site energy generation aimed at creating single NetZero building one-at-a-time to an autonomous and adaptive system of buildings aimed at NetZero clusters. Now most of the existing researches are focusing on single building control and operation development. Therefore, existing building energy system models and testbeds are all developed for single building control and operation studies. For example, EnergyPlus (Crawley et al., 2001) is developed to provide detailed building operation simulation. PV panels have been modeled by De Soto et al. (De Soto etc., 2006). Ihm et al. (2004) developed an ice tank thermal energy storage model for EnergyPlus. Although rich literature exists about buildings, HVAC systems, energy generation and storage devices simulation, few publications have been found studying multiple buildings' operation to improve their overall energy efficiency. Hu et al. (Hu et al., 2012) developed a multibuilding simulation model. In that study, however, all the components are modeled using very simplified equations. There is a need to develop simulation testbeds that can be used to simulate real-world NetZero building cluster, which includes multiple buildings, shared energy generation and energy storage devices, and building control and operation systems.

The objective of this study is to take advantage of existing physics based building and energy system simulation models to develop an emulator that can represent a cluster of buildings along with their energy generation and storage devices. Detailed physics based simulation models, such as those developed in EnergyPlus and in another building energy simulation software package B (Beckman etc., 1994) environment, were reviewed and selected. Those models that are detailed enough to be used to assess the effectiveness of various control strategies, and that have been experimentally validated, were selected. After various models were assembled to form an emulator, a proof-of concept emulator operation demonstration was illustrated. In this illustration, only two buildings, one ice tank, one PV panel, and one battery device, are included in the cluster, but the emulator has the flexibility to be extended to include more buildings and more devices.


The emulator framework is illustrated in Figure 1. Overall, there are 4 modules in the emulator: Building module, Ice tank module, PV-Battery module, and Control module. The buildings in the Building module share the PV-Battery and the Ice tank module. Building, energy generation, energy storage and optimization models are developed in different software. EnergyPlus is chosen to simulate the buildings and ice tank thermal storage devices, because it is widely used and validated to provide detailed simulation results at a minimum one minute time step. PV panel power generation and battery system is modeled in energy simulation environment B by its default Types. Control module, which can include any control/operation strategies that need to testbed resides in a computation environment A (MathWorks, 2012), which is a high-level language and interactive environment for numerical computation, visualization, and programming environment. A decentralized cluster operation module is developed in this computation environment. A Virtual TestBed (Wetter and Haves, 2008) is served as a middleware to connect EnergyPlus, energy simulation environment B and computation environment A in which control and operation strategies can be determined.

The emulator can accept control and operation strategies provided from the computation environment A environment (and BACnet in the future) and simulate the operation of various systems and devices included in the cluster. The inputs and the major outputs of the emulator are summarized in Table 1. At each simulation time step, the emulator can provide other real-world-like "noisy" and "noise-free" measurements based on user requirements, such as energy generation and energy storage devices' performance data, etc. The control signals sent from the Control module are building heating and cooling setpoints (Tseph, Tsepc), operation energy generation and storage (Si, Sb, Spv), electricity purchasing and selling (Ep, Es), out of which the first two are sent to the Building module, Si is sent to the Ice tank module, and Sb and Spv are sent to PV-battery module, Eb and Es are sent to power grid module. Meanwhile, building measurements (Eb, Tzone, etc.) are output from Building module, ice tank measurements (SOC_i, Tchlw, Mchlw, and Eice) are output from Ice tank module, PV-Battery measurements (Epv, Ebat, and SOC_b) are output from PV-Battery module, Ptou, Pc and Pe are output from Power grid module. All these outputs will be transferred to Control model to assess the control strategies at each time step and a set of new control signals will then be sent to the cluster.


Building Module

Although the testbed can include any number and any type of EnergyPlus building model, in this paper, simulation models for two different buildings are developed using the EnergyPlus. Two typical medium-size office building models in Philadelphia which were developed in a previous study are used (Hendricken et al. 2012). One of these two buildings is a one-story, 5,000 square feet commercial building and the other one is a three-story, 15,000 square feet commercial building. The window to wall ratio for both buildings is approximately 0.29. The windows are of various single and double pane construction with 0.118 inch (3 mm) and 0.236 inch (6 mm) glass and either 0.236 inch (6 mm) or 0.512 inch (13 mm) argon or air gap. The U-factors of the windows are 1.0 Btu/hr-[ft.sup.2]-[degrees]F (0.173 W/[m.sup.2]-K) and 0.5 Btu/hr-[ft.sup.2]-[degrees]F (0.086 W/[m.sup.2]-K), respectively. Both buildings have deck roofs with R-15 insulation (solar absorptivity of 0.9). The summary of the mechanical systems are tabulated in Table 2. The first building system is a single duct constant-air-volume (CAV) roof-top units (RTUs) system and the other one is a single duct CAV air-handling units (AHUs) system.

PV-Battery module

As shown in Figure 2, the PV-Battery system module contains a PV panel model, a battery model, a power grid model, a local operation controller model and an inverter model. They are modeled individually and are interconnected in energy simulation environment B. The PV panel is modeled by Type 194 in energy simulation environment B (De Soto, etc., 2006), which is based on the diode equivalent circuit model to calculate the power generation rate, outlet current, voltage and so on. In Figure 2, weather condition model reads in weather data, such as outdoor temperature, solar radiation, at regular time intervals from typical meteorological year weather file, and converts them into a desired system of units which generate all the weather variables that PV panel model needs. The weather file used in this module is the same as the one used on Building module, Ice tank module. Both of them are the same TMY weather files for Philadelphia. Of course, such TMY weather file can be replaced with real weather condition files and for other locations. In this project, the PV panel for each building can be in one of the following four states: charging battery, powering building, selling power to grid or being dormant. The state control signals are determined by an operation controller model (a component in B) which is connected with the Control module. A lead-acid storage battery is modeled by a new defined Type 47b in B. It specifies how the battery state of charge varies over time, given the rate of charge and discharge. Similar to PV panel power generator, the states of battery can be in only one of the following three states: charging from the PV panel, charging from the power grid, discharging to power the building.

Ice Tank Thermal Storage Mode

Ice tank thermal energy storage system is another building energy management equipment. Building operator can use the ice tank to shave the high electricity demand from cooling load during peak hours associated with real time electricity price. There are two different ice tank thermal storage objects in EnergyPlus: simple model and detailed model. The detailed ice storage object in EnergyPlus is used in this project. This detailed model allows user defined charging and discharging curves to model a specific ice storage device more closely (DOE, 2013).

In this study, the ice storage system is modeled in an individual EnergyPlus model (Figure 2). Chilled water generated from the ice storage system is sent to the different buildings separately. The ratio of the chilled water mass flow rate for each building is determined by the Control Module. The cooling load of each building is covered firstly by the chilled water from the shared ice storage tank. And the remaining cooling need is satisfied by the base chillers of each building. A dedicated chiller is used to charge the ice storage system. The charging and discharging schedule is controlled by the Control module through dedicated chiller chilled water outlet temperature setpoints (DOE, 2013). The overall schematic of the ice storage system is illustrated in Figure 3 (a). Key parameters of the ice storage system modeled in this study are shown in Table 3. Because it is very difficult to actually pass parameters among different EnergyPlus models, the following schemes are used to mimic the mass and heat transfer between the ice storage system and the buildings described above. A user defined "load profile" model is added in the ice storage tank chilled water loop to represent the chilled water needs from the buildings. The chilled water from ice storage tank will cover the load request in this "load profile", representing the coverage of the building's cooling requests, which is also determined by Control module.

To represent the cooling provided by the ice storage system in each building model, a new user defined component, "ice cooling", is created. The overall function of this user defined component is to cover some part of the cooling load of each building provided by the ice storage tank. Then the remaining cooling needs is covered by the base chillers in the building model. The building chilled water loop configuration is illustrated in Figure 3 (b). Similar to a series chiller configuration, the new "ice cooling" component is in a series configuration before the base chilled. The "ice cooling" outlet chilled water temperature is determined in Eq. 1:

[T.sub.CWout] = [T.sub.CWin] - []/([M.sub.CW] * [CP.sub.CW]) (1)

Where, [T.sub.CWout] is the chilled water temperature at the outlet of the ice cooling component, which is also the chilled water temperature at the inlet of base chiller; Tcwin is chilled water temperature at the inlet of the ice cooling component; [] is the cooling load that needs to be covered by the ice storage system and is determined by Control Module; [] is chilled water mass flow rate in the chilled water loop; CPcw is the specific heat of chilled water.

Virtual power generator model in EnergyPlus

Since the PV panel and battery system is modeled in energy simulation environment B, a virtual power generator model is created in the building models (in EnergyPlus) to represent the amount of electricity passed from the PV-battery module. This virtual power generator will be enforced by "TrackSchedule" scheme (DOE, 2013) in EnergyPlus to generate the same amount of power as that simulated in the PV-Battery module which is used to power building.

Power grid module

In this study, a simplified power grid model is also developed to provide the time-of-use electricity price and to calculate the electricity cost and earning based on the amount of electricity buying from and selling to the power grid. This information will be sent to the Control module to assess and determine an optimized control/operation strategy. The following equations are used in this model:

[C.sub.g] = [m.summation over (i=1)][H.summation over (j=1)] ([E.sub.p,i,j][P.sub.tou,j] - [E.sub.s,i,j][P.sub.s,j]) (2)

Where, [C.sub.g] is the net electricity cost ($), m is the number of the buildings in this cluster emulator, H is the building operation time, and all other variables have been introduced in Table 1. [E.sub.p,i,j][P.sub.tou,j] calculates the energy purchasing cost ($), and [E.sub.s,i,j][P.sub.s,j] calculates energy selling earnings from power grid ($).


In this project, all different modules/models are connected with the computation environment A through a Building Controls Virtual TestBed (Wetter and Haves, 2008), as shown in Figure 4. "EnergyPlus I" and "EnergyPlus II" simulators are the two building models. The "Ice tank" simulator is the shared ice storage tank EnergyPlus model. The "A" simulator includes the Control module and PV-Battery model connector (in energy simulation environment B), which will provide operation and control signals and call model in energy simulation environment B every time step. To be more specific of the connection, an external interface is created in these three EnergyPlus models, which is used to connect EnergyPlus and the virtual test bed C. Once EnergyPlus model and the virtual test bed C are connected, the control signals (Input) for Building Module are transferred though this interface from the virtual test bed C. Meanwhile, the control signals from Control module are sent to the virtual test bed C through a shared socket connection from computation environment A. The energy simulation environment B is connected to the computation environment A through modeling calling commands and simulation results from energy simulation environment B are imported into computation environment A simultaneously. Control signals for PV-Battery model are injected into their models' "dck" files at each time step.


A proof-of-concept case study is conducted to verify the emulator, especially whether inputs and outputs are connected successfully, whether the data is exchanged correctly, and whether the testbed generates reasonable results. The weather file used in this case study is July 15th from the Philadelphia TMY file.

The building heating and cooling setpoints vary based on the following schedule: 1) 68 [degrees]F (20 [degrees]C) and 71.6 [degrees]F (22 [degrees]C), respectively from 8 am to 6 pm, while 64.4 [degrees]F (18 [degrees]C) and 82.4 [degrees]F (28 [degrees]C) for the rest time of the day for building I, 2) 71.6 [degrees]F (22 [degrees]C) and 75.2 [degrees]F (24 [degrees]C), respectively from 8 am to 6 pm, and 64.4 [degrees]F (18 [degrees]C) and 86 [degrees]F (30[degrees]C) for the rest time of the day for building II. The ice tank starts with fully charged status at the beginning of the day and starts to be discharged to provide cooling from 9 am until its depleted and then the dedicated chiller starts to charge the tank at 8 pm. The operation scheme for PV-Battery Module in this test is summarized in Table 4. PV panel will stay dormant at first and then power building until sunset. Battery will be charged by power grid at mid-night and stay full until noon, and then it starts to power building to shave building electricity demand at peak hours until it is deplete.

Figure 5 shows the temperature simulation results of building I. The zone temperature (blue line) is controlled based on the predefined temperature setpoints. Different temperature setpoints have been applied in building II. Similar to the case of building I, the zone temperature is also controlled based on the set point schedules. According to the control signal, sixty percent of PV power is provided to building I, which is the same as it is shown in Figure 6. The shared ice storage tank covers part of the cooling load for building I and building II from 9 am to 21 pm, until it is depleted. Figure 7 shows that 60% of the discharged chilled water is sent to building I as required by the Control module. Although not plotted here, more than 98% of the discharged ice chilled water energy is received by buildings. This verification case shows that the testbed connection and signal passing are successful.


An emulator testbed that allows the simulation of a building cluster as well as their energy generation and storage devices is developed in this study. In this testbed, multiple buildings are connected and are able to share common energy generation devices such as a PV panel as energy generation devices, and battery and ice tank as energy storage system. Various simulation environments, including EnergyPlus, building simulation environment B, are used to model different part of the building cluster and are connected by the virtual testbed C to the computation environment A. A proof-of-concept test case is designed to illustrate the use of this testbed and to verify the testbed results. This emulator testbed is able to assess building cluster operation strategies and provide real-world-like building cluster operation data. Future work of this study will focus on developing the connection of this testbed with a real BACnet interface and further with real building control systems.


Financial support provided by the U.S. National Science Foundation Award 1239247 is greatly appreciated. Special thank goes to MATLAB (A), TRNSYS (B), and BCVTB (C) for the software use in this paper.


Beckman, W. A., Broman, L., et al. (1994). TRNSYS The most complete solar energy system modeling and simulation software. Renewable Energy, 5(1-4), 486-488.

Crawley, D. B., Lawrie, L. K., Winkelmann, F. C., Buhl, W. F., Huang, Y. J., Pedersen, C. O., . . . Glazer, J. (2001). EnergyPlus: creating a new-generation building energy simulation program. Energy and Buildings, 33(4), 319-331.

DOE, U. S. (2011). Buildings Energy Data Book, Retrieved 07.01, 2013.

DOE US, EnergyPlus Engineering Reference. Input and Output Reference, 2013: p. 6.

De Soto, W., Klein, S., & Beckman, W. (2006). Improvement and validation of a model for photovoltaic array performance. Solar Energy, 80(1), 78-88.

Henze, G. P., D. E. Kalz, et al. (2005). Experimental analysis of model-based predictive optimal control for active and passive building thermal storage inventory. HVAC&R Research 11(2): 189-213.

Hu, M., J. D. Weir, et al. (2012). Decentralized operation strategies for an integrated building energy system using a memetic algorithm. European Journal of Operational Research 217(1): 185-197.

Ihm, P., Krarti, M., & Henze, G. P. (2004). Development of a thermal energy storage model for EnergyPlus. Energy and Buildings, 36(8), 807-814.

Katipamula, S. and M. R. Brambley (2005). Review Article: Methods for Fault Detection, Diagnostics, and Prognostics for Building Systems--A Review, Part I. HVAC&R Research 11(1): 3-25.

Liam Hendricken, Kevin Otto, et al. (2012). Capital Costs and Energy Savings Achieved by Energy Conservation Measures for Office Buildings in the Greater Philadephia Region. SimBuild Madison, WI.

MathWorks (2012), MATLAB Release 2012b, The MathWorks, Inc., Natick, Massachusetts, United States.

Wetter, M. and P. Haves (2008). A Modular building controls virtual test Bed for the Integration of heterogeneous systems. Proceedings of 3rd Nation Conference of IBPSA-USA SimBuild Berkeley, California.

Xiwang Li Student Member ASHRAE

Jin Wen Member ASHRAE

Teresa Wu

Xiwang Li is a Ph.D student, Jin Wen is an associate professor in the Department of Civil, Architectural and Environmental Engineering, Drexel University, Philadelphia, PA. 19104. Teresa Wu is an associate professor of Industrial Engineering at Arizona State University, Tempe AZ. 855287
Table 1. Building Cluster Input-Output Summary

Sub-Model           Input

Building model      Tseph (Heating setpoint)
                    Tsepc (Cooling setpoint)
Ice tank            Si (State of ice tank storage)
Battery model       Sb (State of battery)
PV model            Spv (State of PV panel)
Power grid model    Ep (Power buying from power grid)
                    Es (Power selling to power grid)

Sub-Model           Major Output

Building model      Eb (Building energy consumption), Edc (Dedicated
                    chiller energy consumption), Ebc (Base chiller
                    energy consumption), Tzone (Building zone
                    temperature), Hzone (Building zone humidity)
Ice tank            SOC_i (ice tank state of charge), Eice (charging and
                    discharging rate),
                    Tchlw (chilled water in/out temperature), Mchlw
                    (chilled water in/out temperature flow rate)
Battery model       SOC_b (battery state of battery), Ebat (charging and
                    discharging rate),
                    Ibat (charging and discharging current)
PV model            Epv (Power generation)
Power grid model    Ptou(Time-of-use price), Pc(electricity costs),
                    Pe(electricity earnings)

Table 2. Building Mechanical Systems

Component             Building I         Building II

Main Cooling Coil     DX, COP 3          Chilled water
Main Heating Coil     Hot water          Hot water
Zone Reheat           Hot water          Eclectic
Heat Plant            Central Boiler     Central Boiler

Table 3. Ice Thermal Storage Tank and Dedicated Chiller Parameter

Parameter                      Value             Unit

Ice Tank Capacity              15.8(0.2)         Tonh(GJ)
Tank Loss Coefficient           0.0003           --
Freezing Temperature           32(0)             F(C)
Dedicated Chilled Capacity      2.1E+4 (7000)    Btu/h (W)
Dedicated Chilled COP           3.2              --

Table 4. PV-Battery Module Testing Operation Scheme

Time        PV Panel Operation             Battery Operation

 0-8 am     Dormant                        Charging from power grid and
                                           stay full
 8-12 pm    Power building                 Stay full
12-15 pm    Power building                 Power building until deplete
15-20 pm    Power building until sunset    Power building until deplete
20-24 pm    Dormant                        Stay deplete
COPYRIGHT 2014 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 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Li, Xiwang; Wen, Jin; Wu, Teresa
Publication:ASHRAE Conference Papers
Date:Jun 22, 2014
Previous Article:A Tool for Evaluating Fault Detection and Diagnostic Methods for Fan Coil Units.
Next Article:Development of a Probabilistic Graphical Energy Performance Model for an Office Building.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters