1、246 2010 ASHRAEABSTRACTThis paper investigates anticipatory control strategies ina house designed to have approximately net zero averageannual energy consumption. These strategies are particularlyuseful in the case of optimized solar buildings, which includeone or more of the following features: (a)
2、 passive solar design;(b) active systems for collection and control of solar energy(BIPV or BIPV/T systems, solar thermal collectors, motorizedblinds); and (c) thermal energy storage systems (water tanks,phase change material containers). At the supervisory controllevel, prediction of future conditi
3、ons can be employed to opti-mize the collection, storage and utilization of solar andgeothermal energy. At a lower control level, predictive controlassists in dealing with the discrepancies between the timeconstants of the building structure and its HVAC system, allow-ing the prescribed set-points t
4、o be reached when desired. Thispaper presents results of simulations which model the perfor-mance of a net-zero energy solar demonstration house usinganticipatory control techniques. A relatively simple thermalnetwork model is used throughout the buildings design and inthe development of the control
5、 strategies.INTRODUCTIONNet-Zero Energy Solar BuildingsBuildings using both passive solar design techniques andactive solar technologies integrated into the building envelopefor collecting solar energy as electricity, heat and daylight aredescribed here as “solar-optimized buildings”. Through opti-m
6、ized design and operation, solar buildings may achieve net-zero annual energy consumption (i.e., they generate as muchenergy as they consume over the course of a year). This isusually accomplished by connecting the building to the elec-trical utility grid: when the generation of the buildings renew-
7、able electricity system (usually photovoltaic panels) exceedsthe buildings consumption, the surplus is delivered to the grid.Conversely, when the power generation is insufficient for thebuildings needs, electric power is drawn from the utility grid.This configuration permits the use of the utility g
8、rid as an elec-tric energy storage system. Net-zero energy solar buildings(NZESB) have recently become the focus of coordinatedinternational research efforts (IEA-SHC 2008).Anticipatory Control Strategies and Their Application to Solar-Optimized BuildingsAnticipatory control strategies for buildings
9、, based onweather and load forecasting, have been proposed and studiedas a suitable alternative to conventional control for the pastquarter century (Winn and Winn 1985; Scartezzini et al. 1987;Kintner-Meyer and Emery 1995; Henze et al. 1997; Henzeet al. 2005). In the 80s and early 90s, the lack of r
10、eadilyavailable weather forecasts represented a limitation for thedevelopment of these control strategies. This difficulty hasbeen addressed by approaches such as manually introducinglimited weather data complemented with historical records(Chen 1997), or by stochastic methods estimating the likeli-
11、hood of a future weather pattern based on the current condi-tions (Nygrd-Ferguson and Scartezzini 1992). Today, theavailability of accurate weather forecasts online has opened upnew possibilities for the use of anticipatory control strategiesin buildings.Solar-optimized buildings have several featur
12、es thatincrease the potential of anticipatory control strategies for coldclimates:Investigation of Anticipatory Control Strategies in a Net-Zero Energy Solar HouseJos A. Candanedo Andreas K. Athienitis, PhD, PEngStudent Member ASHRAE Member ASHRAEJ.A. Candanedo is a PhD Candidate at the Department o
13、f Building, Civil and Environmental Engineering at Concordia University in Montral,Canada. A.K. Athienitis is a Professor in the Department of Building, Civil and Environmental Engineering, and the Scientific Director of theCanadian Solar Buildings Research Network.OR-10-027 2010, American Society o
14、f Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permis
15、sion. ASHRAE Transactions 247First, passive solar design in heating-dominated climatesrequires proper orientation and geometry, adequate insu-lation, thermal mass and air-tightness, as well as largerthan usual fenestration areas with high solar heat gaincoefficient and low thermal conductance. These
16、 charac-teristics reduce the impact of exterior air temperatureand make solar heat gains the most important factoraffecting room interior temperature fluctuations. Predic-tive control can help to maximize the usefulness of thesolar heat gains while preventing overheating throughthe use of the buildi
17、ngs thermal mass to store thermalenergy. Predictive control can also help in reducing elec-trical peak loads. Although predictive control has beenmostly applied in cooling-dominated buildings (Braunet al. 1990), it is also appropriate when electric energy isused directly or indirectly for heating.Se
18、cond, solar-optimized buildings include technologiesthat allow active collection of solar energy such asbuilding-integrated photovoltaic (BIPV) modules, BIPV/thermal systemsused for electricity generation and col-lection of thermal energy (Bazilian et al. 2001)solarthermal collectors and solar-assis
19、ted heat pumps.Although these technologies permit collecting largeamounts of energy, their operation often requires a sig-nificant energy input (e.g., heat pump compressorsrequire electric power in the order of kW). In conse-quence, their use should be planned in order to minimizenet energy consumpt
20、ion and possibly peak loads as well.Other technologies, like switchable glazing and motor-ized blinds, allow a certain degree of control over solarheat gains, and therefore a certain control of the “charg-ing” of the distributed thermal mass of the building.Third, thermal energy storage (TES) system
21、s, such aswater tanks or phase change materials (PCMs), may sig-nificantly help to reduce the problem of mismatchbetween the times of energy collection and use. Ananticipatory control system becomes essential to handlethe energy storage inventory.In summary, anticipatory control techniques may be mo
22、stuseful in solar-optimized buildings, as they considerablyreduce the impact of solar irradiance variations. Large-scaleimplementation of grid-connected solar-optimized buildingsmay offer significant benefits for electric utilities in terms ofload management.Figure 1A shows an example of a solar-opt
23、imized house.The passive solar design is largely determined by geometricparameters such as window-to-wall areas, “compactness”(ratio of exposed surface area to heated volume) of the build-ing construction, and position and dimension of the over-hangs. Passive solar design is complemented by active s
24、olartechnologies. These active systems may be components suchas building integrated photovoltaic/thermal (BIPV/T) sys-tems, solar thermal collectors and controllable motorizedblinds. Thermal energy storage (TES) is carried out passively(i.e., storage of the solar heat gains in the buildings thermalm
25、ass) as well as actively in the TES device and the DHW tank.It may also be possible to actively charge the thermal mass ofthe building, for example, with a radiant floor heating (RFH)system. The numerous possibilities for thermal energy trans-fer and storage between the components of a solar-optimiz
26、edhouse (Figure 1B) make a predictive control strategy highlydesirable.Applications of Anticipatory ControlControl may refer to two distinct, but closely related,control levels (ASHRAE 2007):Supervisory control. This is usually done by establish-ing a schedule of set-points for room temperatures and
27、 othervariables (e.g., humidity, state of charge of a TES device).Conventional approaches include maintaining a fixed set-point, using a different nocturnal set-point, or using severalset-points throughout the day. An estimation of future loadsFigure 1 (A) Generic solar-optimized house concept and (
28、B) possibilities of thermal energy storage and energy transfer. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmissi
29、on in either print or digital form is not permitted without ASHRAEs prior written permission. 248 ASHRAE Transactions(especially over a relatively long prediction horizon), can bevery useful in planning the collection, storage and delivery ofenergy. One way to accomplish this is to use optimal contr
30、oltheory, a group of mathematical techniques used to find thecontrol actions that will optimize an objective function(usually energy consumption, peak demand or cost) within aset of constraints. Optimal control techniques have beenproposed to manage the building thermal mass (Braun 1990;Morris et al
31、. 1994), the use of ice storage systems (Henze1997), or the coordination of both passive and active storage(Kintner-Meyer and Emery 1995; Henze et al. 2004). Similarmethods using simple model-based approach and system-identification have recently been proposed by Lee and Braun(2004, 2006, 2008).Lowe
32、r control level. If no corrective measures are taken,the differences between the time constants of the buildingsthermal mass, the HVAC system and the instrumentation willintroduce significant time delays (on the order of hours) whenattempting to follow the set-points prescribed by the supervi-sory c
33、ontrol level (Athienitis et al. 1990). Although close set-point tracking is desirable, a certain degree of fluctuation istolerated if human comfort is maintained. A technique knownas “model predictive control”, which uses a dynamic model ofthe system and expected inputs, has been used by Dexter andH
34、aves (1989), Chen and Athienitis (1996), and Chen (2002).Adaptive control based on neural networks has also been usedin lower-level control (Curtiss et al. 1996; Argiriou et al.2000).There have been relatively few investigations on theapplication of optimal control to residential solar buildings(Win
35、n and Winn 1985; Nygrd-Ferguson and Scartezzini1989, 1992; Chen 2001). Most work on the management ofpassive and active storage has dealt with large commercialbuildings, which are mostly cooling-dominated. Solar-opti-mized detached residential buildings, especially in coolclimates, are often heating
36、-dominated, and therefore manage-ment of the TES devices must be incorporated in the controlstrategies. Active envelope components such as motorizedblinds and renewable energy systems should also be consid-ered within the overall control strategy. Frequently updated,increasingly accurate and informa
37、tive weather forecasts maynow be readily incorporated in these control strategies (CMC2007). Finally, anticipatory control can significantly contrib-ute to reducing the dependence on external energy sources,thus contributing to the net-zero energy goal.This paper presents the result of simulations c
38、arried out tomodel the use of anticipatory control strategies for a case studyof a solar-optimized house.CASE STUDYThe simulations presented here correspond to the Alston-vale Net Zero Energy Solar House (ANZESH), a solar-opti-mized house currently in an advanced phase of construction inthe town of
39、Hudson (4527N, 7408W) in the vicinity ofMontral (Figure 2). This 2470 ft2(230 m2) house wasdesigned as an entry of the EQuilibrium Initiative, a Canadiandesign competition of advanced residential buildings (CMHC2008). Passive solar design plays an important role in thishouse. Table 1 presents the de
40、sign insulation values, and Table2 gives the fenestration areas in each faade.Most of the main floor of the house consists of a concreteslab-on-grade 6 in (15 cm) thick. The rest of the habitablespace of the floor consists of a 2-in (5-cm) layer of concrete ontop of a wooden structure. A radiant flo
41、or heating system willbe used to deliver heat to the space.The simulations performed during the design processwere carried out using several software tools. HOT2000(NRCan 2008a) was used to perform a preliminary estima-tion of the heating and cooling loads, and to choose the levelsof insulation of t
42、he building envelope. The auxiliary heatingTable 1. Building Envelope Values of theAlstonvale Net Zero Energy Solar HouseBuilding Envelope ComponentR-Value(Fhft2Btu1)RSI (m2KW1)Exterior walls 32 5.6Ceiling 68 12.0Concrete slab 26 4.6Basement floor 26 4.6Foundation walls 32 5.6Table 2. Fenestration A
43、reas in theAlstonvale Net Zero Energy Solar HouseTotal Faade AreaWindowAreaWindow/Wall RatioSouth 1,313 ft2(122 m2) 538 ft2(50 m2) 41%North 1,313 ft2(122 m2) 0.0 ft2(0.0 m2) 0%East 721 ft2(67 m2) 59.2 ft2(5.5 m2) 8%West 602 ft2(56 m2) 60.2 ft2(5.6 m2) 10%Figure 2 Construction of the Alstonvale Net Z
44、ero EnergySolar House. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not
45、 permitted without ASHRAEs prior written permission. ASHRAE Transactions 249energy for the house was calculated to be 22,970 MJ(21,768 kBtu) per year. The annual cooling load, accordingto HOT2000, has been estimated as 808 MJ (765 kBtu) for asummer set-point of 25.6C (78.1F). A solar chimney(which c
46、an be seen in Figure 2) has been installed to enhancenatural convection effects and to eliminate the need formechanical cooling.The principal energy system of the ANZESH is its BIPV/Troof (Figure 3). Forty-two (42) photovoltaic panels with anominal peak generating capacity of 7.35 kWpat standard tes
47、tconditions cover the lower two-thirds of the total roof surface(1130 ft2, 105 m2). Calculations carried out with RETScreen(NRCan 2008b) indicate that this system can provide a totalyearly generation of 9,014 kWh. Rectangular glazing boardsoccupy the area of the roof not covered by the photovoltaicp
48、anels. The original design of the house included a smallerphotovoltaic generation capacity (5.8 kW); it was increased toprovide electric energy supply for an electric vehicle (Poghar-ian et al. 2008). The project now also includes a greenhouse forlocal food production.A variable speed fan is used to
49、 draw air through the chan-nel formed by the photovoltaic panels and the glass section (ontop) and an insulated absorber surface (bottom). The airmoving through the channel extracts thermal energy from thePV panels that would otherwise be lost to the exterior andincreases the electrical efficiency of the panels; its temperaturerises as it moves up the roof. Since most of the incident solarradiation can pass through the glazing and be received directlyby the absorber surface, the glazing section significantlyincreases the air temperature. The air flow rates
