- Email: [email protected]
Active Residential Load Management Based on Dynamic Real Time Electricity Price of Carbon Emission
Available online at www.sciencedirect.com Available online at www.sciencedirect.com
ScienceDirect ScienceDirect
Energy Procedia 00www.sciencedirect.com (2018) 000–000 Available online Available online atatwww.sciencedirect.com Energy Procedia 00 (2018) 000–000
ScienceDirect ScienceDirect
www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia
Energy Procedia 152 Energy Procedia 00(2018) (2017)1027–1032 000–000 www.elsevier.com/locate/procedia
Applied Energy Symposium and Forum 2018: Low carbon cities and urban energy systems, Applied Energy Symposium andSymposium Forum 2018: Low carbon cities andcarbon urbancities energy systems, CUE2018-Applied Energy and Forum 2018: Low and CUE2018, 5–7 June 2018, Shanghai, China 5–7 June Shanghai, ChinaChina urban CUE2018, energy systems, 5–7 2018, June 2018, Shanghai,
Active Residential Load Management Based on Dynamic Real Time Active Residential Load Management on Dynamic Real Time The 15th International Symposium onBased District Heating and Cooling Electricity Price of Carbon Emission AssessingElectricity the feasibility ofof using the heat demand-outdoor Price Carbon Emission a Wei Zenga,a,*, Jan von Appenbb, for Patrick Selzambb, Mindistrict Sunaa, Boheat Chendemand , Wei Heaa,forecast Ning Xuaa temperature function a long-term a Wei Zeng *, Jan von Appen , Patrick Selzam , Min Sun , Bo Chen , Wei He , Ning Xu 330096, Chinac a,b,c State Grid Jiangxi a Electric Powera Research Institute,Nanchang, b Jiangxi Electric Powerand Institute,Nanchang, 330096, I. Andrić *,State A.Grid Pina , P. Ferrão ,Research J.Energy Fournier ., B. Lacarrière O. Le Correc Fraunhofer Institute for Energy Economics System Technology, Kassel, China 34117,,Germany a
b
a
Fraunhofer Institute for Energy Economics and Energy System Technology, Kassel, 34117, Germany 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 b
a
Rooftop photovoltaic, residential energy storage and other distributed power sources are put into operation on a large scale at the Rooftopuser photovoltaic, residential energy storage distributed powerThe sources are put into operation on a large scale at the power side, providing energy support forand theother electricity demand. demand response mechanism has become an power user side, providing resource energy for support for the electricity The Indemand response hasproportion become an indispensable and important the stable operation of thedemand. power grid. the current power mechanism grid with high of Abstract indispensable and (RE), important for theelectricity stable operation the power grid. In the current power grid withthe high proportion of renewable energy due resource to the existing market of lacking the effective interaction with end-user, electricity price renewable energy (RE), due to the existing electricity market lacking the effective interaction with end-user, the electricity price does not reflect the external characteristic of the RE and cannot embody the real relationship of power supply and consumption. District heating networks are commonly addressed in the literature as one of the most effective solutions for decreasing the doeseffect not reflect the external characteristicofofintermittent the RE andRE cannot embody realfirstly relationship of power supplya and The of increasing the consumption is limited. Thisthe paper innovatively proposes low consumption. carbon-based greenhouse gas emissions from the building sector. These systems require high investments which are returned through the heat The effectelectricity of increasing the consumption of intermittent RE is limited. This firstly innovativelyofproposes a low real-time pricing mechanism (LCEP) that effectively reflects thepaper external characteristics RE. Then, thecarbon-based flexibility of sales. Due to the changed climate conditions and building renovation policies, heat demand in the future could decrease, real-time electricity pricingismechanism (LCEP)and thatthe effectively characteristics of RE. scheduling. Then, the flexibility of the demand side resource fully exploited, LCEP is reflects used asthe theexternal adjustment lever of energy An energy prolonging the investment return period. the demand side resource is fully exploited, and the LCEP is used as the adjustment lever of energy scheduling. An energy management optimization strategy based on mixed linear integer programming is proposed to integrate all the flexible resources in The main scope of this paper is to assess the feasibility of using the heat demand – outdoor temperature function for heat demand management optimization strategy based on mixed linear integer the programming is proposed to integrate all the flexible resources in the home to be intelligent controlled. Through detailed analysis, effectiveness of the proposed strategy is verified. forecast. The district of Alvalade, located in Lisbon (Portugal), was used as a case study. The district is consisted of 665 the home to be intelligent controlled. Through detailed analysis, the effectiveness of the proposed strategy is verified. buildings that vary in both construction period and typology. Three weather scenarios (low, medium, high) and three district Copyright © 2018 Elsevier Ltd. All rights reserved. renovation scenarios wereLtd. developed (shallow, Copyright © 2018 2018 Elsevier Elsevier All rights rights reserved.intermediate, deep). To estimate the error, obtained heat demand values were Copyrightand © Ltd. All reserved. Selection peer-review under responsibility ofof thethe scientific committee of Applied Energy SymposiumEnergy and Forum 2018: Low Selection peer-review responsibility scientific committee of theand CUE2018-Applied comparedand with results fromunder a dynamic heat demand model, previously developed validated by the authors. Symposium and Selection andand peer-review under responsibility of the scientific committee of Applied Energy Symposium and Forum 2018: Low carbon cities urban energy systems, CUE2018. Forum 2018: Low carbon cities and urban energy systems. The results showed that when only weather change is considered, the margin of error could be acceptable for some applications carbon cities and urban energy systems, CUE2018. (the error in annual demand was lower than 20% for all weather scenarios considered). However, after introducing renovation Keywords: Residential Load Management; Real Time Electricity Price; Low Carbon scenarios, the errorLoad value increased up 59.5% (depending weather and renovation scenarios combination considered). Keywords: Residential Management; RealtoTime Electricity Price; on Lowthe Carbon 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.: +0-086-0791-88646884; fax: +0-086-0791-88646884. * Corresponding author. Tel.: +0-086-0791-88646884; fax: +0-086-0791-88646884.
E-mailThe address: [email protected] © 2017 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 Cooling. Keywords: Heat demand; Forecast; Climate change 1876-6102 Copyright © 2018 Elsevier Ltd. All rights reserved. 1876-6102and Copyright © 2018 Elsevier Ltd. All of rights reserved. committee of the Applied Energy Symposium and Forum 2018: Low carbon cities Selection peer-review under responsibility the scientific Selection peer-review responsibility of the scientific committee of the Applied Energy Symposium and Forum 2018: Low carbon cities and urbanand energy systems, under CUE2018. and urban energy systems, CUE2018. 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 Copyright © 2018 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the CUE2018-Applied Energy Symposium and Forum 2018: Low carbon cities and urban energy systems. 10.1016/j.egypro.2018.09.114
1028 2
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
1. Introduction With the deployment of pilots and popularization of distributed generation equipment such as small wind turbines and rooftop photovoltaics on the residential side, Tesla Power Wall has represented the emergence of new home energy storage technologies and the intelligentization of load devices, there have been major changes in the power consumption [1]. Users are no longer just acting as a consumer but are shifting to a system with self-scheduling capabilities. The user side will have more diversified power options to respond to grid regulation information. How to make full use of distributed generation, adjust the structure of electricity consumption, improve the experience, and save electricity expenses have become the focus of attention. A large amount of research work has been reported on residential load management which usually optimize the utilization of power load in the home. Shen[2] establishes the optimal control model of energy management at home, and proposes an intelligent solution strategy based on particle swarm optimization. Gao[3] proposes an inter-door energy scheduling algorithm to ensure the stable operation of the power grid. Zhuang[4] first introduce a general architecture of energy management system in a home area network based on the smart grid and then propose an efficient scheduling method for home power usage. Yang[5] builds household electricity response mode flow diagram and its mathematical model, then adopts unit commitment of genetic algorithm to achieve household electricity optimization. Sierra[6] design a fuzzy controller to realize the minimization of energy consumption based on existing residential energy management system. Pipat[7] proposes a home energy management algorithm based on demand response for high energy consumption users, ensuring that the total power consumption is lower than a certain level by the priority effect. Jia[8] introduces a home energy dynamic scheduling system based on decision tree classification algorithm. In general, the price method of current demand side management takes time-of-use price, peak price and other forms, these electricity price’s changing period and amplitude generally is more fixed, and its flexibility is not enough, moreover, all focus on reducing the peak load of the grid. With the rapid development of renewable energy (RE), the existing electricity market lacks the effective interaction with the user, the flexibility is poor, and the electricity price does not reflect the external characteristic of the RE, often causing the embarrassing situation that power supply is insufficiency in grid peak, but there is a large amount of electricity waste in the grid valley. The electricity price cannot embody the real relationship of power supply and consumption [9]. How to adopt market-based means and price mechanisms, guide the user side to dynamically adjust the electricity demand through real-time power price signals and realize highly coordination between low-carbon energy supply and demand, are issues that need to be addressed. Firstly, in Section 1, this paper innovatively proposes low-carbon realtime electricity prices based on carbon emissions, and then fully exploits the flexibility of demand-side resources on the load side. Through price signal transmission, this paper studies load management technology based on low-carbon real-time electricity prices, integrating the power generation, power load, and energy storage devices in the home to intelligent optimal dispatch. Section 2 introduces the low-carbon real-time pricing algorithm; Section 3 presents the load management strategy; Section 4 performs detailed calculation and analysis; Section 5 analyzes the impact of different electricity price components; Section 6 summarizes the paper. 2. Low Carbon Real Time Electricity Price Since increasing generation from volatile RE causes higher demand for flexibility in the energy system, flexibilization of household customers' consumption patterns can be an important contribution. Utilization of this potential can be enabled by variable electricity rates with elements like RTP. Based on the spot market, considering the supply and demand level of power generation side and demand side, the variability of grid fees and RE surcharge, we propose a dynamic low carbon real-time pricing (LCEP) mechanism and algorithm based on carbon emission. (1) Spot Market Pricing Only the generation units with the lowest marginal costs have to be used to meet a given demand. The most expensive generation unit, which is needed to meet the given demand, respectively in this case the consumption determines the price for all other generators. This so called market clearing price (MCP) is exactly the marginal costs of the most expensive needed generation unit for a specific period. MCESU =FCESU + VOCESU +CAESU × CC (1)
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
1029 3
Where, MC denotes marginal costs, FC denotes fuel costs, VOC denotes variable operating costs, CA denotes carbon amounts, CC denotes carbon costs, ESU denotes energy source unit. (2) Variable Grid Fees Variable grid fees (VGF) are an alternative approach, instead of power prices, to motivate the grid customers to adjust their energy consumption to the requirements of the distribution grid. The recommended formula for the dynamic processing of the grid fees Mid-Price based on spot prices. The multiplicative factor has been parameterized approximately between the values 0 and 1 (depends on spot prices and the height of the concerned price parameters). VGF =SPj ⋅ MF −
∑
j
SPj ⋅ MF
QMGFi
+ MGF
(2)
Where, VGF denotes variable grid fees, SP denotes Spot Price, MF denotes multiplicative factor, MGF denotes mid-price grid fee, QMGF denotes quantity of hourly Mid-Price Grid Fees, i denotes index for each hour of the year, j denotes index for each hour of the year with Mid-Price Grid Fees. (3) Variable RES Surcharge The Variable RES Surcharge (VRS) is an approach to revitalize the spot market and give an incentive to improve the system integration of renewable energies. The idea of VRS is to couple the RES surcharge to the day-ahead spot prices with a multiplicative factor that has to be calculated based on historical data. This factor should be defined on a yearly basis such as to cover the yearly costs of the RES support scheme. VRSi =SPi ⋅ MF − SPi (3) Where, VRS denotes variable RES surcharge, SP denotes spot price, MF denotes multiplicative factor, i denotes index for each hour of the year. So the LECP price on the end-use power consumption side can be calculated as: LCEPi = MCESU +VGF +VRSi (4) 3. Load management strategies Load management strategies are defined as changes in electric usage by end-use customers from their normal consumption patterns in response to changes of the electricity price over time, or to incentive payments designed to induce lower electricity use at times of high wholesale market prices or when system reliability is jeopardized. The mechanism used in this study is price-based load control. To respond to this signal, customers must modify their load consumption, by using a battery energy storage system (BESS) that can time-decouple load power supply from power withdrawal from the grid. The objective of each household is to reduce its annual electricity procurement cost that is determined by the hourly electricity price and the hourly energy withdrawal from the grid. Since the levels of the hourly electricity price are related to the “carbon content” of the distributed electricity, concentrating withdrawals from the grid in the hours with lower prices allows for a reduction of the CO2 emissions. In the following, we focus on using PV with a BESS on the customer side. According to the price signal, the customer does not modify his/her electricity consumption habits (thus keeping the same level of comfort) and uses a BESS to change his/her power withdrawal profile from the grid, according to the electricity price profile. The BESS power set-point is calculated instant by instant only on the basis of the hourly price. Mathematically, the power set-point for the BESS can be expressed as follows: PS AVG − PS ( h ) PS min if PS ( h ) < PS AVG , PBESS ( t ) = − PBESSMAX PS ( h ) PS AVG − PSmin otherwise, PBESS ( t ) = −
PS ( h ) PS MAX
Pload ( t )
PS ( h ) − PS AVG
(5)
PS MAX − PS AVG
where: PBESS ( t ) is the power exchanged by the BESS, PS ( h ) is the price signal, i.e. the electricity price in the hour h including the time instant t, PBESS is the maximum charging/discharging (nominal) power of the BESS, PS AVG is MAX
the average value of the price signal over the day, PS MAX and PSmin are the maximum and minimum value of the price
signal over the day, Pload ( t ) is the power required by the loads in the household.
There are some constraints that must be taken into account. The power withdrawn from the grid 𝑃𝑃𝑃𝑃𝐺𝐺𝐺𝐺 (t) must be less than or equal to the maximum contractual power PCONTR ; moreover, no power injections from the household to the
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
1030 4
grid are allowed.
0 ≤ PG ( t ) ≤ PCONTR
The power exchanged by the BESS must be less than or equal to its nominal power PBESS
(6) MAX
(this constraint is
automatically met by Eq. 5 in the charging phase; in the discharging phase the constraint is binding if Pload ( t ) > PBESS PBESS ( t ) ≤ PBESSMAX
The State of Charge (SoC) of the BESS must be kept in the allowed range, to preserve the battery life. 20% ≤ SoC ( t ) ≤ 100%
MAX
).
(7) (8)
The exchange profile with the grid of the household and the related economic value and CO2 emissions are calculated, in order to reduce the overall purchase electricity costs. The control strategy used in this case study is: the BESS is charged only with the PV plant production in excess of the load; and the BESS is discharged according to the price signal. The basic algorithm is as following: if PPV(t) > Pload(t) if SoCmin ≤ SoC(t) < 100% PBESS(t) = PPV(t) – Pload(t) //charge the storage else PBESS(t) = 0 //battery full: do nothing endif else if PPV(t) == Pload(t) PBESS(t) = 0 //do nothing else if PPV(t) < Pload(t) if SoCmin < SoC(t) ≤ 100% and PS(h) > PSAVG PS ( h ) PS ( h ) − PS AVG // discharge the storage PBESS ( t ) = Pload ( t ) PS MAX PS MAX − PS AVG else PBESS(t) = 0 //do nothing endif endif
where: PPV is the power produced by the PV plant, Pload is the power consumed by the load, PBESS is the power exchanged by the BESS, SoC is the state of charge of the BESS, 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃(ℎ) is the price signal, i.e. the electricity price in the hour h including the time instant t, 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐺𝐺𝐺𝐺 is the average value of the price signal over the day, 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑀𝑀𝑀𝑀𝐴𝐴𝐴𝐴𝑀𝑀𝑀𝑀 is the maximum value of the price signal over the day, 𝑃𝑃𝑃𝑃𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 (𝑡𝑡𝑡𝑡) is the power required by the loads in the household. 4. Case Study
A mixed integer linear optimization problem (MILP) is formulated to minimize the cost of daily energy withdrawal by computing at midnight the optimal BESS power set-point for the whole day. Python and the Pyomo tool box [10] are chosen simulation environment for the described problem. Such MILP problems are well explored and a couple of open-source solves such as GLPK[11] are available. The household price profiles for the following analysis are: (1) Spot prices incl. a high carbon price with variable grid fees and variable surcharges; (2) Spot prices incl. a high carbon price with normal grid fees and variable surcharges; (3) Spot prices incl. a high carbon price with normal grid fees and normal surcharges; (4) Spot prices incl. a high carbon price with variable grid fees and normal surcharges; (5) Spot prices incl. a low carbon price with variable grid fees and variable surcharges; (6) Spot prices incl. a low carbon price with normal grid fees and variable surcharges ; (7) Spot prices incl. a low carbon price with normal grid fees and normal surcharges; (8) Spot prices incl. a low carbon price with variable grid fees and normal surcharges The 2016 annual power generation and consumption data of Jiangxi province power grid is used to calculate LCEP price. And the SCUDO7 software [12] is utilized to simulate the typical household load profiles. The case focus is on the annual savings related to a reduction of electric energy purchase costs and the annual reduction of CO2 emissions. As far as BESS capacity is concerned, 10 different values (from 1 kWh to 10 kWh) have been considered. A roundtrip efficiency of 85% has been assumed for the BESS. And we assume that the household is equipped with a 4.5 kW peak power photovoltaic system.
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
Fig.1. Annual energy purchase savings.
1031 5
Fig.2. Annual energy purchase percentage savings.
It can be seen that, with BESS size increases from blue to yellow, the larger the BESS, the higher the saving, and also in this case a saturation is reached, due to the saturation of the percentage of self-consumption. With respect to the strategy based on self-consumption maximization the saving is slightly higher with the smallest BESS, while with the largest ones the saving is typically lower. With the largest BESS sizes, savings reach values around 80%.
Fig.3. Annual reduction of CO2 emissions.
Fig.4. Annual percentage reduction of CO2 emissions.
The results of CO2 emissions reduction are less dependent on the price signal than the results of cost savings. Again, the larger the BESS, the higher the saving. With respect to the self-consumption maximization strategy reduction of CO2 emissions is typically slightly higher with the smallest BESS, while with the largest ones in some cases the reduction is lower. 5. Impact of Variable Grid Fees In a scenario with high share of fluctuating renewable energies, a lot of surplus energy is likely to be produced. The surplus energy is an indicator for RES curtailment. E.g. a high RES surplus increases the likelihood of local grid integration challenges, might result in RES curtailment, if the distribution system operator has no other option, and cannot reinforce the grid on short notice. Based on the proposed LCEP mechanism in Section 2, it can assist in setting appropriate price signals to incentivize additional load consumptions and thus reduce the surplus energy. Through simulation, the results show that a household with a fix price requires 3,845 kWh of grid consumption in times of surplus energy situations, which is 34.39% of the annual consumption (11,182 kWh). A household with a BESS, which gets the LCEP price profile based on spot prices (without VGF), is able to increase its consumption during surplus energy situations to approximately 52% of the annual consumption. A household with a BESS, which it operates according to the LCEP price profile with an additional variable grid fee, can further even increase its consumption during surplus energy situations (approximately 57% of the yearly load demand). The analysis demonstrates that the LCEP price mechanism and especially variable grid fees have a positive impact to reduce surplus energy situations and thus would lead to less RES curtailment. In the given example, the variable tariffs based on spot prices without variable grid fees are also good incentives to reduce surplus energy.
1032 6
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
6. Conclusion In the paper presented load management method based on LCEP mechanism has a positive impact to reduce carbon emissions by shifting flexible consumption devices and energy storages for residential customers, and allows cost savings. Adding variable grid fees can provide an additional incentive to further increase consumption during these critical hours. Hence, the mechanism supports grid integration while allowing economic benefits at the consumption side due to lower overall electricity costs. Thus, the LCEP mechanism provides a good signal to mitigate RES curtailment and can facilitate RES grid integration. In addition, a sensitivity analysis regarding the impact of the BESS size on potential savings is provided. The marginal additional benefit steadily decreases with each additionally installed kWh of BESS. Furthermore, investigations should be conducted regarding the cost-benefit of such a system when also including investment costs. In the further research, to further validate the effectiveness of the LCEP mechanism, in-depth simulations that adapt a dynamic market and grid model could help to verify the presented results and help to fine-tune the mechanism to avoid to undesired interactions. 7. Reference [1] Li Jinfeng. Smart home energy efficiency management system considering household distributed photovoltaic power station. Dissertation, Shanghai Jiao Tong University, 2015. [2] Shen Jiajian, Yang Genke, Pan Changchun. Dynamic Model and Optimization of Hybrid Energy Management in Smart Home. Microcomputer Applications, 2012, Vol.28, No.4, 9-14. [3] Gao Siyuan. Research on energy scheduling algorithm for smart home. Dissertation, Hebei University of Engineering, 2015. [4] Zhao Z, Lee W C, Shin Y, et al. An Optimal Power Scheduling Method for Demand Response in Home Energy Management System. IEEE Transactions on Smart Grid, 2013,4(3):1391-1400. [5] Yang Xiaodan, Li Yang. Research on household electricity response mode. Power System Protection and Control, 2014, Vol.42, No.12, 51-56. [6] Sierra E, Hossian A, Britos P, et al. Fuzzy Control for Improving Energy Management Within Indoor Building Environments. Electronics, Robotics＆Automotive Mechanics Conference. IEEE Computer Society, 2007:412-416. [7] Pipattanasompom M, Kuzlu M, Rahman S. An Algorithm for Intelligent Home Energy Management and Demand Response Analysis. IEEE Transactions on Smart Grid, 2012,3(4):2166-2173. [8] Jia Zhigang, He Rong, Li Renfa, Zeng Gang. Dynamic scheduling system of home energy based on decision tree classification algorithm. Application Research of Computer, 2016,33(9). [9] K. M. Tsui, and S. C. Chan. Demand Response Optimization for Smart Home Scheduling under Real-Time Pricing. IEEE Transactions on Smart Grid, Dec. 2012, 3(4): 1812-1821. [10] W. Hart, J.-P. Watson and D. Woodruff. Pyomo: modeling and solving mathematical programs in Python. Mathematical Programming Computation, no. 3, pp. 219-260, 2011. [11] GLPK. https://www.gnu.org/software/glpk/,2018-3-31. [12] Simone Maggiore. Evaluation of the effects of a tariff change on the Italian residential customers subject to a mandatory time-of-use tariff. ECEEE Summer Study Proceedings, 2013, June 1909-1918.
ScienceDirect ScienceDirect
Energy Procedia 00www.sciencedirect.com (2018) 000–000 Available online Available online atatwww.sciencedirect.com Energy Procedia 00 (2018) 000–000
ScienceDirect ScienceDirect
www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia
Energy Procedia 152 Energy Procedia 00(2018) (2017)1027–1032 000–000 www.elsevier.com/locate/procedia
Applied Energy Symposium and Forum 2018: Low carbon cities and urban energy systems, Applied Energy Symposium andSymposium Forum 2018: Low carbon cities andcarbon urbancities energy systems, CUE2018-Applied Energy and Forum 2018: Low and CUE2018, 5–7 June 2018, Shanghai, China 5–7 June Shanghai, ChinaChina urban CUE2018, energy systems, 5–7 2018, June 2018, Shanghai,
Active Residential Load Management Based on Dynamic Real Time Active Residential Load Management on Dynamic Real Time The 15th International Symposium onBased District Heating and Cooling Electricity Price of Carbon Emission AssessingElectricity the feasibility ofof using the heat demand-outdoor Price Carbon Emission a Wei Zenga,a,*, Jan von Appenbb, for Patrick Selzambb, Mindistrict Sunaa, Boheat Chendemand , Wei Heaa,forecast Ning Xuaa temperature function a long-term a Wei Zeng *, Jan von Appen , Patrick Selzam , Min Sun , Bo Chen , Wei He , Ning Xu 330096, Chinac a,b,c State Grid Jiangxi a Electric Powera Research Institute,Nanchang, b Jiangxi Electric Powerand Institute,Nanchang, 330096, I. Andrić *,State A.Grid Pina , P. Ferrão ,Research J.Energy Fournier ., B. Lacarrière O. Le Correc Fraunhofer Institute for Energy Economics System Technology, Kassel, China 34117,,Germany a
b
a
Fraunhofer Institute for Energy Economics and Energy System Technology, Kassel, 34117, Germany 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 b
a
Rooftop photovoltaic, residential energy storage and other distributed power sources are put into operation on a large scale at the Rooftopuser photovoltaic, residential energy storage distributed powerThe sources are put into operation on a large scale at the power side, providing energy support forand theother electricity demand. demand response mechanism has become an power user side, providing resource energy for support for the electricity The Indemand response hasproportion become an indispensable and important the stable operation of thedemand. power grid. the current power mechanism grid with high of Abstract indispensable and (RE), important for theelectricity stable operation the power grid. In the current power grid withthe high proportion of renewable energy due resource to the existing market of lacking the effective interaction with end-user, electricity price renewable energy (RE), due to the existing electricity market lacking the effective interaction with end-user, the electricity price does not reflect the external characteristic of the RE and cannot embody the real relationship of power supply and consumption. District heating networks are commonly addressed in the literature as one of the most effective solutions for decreasing the doeseffect not reflect the external characteristicofofintermittent the RE andRE cannot embody realfirstly relationship of power supplya and The of increasing the consumption is limited. Thisthe paper innovatively proposes low consumption. carbon-based greenhouse gas emissions from the building sector. These systems require high investments which are returned through the heat The effectelectricity of increasing the consumption of intermittent RE is limited. This firstly innovativelyofproposes a low real-time pricing mechanism (LCEP) that effectively reflects thepaper external characteristics RE. Then, thecarbon-based flexibility of sales. Due to the changed climate conditions and building renovation policies, heat demand in the future could decrease, real-time electricity pricingismechanism (LCEP)and thatthe effectively characteristics of RE. scheduling. Then, the flexibility of the demand side resource fully exploited, LCEP is reflects used asthe theexternal adjustment lever of energy An energy prolonging the investment return period. the demand side resource is fully exploited, and the LCEP is used as the adjustment lever of energy scheduling. An energy management optimization strategy based on mixed linear integer programming is proposed to integrate all the flexible resources in The main scope of this paper is to assess the feasibility of using the heat demand – outdoor temperature function for heat demand management optimization strategy based on mixed linear integer the programming is proposed to integrate all the flexible resources in the home to be intelligent controlled. Through detailed analysis, effectiveness of the proposed strategy is verified. forecast. The district of Alvalade, located in Lisbon (Portugal), was used as a case study. The district is consisted of 665 the home to be intelligent controlled. Through detailed analysis, the effectiveness of the proposed strategy is verified. buildings that vary in both construction period and typology. Three weather scenarios (low, medium, high) and three district Copyright © 2018 Elsevier Ltd. All rights reserved. renovation scenarios wereLtd. developed (shallow, Copyright © 2018 2018 Elsevier Elsevier All rights rights reserved.intermediate, deep). To estimate the error, obtained heat demand values were Copyrightand © Ltd. All reserved. Selection peer-review under responsibility ofof thethe scientific committee of Applied Energy SymposiumEnergy and Forum 2018: Low Selection peer-review responsibility scientific committee of theand CUE2018-Applied comparedand with results fromunder a dynamic heat demand model, previously developed validated by the authors. Symposium and Selection andand peer-review under responsibility of the scientific committee of Applied Energy Symposium and Forum 2018: Low carbon cities urban energy systems, CUE2018. Forum 2018: Low carbon cities and urban energy systems. The results showed that when only weather change is considered, the margin of error could be acceptable for some applications carbon cities and urban energy systems, CUE2018. (the error in annual demand was lower than 20% for all weather scenarios considered). However, after introducing renovation Keywords: Residential Load Management; Real Time Electricity Price; Low Carbon scenarios, the errorLoad value increased up 59.5% (depending weather and renovation scenarios combination considered). Keywords: Residential Management; RealtoTime Electricity Price; on Lowthe Carbon 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.: +0-086-0791-88646884; fax: +0-086-0791-88646884. * Corresponding author. Tel.: +0-086-0791-88646884; fax: +0-086-0791-88646884.
E-mailThe address: [email protected] © 2017 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 Cooling. Keywords: Heat demand; Forecast; Climate change 1876-6102 Copyright © 2018 Elsevier Ltd. All rights reserved. 1876-6102and Copyright © 2018 Elsevier Ltd. All of rights reserved. committee of the Applied Energy Symposium and Forum 2018: Low carbon cities Selection peer-review under responsibility the scientific Selection peer-review responsibility of the scientific committee of the Applied Energy Symposium and Forum 2018: Low carbon cities and urbanand energy systems, under CUE2018. and urban energy systems, CUE2018. 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 Copyright © 2018 Elsevier Ltd. All rights reserved. Selection and peer-review under responsibility of the scientific committee of the CUE2018-Applied Energy Symposium and Forum 2018: Low carbon cities and urban energy systems. 10.1016/j.egypro.2018.09.114
1028 2
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
1. Introduction With the deployment of pilots and popularization of distributed generation equipment such as small wind turbines and rooftop photovoltaics on the residential side, Tesla Power Wall has represented the emergence of new home energy storage technologies and the intelligentization of load devices, there have been major changes in the power consumption [1]. Users are no longer just acting as a consumer but are shifting to a system with self-scheduling capabilities. The user side will have more diversified power options to respond to grid regulation information. How to make full use of distributed generation, adjust the structure of electricity consumption, improve the experience, and save electricity expenses have become the focus of attention. A large amount of research work has been reported on residential load management which usually optimize the utilization of power load in the home. Shen[2] establishes the optimal control model of energy management at home, and proposes an intelligent solution strategy based on particle swarm optimization. Gao[3] proposes an inter-door energy scheduling algorithm to ensure the stable operation of the power grid. Zhuang[4] first introduce a general architecture of energy management system in a home area network based on the smart grid and then propose an efficient scheduling method for home power usage. Yang[5] builds household electricity response mode flow diagram and its mathematical model, then adopts unit commitment of genetic algorithm to achieve household electricity optimization. Sierra[6] design a fuzzy controller to realize the minimization of energy consumption based on existing residential energy management system. Pipat[7] proposes a home energy management algorithm based on demand response for high energy consumption users, ensuring that the total power consumption is lower than a certain level by the priority effect. Jia[8] introduces a home energy dynamic scheduling system based on decision tree classification algorithm. In general, the price method of current demand side management takes time-of-use price, peak price and other forms, these electricity price’s changing period and amplitude generally is more fixed, and its flexibility is not enough, moreover, all focus on reducing the peak load of the grid. With the rapid development of renewable energy (RE), the existing electricity market lacks the effective interaction with the user, the flexibility is poor, and the electricity price does not reflect the external characteristic of the RE, often causing the embarrassing situation that power supply is insufficiency in grid peak, but there is a large amount of electricity waste in the grid valley. The electricity price cannot embody the real relationship of power supply and consumption [9]. How to adopt market-based means and price mechanisms, guide the user side to dynamically adjust the electricity demand through real-time power price signals and realize highly coordination between low-carbon energy supply and demand, are issues that need to be addressed. Firstly, in Section 1, this paper innovatively proposes low-carbon realtime electricity prices based on carbon emissions, and then fully exploits the flexibility of demand-side resources on the load side. Through price signal transmission, this paper studies load management technology based on low-carbon real-time electricity prices, integrating the power generation, power load, and energy storage devices in the home to intelligent optimal dispatch. Section 2 introduces the low-carbon real-time pricing algorithm; Section 3 presents the load management strategy; Section 4 performs detailed calculation and analysis; Section 5 analyzes the impact of different electricity price components; Section 6 summarizes the paper. 2. Low Carbon Real Time Electricity Price Since increasing generation from volatile RE causes higher demand for flexibility in the energy system, flexibilization of household customers' consumption patterns can be an important contribution. Utilization of this potential can be enabled by variable electricity rates with elements like RTP. Based on the spot market, considering the supply and demand level of power generation side and demand side, the variability of grid fees and RE surcharge, we propose a dynamic low carbon real-time pricing (LCEP) mechanism and algorithm based on carbon emission. (1) Spot Market Pricing Only the generation units with the lowest marginal costs have to be used to meet a given demand. The most expensive generation unit, which is needed to meet the given demand, respectively in this case the consumption determines the price for all other generators. This so called market clearing price (MCP) is exactly the marginal costs of the most expensive needed generation unit for a specific period. MCESU =FCESU + VOCESU +CAESU × CC (1)
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
1029 3
Where, MC denotes marginal costs, FC denotes fuel costs, VOC denotes variable operating costs, CA denotes carbon amounts, CC denotes carbon costs, ESU denotes energy source unit. (2) Variable Grid Fees Variable grid fees (VGF) are an alternative approach, instead of power prices, to motivate the grid customers to adjust their energy consumption to the requirements of the distribution grid. The recommended formula for the dynamic processing of the grid fees Mid-Price based on spot prices. The multiplicative factor has been parameterized approximately between the values 0 and 1 (depends on spot prices and the height of the concerned price parameters). VGF =SPj ⋅ MF −
∑
j
SPj ⋅ MF
QMGFi
+ MGF
(2)
Where, VGF denotes variable grid fees, SP denotes Spot Price, MF denotes multiplicative factor, MGF denotes mid-price grid fee, QMGF denotes quantity of hourly Mid-Price Grid Fees, i denotes index for each hour of the year, j denotes index for each hour of the year with Mid-Price Grid Fees. (3) Variable RES Surcharge The Variable RES Surcharge (VRS) is an approach to revitalize the spot market and give an incentive to improve the system integration of renewable energies. The idea of VRS is to couple the RES surcharge to the day-ahead spot prices with a multiplicative factor that has to be calculated based on historical data. This factor should be defined on a yearly basis such as to cover the yearly costs of the RES support scheme. VRSi =SPi ⋅ MF − SPi (3) Where, VRS denotes variable RES surcharge, SP denotes spot price, MF denotes multiplicative factor, i denotes index for each hour of the year. So the LECP price on the end-use power consumption side can be calculated as: LCEPi = MCESU +VGF +VRSi (4) 3. Load management strategies Load management strategies are defined as changes in electric usage by end-use customers from their normal consumption patterns in response to changes of the electricity price over time, or to incentive payments designed to induce lower electricity use at times of high wholesale market prices or when system reliability is jeopardized. The mechanism used in this study is price-based load control. To respond to this signal, customers must modify their load consumption, by using a battery energy storage system (BESS) that can time-decouple load power supply from power withdrawal from the grid. The objective of each household is to reduce its annual electricity procurement cost that is determined by the hourly electricity price and the hourly energy withdrawal from the grid. Since the levels of the hourly electricity price are related to the “carbon content” of the distributed electricity, concentrating withdrawals from the grid in the hours with lower prices allows for a reduction of the CO2 emissions. In the following, we focus on using PV with a BESS on the customer side. According to the price signal, the customer does not modify his/her electricity consumption habits (thus keeping the same level of comfort) and uses a BESS to change his/her power withdrawal profile from the grid, according to the electricity price profile. The BESS power set-point is calculated instant by instant only on the basis of the hourly price. Mathematically, the power set-point for the BESS can be expressed as follows: PS AVG − PS ( h ) PS min if PS ( h ) < PS AVG , PBESS ( t ) = − PBESSMAX PS ( h ) PS AVG − PSmin otherwise, PBESS ( t ) = −
PS ( h ) PS MAX
Pload ( t )
PS ( h ) − PS AVG
(5)
PS MAX − PS AVG
where: PBESS ( t ) is the power exchanged by the BESS, PS ( h ) is the price signal, i.e. the electricity price in the hour h including the time instant t, PBESS is the maximum charging/discharging (nominal) power of the BESS, PS AVG is MAX
the average value of the price signal over the day, PS MAX and PSmin are the maximum and minimum value of the price
signal over the day, Pload ( t ) is the power required by the loads in the household.
There are some constraints that must be taken into account. The power withdrawn from the grid 𝑃𝑃𝑃𝑃𝐺𝐺𝐺𝐺 (t) must be less than or equal to the maximum contractual power PCONTR ; moreover, no power injections from the household to the
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
1030 4
grid are allowed.
0 ≤ PG ( t ) ≤ PCONTR
The power exchanged by the BESS must be less than or equal to its nominal power PBESS
(6) MAX
(this constraint is
automatically met by Eq. 5 in the charging phase; in the discharging phase the constraint is binding if Pload ( t ) > PBESS PBESS ( t ) ≤ PBESSMAX
The State of Charge (SoC) of the BESS must be kept in the allowed range, to preserve the battery life. 20% ≤ SoC ( t ) ≤ 100%
MAX
).
(7) (8)
The exchange profile with the grid of the household and the related economic value and CO2 emissions are calculated, in order to reduce the overall purchase electricity costs. The control strategy used in this case study is: the BESS is charged only with the PV plant production in excess of the load; and the BESS is discharged according to the price signal. The basic algorithm is as following: if PPV(t) > Pload(t) if SoCmin ≤ SoC(t) < 100% PBESS(t) = PPV(t) – Pload(t) //charge the storage else PBESS(t) = 0 //battery full: do nothing endif else if PPV(t) == Pload(t) PBESS(t) = 0 //do nothing else if PPV(t) < Pload(t) if SoCmin < SoC(t) ≤ 100% and PS(h) > PSAVG PS ( h ) PS ( h ) − PS AVG // discharge the storage PBESS ( t ) = Pload ( t ) PS MAX PS MAX − PS AVG else PBESS(t) = 0 //do nothing endif endif
where: PPV is the power produced by the PV plant, Pload is the power consumed by the load, PBESS is the power exchanged by the BESS, SoC is the state of charge of the BESS, 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃(ℎ) is the price signal, i.e. the electricity price in the hour h including the time instant t, 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐺𝐺𝐺𝐺 is the average value of the price signal over the day, 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑀𝑀𝑀𝑀𝐴𝐴𝐴𝐴𝑀𝑀𝑀𝑀 is the maximum value of the price signal over the day, 𝑃𝑃𝑃𝑃𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 (𝑡𝑡𝑡𝑡) is the power required by the loads in the household. 4. Case Study
A mixed integer linear optimization problem (MILP) is formulated to minimize the cost of daily energy withdrawal by computing at midnight the optimal BESS power set-point for the whole day. Python and the Pyomo tool box [10] are chosen simulation environment for the described problem. Such MILP problems are well explored and a couple of open-source solves such as GLPK[11] are available. The household price profiles for the following analysis are: (1) Spot prices incl. a high carbon price with variable grid fees and variable surcharges; (2) Spot prices incl. a high carbon price with normal grid fees and variable surcharges; (3) Spot prices incl. a high carbon price with normal grid fees and normal surcharges; (4) Spot prices incl. a high carbon price with variable grid fees and normal surcharges; (5) Spot prices incl. a low carbon price with variable grid fees and variable surcharges; (6) Spot prices incl. a low carbon price with normal grid fees and variable surcharges ; (7) Spot prices incl. a low carbon price with normal grid fees and normal surcharges; (8) Spot prices incl. a low carbon price with variable grid fees and normal surcharges The 2016 annual power generation and consumption data of Jiangxi province power grid is used to calculate LCEP price. And the SCUDO7 software [12] is utilized to simulate the typical household load profiles. The case focus is on the annual savings related to a reduction of electric energy purchase costs and the annual reduction of CO2 emissions. As far as BESS capacity is concerned, 10 different values (from 1 kWh to 10 kWh) have been considered. A roundtrip efficiency of 85% has been assumed for the BESS. And we assume that the household is equipped with a 4.5 kW peak power photovoltaic system.
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
Fig.1. Annual energy purchase savings.
1031 5
Fig.2. Annual energy purchase percentage savings.
It can be seen that, with BESS size increases from blue to yellow, the larger the BESS, the higher the saving, and also in this case a saturation is reached, due to the saturation of the percentage of self-consumption. With respect to the strategy based on self-consumption maximization the saving is slightly higher with the smallest BESS, while with the largest ones the saving is typically lower. With the largest BESS sizes, savings reach values around 80%.
Fig.3. Annual reduction of CO2 emissions.
Fig.4. Annual percentage reduction of CO2 emissions.
The results of CO2 emissions reduction are less dependent on the price signal than the results of cost savings. Again, the larger the BESS, the higher the saving. With respect to the self-consumption maximization strategy reduction of CO2 emissions is typically slightly higher with the smallest BESS, while with the largest ones in some cases the reduction is lower. 5. Impact of Variable Grid Fees In a scenario with high share of fluctuating renewable energies, a lot of surplus energy is likely to be produced. The surplus energy is an indicator for RES curtailment. E.g. a high RES surplus increases the likelihood of local grid integration challenges, might result in RES curtailment, if the distribution system operator has no other option, and cannot reinforce the grid on short notice. Based on the proposed LCEP mechanism in Section 2, it can assist in setting appropriate price signals to incentivize additional load consumptions and thus reduce the surplus energy. Through simulation, the results show that a household with a fix price requires 3,845 kWh of grid consumption in times of surplus energy situations, which is 34.39% of the annual consumption (11,182 kWh). A household with a BESS, which gets the LCEP price profile based on spot prices (without VGF), is able to increase its consumption during surplus energy situations to approximately 52% of the annual consumption. A household with a BESS, which it operates according to the LCEP price profile with an additional variable grid fee, can further even increase its consumption during surplus energy situations (approximately 57% of the yearly load demand). The analysis demonstrates that the LCEP price mechanism and especially variable grid fees have a positive impact to reduce surplus energy situations and thus would lead to less RES curtailment. In the given example, the variable tariffs based on spot prices without variable grid fees are also good incentives to reduce surplus energy.
1032 6
Wei Zeng et al. / Energy Procedia 152 (2018) 1027–1032 Author name / Energy Procedia 00 (2018) 000–000
6. Conclusion In the paper presented load management method based on LCEP mechanism has a positive impact to reduce carbon emissions by shifting flexible consumption devices and energy storages for residential customers, and allows cost savings. Adding variable grid fees can provide an additional incentive to further increase consumption during these critical hours. Hence, the mechanism supports grid integration while allowing economic benefits at the consumption side due to lower overall electricity costs. Thus, the LCEP mechanism provides a good signal to mitigate RES curtailment and can facilitate RES grid integration. In addition, a sensitivity analysis regarding the impact of the BESS size on potential savings is provided. The marginal additional benefit steadily decreases with each additionally installed kWh of BESS. Furthermore, investigations should be conducted regarding the cost-benefit of such a system when also including investment costs. In the further research, to further validate the effectiveness of the LCEP mechanism, in-depth simulations that adapt a dynamic market and grid model could help to verify the presented results and help to fine-tune the mechanism to avoid to undesired interactions. 7. Reference [1] Li Jinfeng. Smart home energy efficiency management system considering household distributed photovoltaic power station. Dissertation, Shanghai Jiao Tong University, 2015. [2] Shen Jiajian, Yang Genke, Pan Changchun. Dynamic Model and Optimization of Hybrid Energy Management in Smart Home. Microcomputer Applications, 2012, Vol.28, No.4, 9-14. [3] Gao Siyuan. Research on energy scheduling algorithm for smart home. Dissertation, Hebei University of Engineering, 2015. [4] Zhao Z, Lee W C, Shin Y, et al. An Optimal Power Scheduling Method for Demand Response in Home Energy Management System. IEEE Transactions on Smart Grid, 2013,4(3):1391-1400. [5] Yang Xiaodan, Li Yang. Research on household electricity response mode. Power System Protection and Control, 2014, Vol.42, No.12, 51-56. [6] Sierra E, Hossian A, Britos P, et al. Fuzzy Control for Improving Energy Management Within Indoor Building Environments. Electronics, Robotics＆Automotive Mechanics Conference. IEEE Computer Society, 2007:412-416. [7] Pipattanasompom M, Kuzlu M, Rahman S. An Algorithm for Intelligent Home Energy Management and Demand Response Analysis. IEEE Transactions on Smart Grid, 2012,3(4):2166-2173. [8] Jia Zhigang, He Rong, Li Renfa, Zeng Gang. Dynamic scheduling system of home energy based on decision tree classification algorithm. Application Research of Computer, 2016,33(9). [9] K. M. Tsui, and S. C. Chan. Demand Response Optimization for Smart Home Scheduling under Real-Time Pricing. IEEE Transactions on Smart Grid, Dec. 2012, 3(4): 1812-1821. [10] W. Hart, J.-P. Watson and D. Woodruff. Pyomo: modeling and solving mathematical programs in Python. Mathematical Programming Computation, no. 3, pp. 219-260, 2011. [11] GLPK. https://www.gnu.org/software/glpk/,2018-3-31. [12] Simone Maggiore. Evaluation of the effects of a tariff change on the Italian residential customers subject to a mandatory time-of-use tariff. ECEEE Summer Study Proceedings, 2013, June 1909-1918.