^{1}

^{2}

^{1}

^{1}

^{1}

^{1}

^{1}

Northern China has rich wind power and photovoltaic renewable resources. Combined Heat and Power (CHP) Units to meet the load demand and limit its peaking capacity in winter, to a certain extent, it results in structural problems of wind-solar power and thermoelectric. To solve these problems, this paper proposes a plurality of units together to ensure supply of heat load on the premise, by building a thermoelectric power peaking considering thermal load unit group dynamic scheduling model, to achieve the potential of different thermoelectric properties peaking units of the excavation. Simulation examples show, if the unit group exists obvious relationship thermoelectric individual differences, the thermal load dynamic scheduling can be more significantly improved overall performance peaking unit group, effectively increase clean energy consumptive.

According to the planning target is expected to “13th Five-Year” at the end, hydropower, wind power, photovoltaic power generation of state Grid installed capacity will be increased to 270 million kilowatts, 220 million kilowatts and 110 million kilowatts, a total of 600 million kilowatts, wind and other new energy consumptive issues will be more prominent. The scenery of the consumptive problem, to a certain extent trapped in the uncoordinated system integration mode of intermittent renewable energy and traditional energy. With China’s “Three North” area as an example, the winter scenery resources is very rich, but the power peaking resources can participate in the scenery consumption，which is further compressed because the CHP power determined by heat load. On the other hand, compared with the thermoelectric power generation, CHP has obvious economic advantages; CHP units will be further increased. In order to further improve the energy efficiency of the proportion a large number of existing research show that two of coexistence will give the power peaking difficult [

The paper [

Due to “power determined by heat” mode, CHP units must ensure the supply of heat load, which limits the peak load capacity. For the operation of the power grid, the power range of the CHP unit considering the heat load demand is the boundary condition of the economic dispatch. In the actual production of electric power for some reason, a thermoelectric unit limit on the difficult to adjust, on the other hand, the heat load of a region in different units between the distribution did not take into account the overall peak load regulation, so the actual peak load capacity of CHP units should have greater potential. By paper [

Thermo-electric relationship curve is the key for analyzing units’ heat load supply and electric power regulation, thus it’s proper to start with the relationship for the study of units’ peak-load regulating performance. It’s illustrated in _{min} and upper bound P_{max} respectively. Besides, the regulation interval P_{ran} = P_{max} − P_{min}. In all, the greater the capacity of suction Q, the smaller the regulation interval P_{ran}.

As the bound lines are a series of irregular curves, for convenience of model building later, the bound curves are piecewise linearized, as in

where a_{1}, a_{2}, b_{1}, b_{2} are known constants.

In air extraction interval 1, when the capacity of suction Q is given, expression of limits regulation interval P_{1,ran} is as Equation (2).

When traditional thermal units need to be transformed to CHP units, as the traditional thermal power units’ performance varies, thermoelectric relationship curves have also their differences in some constant. What’s more, by equipped with different thermal storage devices, the curves can be adjusted later [

In the traditional economic dispatch of thermal power units, coal consumption (operational cost) of units is normally expressed as a quadratic curve, while the operational cost of CHP units could be expressed as a quadratic curve of electric power and heat power of the units as in Equation (3).

where c is a constant term, a_{p} and b_{q} are linear and quadratic coefficient of the heat power, d_{qp} is the thermoelectric coupling coefficient. For one unit, when the thermoelectric coupling optimization objective only contains the supply of heat or electricity, its operational state is the most economic under rated power through the analysis of the quadratic curves. However, if the supply cost of heat and electricity are considered simultaneously, the determination of optimal value is relatively complicated. The linearization or proper decoupling method is the key of solving the problem.

CHP optimal dispatch problem is normally the realization of the minimum of whole operation cost while satisfying various kinds of related constraints of the units operation, heat grid and electric grid.

where, in Equation (4), f_{i} is the operational cost of unit i, constraints in Equation (5) mainly include electricity balance, heat balance and various kinds of inequality constraints. The model has been illustrated in literature [

The main purpose of this paper is to improve the power grid peak-load regulation performance and promote wind/photovoltaic clean energy accommodation by distributing the proper heat load. The peak-load regulation performance could be considered in two aspects, one is to maximize the electric power regulation interval of the whole CHP units, the other is to minimize the units’ downward output ability that is to release more generation space in load valley periods. In the work of power grid dispatching schedule, these two aspects may all affects its peak-load regulation performance, while in periods of electric power load valley, the latter has a more direct influence of wind/photovoltaic accommodation. As illustrated above in 2.1, when there are several CHP units whose heat and electric bound curves seem inconformity, there is a heat load distribution scheme to maximize units’ integral electric power regulation interval P_{sum,ran} or minimize units’ integral electric power output while satisfying the heat load requirement. It’s normally conducted as dynamic economic dispatch according to the load requirements in different time intervals, what’s more, heat load is also affected by various kinds of weather situation in the outside world in the dispatching cycle. Thus this paper builds the dynamic load distribution model of the heat load requirement in the dispatching cycle. The optimization objectives can be chosen as Equations (6) or (7) according to requirements.

where P_{i,t,}_{ran}, is the limit regulation ability of unit i when the heat load Q is given, N is the total number of thermal power generation units, and T is the total time intervals.

where P_{i,t,}_{min} is the minimum electric power of unit i at time t when the unit’s heat load Q is given.

As the load regulation pressure normally exists in electric load valley periods, the model is simplified for pertinence and the optimal objective is focused on specific time intervals.

The equations are an exception to the prescribed specifications of this template. You will need to determine whether or not your equation should be typed using either the Times New Roman or the Symbol font (please no other font). Equations should be edited by Mathtype, not in text or graphic versions. You are suggested to use Mathtype 6.0 (or above version).

1) Heat load requirement constraints

CHP units should first meet the heat load requirement.

where Q_{load,t} is the total amount of heat load at time t and Q_{i,t} is the heat power of unit i at time t.

2) Electric load requirement constraints

In economic dispatch with heat load, electric load requirement of power grid is no need. However, different electric load requirements have directly influenced the supply of heat load. Thus electric load constraints would be considered in form of load requirement intervals.

where P_{load,t,min} and P_{load,t,max} are the lower and upper bounds of electric load requirements respectively. P_{i,}_{min,t} and P_{i,}_{max,t} are the lower and upper bounds of electric power respectively when the capacity of suction is given. The meaning of the equation is that all the units’ electric power sum should meet the limits bound constraints of load.

3) Thermoelectric coupling constraints

As the thermoelectric relationship curve in

4) Heat power operation constraints

As constrained by the unit’s operational condition, units’ heat power needs to meet heat power operation constraints.

where Q_{i}_{,max} and Q_{i}_{,min} are the maximum and minimum capacity of suction, Q_{i}_{,up} is the added capacity limit of suction in adjacent time intervals of unit i and Q_{i}_{,down} is the reduced capacity limit of suction in adjacent time intervals of unit i which can be perceived as heat power climbing constraints.

5) Electric power operational constraints

The units should also meet the electric power operational constraints.

In Equation (11), P_{i}_{,min} and P_{i}_{,max} are the lower and upper output limit of unit I respectively, P_{i,}_{up} and P_{i}_{,down} are the upper and lower limit of electric power climbing rate of unit i respectively.

In practice, a CHP unit’s thermoelectric relationship curve is not described with two linear curves in one pump interval but maybe a series of irregular curves as in

To solve the problems above, the model should be improved. The method adopted in this paper is to transfer the problem into MIP with 0 - 1 binary variables by piece-wise linearization. The strategy built here is as follows.

1) First, to piece-wisely linearize the thermoelectric relationship curve according to its calculation accuracy acquirement and the unit is split into various virtual units according to total sectional number. The relationship curve in figure 1 is split into 4 virtual units and their thermoelectric operational performances are corresponding to pump interval 1 - 4.

2) To build a 0 - 1 binary variable S_{f,t} for the split virtual units. As for capacity of suction in each time interval can only drop in one pump interval, thus the split units should meet the following virtual on-off constraints.

where S_{f,t} is the on-off state of split virtual unit f and I is the former physical unit I.

3) Through adding 0 - 1 binary variable of virtual units, the objective function and constraints in 2.1 and 2.2 are improved, i.e. Equation (7) is changed into Equation (13).

In Equation (13), N^{’} is the total units number including virtual units, for non-virtual units, S_{i,t} = 1.

For constraints, i.e. Equation (10) is changed into Equation (14).

Other expression can also been built in the model according to similar method. As for model solving, it’s a classical MIP which can be solved by mature commercial programming solver CPLEX.

In order to verify the correctness of the proposed model and the degree of improvement to the peak load capacity of power network, this paper constructs a 10 machine system. Among them, the unit 1 and unit 2 are made up of two pumping intervals, the unit 3 - 8 are composed of a pumping interval, the thermoelectric power of the unit 9 is a simple linear relationship, the unit 10 is composed of the of 4 pumping intervals (due to space limitations, the specific parameters of the unit is no longer listed). According to the modeling method proposed in this paper, it is divided into 15 units, including the virtual unit, and the model is solved on the Xeon E7420 2.13 GHz server, and calculation time is less than 1 second. The thermoelectric parameter of the unit is shown in

In order to validate the effect of peak regulation performance, an example to the overall power output is minimized as the optimization objective, the results of the calculation of the first time section are shown in

Now the lower power limit of total ten units is 1337 MW.

If the heat load can be allocated according to the heating capacity of the unit, then solve the electricity power adjustment interval of each unit according to the distribution of heat load, the results are shown in

Unit name | Lower power limit | Upper power limit | Pumping quantity |
---|---|---|---|

Unit 1 | 75 | 142 | 60 |

Unit 2 | 80 | 136 | 100 |

Unit 3 | 112 | 126 | 190 |

Unit 4 | 150 | 300 | 100 |

Unit 5 | 150 | 250 | 300 |

Unit 6 | 180 | 280 | 300 |

Unit 7 | 180 | 250 | 300 |

Unit 8 | 150 | 300 | 100 |

Unit 9 | 100 | 100 | 200 |

Unit 10 | 160 | 280 | 150 |

Unit name | Lower power limit | Upper power limit | Pumping quantity |
---|---|---|---|

Unit 1 | 84 | 130 | 128 |

Unit 2 | 87 | 130 | 128 |

Unit 3 | 100 | 134 | 140 |

Unit 4 | 178 | 289 | 210 |

Unit 5 | 173 | 273 | 210 |

Unit 6 | 166 | 289 | 210 |

Unit 7 | 166 | 272 | 210 |

Unit 8 | 177 | 272 | 210 |

Unit 9 | 141 | 141 | 142 |

Unit 10 | 187 | 274 | 212 |

Now the lower power limit of total ten units is 1459 MW. Compare the two methods, the optimization model proposed in this paper can make the electric power down 122 MW, it can improve the peak depth of 8.36%. In the period of a lot of wind power, if there is a large number of abandoned wind, by the way you can enhance the space wind power 122 MW. When the size of CHP unit is large this mode can bring considerable increase peak regulation capacity.

In fact, the thermal load distribution is the most economical choice according to model which formula (4 - 5) show, this paper will make economic comparison according to two ways, which is the load mode and the heat capacity mode. The coefficients in the formula (3) are based on the data in paper [

In accordance with the heat load requirements shown in

It can be seen from the analysis that the principle of load distribution depends on the slope of the thermoelectric relationship a, when the a is greater that a small amount of electricity power limit can be realized with large heat load fluctuation change, when take the strongest peak load capacity as the optimization objective, the smaller the slope a is, the more suitable for the base heat load.

If you want to achieve a more detailed allocation of heat load, in accordance with the model proposed in this paper, only need to increase the divided number of CHP unit pumping interval. When the thermal power relationship curves of all the units are divided into enough piecewise linear functions, the optimization results can be obtained with high accuracy. Of course, this measure will also increase the amount of calculation model.

Because the demand of electric load will directly affect the heat load of the unit,

Unit number | Electric load demand | peak regulation mode heat load | heat capacity mode heat load | peak regulation mode cost | heat capacity mode cost load |
---|---|---|---|---|---|

1 | 84 | 60 | 128 | 4628 | 5474 |

2 | 87 | 100 | 128 | 5162 | 5547 |

3 | 100 | 190 | 140 | 6915 | 6055 |

4 | 178 | 100 | 210 | 7596 | 9688 |

5 | 173 | 300 | 210 | 11,760 | 9522 |

6 | 166 | 300 | 210 | 11,511 | 9293 |

7 | 166 | 300 | 210 | 11,511 | 9293 |

8 | 177 | 100 | 210 | 7566 | 9655 |

9 | 141 | 200 | 142 | 8295 | 7202 |

10 | 187 | 150 | 212 | 8742 | 10,036 |

it is necessary to analyze the influence of the change of electric load on the optimal allocation of heat load. For the electric load demand curve in

For the CHP unit limit the peak regulation capability of the unit to ensure the supply of heat load, which causing the problem of wind and light insufficient consumptive, this paper put forward to supply multiple units together to ensure the heat load, give full play to the potential of different of peak thermoelectric properties set. In this paper, CHP power group thermal load dynamic scheduling model considering peak load regulation, and puts forward solutions to the irregular power curve. The simulation example shows that if there are obvious differences between the relationship between the thermal power unit inside the individual heat load dynamic scheduling method of multi-unit can ensure the load requirements effectively improving the peak load capacity of CHP unit. Of course, due to the purchase of thermal power peaking at the expense of benefit is a need for further study of economic problems.

Thanks for support from the science and technology program of State Grid Corporation of China (SGCC).

Liu, C., Men, D.Y., Xu, D., Ding, Q., Huang, G.D., Dai, S. and Zhou, J.Y. (2017) Optimization of Minimum Power Output for Combined Heat and Power Units Considering Peak Load Regulation Ability. Energy and Power Engineering, 9, 452-463. https://doi.org/10.4236/epe.2017.94B051