Optimal scheduling for unit commitment considering wind power consumption and natural gas peak-shaving

Abstract

The mismatch between the rate of the new energy development and the system’s peak-shaving capacity has resulted in severe wind abandonment. Based on the grid connection of wind power and natural gas peak-shaving, a model of unit commitment considering wind power consumption and natural gas peak-shaving and taking into account a combination of system economics and wind power consumption capability is designed. Natural gas peak-shaving is added to improve the system’s peak-shaving capacity, and a wind abandonment penalty constraint is added to reduce the amount of wind abandoned by the system, and the model is solved by an improved genetic algorithm. Finally, to verify how wind power and natural gas peak-shaving impact unit commitment, the IEEE-30 node system is used. The results show that natural gas peak-shaving reduces system operating costs and improves the safety of the system. This model ensures the economics of system operation while positively promoting wind power consumption effectively and reasonably.

Keywords

Unit commitment optimal scheduling wind power consumption natural gas peak-shaving improved intelligent algorithm

1. Introduction

With the objective of reaching peak carbon and carbon neutrality, the total amount of installed wind power in China has shown an extraordinary and leapfrog development [1]. However, due to wind power’s stochastic, indirect, and fluctuating characteristics, it leads to the optimal scheduling for unit commitment becoming more complex and wind abandonment becoming more severe when wind power is connected on a large scale [2, 3]. Insufficient peak-shaving capacity is a major cause of wind abandonment [4]. As a result of wind power’s uncertain nature and anti-peak regulations [5], the mismatch between load and output is prominent. But the power structure of China, which is dominated by thermal power generation, has not been broken for the time being, and there is a lack of flexible and fast power peak-shaving response in power systems containing wind power [6]. How to find a method for increasing wind power consumption, optimise the unit commitment scheduling scheme and ensure the economy and safety at the same time has become an important issue that needs to be resolved urgently.

Traditional peak-shaving methods include thermal and hydroelectric power [7, 8], but the use of thermal or hydroelectric power for peak-shaving has not been effective due to economic and environmental factors [9]. Currently, more research is being done on the use of pumped storage power plants for peak shifting [9, 10]. However, the establishment of pumped storage power stations is highly dependent on geographical location, and the plains and inland arid areas are not suitable for pumped storage power stations. As gas-fired power generation has the characteristics of fast start-stop speed and low pollution emissions, natural gas peak-shaving is a safe, convenient, economic and environmentally friendly peak-shaving method that is not affected by geographical location [11]. Combined wind power and natural gas peak-shaving generation can relieve thermal power unit peak-shaving pressure and reduce the effects of wind power grid integration effectively [12]. Natural gas-fired power generation has been studied for its potential to be integrated into the energy system as a means of peak-shaving emission. The literature13 has used natural gas hydrate as a medium to validate the advantages of natural gas-fired power generation for both long and short-term peaking needs. The literature 14 investigated the current situation of gas-fired power stations participating in grid peak-shaving in China and studied the economy and low-carbon environmental protection of gas-fired power stations in grid peaking. The literature 15 presents a unit model that considers natural gas delivery constraints and standby dispatch, and demonstrates the effectiveness of natural gas-fired power generation in reducing the cost of thermal power units. The literature 16 proposes a new decomposition technique based on Benders feasible cuts, which reduces the overall operating cost of power systems containing gas-fired. The literature 17 investigated the reduction of power system operating costs by increasing the power output of gas-fired systems during peak demand hours in lieu of more costly units. Although all of the above literature affirms the positive impacts of natural gas on the power system’s peak-shaving capability, and also researches the joint operation of natural gas peak-shaving and wind power generation, most of the objectives are single consideration of the overall economy or environmental protection factors, and do not take into account the promotion of wind energy consumption as being a factor.

In view of this, this paper establishes a unit commitment model considering the promotion of wind power consumption and natural gas peak-shaving by adding a wind abandonment penalty factor to the traditional model and adding wind power output constraints and natural gas peak-shaving constraints to the constraints, thus reducing the cost of the whole unit. The model is solved using an improved genetic algorithm. The role of natural gas peak-shaving in improving the stability and economics of the power system and the effects achieved by promoting wind energy consumption are analysed.

2. Model of the unit commitment

2.1 Target function

The model does not pursue full wind power consumption, but allows for a certain amount of wind abandonment, weighing a combination of factors. An abandonment penalty cost is introduced into the objective function in order to promote wind power consumption. This system’s total cost is comprised of two components, namely the thermal power plant’s operating cost and the penalty for wind abandonment:

$\displaystyle\min F=F_{r}+F_{p}$ (1)

where $F_{r}$ is the cost of operating a conventional thermal power plant, including the burning coal’s cost and the start-up and shutdown cost of the unit:

$\displaystyle F_{r}=\sum\limits_{t=1}^{T}{\sum\limits_{i=1}^{N}{[f(p_{\textit{% it}})+S_{i}u_{\textit{it}}(1-u_{i(t-1)})]}}$ (2)

where $f(p_{\textit{it}})$ is the cost of coal combustion; $p_{\textit{it}}$ is the output of unit $i$ at time $t$ ; $S_{i}$ is the start-up cost of unit $i$ ; and $u_{\textit{it}}$ is the current operating status of unit $i$ at time $t$ , using 0 for off and 1 for on.

The cost of coal combustion can be expressed as a quadratic function:

$\displaystyle f(p_{\textit{it}})=a_{i}p_{\textit{it}}^{2}+b_{i}p_{\textit{it}}% +c_{i}$ (3)

where $F_{p}$ is the penalty cost for wind abandonment. The wind abandonment is the difference between the forecasted wind output and the actual wind output at each given moment in time.

$\displaystyle F_{p}=\partial_{w}\sum\limits_{t\in T}{(P_{t}^{\textit{pre}}}-P_% {t}^{w})$ (4)

where $\partial_{w}$ is the set penalty factor for wind power, $P_{t}^{\textit{pre}}$ is the forecast wind power output forecast at time $t$ , $P_{t}^{w}$ is the actual wind power output at time $t$ .

2.2 Binding conditions

2.2.1 Constraints of thermal power units

In terms of thermal power units output constraints, the upper and lower limits are as follows [18]:

$\displaystyle P_{i\min}u_{\textit{it}}\leqslant P_{\textit{it}}\leqslant P_{i% \max}u_{\textit{it}}$ (5)

where $P_{i\max}u_{\textit{it}}$ and $P_{i\min}u_{\textit{it}}$ are the maximum and minimum technical output of unit $i$ .

Thermal power units are constrained by the following climbing constraints:

$\displaystyle-D_{{R}i}\leqslant p_{\textit{it}}-p_{i(t-1)}\leqslant U_{{R}i}% \quad u_{\textit{it}}u_{i(t-1)}=1$ (6)

where $D_{{R}i}$ and $U_{{R}i}$ are the downstream and upstream climbing thresholds of unit $i$ .

Time constraints for starting up and shutting down thermal power units:

$\displaystyle T_{\textit{it}}^{\textit{on}}\geqslant T_{i\min}^{\textit{on}},T% _{\textit{it}}^{\textit{off}}\geqslant T_{i\min}^{\textit{off}}$ (7)

where $T_{\textit{it}}^{\textit{on}}$ is the continuous start time of unit $i$ from the last change state to the moment $t$ ; $T_{\textit{it}}^{\textit{off}}$ is the continuous stop time of the unit $i$ from the last change state to the moment; $T_{i\min}^{\textit{on}}$ is the minimum start time of the unit $i$ ; $T_{i\min}^{\textit{off}}$ is the minimum stop time of the unit $i$ .

2.2.2 Operating constraints of the system

Constraints on power balance:

$\displaystyle\sum\limits_{i=1}^{N}{P_{i,t}+P_{t}^{W}}+P_{{G}t}=L_{t}+L_{% \textit{OSS}t}$ (8)

where $P_{t}^{W}$ is the power output of wind energy, $P_{{G}t}$ is the power output of natural gas peak-shaving; $L_{t}$ is the system load size and $L_{\textit{OSS}t}$ is the system’s network loss.

Nodal power balance constraint:

$\displaystyle L_{t}=\sum\limits_{k=1}^{N_{B}}{P_{k,t}}$ (9)

where $N_{B}$ is the nodes total number and $P_{k,t}$ is the active power of node $k$ at moment $t$ .

Nodal voltage amplitude constraint [18]:

$\displaystyle V_{k}^{\min}\leqslant V_{k,t}\leqslant V_{k}^{\max}$ (10)

$V_{k,t}$ is the magnitude of the voltage at node $k$ at moment $t$ .

Trend Bound:

$\displaystyle P_{k,t}=V_{k,t}\sum\limits_{j\in k}{V_{j,t}(G_{kj}\cos\theta_{kj% ,t}+B_{kj}\sin\theta_{kj,t})}$ (11) $\displaystyle Q_{k,t}=V_{k,t}\sum\limits_{j\in k}{V_{j,t}(G_{kj}\sin\theta_{kj% ,t}+B_{kj}\cos\theta_{kj,t})}$ (12) $\displaystyle C_{kj,t}\leqslant C_{kj}^{\max},C_{kj,t}=\sqrt{(P_{k,t})2+(Q_{k,% t})^{2}}$ (13)

where $P_{k,t}$ and $Q_{k,t}$ are the active and reactive power of node $k$ at time $t$ ; $G_{kj}$ and $B_{kj}$ are the real and imaginary parts of node derivative matrix $k$ rows $j$ columns; $\theta_{kj,t}$ is the phase angle difference between node $k$ and node $j$ at time $t$ ; $C_{kj,t}$ is the tide magnitude of node $k$ and node $j$ at time $t$ .

Rotating standby constraint:

$\displaystyle\sum\limits_{i=1}^{N}{P_{i}^{\max}I_{i,t}+P_{{G}t}\geqslant L_{t}% +S_{{R}t}}$ (14) $\displaystyle\sum\limits_{i=1}^{N}P_{i}^{\min}I_{i,t}+P_{{G}t}\leqslant L_{t}+% S_{{R}t}$ (15)

where $S_{{R}t}$ is the rotating reserve power of the system at time $t$ , $P_{i}^{\max}$ and $P_{i}^{\min}$ are the maximum and minimum power of the thermal power unit $i$ .

Output constraint for wind power:

A wind power station’s output at any given point in time should not exceed the forecasted amount.

$\displaystyle 0\leqslant P_{t}^{W}\leqslant P_{t}^{\textit{pre}}$ (16)

where $P_{t}^{\textit{pre}}$ is the predicted value of wind power at time $t$ .

2.2.3 Operational constraints on natural gas generation

$\displaystyle P_{G}^{\min}\leqslant P_{{G}t}\leqslant P_{G}^{\max}$ (17)

where $P_{G}^{\max}$ and $P_{G}^{\min}$ are the maximum and minimum values for natural gas-fired electricity generation respectively.

Natural gas supply and demand, as well as natural gas inventory, are taken into account to determine the maximum daily generation capacity of natural gas in areas with low natural gas supply:

$\displaystyle\int_{t=1}^{T}{P_{{G}t}}(t){d}t\leqslant E_{G}^{\max}$ (18)

3. Model solving process

In this paper, a unit commitment model considering the promotion of wind power consumption and natural gas peak-shaving is developed as a large-scale, discrete, non-convex, non-linear model. Genetic algorithms tend to converge prematurely and have weak hill-climbing abilities, resulting in non-optimal results. The model is solved using an improved genetic algorithm in this paper. Improved genetic algorithms include the following elements:

(1)
A special coding method is used, with a floating point code to represent the unit output size and a binary code to represent the unit start-stop status information.
(2)
The fitness function is ${\min}f\left(x\right)={\max}\left({-f\left(x\right)}\right)$ , in order to make the fitness result of all individuals positive or 0.
(3)
The variable crossover operator and the variation operator are set according to the number of genetic evolution. At the early stage of evolution, the crossover probability is increased to expand the global search range. At the later stage of evolution, the population may fall into a local optimum due to its genetic similarity, and the variation probability is appropriately increased to encourage new individuals to be generated and find the optimal solution.

Table 1
Six thermal power units’ specific parameters

Uint $P_{\max}$ /MW $P_{\min}$ /MW $a$ /($/(MW ${}^{2}\cdot$ h)) $b$ /($/(MW $\cdot$ h)) $c$ /($/h) $D_{R}$ /MW $U_{R}$ /MW $S$ ($)

1 450 120 0.00048 16.2 1000 200 200 9000

2 200 90 0.0014 17.53 800 100 100 18000

3 130 20 0.002 16.6 680 60 60 11000

4 80 25 0.004 19.7 450 90 90 3400

5 55 10 0.0071 22.26 370 40 40 600

6 40 12 0.0008 27.74 480 40 40 400

Figure 1.
IEEE-30 nodal system.

4. Case analysis

Uint	$P_{\max}$ /MW	$P_{\min}$ /MW	$a$ /($/(MW ${}^{2}\cdot$ h))	$b$ /($/(MW $\cdot$ h))	$c$ /($/h)	$D_{R}$ /MW	$U_{R}$ /MW	$S$ ($)
1	450	120	0.00048	16.2	1000	200	200	9000
2	200	90	0.0014	17.53	800	100	100	18000
3	130	20	0.002	16.6	680	60	60	11000
4	80	25	0.004	19.7	450	90	90	3400
5	55	10	0.0071	22.26	370	40	40	600
6	40	12	0.0008	27.74	480	40	40	400

For verifying the role played by natural gas peak-shaving in enhancing power system stability and economy, as well as for encouraging wind power consumption, the model was simulated and analysed using the IEEE-30 nodal system, as shown in Fig. 1. Two cases are designed, case 1 includes 6 thermal and wind power units in the node system, and case 2 adds 1 natural gas peak-shaving unit to algorithm 1. Each scheduling cycle is set to 24 periods of 1 hour each. The six thermal power units’ specific parameters are shown in Table 1. The load forecast values and wind power are shown in Fig. 2. The maximum output power of the natural gas unit is 80 MW and the minimum output power is 0 MW. The maximum daily gas supply is 2.4 $\times$ 10^{3 / m^3}. The abandonment penalty factor is taken as 300.

Figure 2.

Load forecast values and wind power.

Figure 3.

Output per time without natural gas peak-shaving.

4.1 24-hour unit commitment results

Based on the constructed power system model and parameters, simulations of cases 1 and 2 were conducted, and as shown in Fig. 3, output per time without a natural gas peak-shaving was obtained, and output per time with a natural gas peak-shaving was obtained, as shown in Fig. 4.

Figure 4.

Output per time with natural gas peak-shaving.

The start-stop status information for each unit at each moment in the system without natural gas peak-shaving as shown in Fig. 5. And the start-stop status information for each unit at each moment in the system with natural gas peak-shaving, as shown in Fig. 6.

Figure 5.

Start-stop status information without natural gas peak-shaving.

Figure 6.

Start-stop status information with natural gas peak-shaving.

Figure 7.

The output of natural gas generation.

Figure 8.

The output of wind power.

4.2 Analysis of natural gas peak-shaving power

In the case with natural gas peak-shaving, Fig. 7 shows the output of natural gas generation. The periods 1–3 and 22–24 are in the load trough period, and the thermal unit output decreases, but the wind power generation increases as well, and the thermal unit output decreases as well, at this time the natural gas turbine is shut down, giving priority to wind power consumption, reducing the amount of abandoned wind. During the periods 4–10, it is in the steady riser stage, but since there is not enough wind power generation, the natural gas unit begins to peak at a small output to maintain system stability. The periods 12–18 is the peak load period, and since wind output is higher during this period, the natural gas unit is shut down to promote the use of wind power. During the 19–21 period, wind output declines, requiring natural gas units to start up quickly to fill the output gap to ensure power system security and stability and meet the peak-shaving demand.

In the power system with wind power generation, the addition of natural gas peak-shaving can effectively reduce the pressure on thermal power units, increase the flexibility and rapidity of peaking, improve the system peaking performance, and ensure the stability of the power system.

4.3 Analysis of the promotion of wind power consumption

There are two graphs in Fig. 8, one without the natural gas peak-shaving, and one with the natural gas peak-shaving. The wind power output curves with and without the natural gas peak-shaving are represented in Fig. 9. Based on an analysis of the cumulative wind abandonment value, as shown in Fig. 9, natural gas units can significantly reduce the amount of wind abandonment and improve the new energy consumption capacity after participating in a peaking operation.

4.4 Economic analysis

A comparison of the economics of operating the two systems with and without natural gas peak-shaving, as shown in Table 2, shows that the inclusion of natural gas peak-shaving reduces the number of thermal unit start-ups and stops, saves manpower and material resources, and reduces the cost of thermal unit operation. In addition, due to the reduction of wind abandonment, the wind abandonment penalty is reduced. Taken together, natural gas peak-shaving reduces unit operating costs and facilitates efficient operation of the system as a whole.

Table 2
Comparison of the economics of operating the two systems

Case	Number of starts and stops	$F_{r}$ (10⁴$)	$F_{p}$ (10⁴$)	$F$ (10⁴$)
Excluding natural gas peak-shaving	14	48.41	20.62	69.03
Including natural gas peak-shaving	10	41.47	4.32	45.79

Figure 9.

Comparison of the cumulative wind abandonment value.

5. Conclusions

As a means of improving the unit operation economy and promoting new energy consumption, the paper proposes adding the natural gas peak-shaving to wind power generation systems, in order to reduce the total cost of thermal power units and add a penalty for wind abandonment, establishes a unit commitment model considering the promotion of wind power consumption and natural gas peak-shaving. The algorithm is solved using a modified genetic algorithm that avoids premature convergence of the results to obtain an optimal solution. In this paper, natural gas peak-shaving is analysed through case analyses in order to identify the effects on the system’s economy and wind power consumption. Following are the conclusions:

(1)

When natural gas peak-shaving units are added to a system with new energy generation, the fast and convenient peak-shaving characteristics of natural gas generation can increase system flexibility. Thermal power units can be started up and shut down less frequently, which lowers their operating costs.

(2)

As natural gas peak-shaving is added, it will have a positive influence on the system’s energy consumption, and it can also reduce the amount of abandoned wind, as the system’s peak-shaving capacity is increased.

Footnotes

Acknowledgments

The work has been partially funded by “National Natural Science Foundation of China” (62076152), “Innovation Ability Improvement Project of Technology-based SMEs in Shandong Province” (2023TSGC0966), and “Zibo Science and Technology-based Small and Medium-sized Enterprises Innovation Capacity Enhancement Project Program” (2023tsgc0013).

References

Chen

Xiao

Zhang

Hao

, et al. Cost dynamics of onshore wind energy in the context of China’s carbon neutrality target. Environmental Science and Ecotechnology. 2024; 19: 100323-100323. doi: 10.1016/J.ESE.2023.100323.

Xiao

Zheng

. New power systems dominated by renewable energy towards the goal of emission peak & carbon neutrality: contribution, key techniques, and challenges. Advanced Engineering Sciences. 2022; 54(01): 47-59. doi: 10.15961/j.jsuese.202100656.

Melamed

Ben-Tal

Golany

. A multi-period unit commitment problem under a new hybrid uncertainty set for a renewable energy source. Renewable Energy. 2018; 118: 909-917. doi: 10.1016/j.renene.2016.05.095.

Zhao

, et al. Analysis of power system peaking capacity and economy considering uncertainty of wind and solar output. Power System Technology. 2022; 46(05): 1752-1761. doi: 10.13335/j.1000-3673.pst.2021.0814.

Sun

Zheng

Ruan

, et al. Multi-energy cooperative primary frequency regulation analysis of a hybrid plant station for wind power and hydrogen production based on ensemble empirical-mode decomposition algorithm. Applied Sciences. 2023; 13(22): 12394. doi: 10.3390/APP132212394.

Wang

Niu

Chen

, et al. Research on day-ahead optimal dispatching of high-proportion renewable energy power system considering deep peak load regulation of thermal power. Acta Energiae Solaris Sinica. 2023; 44(01): 493-499. doi: 10.19912/j.0254-0096.tynxb.2021-0868.

Chang

Zou

, et al. Design and performance analysis of deep peak shaving scheme for thermal power units based on high-temperature molten salt heat storage system. Energy. 2024; 288: 129557. doi: 10.1016/J.ENERGY.2023.129557.

Liu

Liao

Wang

Jin

, et al. Short-term operation of cascade hydropower system sharing flexibility via high voltage direct current lines for multiple grids peak shaving. Renewable Energy. 2023; 213: 11-29. doi: 10.1016/J.RENENE.2023.05.095.

Wen

Zhen

, et al. Research framework and evolution paradigm of power grid with high proportion of high power toward carbon neutrality target. Automation of Electric Power Systems. 2023; 47(4): 1-9.

10.

Zhang

Tan

Zhang

Wang

Zhao

, et al. Natural gas market and underground gas storage development in China. Journal of Energy Storage. 2020; 29: 101338. doi: 10.1016/j.est.2020.101338.

11.

Barsali

Ciambellotti

Giglioli

Paganucci

Pasini

, et al. Hybrid power plant for energy storage and peak shaving by liquefied oxygen and natural gas. Applied Energy. 2018; 228: 33-41. doi: 10.1016/j.apenergy.2018.06.042.

12.

Zhang

Han

Yang

, et al. Optimization of integrated energy system considering transmission and distribution network interconnection and energy transmission dynamic characteristics. International Journal of Electrical Power and Energy Systems. 2023; 153: 109357. doi: 10.1016/J.IJEPES.2023.109357.

13.

Chen

Yuan

Wang

, et al. A novel conceptual design of LNG-sourced natural gas peak-shaving with gas hydrates as the medium. Energy. 2022; 253: 124169. doi: 10.1016/J.ENERGY.2022.124169.

14.

Liu

Yang

, et al. Development path of China’s gas power industry under the background of low-carbon transfor-mation. Natural Gas Industry B. 2021; 8(06): 576-587. doi: 10.1016/J.NGIB.2021.11.005.

15.

Jiang

Yuan

Zhang

Bai

Chen

, et al. Exploiting flexibility of combined-cycle gas turbines in power system unit commitment with natural gas transmission constraints and reserve scheduling. International Journal of Electrical Power and Energy Systems. 2021; 125: 106460. doi: 10.1016/j.ijepes.2020.106460.

16.

Fallahi

Maghouli

. Integrated unit commitment and natural gas network operational planning under renewable generation uncertainty. International Journal of Electrical Power and Energy Systems. 2020; 117: 105647. doi: 10.1016/j.ijepes.2019.105647.

17.

Cheng

Neha

Sambuddha

, et al. Modeling the operational flexibility of natural gas combined cycle power plants coupled with flexible carbon capture and storage via solvent storage and flexible regeneration. International Journal of Greenhouse Gas Control. 2022; 118: 103686. doi: 10.1016/J.IJGGC.2022.103686.

18.

Xia

Gong

Wang

Zhang

Shahidehpour

Ding

, et al. Multitime Scale Coordinated Scheduling for the Combined System of Wind Power, Photovoltaic, Thermal Generator, Hydro Pumped Storage, and Batteries. IEEE Transactions on Industry Applications. 2020; 56(3): 2227-2237. doi: 10.1109/TIA.2020.2974426.

Optimal scheduling for unit commitment considering wind power consumption and natural gas peak-shaving

Abstract

Keywords

1. Introduction

2. Model of the unit commitment

2.1 Target function

2.2.1 Constraints of thermal power units

4.3 Analysis of the promotion of wind power consumption

4.4 Economic analysis

Table 2 Comparison of the economics of operating the two systems

Footnotes

Acknowledgments

References

Table 2
Comparison of the economics of operating the two systems