From the perspective of a community energy operator, a two-stage optimal scheduling model of a community integrated energy system is proposed by integrating information on controllable loads. The day-ahead scheduling analyzes whether various controllable loads participate in the optimization and investigates the impact of their responses on the operating economy of the community integrated energy system (IES) before and after; the intra-day scheduling proposes a two-stage rolling optimization model based on the day-ahead scheduling scheme, taking into account the fluctuation of wind turbine output and load within a short period of time and according to the different response rates of heat and cooling power, and solves the adjusted output of each controllable device. The simulation results show that the optimal scheduling of controllable loads effectively reduces the comprehensive operating costs of community IES; the two-stage optimal scheduling model can meet the energy demand of customers while effectively and timely suppressing the random fluctuations on both sides of the source and load during the intra-day stage, realizing the economic and smooth operation of IES.

Community integrated energy system combines controllable new energy units, gas units, energy storage devices, and cooling, heat, and electrical loads to form a multi-energy complementary energy supply and demand system, which can meet various load demands while improving energy utilization. However, the complex multi-energy structure in integrated energy system (IES) and the response rates of different energy sources have made the operation and scheduling of IES more difficult [

The optimal scheduling of community IES is one of the research priorities in the construction of IES, and its main idea is the coordinated management of loads and power sources in the system based on the load and energy supply output forecasts. Therefore, the references [

In addition, since the prediction accuracy of wind turbines as well as load increases with decreasing time scale [

Based on the above analysis, a two-stage optimal scheduling model for community IES is established. The main contributions of this paper are as follows:

(1) A complete controllable load model is established at the day-ahead stage, taking the comprehensive operating cost of economy and environment as the optimization objective, focusing on the impact of the shifting, curtailment, and transferring of demand-side loads on the system scheduling, which improves the economy of the system operation.

(2) The short-term forecast of load and the uncertainty of wind turbine output are considered in the intra-day stage, and a two-stage rolling optimal dispatch model considering the differentiation of the response rates of electricity, heat and cooling is established, which effectively suppresses the output fluctuation of the system turbines and ensures the smooth operation of the system.

The rest of this article is organized as follows. The structure and controllable load model of IES are presented in

The community IES energy supply and demand network are formed by referring to the energy hub model [

Considering that the fuel cell (FC) is mainly responsible for electric energy dispatch in intra-day dispatch, this paper does not consider its heat energy utilization, and the power output satisfies the following constraint:

where

The gas consumption of FC can be expressed as:

where

The remaining equipment models include those for MT, GB, and LBR, which are described along with operational limitations in reference [

The loads in community IES can be classified into the base load, shiftable load, transferable load, and curtailable load according to the energy-using characteristics.

The scheduling cost

where _{1}, _{2}] denotes the energy usage time period of the load before the shift, [_{sh−}, _{sh+}] denotes the interval of the shift time period, and

The scheduling cost

where _{3}, _{4}] denotes the energy usage time period of the load before the transferring, [_{tr−}, _{tr+}] denotes the interval of the transferring time period, and

The power for each time period after the transferring shall satisfy the following constraint:

where

To avoid frequent start-up and shutdown of power-using equipment, the minimum duration of operation for transferable loads is constrained:

where

The scheduling cost

Considering that frequent reductions affect users’ comfort, the upper limit of reductions and the length of reductions in a dispatch cycle should be constrained:

1) Maximum number of reduction limit:

where

2) Maximum and minimum reduction time:

where

According to the different time scales, the community IES scheduling is divided into two scheduling stages: day-ahead and intra-day stage, as shown in

In the day-ahead stage, the day-ahead plan values for each unit operation and controllable load adjustment are obtained based on the day-ahead short-term forecasted electric, heat, and cooling loads and wind turbine output, with the objective of minimizing the comprehensive operating cost of the community IES for one dispatching cycle and meeting the operating constraints.

In the intra-day stage, the upper-level heat and cooling energy dispatching model and the lower-level electric energy dispatching model are divided into the upper-level heat and cooling energy dispatching model and the lower-level electric energy dispatching model, based on the day-ahead dispatching plan, and the wind turbine output and the load power ultra-short-term forecast information are updated on a rolling basis, and the day-ahead plan value is revised in an incremental balancing manner to adjust the unit output. The above two stages are mixed integer linear programming problems, and Cplex solver is used to solve the output of each unit.

In the day-ahead scheduling model, the scenario generation and reduction method are used to deal with the uncertainty of wind turbines, and historical data can be used to determine the short-term forecasted power expectation μ_{w} of wind turbines, and the power error of wind turbines is assumed to satisfy the normal distribution of N (0, σ^{2}), then the day-ahead prediction of wind turbines is the sum of the expectation and the prediction error. A large number of original scenario sets of wind turbine power obeying probability distribution are generated by Latin Hypercube Sampling (LHS) [_{m}, (m = 1,2, …,10) are obtained. Finally, the probabilities of the above scenarios and the corresponding power are multiplied and summed to obtain the power curve with the uncertainty of wind turbine output.

In the day-ahead scheduling plan, both economic and environmental factors are considered to affect the operating cost of IES, and the environmental indicators are converted into economic indicators, and the minimum integrated operating cost of IES in one scheduling cycle is taken as the objective function of day-ahead scheduling, and the integrated operating cost F^{DH} is:

where _{2} emissions.

where ^{3}.

where

where

where _{2} emission, ¥; _{2} emission factors per unit of power output of conventional coal-fired power plants, the CO_{2} emission factors per unit of power output of MT, the CO_{2} emission factors per unit of heat power output of GB, and the CO_{2} emission factors per unit of power output of FC, ¥/kg, respectively; _{2} emission factors per unit of power output of MT, heat power output of GB, and power output of FC, respectively, at time _{2} emission data and penalty cost are shown in reference [

The stable operation of IES in the community needs to meet certain constraints, which can be divided into equality constraints and inequality constraints.

1) The electric balance equation of each period of the system is:

where

2) Heat load supply and demand balance:

where

where

3) Cooling load supply and demand balance:

where

In actual operation, MT units are subject to the following climbing constraints:

where

The day-ahead plan determines the operation state of the shiftable load, transferable load, and curtailable load in the intra-day plan, and no optimization is necessary. Given the frequent power fluctuations in the intra-day operation plan, as well as the energy storage equipment’s service life and the fact that it has reached the maximum number of energy storage and discharge in the day-ahead plan, participation of energy storage equipment in the intra-day dispatch plan is not considered for the time being.

The upper-level scheduling model is used to smooth out cooling and heat power fluctuations with a slower response rate, with a scheduling time domain of 2 h and a control time domain of 1 h. The lower-level scheduling model is used to smooth out electrical power fluctuations with a faster response rate, with a scheduling time domain of 1 h and a control time domain of 15 min.

The intra-day scheduling strategy is shown in

1) The upper-level rolling optimization objective function

In the upper-level scheduling strategy, the output capacity of each equipment is adjusted according to the fluctuation of hot and cooling loads, and its objective function is:

where

2) Upper-level optimization constraints

(1) Heat energy balance constraint

where

(2) Cooling energy balance constraint

where

(3) Unit constraints

1) Lower-level rolling optimization objective function

In the lower-level dispatching strategy, the output of the power supply equipment is adjusted according to the fluctuation of the electric load and in conjunction with the adjusted output plan of the upper-level combined heat and cooling supply equipment, with the minimum system dispatching cost as the objective function:

where

2) Constraints on lower-level optimization

(1) Constraint on electrical power balance:

where

(2) The power fluctuations in the intra-day stage should be kept to the contact line with the grid in order to preserve the stability of the external grid:

(3) FC constraint

In this paper, a typical summer day in a community is selected for the calculation. The energy structure of the community is shown in _{w}, the natural gas price is 2.5 ¥/m^{3}, the rated capacity of the energy storage equipment is 300 kWh. The parameters for controllable load scheduling are shown in

In order to verify the advantages of two-stage rolling optimal scheduling considering controllable loads of heat, cooling and electricity, the following cases are set up for comparison:

Case 1: Comprehensive demand response is not considered, and only day-ahead optimal scheduling is performed.

Case 2: Comprehensive demand response is considered and only day-ahead optimization is performed.

Case 3: Comprehensive demand response is considered, and two-stage rolling optimization scheduling is carried out.

All the above 3 cases have the same conditions except for different optimization scheduling methods. The computer processor used in this paper is Intel(R) Core (TM) i5-8300H CPU and the simulation software used is MATLAB version 2018b. The simulation time for the case 3 is 63.08 s.

1) Comprehensive cost analysis:

The combined operating costs of the system under case 1 and case 2 are shown in

Case | Economic costs/¥ | Environmental costs/¥ | Comprehensive costs/¥ |
---|---|---|---|

Case 1 | 6591.5 | 810.1 | 7401.6 |

Case 2 | 6027.2 | 681.9 | 6719.1 |

As can be seen from _{2} emissions and hence lower fuel costs. In particular, the power output of FC has been maintained at the maximum power allowed during the day-ahead period because of the reserve capacity and the low cost of power generation.

2) Controllable load analysis before and after demand response:

From the load curves before and after demand response in

The distribution of the controllable load prior to and during the demand response is shown in

Parameters | Case 1 | Case 2 | |
---|---|---|---|

Shiftable electrical load | Time/t | 17:00–22:00 | 05:00–10:00 |

Power/kW | 54/61/97/82/63 | 54/61/97/82/63 | |

Transferable electrical load | Time/t | 12:00–16:00 | 04:00–09:00 |

Power/kW | 60/69/56/52 | 50/50/50/50/37 | |

Curtailable electrical load | Time/t | 10:00–15:00 | 10:00–15:00 |

18:00–21:00 | 18:00–21:00 | ||

Power/kW | 71/74/75/75/71 | 21/22/22/22/21 | |

61/55/52 | 18/16/16 | ||

Shiftable heat load | Time/t | 16:00–21:00 | 04:00–09:00 |

Power/kW | 23/31/42/39/32 | 23/31/42/39/32 | |

Curtailable heat load | Time/t | 10:00–15:00 | 10:00–15:00 |

18:00–21:00 | 18:00–21:00 | ||

Power/kW | 14/15/17/17/17 | 6/6/7/7/7 | |

18/17/14 | 7/7/6 | ||

Shiftable cooling load | Time/t | 12:00–19:00 | 00:00–07:00 |

Power/kW | 168/128/144/ | 168/128/144/ | |

112/96/176/168 | 112/96/176/168 | ||

Curtailable cooling load | Time/t | 10:00–15:00 | 10:00–15:00 |

18:00–21:00 | 18:00–21:00 | ||

Power/kW | 42/43/43/43/42 | 17/17/17/17/17 | |

55/54/35 | 22/21/14 |

The intra-day optimal dispatch is case 3, and the optimization results are shown in

From

From

In this paper, a two-stage optimal dispatch model of community IES with controllable loads is proposed for community IES, considering the differences in response rates of various controllable loads and hot and cooling electricity on different time scales. The inferences that can be made are as follows:

1) The participation of controllable loads of cooling, heat, and electricity in optimal scheduling can effectively reduce the comprehensive operating costs of community IES.

2) The intra-day two-stage rolling optimization model is capable of dispatching heat, cooling energy, and electric energy at different time scales, respectively, enabling the system to timely and effectively smooth out the prediction errors on both sides of the source and load, and realize the stable operation of community IES and connected large grids.

In this paper, we have considered the impact of load demand response on the optimal scheduling of the system, but the implementation of demand side management will change the way of using energy, resulting in customer dissatisfaction. Therefore, not only economic issues but also customers’ satisfaction needs to be considered in future work studies.

The wind turbine

The micro gas turbines

The gas boilers

The electric boiler

The electric refrigerator

The absorption refrigerator

The waste heat boiler

The heat storage tank

The cooling storage tank

The energy storage

_{FC}

The output power of FC

The minimum output power of FC

The maximum output power of FC

_{FC}

The backup factor

_{NG}

Natural gas’s low-level heat value

_{FC}

FC’s power generation efficiency

None.

This work was supported in part by the National Natural Science Foundation of China (51977127), Shanghai Municipal Science and Technology Commission (19020500800), and “Shuguang Program” (20SG52) Shanghai Education Development Foundation and Shanghai Municipal Education Commission.

The authors confirm contribution to the paper as follows: study conception ancdesign: Ming Li; data collection: Rifucairen Fu; analysis and interpretation of results: Tuerhong Yaxiaer; draft manuscript preparation: Yunping Zheng. All authors reviewed the results and approved the final version of the manuscript.

Data supporting this study are included within the article.

The authors declare that they have no conflicts of interest to report regarding the present study.

Type | |||||
---|---|---|---|---|---|

Shiftable electrical load | 17:00–22:00 | 05:00–23:00 | 0.15 | ||

Shiftable heat load | 16:00–21:00 | 04:00–24:00 | 0.10 | ||

Shiftable cooling load | 12:00–19:00 | 00:00–21:00 | 0.10 | ||

Type | |||||

Transferable electrical load | 12:00–16:00 | 05:00–23:00 | 30-50 | 2 | 0.15 |

Type | |||||

Curtailable electrical load | 2 | 4 | 8 | 0.20 | |

Curtailable heat load | 2 | 4 | 8 | 0.15 | |

Curtailable cooling load | 2 | 4 | 8 | 0.15 |

Equipment | Power maximum/kW | Power minimum/kW |
---|---|---|

Electrical/heat power of MT | 200/454 | 0/0 |

Heat power of GB | 1500 | 0 |

Cooling power of LBR | 1500 | 0 |

Electrical power of EB | 100 | 0 |

Electrical power of EC | 100 | 0 |

Electrical power of FC | 100 | −60 |

Grid | 450 | −450 |

Energy storage equipment | 70 | −70 |

WT | Predicted value | 0 |