You are on page 1of 10

Energy 132 (2017) 126e135

Contents lists available at ScienceDirect

journal homepage:

An optimisation algorithm for distributed energy resources
management in micro-scale energy hubs
n-Blay*, Guillermo Escriva
Carlos Rolda -Escriva
, Carlos Rolda 
n-Porta, Carlos Alvarez-Bel
Institute for Energy Engineering, Universitat Polit
ecnica de Val
encia, Camino de Vera, s/n, edificio 8E, escalera F, 5a planta, 46022 Valencia, Spain

a r t i c l e i n f o a b s t r a c t

Article history: In this paper, a new algorithm for optimal management of distributed energy resources in facilities with
Received 12 October 2016 distributed generation, energy storage systems and specific loads e energy hubs e is shown. This
Received in revised form method consists of an iterative algorithm that manages optimal energy flows to obtain the minimum
1 March 2017
energy cost based on availability of each resource, prices and expected demand. A simulation tool has
Accepted 6 May 2017
been developed to run the algorithm under different scenarios. Eight different scenarios of an energy hub
Available online 7 May 2017
have been simulated to illustrate the operation of this method. These scenarios consist of a demand curve
under different conditions related to the existence or absence of renewable energy sources and energy
Distributed energy resources
storage systems and different electricity tariffs for grid supply. Partial results in the iterative process of
Renewable energy sources the developed algorithm are shown and the results of these simulations are analysed. Results show a
Energy management system good level of optimisation of energy resources by means of optimal use of renewable energy sources and
Energy resources management optimal management of energy storage systems. Moreover, the impact of this optimised management on
Energy hub operation carbon dioxide emissions is analysed.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction the lowest energy costs.
A facility with multiple energy sources that has energy pro-
Energy consumption has been growing over recent decades. In duction, conversion and storage technologies (RESs, batteries, ice
order to minimise the dependence on fossil fuels, new energy re- storage, hydrogen cells and so on) to supply electricity (and other
sources are being integrated in energy systems [1], such as services such as heating or cooling) is widely called an energy hub
renewable energy sources (RESs) e.g. wind, solar, biomass and so [8]. From the perspective of end users in energy hubs, optimal
on, and energy storage systems (ESSs). Also, new regulations have management of their available resources consists of controlling all
to be developed to support this integration [2]. Optimal manage- energy flows in their facilities (between power grid, distributed
ment and control of the available resources is a key issue to be generation resources, ESSs and loads) to minimise the total energy
addressed and many research studies have been developed. Some costs. Therefore, these facilities require reliable energy manage-
of these works study optimal planning for new generation facilities ment systems (EMS) with real-time data acquisition and processing
[3]. In other works, optimal management is studied focusing on from energy resources and external variables (e.g. temperature,
management of loads [4], i.e. demand side management (DSM). wind speed or energy purchase prices) and sophisticated algo-
Some research studies are focused on DSM of individual facilities rithms to achieve an optimal management of the available re-
[5]. Conversely, in other studies, management of loads is studied at sources along the time.
aggregator level, to optimise the global benefit [6]. Nevertheless, Over the past years, various algorithms to find an optimal so-
the optimisation problem may be studied from the perspective of lution of energy operation in energy hubs have been proposed. For
energy generation facilities management [7], since modern facil- example, Ref. [9] shows a multi-objective optimisation method
ities include more and more distributed energy resources (DERs) applied to manage several generators in a microgrid. So, some al-
that must be managed to get the maximum benefit from them, i.e. gorithms have been developed to optimise overall costs by man-
aging ESSs to purchase energy for microgrids with several available
sources [10]. Many studies indicate that the impact of RESs is
greater when ESSs are installed, as the unpredictability of these
* Corresponding author.
resources has a lower impact on the optimal solution [11]. In fact,
E-mail address: (C. Rolda
0360-5442/© 2017 Elsevier Ltd. All rights reserved.

[12]. However. only linear functions are used in this algorithm. depending on the charging of the available resources in energy hubs. Methodology of loads. reliability of supply. Others are focused on optimal planning for smart grids. Therefore. So. a simulation step size t is chosen 3 defines the scenarios to be simulated in order to show the and a set of N identical intervals are taken of a length . Some advantages of DEROP on the raw material and time. a variable cost algorithm.. Moeini-Aghtaie ficiencies. energy As aforementioned. However. for PV panels and wind generator. only energy costs are taken into ac- and with different purposes. most algorithms to optimise management transmission. in which some aspects might be relevant. by developing demand response programs and efficient management 2. Rastegar and functions. DEROP is easy to be maintenance operations are considered since these resources do implemented as an online service for end users. From 1 to n pout effective power that is extracted from batteries when they are discharged Parameters. n-Blay et al. Ref. DEROP algorithm calculates the cost of the total consumption In this paper. [13]. this paper shows DEROP algorithm. fixed demand curve. despite the interest of researchers in finding an analysed and discussed in this section. tjþ1  Superscripts pin effective power that batteries receive when they are ðiÞ iteration in the DEROP algorithm. Fig. tjþ1  PV photovoltaic RES renewable energy source C total cost of energy during the simulated period UPV cnica de Vale Universitat Polite ncia Ajk mean available power of resource k during the interval ½tj . not need any raw material. length of each simulation interval t0 initial time in the simulation period Acronyms tj instant in which the jth simulation interval ends (j > 0) DER distributed energy resource dj total power demand at instant tj DEROP distributed energy resources optimisation (name of pjk power provided by resource k at instant tj the proposed algorithm) qjk associated cost of resource k at instant tj DG distributed generation Qj total cost of generated power at instant tj DSM demand side management S schedule consisting of a N  n binary matrix that ESS energy storage system indicates the state of each resource at each simulation LabDER Laboratory of distributed energy resources interval MILP mixed integer linear programming Cj total cost of supplied energy during the interval ½tj . From 0 to N  1 WstðjÞ total energy stored in batteries at instant tj k energy resource index. some conclusions to optimal solution to systems operation.tjþ1  energy received by loads from battery during interval divided ½tj . a new algorithm to achieve an optimal manage. in this experiment. most of the developed this work are drawn in Section 5. which optimises DG and ESSs management allowing non-linear nomic dispatch in a grid with several energy hubs. Each generation resource has a different cost depending optimising the management of ESSs. Section 4 reducing peak load or to help successfully incorporate distributed shows the results of multiple simulations carried out with DEROP generation (DG) as shown in Ref. DEROP is flexible to be the main goal is not to plan the installation of generation or storage used under a wide range of situations with many different DERs systems. 1 shows the concept of an energy hub where Fotuhi-Firuzabad [16] describes an algorithm to achieve optimal DEROP algorithm would enable energy cost minimisation to meet a management of energy resources and loads along a day. From ð0Þ to ðf Þ charged A available power used to charge batteries Subscripts hc efficiency of the battery charging process j time index in the simulation period. with a simple optimise the operation of energy resources in an existing facility. As regards ESSs. DEROP algorithm and its features are shown in this section. For example. and functions pbat net power that batteries lose when they are discharged T Duration of the simulation period hd efficiency of the battery discharging process N Number of intervals in which the simulation period is Wout½tj . modifications in the energy supply schedule attempting to reduce The main target of the new DERs optimisation algorithm (DEROP) is this cost until reaching an optimal situation where no improvement to minimise energy costs by maximising RESs generation and is possible. but they assume linearity in costs and ef- installed and diesel generators in an energy hub. This guarantees that the algorithm will This paper is organised as follows. [15] shows an optimisation algorithm to compute the eco. Section For a simulation of length T. such as topology. which has been developed to achieve an optimal control may be associated with these systems. variables. All aspects related to and discharging profiles. Rolda 127 Nomenclature f last iteration of DEROP algorithm t simulation step size. Others are focused on DG shows an algorithm to optimise the capacity of batteries to be resources management. Due to this reason. from grid supply under a set of conditions and it iterates with some ment of the available energy resources in energy hubs is proposed. only costs for costs and efficiency of each resource. optimisation methods are focused on DSM as in Ref. algorithm and the developed tool. Section 2 describes DEROP maximise the use of these sources. Finally. / Energy 132 (2017) 126e135 C. and so on. The aim of this algorithm is to are that it allows considerable energy cost reductions. As procedure and a fast computation algorithm. including non-linear functions for count. et al. [14] of energy resources are focused on DSM. These results are compared. tjþ1  n Number of energy resources in the energy hub (generation and storage resources) ESSs are key elements to reduce energy costs in facilities by performance of DEROP algorithm and the simulation tool.

the problem is formulated as follows: that provides an optimal solution with little computational effort. a proposed schedule SðiÞ is simulated. cost pjk being the total power provided by resource k in a set of n ðiÞ Cj is computed (11). expression (3) could be linearised ! and a mixed integer linear programming (MILP) method would ZT X N1 X n X N1 ðiÞ ðiÞ ðiÞ provide an accurate solution.tjþ1  . …. along with k¼1 the operating rules programmed into the software. tjþ1 . tjþ1 . that will be: and the constraint addressed in (2). ðiÞ ðiÞ Qj ¼ pjk $qjk (3) Throughout the entire simulation period the total energy cost k¼1 C ðiÞ . ck2½1.tjþ1  ¼ pout $dt (7) tj And the energy stored in battery at the end of interval ½tj . the effective po. N  1 a state (ON/OFF) associated X n dj ¼ pjk (2) with each element of the simulated facility. when battery is being charged. the effective power energy is provided by power grid. With the prices of each resource at these intervals. C ðiÞ ¼ Qj $dtz pjk $qjk $t ¼ Cj (12) j¼0 k¼1 j¼0 MILP algorithms have been used at microgrid level [18] and at 0 district-scale level [19] by other researchers. determine en- ergy flows. hc ¼ hc ðAÞ (9) hd ¼ hd ðpout Þ (10) T t¼ (1) To calculate the optimal schedule through an iterative proced- N ure.   X N1 To clarify this. in each iteration i. pjn (13) wer that it receives (pin ) is not 100% of the available power (A): j¼0 pin ¼ hc $A (4) subject to pjk  Ajk . These states. pj2 . only accepting those that reduce C ðiÞ . Therefore. j2½0. DEROP algorithm begins with the steps in which the greatest Ztjþ1 potential savings may be obtained and it analyses possible alter- Wstðjþ1Þ ¼ WstðjÞ þ pin $dt (5) natives. real efficiencies would not be constant: Fig. when battery is being discharged. Wstðjþ1Þ . with tj DEROP algorithm. Therefore. Wstðjþ1Þ . minimise Cj pj1 . Wout½tj . with the purpose of gradually . Energy hub concept to implement DEROP algorithm in order to optimise DERs. resulting in new schedules SðiÞ . and its The resulting intervals are energy cost C ðiÞ is calculated as explained below. N  1 (14) This has a direct impact on the energy stored in battery at the end of interval ½tj . That is. total cost of generated power in iteration i for that Cj ¼ Qj $dtz pjk $qjk $t (11) interval ½tj . tjþ1 . allowing non-linear functions for costs or efficiencies. / Energy 132 (2017) 126e135 C. tjþ1 . due to the Assuming the step size t is constant (intervals may occasionally low value of n in real cases. the amount of energy received by loads from battery during interval ½tj . N  1 : tjþ1 ¼ tj þ tg. Rolda pout ¼ hd $pbat (6) In this case. At each instant tj . as proposed in other studies [17]. The simulated f½tj . tjþ1 . the total schedule consists of a N  n binary matrix that indicates the state of power demand dj is satisfied as each resource at each simulation interval. n. cj2½0. the authors use an iterative algorithm have different durations). Ajk being the available power of resource k during interval ½tj . tjþ1 . will be: Ztjþ1 Wout ½tj .128 n-Blay et al. j2½0. in which all Similarly. this schedule is modified step by extracted from it (pout ) is not 100% of the power that it gives (pbat ): step. At each interval ½tj . tjþ1  will be tj k¼1 X n Note that these prices are not necessarily linear functions. (6) and (8). each resource k has an associated cost Ztjþ1 ðiÞ ðiÞ X n ðiÞ qjk . available resources. each schedule has ðiÞ ðiÞ for each interval ½tj . is obtained as Under certain simplifications. 1. will be: Ztjþ1 pout Wstðjþ1Þ ¼ WstðjÞ  $dt (8) hd tj In expressions (4). tjþ1 . starting from an initial schedule Sð0Þ . However.

then Sðiþ1Þ is accepted. 2. in each iteration i. proposal Sðiþ1Þ is reached. leaving schedule Sðf Þ as the optimal selling is considered because it is not covered by current Spanish one. Implementation features power supplied by grid is decreased during the most expensive ðiÞ ðiÞ available interval ½tj . the grid power reduction attempted in tj does not related to management of available resources in order to establish produce any benefit.g. LabDER is a laboratory ðiÞ ðiÞ with various DERs (solar. ðiÞ The EMS of this energy hub was developed to design experiments Otherwise. the algorithm attempts to times to reduce power that enable a new state with a schedule exploit all RESs charging batteries with surplus generation (surplus Sðf þ1Þ such that C ðf þ1Þ < C ðf Þ . n-Blay et al. the algorithm optimises the usage of RESs by managing ESSs. so the algorithm goes back to the previous the basis of a control strategy that optimises energy costs. a new schedule fuel or biogas. If C ðiþ1Þ < C ðiÞ . wind power. among Fig. a generator that works with some point. 2. laws [21]). . 2. At algorithm ends when in iteration i ¼ f þ 1 there are no available the beginning of the iterative process. In this stage. tbþ1  with b < j.1. DEROP ation of the algorithm. is cancelled (marked as not available) for next iteration. / Energy 132 (2017) 126e135 C. batteries and grid supply) and some loads such as heaters or other consumption of the laboratory (e. with associated cost C ðf Þ being the minimum possible cost for regulation related to self-consumption [20] and electricity sector that scenario. tjþ1  and battery power is supplied during The algorithm has been tested in the Laboratory of DERs (Lab- that time period. DEROP algorithm diagram. DEROP consists of two separated stages. it is charged in the cheapest available interval ½tb . When demand is successfully supplied. The diagram of DEROP algorithm is shown in Fig. Rolda 129 ðiÞ ðiÞ reducing initial cost C ð0Þ up to a situation where schedule Sðf Þ has state (schedule SðiÞ ) undoing the last changes and interval ½tj . If battery requires more charge to meet demand at DER) at Universitat Politecnica de Vale ncia. tjþ1  the minimum possible associated cost C ðf Þ where f is the last iter. During the second stage. lighting) [22].

Scenarios identified as B1 and B2 add the DEROP algorithm in continuous mode. en- Table 1 ergy will be optimally managed for the next hours. several real buildings with some RESs. it takes up to 10 h to in this paper to test and show the performance of DEROP algorithm. shoulder and valley periods. every new measured value.g. initial state of ESSs) are the inputs to the developed simulation than the others. thus achieving highly accurate results. wind generator. This software uses a set of rules of installed equipment price of energy is particularly high and for electric power supply operation (panels. a charging rate of Using this simulation tool. energy hub's management. These scenarios correspond to a real power or rental of measuring equipment are not covered. In valley period.130 n-Blay et al. shoulder and cantly reduced. Rolda other purposes.7% per hour has been assumed. may be used to are intended to illustrate the second phase of the algorithm for the improve forecasts during the next time intervals. new To reach this algorithm certain assumptions that reduce the rules and conditions may be added easily. ESSs (battery) and two Secondly. only operating costs of the facility are considered. three different speeds have been considered ESSs and power grid connection. If stored energy is less than 38% Worksheets to show graphical results and handle all the data. valley. During shoulder period. such as the availability or absence of RESs. The real consumption curve shapes of a building simulated with the optimal schedule of each scenario that provides the lowest energy controllable loads in LabDER. sources (like the biomass generator or the hydrogen cell that are To calculate CO2 emissions associated with the proposed not being used at the moment) and more loads in parallel. only maintenance costs are taken into . which will be effectively achieved only if discrimination. VBA of the effective capacity. 8 scenarios have of electricity bills. the assumed value is 0. The simulations selected for this work show that. whereas scenarios X2 have a tariff with hourly able resources optimally. D) have a tariff with no time the aim of this study is to achieve a methodology to control avail. wind. C2 Yes Yes No D2 Yes Yes Yes cles. energy costs are signifi. so only 3. Scenarios definition the purchase price of energy for every moment of the day and DERs costs are used when computing the total costs. / Energy 132 (2017) 126e135 C. for every simulation interval (they are null by default). Some methodologies to improve so its costs (reached in iteration i ¼ 0) are used to be compared forecasts have been developed in other studies [24]. the assumed charging rate is 13% per calculates the use of each DER and its cost in each simulation step. Running with other scenarios. Thus. account. When a schedule. simulated they are taken into account when analysing the results to draw with the manageable loads in LabDER under different conditions. when a battery is discharged too C1 No Yes No quickly it has lower efficiency). emissions and costs must be introduced peak. inaccurate the forecasts have been in the previous hours. …. a value of 1 ton of CO2 per (e. However. MWh is estimated. if the forecasts of Related to emissions. These assumptions are explained below. mean values in kg/kWh have been taken from Ref. Finally. restrictions.e. with hourly discrimination. This source is reserved for situations where the software. This charging profile has been ESS and grid supply have been simulated under different situations proposed by means of real experiments data (although it can be and some of these simulations carried out in LabDER are presented controlled in LabDER). for RESs. an 3. The developed tool is prepared to allow more generation re- mise these costs. Also. and problems. costs associated with energy supply and executes DEROP to opti. However. when the charge level exceeds 77%.26 tons of CO2 per MWh is Electricity prices are obtained from the System Operator's assumed. the considered value is 0. Under these conditions. planning of CO2 per MWh. a charging rate of 20% of this capacity per code to execute DEROP algorithm and SQL Databases connections hour has been assumed.22 tons of demand needs for the selected facility during the simulated period CO2 per MWh. This program 77% of the effective capacity. As previously stated. flows. other concepts To show the performance of DEROP algorithm. This tool is based on MS Excel depending on the level of storage. Scenario name Tariff with hourly discrimination RESs ESS The simulation tool has been designed to include several oper- ation profiles for batteries. In the simulations A2 Yes No No carried out to test the algorithm with real data. To charge battery. gener. moderate costs have B2 Yes No Yes been assumed for ESSs assuming reasonable operation speed cy. such as performance reduction due to A1 No No No B1 No No Yes losses or speed of cycles (i. A1 and A2 are base scenarios without RESs or ESSs. no matter how Features of the simulated scenarios. since the purpose of this research was to propose an al- In order to facilitate the task of proposing and simulating this gorithm for optimal management of an existing hub with specific algorithm. First. in these simulations due to the fact that it is much more expensive ator. During peak period a value of 0. a quarter-hourly energy curve for a whole day) and the ex. These new features are complexity of the problem and do not affect the optimality of expected to improve the results of the optimisation process [25]. depending on the time-of-use that defines peak. every resource D1 No Yes Yes (including battery) has been assigned a cost. Scenarios X1 (where X ¼ A. some conclusions. For the fuel generator. Table 1 shows the 8 scenarios. costs and emissions in a facility with several DG resources. These scenarios both in consumption and generation systems. good forecasts are available. such as the cost associated with the contracted been defined and simulated. the authors have developed a tool to simulate energy RESs already installed. running DEROP every step. inverter. no emissions grid) through which it computes actual energy flows and overall are considered. three different values have been assumed all resources and demand are accurate. Similarly. All the data related to costs. When this methodology is implemented for an algorithm. [23]). When the charge level is between 38% and to read real data from an energy hub control system. charger. although at first. hour. all the simulations have been carried out with real different electric tariffs: a tariff with no time restrictions and a tariff data (solar power and wind power generation curves in LabDER. For RESs (PV panels and wind generator). grid. complete a full charge. generator. [23] in new resource is defined. DEROP algorithm have been taken. real hourly purchase prices of energy cost has been obtained with the developed tool using DEROP from Ref. forecasts might be inaccurate. possibility of storing energy to the previous ones. this energy source is not used pected availability of each energy source (solar. although curve shape of an academic building at UPV for a full day.25 tons of website [23] every day.

3. n-Blay et al. sce.7 kg of CO2. In this scenario. narios D1 and D2 are complete scenarios. the initial scenarios C1 and C2 add the existence of RESs to scenarios A1 and schedule Sð0Þ consists of using grid supply for the entire simulation A2. Rolda 131 two types of tariff analysed in this study. To perform this simulation the selected step size is scheduled in expensive intervals. Fig. Energy cost in this iteration is V 9.9 kg of CO2. In iteration i ¼ 1 the available RESs are used and the surplus which illustrate the whole process of DEROP algorithm and allow generation is used to charge batteries. which is a very scenario has an electricity tariff with no time restrictions. tjþ1  with the highest purchase price (which takes 4. demand of 71.96 and the total emissions are 11. Results in iteration i ¼ 1 in scenario D1. 5. The associated emissions are 17. with a total to this facility. As an example.18. Finally. Their simulation allows the analysis of the contribution of RESs period (one day). Scenarios simulation place during the interval 21:00e21:15) and grid supply is discon- nected during that time interval. until a moment when battery is 15 min. / Energy 132 (2017) 126e135 C. To complete the second iteration DEROP algorithm looks for interval the ½tj . important task in this kind of systems [10]. As a result of this iteration the The described scenarios have been simulated using the devel- energy cost has decreased until V 5. there are RESs and an ESS (batteries).6 kg of CO2. in iteration i ¼ 0. This mise ESSs usage to minimise total costs of energy. On the other hand. with no possibility of storing energy.87 kWh. the simulation of scenario D1 is detailed 11. At the beginning of the simulation. 3. The result of this step is the comparison of optimal costs and emissions with other shown in Fig. In the next iterations new grid disconnections are in this section. with all the resources. This phase aims to opti. . The total cost of supplied energy in this case is V scenarios.94 and total emissions are oped tool.

40%. and D2) provide greater savings than the sum of the savings ob- compared to results of scenario A1. However.132 n-Blay et al.08 supply failures take place at that time.06% is achieved (scenario D1). great saving of CO2 emissions. The simulated building has a consumption profile in which the Fig.03 energy has been reduced to V 5. Simulation results in the final iteration (i ¼ 35) in scenario D1.22 and total emissions have Solar generation (kWh) 15. nomic savings up to 18% higher than BM1. savings of 44. Continuing with this procedure the optimal situation is reached tariff with hourly discrimination is not profitable.05% (scenario B2) and the RESs below in order to analyse the performance of DEROP algorithm in produce a saving of 37. due to the low in the iteration i ¼ 35 when there is no available time interval to night-time consumption. 5 shows the evolution of overall energy costs throughout the (which achieves savings of 54. the total cost of Emissions (kg CO2) 10. a significant economic saving is obtained in scenario D1. battery cannot meet demand during the interval 21:15e21:30 and Cost (V) 5. a synergy 4. the reduction obtained in C1 compared to A1 is 34. there has been a tained with each resource separately. This is possible because batteries allow the storage of surplus generation and through an optimal management. 4.3 kg of CO2.2 kg of CO2.97 decreased up to 10. Table 4 shows the results of both comparison with the energy supply from grid (scenario A2). 4 and results are discrimination more profitable than the tariff with no time re- summarised in Table 2. to charge ESS during the cheapest hours and disconnect the grid savings up to 54.22 battery is charged during the cheapest interval (during 00:00e00:15). a reduction of are up to 14% higher with DEROP and energy savings from the grid 44. So. Rolda not able to meet demand. This may be observed by comparing scenario D2 to A2 Fig.40% of the total energy costs are obtained in during the most expensive hours.11 19 a fraction of grid supply is disconnected during 10:45e11:00 and Demand (kWh) 71. based on scenario A2.86%. In iteration i ¼ 18. However. In D2 (Fig. When battery is used. scenario D2 different situations.66%). As for methods for scenario D2 compared to A2.1. CO2 savings due to the impact of RESs.21 and overall emissions of 10. The basis of BM1 is savings are achieved in D1 and D2. to complete iteration i ¼ Wind generation (kWh) 11.69% (scenario C2). 6).40%) and D1 to A1 (which achieves entire simulation.87 Power grid supply (kWh) 42. strictions. The greatest been used to compare the results with DEROP. This is because the total energy cost in disconnect grid supply and meet demand with batteries obtaining scenario A1 is lower than the cost in A2. which are higher than the sum of both savings. Scenarios results and discussion between both resources takes place. The total cost is V 5. That is. In addition. DEROP achieves eco- emissions. Results of all scenarios are compared the batteries produce a saving of 14. as it was expected. In addition. This situation is shown in Fig. At this point. if a new fraction of Table 2 grid supply is disconnected (during the interval 10:45e11:00) Final simulation results in scenario D1. It must be noted that scenarios that have all the resources (D1 rithm. / Energy 132 (2017) 126e135 C. The result of this iteration is a total cost of V 5. Therefore. as a result of the optimal management using DEROP algo.08 and the total emissions agement of resources through DEROP makes the tariff with hourly are 10 kg of CO2. An additional algorithm called Basic Management 1 (BM1) has Table 3 shows the results of all simulated scenarios. are up to 10% higher. . optimal man- an economic benefit. achieves savings of 54.

81 Number of iterations (N) 0 23 1 35 0 15 1 32 Fig.11 11.97 0. Evolution of costs in scenario D1.48 47.65 5.11 0.68 9.00 15.87 71. n-Blay et al. .87 71.30 4.00 0.00 0.87 69. Scenario A1 B1 C1 D1 A2 B2 C2 D2 Cost (V) 9.00 11.93 11.87 71.85 11.19 47.61 Emissions (kg CO2) 17.97 15.87 71.93 16.18 8.87 71.08 10.11 Demand (kWh) 71.87 71.00 15.11 11. Rolda 133 10 9 8 7 6 Cost (€) 5 4 3 2 1 0 5 10 15 20 25 30 35 Iteration Fig.87 71.69 6.95 Solar generation (kWh) 0.00 0. / Energy 132 (2017) 126e135 C.03 17. 5.22 71.93 16. 6.87 Power grid supply (kWh) 71.68 10.11 8.00 11.97 Wind generation (kWh) 0.56 41. Table 3 Simulation results comparison between all scenarios.00 0.96 5.56 42.97 15.87 69. Simulation results in scenario D2 (tariff with hourly discrimination).

6175785. be compared with different tariffs or resources.51 Energy Rev May 2015. Optimal management of wind and solar energy resources. In addition. Cost savings (%) e 36. Xiao Yunhan.87 71.2015. Appl Energy 15 April 2014. Acknowledgement [20] R.  n emergentes. http:// dx. In addition.enbuild. Energy October 2015.20 41. Mendes Gonçalo. Ahmadi Abdollah. Lambotharan S. ISSN mising CO2 emissions. June 2011.81 j.95 encouraging renewable energy integration & best practices.03. resource and the expected demand as inputs.doi. defining the size of RESs.renene.1016/ Power grid supply (kWh) 71. hub under different conditions have been simulated. [22] Perez-Navarro A.10.doi. MILP (mixed-integer linear programming) approach for residential with no time restrictions. tion of renewable energy resources. Energía y Turismo de Espan ~ a. ogy shown in this study. These differences are significantly increased with smaller [6] Hansen TM. Renew Energy February 2016. Dehghanian P. Vale Zita A.97 15. http://dx. [15] Moeini-Aghtaie M.179.1109/ Cost (V) 10.5(no. thermal energy management of a residential energy hub.014. scenario D1 has required 35 linear programming based optimization for home demand-side management iterations to reach the optimum.doi. Conclusions 10. [11] Kienzle F. Fotuhi-Firuzabad Mahmud.1016/ in smart grid. ISSN 0306-2619. Smart distribution: coupled microgrids. Honarmand Masoud.1109/TSTE. Ai Qian.08.11 Usaola-García Julio. D.2014. Khodr HM. http://dx. (DEROP) is shown. Review of policies Emissions (kg CO2) 17. The goal of DEROP is to minimise the cost of http://dx.renene.013. Zurich: ETH (Swiss Federal Institute of Technology). Renew Energy May 2016. Experimental verification of hybrid renewable systems as feasible en- cnica de Vale de la Universitat Polite ncia (Spain) and the GV/2015/ ergy PhD thesis.doi. the Programa de Apoyo a la Investigacion y Desarrollo (PAID-06-12) et al. gramming) model for optimal design of district-scale distributed energy resource systems.035.11 11. a tool has been developed to simulate DEROP in energy hubs constrained by wind availability.41 4.45:249e62. / Energy 132 (2017) 126e135 C.1016/j. consumption) DEROP achieves higher savings than with the tariff [18] Wouters Carmen. de 9 de octubre sobre autoconsumo del Ministerio de Industria.2015. However. James Adrian M. the Minimizing the energy cost for microgrids integrated with renewable energy resources and conventional generation using controlled battery energy stor- algorithm is flexible because it allows you to introduce as many age.97:646e55. 1074e82. de 26 de diciembre. DEROP may be used for other purposes such as mini.85:30e44.2015. http://dx. Marnay Chris. agent genetic algorithm: an online probabilistic view on economic dispatch of In addition. economic and environmental impact of RESs. Ahmadi Abdollah. Optimal deployment of thermal energy storage under diverse In this paper. Alammari Rashid AM. [8] Geidl M.93 12. 4):1785e94. distributed energy system planning e a South Australian case-study.05. Appl Energy 1 February 2016.doi. ISSN 0378-7788.1016/j. Roche R. In addition. and demand-side management systems under uncertainty. This research work has been made possible with the support of [21] Ley 24/ 5.73 41. while D2 has required 32 itera. Through an iterative http://dx. 2):194e202. Tang Jie. vol. ISSN 0960-1481. ISSN Siegel 1): hourly discrimination is not profitable (for example. 2007. the tariff with integer linear programming. Solar generation (kWh) 0 15.75:598e606.02. Stadler Michael. ISSN 0960-1481. DC: IEEE PES Innovative Smart Grid Technologies (ISGT). Gandomkar Majid.doi. Energy Build 1 March 2015. ISSN 0306-2619.doi. Optimal procedure. and so 71. [3] Thomaidis Nikolaos S. 900/2015. May 2016. Correcher A. Srinivasan D. http://dx. This optimization for an aggregator-based resource allocation in the smart grid. IEEE Trans Smart Grid Sept. 2012. http://dx.6(no.051. Andersson G. http://dx.60 54. An integer Regarding the number of iterations. An energy integrated. Demand (kWh) 71. Integrated modeling and optimization of multi-carrier energy sys- installed 068-Ayudas para la realizacio n de proyectos de IþD para grupos 10.1016/j.1016/j. Santos-Alamillos Francisco J. In: Proceedings of the Scenario A2 (base) D2 (BM1) D2 (DEROP) IEEE.doi. Pozo-Va zquez David. Zhang Shijie.1109/TSG. Agelidis Vassilios G. p.2015.07. Grid supply savings (%) e 32. http://dx. a new algorithm for optimal management of DERs economic and climate conditions.043. Fraga Eric S. http://dx. 2014.87 48. efficiencies. [5] Zhu Ziming.2012. it has [17] Morais Hugo. Res February 2016.doi. 2015.2106228. 50%) may be achieved with the application of DEROP. generation forecasts of each response and energy storage system.1016/j. Chin Woon Hau. Savkin Andrey V.1016/j. 2):699e708.039.61 JPROC. Jadid Shahram.2271517.17 44. Lai Judy. Abbaspour A. Shun TZ. http://dx. Energy 1 Finally.doi. [1] Lasseter RH.1016/ ISSN 0305-0548.031. http://dx. ISSN 0378-7788.11 6. tions. [7] Lv Tianguang.cor.apenergy. storage. 99.97 01. Rolda Table 4 References Comparison of results obtained with DEROP and BM1.1109/ discrimination than with the tariff with no time restrictions. [14] Abedini Mohammad.renene. [12] DeForest Nicholas. Ariza HE.2014.2015. Moradi Mohammad H.2015. http://dx. Valuing investments in multi-energy con- version. 2012.134 n-Blay et IEEE Trans Sustain Energy April 6. multi- Renew Energy March 2015. p. ISSN 1364-0321. As a case study. 3): 1244e52.83 [4] Logenthiran T. Optimal electrical and energy supply in energy 8 scenarios of an energy 2014. DEROP calculates optimal energy flows to obtain the management of microgrids including renewable energy sources using GPSO- minimum cost. www.02. In any TSG. Heuristic simulation step sizes (the step used in this study is 15 min).52 9. scheduling 0360-5442.047.2015.90:65e75. 2013. the results depend on the demand profile and the 408e22. Fotuhi-Firuzabad M. de investigacio .12.1016/j.016.1109/ calculation. Mahdi Hosseinian S. different scenarios of an energy hub may [10] Khalid Muhammad. the synergy between RESs and ESSs has been proven.enbuild. Feng Wei.2016.2(no. Maciejewski AA. a priori. Demand side management in smart grid using Heuristic Optimal scheduling of a renewable micro-grid in an isolated load area using mixed- been observed that in situations where. http://dx.10.apenergy. [23] Spanish System Operator Website (Red Ele ctrica de Espan ~ a).01.doi. optimising contracted power.doi.87 Comput Oper.163: Logically. since it allows the use of non-linear 10.2013. Interactive energy management of networked case. http://dx. 2015. Suryanarayanan S. June 2015.3(no. http://dx. Washington.1016/ RESs as needed and use their cost and efficiency curves in the the described methodology allows to evaluate the Short-term resource scheduling of a renewable energy based micro grid. low night-time 151e6. [13] Brahman Faeze. Renew Energy January 2010. ISSN 0360-5442. simple and easy to implement. it is concluded that significant cost savings (over Build 15 November 2015. Faria Pedro.1016/j. With the methodol. [19] Yang Yun.1016/j. IEEE Trans Sustain Energy April different types of In addition. it is flexible. Multi- functions for costs and efficiencies during the calculation. del Sector Ele ctrico. Load management in a residential energy hub with renewable distributed energy resources. K adar P e ter.2011. Rezvani Alireza.doi. integrating demand DEROP uses prices. Fan Zhong.doi. Renew Sustain Saved emissions (%) e 30.doi. http://dx. Energy the simulations.40 [2] Abdmouleh Zeineb. ISSN 0960-1481. [9] Izadbakhsh Maziar.renene.1109/TSTE. Gastli Adel.doi. indicates that DEROP optimises faster with the tariff with hourly IEEE Trans Smart Grid July 2015.2399359. As a result of [16] Rastegar Mohammad.doi. ISSN 0960-1481.doi.2012.030.renene.2195686. Alfonso D.2011.doi. the algorithm is fast. manage resources in an energy hub and achieve significant savings. GM algorithm. it may be concluded that DEROP can no. The algorithm is fast.90(Part 2):1901e15. since the simulations shown in this study microgrids-based active distribution system considering large-scale integra- take about 1 s in Matlab (in VBA they take a few seconds). Ahcin P. Renew Energy November 2016.2016. An MILP (mixed integer linear pro- maintenance tasks. Escriva-Escriv a G. Ca rcel 90:430e9. Wind generation (kWh) 0 11.rser. 1e5.107:234e42.

2012. ISSN 0960-1481. http://dx.36(Issue 5) Hamed Zeinoddini Meymand. Rolda 135 ree.60:38e46.2010. n-Blay et al. http:// 11. Rold an. / Energy 132 (2017) 126e135 C.027.enbuild. [24] Roldan-Blay Carlos. Upgrade of an artificial neural network of distribution network with regard to fuel cell power temperature curve model.009. dx. ISSN 0378-7788.1016/j.1016/j. Escriva -Escriva   Guillermo. Energy Build May Rodríguez-García Javier.12. A practical multi-objective PSO algorithm for optimal operation management Porta Carlos. Hasan Doagou Mojarrad. . Renew Energy prediction method for electrical consumption forecasting using an hourly May 2011. [25] Niknam Taher. Alvarez-Bel Carlos.