ScienceDirect ScienceDirect Energy Procedia 00 (2017) 000–000 Energy Procedia 00 (2017) 000–000
Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com Available
ScienceDirect ScienceDirect
www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia
Energy (2017) 000–000 463–468 EnergyProcedia Procedia122 00 (2017) www.elsevier.com/locate/procedia
CISBAT 2017 International Conference – Future Buildings & Districts – Energy Efficiency from CISBATNano 2017toInternational – Future Buildings & Districts – Energy Efficiency from Urban Scale,Conference CISBAT 2017 6-8 September 2017, Lausanne, Switzerland Nano to Urban Scale, CISBAT 2017 6-8 September 2017, Lausanne, Switzerland
Potential and optimization of a price-based control strategy for International Symposium on District Heating and Cooling PotentialThe and15th optimization of a price-based control strategy for improving energy flexibility in Mediterranean buildings improvingthe energy flexibility in Mediterranean buildings Assessing feasibility of using thea heat demand-outdoor a,b, a,b Thibault Q. Péan *, Jaume Salom , Joana Ortiz a a,b temperature function district forecast Thibault Q.for Péanaa,b,long-term *, Jaume Salom , Joanaheat Ortizdemand Catalonia Institute for Energy Research (IREC), Jardins de les Dones de Negre 1, 08930 Sant Adrià de Besòs (Barcelona), Spain a a
a
Universitat Politècnica de Catalunya C/Pau 5,08028Barcelona, Catalonia Institutea,b,c for Energy Research Jardins de les Dones de Gargallo Negreb 1, 08930 Sant Adrià deSpain Besòs (Barcelona), Spain a(IREC), a(UPC), c c b Universitat Politècnica de Catalunya (UPC), C/Pau Gargallo 5,08028Barcelona, Spain
I. Andrić
b
*, A. Pina , P. Ferrão , J. Fournier ., B. Lacarrière , O. Le Corre
IN+ Center for Innovation, Technology and Policy Research - Instituto Superior Técnico, Av. Rovisco Pais 1, 1049-001 Lisbon, Portugal b
Veolia Recherche & Innovation, 291 Avenue Dreyfous Daniel, 78520 Limay, France Abstract c Département Systèmes Énergétiques et Environnement - IMT Atlantique, 4 rue Alfred Kastler, 44300 Nantes, France Abstract The present study proposes the implementation and fine-tuning of a rule-based control strategy aiming to improve the energy The present proposes the implementation and fine-tuning of a rule-based control strategy aiming to improveprice the energy flexibility of study residential buildings in the Mediterranean area. The adopted control reacts to the time-varying electricity signal, flexibility of residential buildings in the Mediterranean area.hot Thewater adopted control reacts to the time-varying price control signal, and modulates the set-point for space heating and domestic production accordingly. A parametricelectricity study on three Abstract and modulates set-point for space hotchoose water production A parametric studyconsisted on three control variables of thethe control algorithm washeating carriedand out,domestic in order to appropriateaccordingly. values. The analysed outcome mainly variables of the control algorithm was carried out, in order to literature choose Theeffective analysed outcome consisted mainly inDistrict the provided flexibility, the on the comfort conditions, andappropriate the evolution the energy use and costs compared to the a heating networks are impact commonly addressed in the as one ofvalues. the ofmost solutions for decreasing greenhouse gasflexibility, emissions from the the building sector. These systems high which arewas returned thewith in the provided the impact oncontrol the comfort conditions, and the evolution the energy use and costs through compared toheat a standard reference case. After tuning strategy, a decrease inrequire energy costsinvestments of of around 22 to 26% observed, along Due to thecase. changed climate conditions and building renovation demand in the could decrease, standard reference tuning controllow strategy, a price decrease in energy costs of heat around 22additionally to 26% wasfuture observed, along with ansales. important shifting of After heating loadsthe towards energy periods. Thepolicies, proposed control caused an increase of prolonging investment return period. an important shifting heating loads towards low energy price periods. The proposed control additionally caused anbuildings increasefor of energy use ofthe 2 to 4%,ofwithout jeopardizing comfort conditions. These results emphasize the potential of residential The main scope of this paper is to assess feasibility of usingThese the heat demand – outdoor function forbuildings heat demand energy flexibility use of 2 toin4%, jeopardizing conditions. results emphasize the temperature potential of residential for thewithout Mediterranean area.thecomfort forecast. The district of Alvalade, located energy flexibility in the Mediterranean area. in Lisbon (Portugal), was used as a case study. The district is consisted of 665 vary in both construction ©buildings 2017 Thethat Authors. Published by Elsevierperiod Ltd. and typology. Three weather scenarios (low, medium, high) and three district renovation scenarios were developed (shallow, © 2017 The Authors. Published by Elsevier Ltd. intermediate, deep). To estimate the error, obtained heat demand values were Peer-review responsibility of Elsevier the scientific © 2017 The under Authors. Published by Ltd. committee ofthe scientific committee of the CISBAT 2017 International Peer-review under responsibility of the scientific committee the CISBAT 2017 International Conference – Future Buildings & compared with results from a dynamic heat demand model,of previously developed and validated by the authors. Conference –under Future Buildings &of Districts – Energy Efficiency from Nano to Urban Scale. Peer-review responsibility the scientific committee ofthe scientific committee of the CISBAT 2017 International Districts – Energy Efficiency from Nano to Urban Scale The results showed that when only weather change is considered, the margin of error could be acceptable for some applications Conference – Future Buildings & Districts – Energy Efficiency from Nano to Urban Scale. (the error in annual demand was lower than 20% for all weather scenarios considered). after introducing Keywords:energy flexibility in buildings; demand-side management; heating control strategy; rule-basedHowever, control; sensitivity analysis. renovation scenarios, the error value increased up to 59.5% (depending on the weather and renovation scenarios combination considered). Keywords:energy flexibility in buildings; demand-side management; heating control strategy; rule-based control; sensitivity analysis. The value of slope coefficient increased on average within the range of 3.8% up to 8% per decade, that corresponds to the decrease in the number of heating hours of 22-139h during the heating season (depending on the combination of weather and renovation scenarios considered). On the other hand, function intercept increased for 7.8-12.7% per decade (depending on the coupled scenarios). The values suggested could be used to modify the function parameters for the scenarios considered, and improve the accuracy of heat demand estimations.
* Corresponding author. Tel.: +34 933562615; fax: +34 933563802. address:author.
[email protected] * E-mail Tel.: +34 933562615; fax: +34 933563802. ©Corresponding 2017 The Authors. Published by Elsevier Ltd. E-mail address:
[email protected] Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and 1876-6102© 2017 The Authors. Published by Elsevier Ltd. Cooling. Peer-review responsibility ofthe scientific committee 1876-6102©under 2017 The Authors. Published by Elsevier Ltd. of the CISBAT 2017 International Conference – Future Buildings & Districts – Energy Efficiency Nano to Urban Scale. committee of the CISBAT 2017 International Conference – Future Buildings & Districts – Peer-review under from responsibility ofthe scientific Keywords: Heat demand; Forecast; Climate change Energy Efficiency from Nano to Urban Scale.
1876-6102 © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and Cooling. 1876-6102 © 2017 The Authors. Published by Elsevier Ltd. Peer-review under responsibility of the scientific committee of the CISBAT 2017 International Conference – Future Buildings & Districts – Energy Efficiency from Nano to Urban Scale 10.1016/j.egypro.2017.07.292
Thibault Q. Péan et al. / Energy Procedia 122 (2017) 463–468 Thibault Q. Péan et al./ Energy Procedia 00 (2017) 000–000
464 2
1. Introduction With the increasing penetration of renewable energy sources (RES), electricity grids are facing new challenges. In particular, the high variability of some of these sources such as solar and wind power can jeopardize the balancing of the networks at every instant.Integrating large shares of variable RES is however necessary in order to achieve the decarbonization of our energy systems agreed upon the Paris agreement to fight climate change [1]. New methods for counteracting the threats of these RES on the stability of the grid are thus needed. In this regard, Demand-Side Management (DSM) solutions are investigated in different fields, as demonstrated by the numerous tasks developed by the International Energy Agency (IEA) within its DSM programme [2]. The present work applies DSM to make the energy consumption of buildings more flexible and thus more grid-supportive. Buildings represent approximately 33% of the global primary energy use [3], therefore driving theseenergy consumers towards more flexibility represents a large potential [4]. In particular, heating, cooling and Domestic Hot Water (DHW) loads can partly be shifted in time to provide this flexibility, since the heat (or cold) can be retained for a certain period within the thermal mass of the building or in other storage means such as water tanks [5]. Several control strategies exist to manage building flexible loads. Rule-based controls (RBC) consist in relatively simple algorithms aiming for instance to shift loads with fixed schedules, shave energy peaks or reduce the energy costs. More advanced strategies such as Model Predictive Control (MPC) make use of an optimization problem to achieve the best performance over a certain time horizon, projecting the behavior of the systems in the future with weather or occupancy forecasts [6]. MPC presents more difficulties and costs in its implementation due to the prior need of a building model for the controller. A well-tuned or predictive RBC can already achieve substantial results, even though they are not entirely optimized [7]. In the present work, an RBC strategy is utilized: it reacts to the time-varying electricity price signal and adapts the heating set-points accordingly, turning the space heating (SH) and DHW needs into flexible loads, shiftable in time. The first objective of the study consists in adjusting three parameters of the applied method with regards to energy costs, flexibility and comfort.These parameters relate to the definition of the price thresholds, and to the amplitude of the set-point modulation. Once the most satisfying combination is found, the potential of a typical Mediterranean building is evaluated, since energy flexibility has still scarcely been investigated in this climate zone. Nomenclature DHW DSM MPC
Domestic Hot Water Demand-Side Management Model Predictive Control
RBC RES SH
Rule-based control Renewable Energy Source Space Heating
2. Methods 2.1. Model of the building and its mechanical systems The chosen case study is a residential flat of 109 m2within a multi-storey building, situated in Barcelona (Spain) and typical of Catalonia’s building habits. For this study, a refurbishment of the building is considered, with an improved insulation of the external walls (12 cm EPS) leading to a U-value of 0.2 W/m2K. The external windows have a U-value of 2.5 to 5.7 W/m2K, and only natural ventilation is considered. The dwelling is occupied by a family of two adults and two children, to whom the occupancy schedule has been adapted. For the production of hot water (SH and DHW), an air-to-water heat pump is implemented (nominal power of 4.3 kW and COP of 3.0 for leaving water at 55°C and outside air temperature at 7°C [8]). For DHW, the heated water is stored in a tank of 250 liters, normally kept at 60°C, and withdrawn according to the standard tapping programme M [9]. For space heating, the heated water is circulated in a circuit of eight radiators, controlled by a central thermostat placed in the living room. More details about the modelling hypothesis can be found in [10,11].
Thibault Q. Péan et al. / Energy Procedia 122 (2017) 463–468 Thibault Q. Péan et al. / Energy Procedia 00 (2017) 000–000
465 3
2.2. Control strategy The control strategy consists in modulating the heating set-points, respectively for SH and DHW. Firstly, thresholds on the varying electricity price are defined: one for low price and the other for high price. They are both calculated using respectively the k-th and (100-k)-th percentiles (respectively noted 𝑃𝑃𝑃𝑃𝑘𝑘𝑘𝑘 and 𝑃𝑃𝑃𝑃100−𝑘𝑘𝑘𝑘 ) of the price data of the past day (k=10% to 50%). When the current electricity price exceeds the high threshold(𝑃𝑃𝑃𝑃 > 𝑃𝑃𝑃𝑃100−𝑘𝑘𝑘𝑘 ), the set-point is decreased to limit the energy use in that period. Conversely, when the current price goes below the lower threshold(𝑃𝑃𝑃𝑃 < 𝑃𝑃𝑃𝑃𝑘𝑘𝑘𝑘 ), the set-point is increased. The magnitude of this modulation is ∆𝑇𝑇𝑇𝑇𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = ±1℃/ ±2℃for SH and∆𝑇𝑇𝑇𝑇𝐷𝐷𝐷𝐷𝑆𝑆𝑆𝑆𝐷𝐷𝐷𝐷 = ±5℃/±10℃ for DHW, realized over a reference case. The reference set-point is 19 or 21°C for SH (respectively during non-occupancy and occupancy periods) and 60°C for DHW. 2.3. Performance indicators To quantify the load-shifting introduced by the aforementioned control strategy, a flexibility factor 𝐹𝐹𝐹𝐹(𝑖𝑖𝑖𝑖) is introduced [12]. 𝐹𝐹𝐹𝐹also depends on the low and high thresholds of the electricity price, hence on the percentile 𝑖𝑖𝑖𝑖 chosen for this definition. It should be noted that the control strategy can be implemented using the percentile 𝑘𝑘𝑘𝑘 but analyzed with a different percentile 𝑖𝑖𝑖𝑖. The expression of 𝐹𝐹𝐹𝐹(𝑖𝑖𝑖𝑖) is shown in Eq. (1).
𝐸𝐸𝐸𝐸𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 − 𝐸𝐸𝐸𝐸ℎ𝑝𝑝𝑝𝑝𝑖𝑖𝑖𝑖ℎ 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 ∫𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃100−𝑖𝑖𝑖𝑖 𝑊𝑊𝑊𝑊ℎ𝑝𝑝𝑝𝑝 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 (1) = 𝐸𝐸𝐸𝐸𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 + 𝐸𝐸𝐸𝐸ℎ𝑝𝑝𝑝𝑝𝑖𝑖𝑖𝑖ℎ 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 ∫ 𝑊𝑊𝑊𝑊 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 + 𝑊𝑊𝑊𝑊 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 ∫ ℎ𝑝𝑝𝑝𝑝 ℎ𝑝𝑝𝑝𝑝 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃100−𝑖𝑖𝑖𝑖 Where 𝑊𝑊𝑊𝑊ℎ𝑝𝑝𝑝𝑝 is the electricity power of the heat pump. This parameter is integrated over low price (𝑃𝑃𝑃𝑃 < 𝑃𝑃𝑃𝑃𝑝𝑝𝑝𝑝 ) or high price periods (𝑃𝑃𝑃𝑃 > 𝑃𝑃𝑃𝑃100−𝑝𝑝𝑝𝑝 ), resulting in the corresponding energy uses 𝐸𝐸𝐸𝐸. 𝐹𝐹𝐹𝐹(𝑖𝑖𝑖𝑖)varies between -1 and 1. A value of 1 signifies a high flexibility, which means that all the energy use has been shifted to the defined low-price periods (i.e. 𝐸𝐸𝐸𝐸ℎ𝑝𝑝𝑝𝑝𝑖𝑖𝑖𝑖ℎ 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 = 0). A value of -1 signifies a low flexibility, hence that the energy use was shifted to the defined high price periods (i.e. 𝐸𝐸𝐸𝐸𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 = 0). Modulating the room temperature set-point might jeopardize the occupants’comfort, due to the higher temperature variations caused indoors. To evaluate if the indoor environment remains in the limits of thermal comfort, the methods from the European standards are utilized [13]. In particular, the percentages of occupancy time spent with the operative temperature within the ranges of Comfort Categories I, II, III and IV are calculated for the living room. The lower thresholds for the three first categories are 20.6, 19.2 and 18.3°C, respectively corresponding to PMV values of 0.2, 0.5 and 0.7 (with the assumptions of 1 clo and 1.2 met). All temperatures below 18.3°C are accounted for in Category IV. Further than flexibility and comfort, the total energy used by the heat pump for providing SH and DHW is calculated over each simulated case. The associated energy cost is also computed, using time-varying price profiles. 𝐹𝐹𝐹𝐹(𝑖𝑖𝑖𝑖) =
2.4. Simulation boundaries and cases Table 1.List of simulated cases. Case nr
Price thresholds percentiles for control
Flexibility factor percentiles
0 1 to 5 6 7 8
0 (reference) k=10 to 50% k=40% k=40% k=40%
i=10 to 50% i=10 to 50% i=40% i=40% i=40%
∆𝑇𝑇𝑇𝑇𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆
0 ±1°C ±1°C ±1°C ±2°C
∆𝑇𝑇𝑇𝑇𝐷𝐷𝐷𝐷𝑆𝑆𝑆𝑆𝐷𝐷𝐷𝐷 0 ±5°C ±5°C ±5°C ±10°C
Price and weather input
Time period of price analysis
Jan. 2015 (a) and Apr. 2015 (b) Jan. 2015 (a) and Apr. 2015 (b) Jan. 2015 (a) and Apr. 2015 (b) Jan. 2015 (a) and Apr. 2015 (b) Jan. 2015 (a) and Apr. 2015 (b)
Past 24h Past 24h Next 24h Past 72h Past 24h
The simulations were carried out in TRNSYS, with time steps of three minutes and over periods of one week; they are listed in Table 1.The weather data was retrieved from a weather station situated in Terrassa (Barcelona, Spain) during the year 2015. A selected week in January 2015 was chosen (cold season), as well as one in April 2015 (midseason). The corresponding price profile of the same weeks was retrieved from the Spanish Transmission System Operator [14]. Two hourly-varying tariffs exist in Spain for small consumers of less than 10 kW contracted power.
Thibault Q. Péan et al. / Energy Procedia 122 (2017) 463–468 Thibault Q. Péan et al./ Energy Procedia 00 (2017) 000–000
466 4
The “2-periods” tariff was chosen over the default one, since it presents larger daily variations and thus higher potential for monetary savings through load-shifting. 3. Results 3.1. Adjusting the thresholds for low and high price The flexibility factor analysis is presented inFig.1. for cases 1 to 5. On the horizontal axis, the different simulated cases correspond to different values of k (10 to 50%) used to define the thresholds for the control; the different series show the different values of i used to define the thresholds of the flexibility factor. The flexibility factor highly depends on both values. It can be observed in January (Fig.1 (a)) that for a fixed case, the value of i=k shows the highest flexibility factor, which is only logical since the evaluation factor then coincides with the actual control implemented. The overall highest flexibility factor of 0.69 is obtained in case 4a, with k=i=40%. In spring season (Fig.1 (b)), this tendency is less clear, but in most cases, the configurations k=ialso present the highest flexibility. The highest values of the flexibility factor are obtained with k=50% (case 5b) and i=10 and 50%. As the heating needs are reduced in mid-season, and thus provide less interest in load-shifting, the control strategy is preferably optimized for the winter season. Therefore, the values k=i=40% are chosen, and applied in the remaining cases.
Flexibility factor
1.0
(a) i=10%
0.6 0.4
i=20%
0.2
i=30%
0.0 -0.2 -0.4
0.8
1a
2a 3a 4a Studied cases (January 2015)
Flexibility factor
1.0 0.8
i=40%
5a
0.6 0.4 0.2 0.0
-0.2
i=50%
-0.4
-0.6
(b)
1b
2b 3b 4b Studied cases (April 2015)
5b
-0.6
Fig.1. Flexibility factor computed for the total electricity use of the heat pump in cases 1 to 5, both in January 2015 (a) and April 2015 (b). Cases 1 to 5 correspond respectively to values of the percentiles k=10% to 50%, applied to the price thresholds for the control strategy.
3.2. Adjusting the time range for the price data analysis Table 2.Results of the simulated case in terms of electricity use and costs.
Case 0a 4a 6a 7a 8a
January 2015 (a) Electricity use Electricity cost HP_Tot Var /ref HP Var /ref [kWh] [%] [€] [%] 111.8 11.5 116.6 4.3% 8.5 -26.0% 119.0 6.5% 8.8 -24.0% 118.5 6.1% 8.4 -27.1% 134.2 20.1% 8.0 -30.8%
Flex. factor [-] -0.21 0.69 0.71 0.70 0.94
Case 0b 4b 6b 7b 8b
April 2015 (b) Electricity use Electricity cost HP_Tot Var /ref HP Var /ref [kWh] [%] [€] [%] 54.5 4.7 55.5 1.9% 3.6 -22.2% 57.3 5.2% 3.9 -17.4% 57.0 4.6% 3.7 -20.2% 65.6 20.4% 3.6 -23.8%
Flex. factor [-] -0.08 0.67 0.48 0.68 0.97
In the reference control, the low and high price thresholds are defined based on the percentiles of the price data recorded in the past day. However, this period of analysis can be extended or moved in time: in case 6, the price data of the future 24 hours are considered, instead of the past 24 hours. This scenario is realistic in the context of the large implementation of smart meters in Europe and Spain [15], which enable communications with the grid and thus the retrieval of the future price data. In case 7, the data of the last 3 days is considered for the definition of the thresholds. In this way, the evolution trend of the price data can be captured on a longer time frame. The other parameters are kept constant (k=i=40%), and therefore cases 6 and 7 are compared with case 4.
Thibault Q. Péan et al. / Energy Procedia 122 (2017) 463–468 Thibault Q. Péan et al. / Energy Procedia 00 (2017) 000–000
467 5
The energy used by the heat pump and the associated monetary costs are presented in Table 2. Unexpectedly, case 6 is outperformed both in winter and spring season, in terms of reductions of the energy costs. In this case, information is provided about the future price, so the control should benefit from this information and store energy accordingly. However, it appears that the proposed rule-based control is unable to anticipate and manage the energy use in an optimized manner. Case 7 provides the highest cost savings in winter (-27.1% compared to the reference case) and Case 4 in spring (-22.2% compared to reference case). However, Case 4 presents a lower increase in the electricity use than Case 7, for this reason the strategy of Case 4 is adopted (i.e. analyzing the price data of only the past day, when defining the price thresholds). 3.3. Adjusting the set-point modulation In Case 8, the amplitude of the set-point modulation is doubled: instead of ∆𝑇𝑇𝑇𝑇𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = ±1℃and ∆𝑇𝑇𝑇𝑇𝐷𝐷𝐷𝐷𝑆𝑆𝑆𝑆𝐷𝐷𝐷𝐷 = ±5℃, the values of ∆𝑇𝑇𝑇𝑇𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = ±2℃ and ∆𝑇𝑇𝑇𝑇𝐷𝐷𝐷𝐷𝑆𝑆𝑆𝑆𝐷𝐷𝐷𝐷 = ±10℃ are utilized. The impact on comfort is investigated in particular, since the higher temperature variations could cause discomfort situations indoors. Results are shown in Fig.2. It appears that the impact on comfort is relatively limited between reference case 0 and case 8: the percentage of time with an operative temperature outside Category II only increases from 3.6% to 4.9% (it is even reduced to 0% in case 4). However, Case 8 provokes a large increase of the energy use (+20.1 to +20.4%, see Table 2), therefore such high setpoint modulation should be avoided. The original values of ∆𝑇𝑇𝑇𝑇𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = ±1℃ and ∆𝑇𝑇𝑇𝑇𝐷𝐷𝐷𝐷𝑆𝑆𝑆𝑆𝐷𝐷𝐷𝐷 = ±5℃ are thus preferred. Furthermore, in spring season, the maximum flexibility for SH has already been reached with ∆𝑇𝑇𝑇𝑇𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = ±1℃, therefore increasing this value is not necessary and does not provide any further improvement. 0%
25%
50%
75%
100%
0%
Cat. I
0a
Cat. II
0b
4a
Cat. III
4b
8a
Cat. IV
8b
25%
50%
75%
100%
Fig.2. Comfort analysis of the simulated cases in January 2015 (a) and April 2015 (b), using the comfort categories defined in the standard [13] for residential buildings. The analysis is done in the dayzone (living room), only during occupancy time.
4. Discussion and conclusions The importance of tuning rule-based control strategies was highlighted in the present work. In fact, such algorithms rely on different parameters whose values can greatly impact the outcome of the strategy [16], as shown in the Results section. A well-tuned RBC can reach high levels of performance, and could compete in some cases with MPC strategies, if the implementation effort is also taken into account in the comparison (MPC requires accurate models that are costly to build). A major issue is that in practice, the parameters of RBC strategies are usually chosen based on rules of thumb, and result in much lower performance than expected. Furthermore, the chosen parameters might not be adapted in every situation, as was shown for instance in section 3.2 for the time range of the price analysis: one value of the parameter is better suited for spring season, while another will perform better in winter season. This reveals the limits of RBC, which can be overcome with optimal control strategies, able to find an optimized solution at every time step. In this study, a rule-based control strategy aimed at improving the energy flexibility of buildings was implemented and tuned. In the most promising configuration, the first step of the control process consists in defining a low threshold on the electricity price, using the 40-th percentile of the price data of the past day, and a high threshold using the 60th percentile. A set-point modulation of ±1°C for SH and ±5°C for DHW is then implemented when these thresholds are passed. This method results in an increase of the flexibility factor from -0.21 to 0.69 in winter and from -0.08 to 0.67 in spring, indicating an important load-shifting towards periods of cheaper energy price. As a result, the energy cost is decreased by 22 to 26%, but additionally causing an increase in the energy use of 2 to 4%. The comfort is maintained at a similar level when the flexibility control strategy is activated. The proposed control therefore presents interesting features for shifting heating loads and activating energy flexibility in buildings, with a limited impact on the energy use and the comfort. It reveals the potential of residential
468 6
Thibault Q. Péan et al. / Energy Procedia 122 (2017) 463–468 Thibault Q. Péan et al./ Energy Procedia 00 (2017) 000–000
buildings, in particular in the Mediterranean area, for this type of application. This study proves the feasibility of loadshifting based on price data at the building level; further developments should also include an aggregation at an upper level, in order to reach a sufficient flexibility to offer to the grid. Furthermore, Model Predictive Control strategies will be developed in order to further optimize the energy management strategy, taking into account information about the future occupancy, price and weather. Acknowledgements This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 675318 (INCITE). Part of this work stems from the activities carried out in the framework of the IEA-EBC Annex 67 (International Energy Agency – Energy in Buildings and Communities program) about Energy Flexibility in Buildings. References [1] UNFCC Conference of the Parties (COP). Adoption of the Paris Agreement.Proposal by the President. Geneva (Switzerland): FCCC/CP/2015/L.9/Rev.1; 2015. [2] International Energy Agency.IEA DSM Task Overview 2017. http://www.ieadsm.org/task-overview/ (accessed April 4, 2017). [3] REN 21. Renewables 2016. Global status report 2016. [4] Esterl T, Kaser S, Stifter M, Kamphuis R, Galus M, Rijneveld A, et al. IEA DSM Task 17 Valuation Analysis of Demand Flexibility in Households and Buildings. 2016. [5] Hedegaard K, Mathiesen BV, Lund H, Heiselberg P. Wind power integration using individual heat pumps - Analysis of different heat storage options. Energy 2012;47:284–93. doi:10.1016/j.energy.2012.09.030. [6] Camacho EF, Bordons C. Model Predictive control. vol. 53. London: Springer London; 2007. doi:10.1007/978-0-85729-398-5. [7] Fischer D, Madani H. On heat pumps in smart grids: A review. Renew Sustain Energy Rev 2017;70:342–57. doi:10.1016/j.rser.2016.11.182. [8] Hitachi Air Conditioning Products Europe. Technical Catalogue - YutakiSeries. 2016. [9] CEN. EN 15316-1: Heating systems in buildings — Method for calculation of system energy requirements and system efficiencies — Part 31 Domestic hot water systems ,characterisation of needs (tapping requirements) 2007:1–20. [10] Ortiz J, Fonseca A, Salom J, Garrido N, Fonseca P, Russo V. Comfort and economic criteria for selecting passive measures for the energy refurbishment of residential buildings in Catalonia. Energy Build 2016;110:195–210. doi:10.1016/j.enbuild.2015.10.022. [11] Ortiz J, Guarino F, Salom J, Corchero C, Cellura M. Stochastic model for electrical loads in Mediterranean residential buildings: Validation and applications. Energy Build 2014;80:23–36. doi:10.1016/j.enbuild.2014.04.053. [12] Le Dréau J, Heiselberg P. Energy flexibility of residential buildings using short term heat storage in the thermal mass. Energy 2016;111:1–5. doi:10.1016/j.energy.2016.05.076. [13] CEN. EN 15251: Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics. 2007. [14] Red Electrica de España. ESIOS – Sistema de información del operador del sistema. n.d. https://www.esios.ree.es/en (accessedApril 4, 2017). [15] European Parliament. Directive 2009/72/EC of the European Parliament and of the Council of 13 July 2009 concerning common rules for the internal market in electricity. Strasbourg, France: 2009. [16] Oldewurtel F, Parisio A, Jones CN, Gyalistras D, Gwerder M, Stauch V, et al. Use of model predictive control and weather forecasts for energy efficient building climate control. EnergyBuild 2012;45:15–27. doi:10.1016/j.enbuild.2011.09.022.
