Optimal configuration for photovoltaic storage system capacity in 5G base station microgrids

0 Introduction

The construction of a new power system is an important support for achieving emission peak and carbon neutrality,and the proportion of new energy will continue to increase.In recent years,investment in new information infrastructure represented by 5G has increased,and the degree of network density and data volume has also increased,resulting in an increase in the power loss of the base station system.The world’s leading communications operators have successively launched a zero-carbon network strategy and intend to deploy distributed photovoltaics on a large scale in 5G base stations.To ensure the stable operation of 5G base stations,communication operators generally configure backup power supplies for macro base stations and approximately 70% of the micro base stations according to the maximum energy demand.Therefore,the battery used for the power backup has a large idle space.If it is combined with a distributed photovoltaic system to form an intelligent photovoltaic storage system,it can maximize the value of energy storage,stabilize the photovoltaic output,and promote the local digestion of new energy[1-2].Existing studies have shown that it is more economical and reasonable to connect large-scale distributed photovoltaic power generation to the distribution network with a microgrid scheme than to connect directly to the grid[3].Therefore,5G macro and micro base stations use intelligent photovoltaic storage systems to form a source-load-storage integrated microgrid,which is an effective solution to the energy consumption problem of 5G base stations and promotes energy transformation.

continue

Techno-economic parameters of the photovoltaicsValue TC,STC25 ℃cPV inv1000 CNY/kW cPV ma10 CNY/kW vsub0.42 CNY/kWh

At present,there are many studies on the energy conservation and emission reduction of base stations,mainly covering two aspects.On the one hand,considering the base station itself,the base station sleep mechanism is used to improve the energy efficiency of the system[4-6].On the other hand,considering the energy use,the concept of a green base station system[7]is proposed,which uses renewable energy or hybrid power to provide energy for the base station system,allowing energy flow between base stations and smart grid[8-11].A reasonable configuration of the capacity of the energy storage unit can improve the stability and security of the power supply of the base station[12]and reduce the economic cost of the microgrid system[13].Many researchers have conducted extensive studies on the optimal configuration of the optical storage microgrid capacity.According to the characteristics of the photovoltaic storage system joint operation,studies[14]have optimized the configuration of different types of batteries with minimum initial investment as the goal.In[15],for multiple photovoltaic storage microgrids in the distribution network,a two-layer optimal configuration method was used to determine the economic scheduling scheme of each photovoltaic storage microgrid and optimize the capacity of photovoltaic storage.

The above-mentioned studies have provided ideas and directions for the research work of this study.In terms of the optimal configuration of a photovoltaic storage microgrid,the constraint condition only considers the technical characteristics of the energy storage unit.However,the backup energy storage of 5G base stations not only has the technical characteristics of energy storage,but also has the characteristics of standby power supply.Therefore,the existing research lacks consideration of the standby power demand of the backup battery of 5G base stations,which is not conducive for ensuring the reliable operation of the 5G base stations.Moreover,there is a lack of research on the flexible scheduling operation that considering backup energy storage of 5G base stations and renewable energy.Therefore,in this study,we construct a new scenario of base station microgrids composed of 5G macro and micro base stations,and the power consumption of the base station microgrid is further reduced using a sleep mechanism,which innovatively combines the communication characteristics of 5G base stations and the backup power demand of the energy storage battery.From the perspective of the interests of both the power grid and base station operators,we construct a two-layer optimal configuration model,and the inner layer optimization considers the energy sharing among the base station microgrids and determines an economic scheduling strategy for each photovoltaic storage system with the goal of minimizing the daily operation cost of the base station microgrid.The outer model aims to minimize the annual average comprehensive revenue of the 5G base station microgrid,while considering peak clipping and valley filling,to optimize the photovoltaic storage system capacity.Finally,the actual operation of the 5G base station micro-network in a region is taken as an example,and the obtained results can serve as a reference for future 5G base station planning and construction and related research.

1 5G base station microgrid structure and mathematical model

1.1 5G base station microgrid structure

The photovoltaic storage system is introduced into the ultra-dense heterogeneous network of 5G base stations composed of macro and micro base stations[16]to form the micro network structure of 5G base stations[17].Photovoltaic power generation is used as a distributed power source,and the backup power storage and photovoltaic power form a photovoltaic storage system.The photovoltaic storage microgrid structure of the grid-connected 5G base station is shown in Fig.1.

Fig.1 Microgrid control architecture of a 5G base station

Photovoltaic power generation is the main power source of the microgrid,and multiple 5G base station microgrids are aggregated to share energy and promote the local digestion of photovoltaics[18].An intelligent informationenergy management system is installed in each 5G base station micro network to manage the operating status of the macro and micro base stations.The intelligent informationenergy management system aggregates and regulates information and power through an aggregated interaction platform for controllable resources.During the peak period of photovoltaic power generation,the power demand of the 5G base station microgrid is first met.If the supply exceeds the demand,the battery energy storage system is charged;if the demand exceeds the supply,priority is given to the battery energy storage and discharged to the microgrid;if the supply remains insufficient,electricity is purchased from the grid to meet the 5G base station microgrid load demand.The architecture of the complete system model is illustrated in Fig.2.

Fig.2 Architecture of the aggregated interaction platform for controllable resources

1.2 Time-space characteristic model of the 5G base station load based on the sleep mechanism

In this study,a macro station Bmand kmicro base stations within its coverage range constitute a base station microgrid,and their load characteristics,namely power consumption,are modeled.The total power consumption in a 5G base station microgrid at time tas in(1).

where Ptotal(t)is the total power consumption of the base station microgrid at time t,Pm(t)is the power consumption of the macro station in the base station microgrid at time t,and Pk(t)is the power consumption of the k-th micro base station within the range Bsat time t.

In a typical base station power consumption model,the power consumption of the base station is not stable at a particular value but changes with the real-time communication traffic load[19].Owing to the behavior of the communication users,the traffic load has the dual characteristics of time and space.The time-domain traffic load model of the base station in the same spatial domain can be fitted using the sinusoidal superposition model[20].The model as in Equation(2):

where Pout(t)is the total communication traffic of the macro base station in the microgrid at time t,and α0 is the average business traffic over time.ωwis the frequency component of the business traffic changes.αwand φware the amplitude and phase,respectively.wis the number of frequency components.

Therefore,the power consumption Pm(t)of the acer station as in Equation(3).

where ξis the proportional coefficient related to the traffic load.ξ Pout(t)is the power consumption generated by the traffic load.P0 is the base power consumption generated by the four base stations when there is no traffic load.

In the 5G base station microgrid,the traffic of the macro and micro base stations exhibits obvious periodicity in time,and the upward and downward trends are in step.Therefore,the flow load of the macro base station is set to Xtimes that of the micro-base station.The change in the service flow of a single micro base station in the 5G base station micro network is random and non-uniformly distributed.According to the lognormal distribution model,the traffic load model of the k-th micro base station at time t as in Equation(4).

The average business traffic of a single micro base station as in Equation(5).

Owing to user social behavior,the standard deviation σvaries with the distribution of the user and business traffic and the type of scenario.The mean value μof the lognormal distribution model at each moment as in Equation(6).

The power consumption of micro base station is mainly basic power consumption.It does not change significantly with the traffic load,and because the micro base station is in the active or dormant state,the power consumption of the k-th micro base station as in Equation(7).

where Pθactive is the power consumption of the micro base station in the active state.Pθsleep is the power consumption of the micro base station in the dormant state.θk={0,1} represents the state of the micro base station.When the micro base station is mounted without traffic,θkis 0,indicating sleep state.When the micro base station has a traffic mount,θkis 1,indicating active state.

The 5G network is always designed with the maximum traffic load that the system can withstand during deployment,which leads to energy waste.The sleep mechanism can further optimize the power consumption of the 5G base station microgrid[21].Every hour,the intelligent information-energy management system wakes up all the macro and micro base stations in the 5G base station micro network,mounts the user traffic,and traverses all the micro base stations in the 5G base station micro network.In addition,if the traffic load of a micro base station is less than the threshold,the other macro and micro base stations in the micro network will not have reached the maximum communication load that can be received,and this micro base station goes to sleep.

The objective function as in Equation(8).

where is the total power consumption of the base station microgrid after sleep,as in Equation(9).

where is the power consumption of the macro base station after sleep,as in Equation(10).

where is the traffic load of the macro base station after the sleep mechanism,as in Equation(11).

where ΔPis the dormant traffic load of the micro base station,that is,the traffic load of the dormant micro base station.If the k-th micro base station is dormant,then ΔP=Pk·out(t).At this time,the value condition of θkas in Equation(12).

where is the sleep threshold of the micro base station.

The constraint conditions are:

The maximum communication load constraint that the macro base station can receive as in Equation(13).

The maximum communication load constraint that the micro base station can receive as in Equation(14).

1.3 Photovoltaic power generation model

The relationship between the photovoltaic power generation power and its rated capacity as in Equation(15).[22]

where PPV,STC is the rated capacity per unit area of the photovoltaic panel.Mis the area of the photovoltaic modules.GCis the regional light intensity(kW/m2).GSTC is the solar radiation intensity under standard test condition ap is the power-temperature coefficient of the panel.TC,STC is the temperature of the panel in the photovoltaic array under standard test condition.TC is the working temperature of the battery panel during the energy conversion process of the photovoltaic cell,as in Equation(16).

where Tis the ambient temperature of the area where the photovoltaic array is located.

1.4 Base station backup energy storage battery model

The charge-discharge power formula of the 5G base station backup energy storage battery as in Equation(17)and Equation(18).

where Eis the rated capacity of the 5G base station backup energy storage battery,SOC t( )is the state of charge at time t,and DESS is the self-discharge coefficient.Pcha(t)and Pdisc(t)are the charging and discharging powers within 1 h,respectively,and P t′( )is the power input from the other 5G base station microgrids,which can be positive or negative.Uc and Ud are the charge and discharge state coefficients respectively,which both are variables of 0 and 1.ηinv is the inverter efficiency,and ηbat is the battery charge and discharge efficiency.

2 Optimal configuration model of a base station microgrid based on energy sharing

2.1 Outer layer optimization configuration model

The outer planning model starts from the base station operator and the power grid and takes the lowest annual average comprehensive cost of the 5G base station microgrid system as the objective function.The decision variables include the configuration capacity of the photovoltaic and energy storage in the microgrid.In this study,5G base station operators are considered as photovoltaic storage system investors,and the electricity cost of the base station microgrid is the total cost of the operators,including the operators’ annual investment and maintenance cost of the energy storage and photovoltaic system,the cost of electricity purchase from the grid,and government subsidies.With respect to the power grid,the participation of the 5G base station microgrids in the power grid interaction introduces the benefits of delayed power grid upgrading.In this study,only typical days are considered,and the typical days of four quarters are selected to represent the entire year.This method can easily obtain data and is simple to calculate,which conforms to the current actual situation.

Considering the lowest annual average comprehensive cost of the 5G base station micro-network as the objective function,as in Equation(19).

2.2 Inner layer optimization of the operation model

The inner model is a daily operation model of multiple 5G base station microgrids based on energy sharing strategies.After the outer planning model determines the capacity of the photovoltaic system and energy storage system,the inner model can optimize the operation of the base station microgrid.The electric power demand,photovoltaic output,and backup energy storage inventory of each 5G base station microgrid are monitored through an intelligent information-energy management system.Considering the backup power demand of the 5G base station’s own backup energy storage,the photovoltaic output of each microgrid is shared through the aggregated interaction platform for controllable resources.The daily operation cost is minimized using an objective function,optimizing the daily operation of the multi-5G base station micro-network energy sharing.

The objective function used to minimize the daily operation cost of the multi microgrid as in Equation(28):

3 Model solving

In the optimal configuration model of the photovoltaic storage system established in this study,the outer planning model adopts a genetic algorithm,the objective function is defined in Equation(19),and the constraint conditions are defined in Equations(26)-(27).The initialization decision variable is the rated capacity of the photovoltaic and energy storage of the base station microgrid,which are transferred to the inner layer.The inner layer optimization model is calculated using the CPLEX solver[24].The objective function is defined in Equation(28),and the constraint conditions are given in Equations(32)-(33)and(37)-(40).The decision variables of inner layer are the energy storage charge and discharge power,photovoltaic output,and load peak-cutting rate of each base station microgrid in each season,which are transferred to the outer layer.The outer layer calculates the annual average comprehensive revenue and fitness value of the 5G base station microgrid,and then obtains the rated capacity of the photovoltaic and energy storage sub-generation of the base station microgrid through genetic operations such as crossover and mutation.Via repeated iterations,the optimal solution of the two-layer optimization model can be obtained.The solution process is illustrated in Fig.3.

Fig.3 Flowchart for solving the bilayer model

4 Case study

4.1 Basic data

According to the actual construction and distribution of 5G in a certain region,2100 5G base station microgrids of three categories were selected for simulation to verify the effectiveness of the capacity optimization configuration model of the optical storage system proposed in this paper.The base power consumption of the Acer station is 0.78 kW,the flow load coefficient is 4.7,the active power consumption of the micro base station is 0.112 kW,and the dormant power consumption of the micro base station is 0.039 kW.The standard deviations of the 5G base station microgrids in the university,park,and business districts are 3.6,1.3,and 2.8,respectively.The typical daily load curves of each type of 5G base station microgrid obtained before and after the hibernation algorithm are shown in Fig.4.Photovoltaic data were obtained from the statistics of the measured satellite data in a certain area.The photovoltaic equipment parameters are listed in Table 1.A lithium battery was used as an example for energy storage equipment,and the equipment parameters are listed in Table 2.The simulation period was 10 years,the discount rate was 0.1,and the annual load growth rate was 1.5%[25].5G base stations implement industrial and commercial electricity prices.The current TOU electricity price policy in a region is as follows:peak period(8:00—12:00 and 21:00—23:00),0.9876 CNY/(kWh);trough period(23:00—7:00 the next day),0.3289 CNY/(kWh);peacetime period(7:00—8:00 and 12:00—19:00),0.6758 CNY/(kWh);critical peak time(19:00—21:00 in January,July,August,and December every year),1.11826 CNY/(kWh).

Table 2 Parameters of the proposed model of energy storage

Techno-economic parameters of energy storageValue cinvESS2000 CNY/kWh cmaESS10 CNY/kWh DESS0.01%ηbat95%SOCmax0.8[26]SOCmin0.2[27]

Fig.4 Typical daily load curves before and after application of the sleep algorithm of the three types of 5G base station microgrids

Table 1 Parameters of the proposed model of photovoltaics

Techno-economic parameters of the photovoltaicsValue PPV,STC0.45 kW/m2 GSTC1 kW/m2 ap-0.0047/℃

4.2 Results analysis

To clarify the impact of photovoltaic system and energy storage on the operation of the 5G base station microgrid system and the base station operator and power grid,one microgrid was selected from each of the three types of 5G base station microgrids,and simulations and analyses were performed for the following four scenarios.Scenario 1:The user does not invest in the configuration of photovoltaics,and the energy storage equipment is configured according to the maximum energy demand;Scenario 2:The user invests in the configuration of photovoltaics,and the energy storage equipment is configured according to the maximum energy demand of the equivalent load;Scenario 3:The user invests in the configuration of photovoltaics,and configures energy storage equipment according to the real-time requirements of the backup power;Scenario 4:Users invest in the deployment of photovoltaic and energy storage equipment,multi-microgrids conduct energy sharing strategies,and configure the photovoltaics and storage according to realtime requirements.Scenarios with the same photovoltaic storage capacity configuration were selected for scenarios 3 and 4 for comparison.

· Analysis of the photovoltaic and energy storage planning results

When the base station operator does not invest in the deployment of photovoltaics,the cost comes from the investment in backup energy storage,operation and maintenance,and load power consumption.Energy storage does not participate in grid interaction,and there is no peak-shaving or valley-filling effect.The comprehensive income of multiple microgrids is negative,and the value is equal to the average annual cost of the base station operators.It can be seen from the third column of Table 3 that after the investment and deployment of photovoltaics,the average annual cost of investment and operation,and maintenance of the base station operators increased,but the cost of load electricity dropped by 40.94%,and the load peak reduction rate was 4.50%.The revenue was positive,up by 143.49%,indicating a large increase.The deployment of distributed photovoltaics in the base station can effectively promote the construction of a zero-carbon network by the base station operators.

Table 3 Comparison of the 5G base station micro-network operation results in different scenarios

Scenario 1 Scenario 2 Scenario 3 Scenario 4 Photovoltaic capacity(kW)0555 Energy storage capacity(kWh)19192121 Average annual investment cost(CNY)9,276.4914,158.85 15,135.32 15,135.32 Average annual maintenance cost(CNY)570720780780 Electricity cost of multiple microgrids(CNY)41,202.95 24,335.71 15,447.65 15,105.86 Government subsidy of multiple microgrids(CNY)0589.631,738.921,767.14 Load peakshaving rate （%）04.55.835.92 Delayed power grid upgrade income(CNY)060,827.9668,687.2 69,087.74 Comprehensive income of multiple microgrids(CNY)-51,049.4 22,203.02 38,878.73 39,833.7

As shown in the third and fourth columns of Table 3,we compare the energy storage equipment configured according to the maximum energy demand of the equivalent load with according to the requirements of the real-time back-up power energy storage equipment configuration and flexible scheduling.For base-station operators,although the energy storage investment and operation and maintenance costs in scenario 3 are relatively high,the cost of electricity purchase from the grid reduced,the utilization rate of the photovoltaic storage system and government subsidies increased,and the average annual cost dropped by 23.30%.For the power grid,the load peak-shaving rate increased by 1.33%,and the load curve became smoother.The comprehensive income of the base station operators and power grids increased by CNY 16,675.71.

From the above comparative analysis results,5G base station operators invest in photovoltaic storage systems and flexibly dispatching the remaining space of the backup energy storage can bring benefits to both the operators and power grids.With the further reduction of energy storage costs in the future,the advantages of investing in photovoltaic storage systems for 5G base station operation will become more obvious.

· Analysis of the effectiveness of energy sharing strategies.

To verify the feasibility and effectiveness of the energy sharing operation strategy proposed in this paper,the optimal planning results obtained in Scenario 4 were adopted,and a typical summer day was considered as an example for the simulation.The state of charge ranges of various types of 5G base station microgrid energy storage and the ratio of the storage capacity to the rated capacity at each moment are shown in Fig.5.

Fig.5 The energy storage and charge state range and capacity ratio diagram of various optical storage systems in Scenario 4

The charging and discharging actions of energy storage meet the requirements of various 5G base stations for microgrid power backup.During the low electricity price period,the 5G base station microgrid purchases electricity from the grid to meet the power demand of the base station.During the normal period of electricity price from 7:00 to 19:00,the photovoltaic output fails to meet the demand of the 5G base station microgrid,and the energy is stored for discharging.During 10:00-17:00,the photovoltaic output meets the requirements of the 5G base station microgrid,and the excess photovoltaic output is used for energy storage charging.From 18:00-23:00,the energy storage is discharged.Fig.6 shows a comparison between the final load curve of scenario 4 and the original load curve.In scenario 4,there is a significant reduction in the peak load and a smoother curve is obtained.

Fig.6 Comparison between the load curve of scenario 4 and the original load curve

It can be seen from the 4th and 5th columns of Table 3 that compared with the case where the energy sharing strategy was not adopted,adopting the energy sharing strategy increased the load peak-cutting rate by 1.5%,and the photovoltaic utilization rate increased,especially in summer,by 6.1%.The cost of the base station operators decreased by 554.43 CNY,and the comprehensive income of the multi-microgrids increased by 954.97 CNY.

In summary,the use of energy sharing strategies for multiple 5G base station microgrids can improve photovoltaic utilization,effectively smooth the load curve,and further increase the comprehensive benefits of the 5G base station microgrids.

5 Conclusion

In this study,for the optimal configuration of a 5G base station microgrid photovoltaic storage system,a twolevel optimization planning model was established,which comprehensively considers the average annual integrated cost of multiple 5G base station microgrids and grids and the daily operating cost of 5G base station microgrids.The analysis results showed the following.

(1)The configuration of the 5G base station microgrid photovoltaic storage system can not only meet the energy storage requirements of the 5G base stations,but also reduce the operating costs of the base station operators under the time-of-use electricity price mechanism.

(2)Access to the 5G base station microgrid photovoltaic storage system can effectively reduce the peak load and delay power grid upgrading.

(3)Access to the 5G base station microgrid photovoltaic storage system based on the energy sharing strategy has a significant effect on improving the utilization rate of the photovoltaics and improving the local digestion of photovoltaic power.

The case study presented in this paper was considered the base stations belonging to the same operator.The price of the photovoltaics and storage for energy sharing between different operators was not discussed in depth.In the future,the configuration results of the photovoltaic storage system of 5G base station microgrids under different investment and operation subjects will be further studied in combination with the development situation of China’s power market.

Acknowledgements

This work was supported by the State Grid Science and Technology Project(KJ21-1-56).

Declaration of Competing Interest

The authors have no conflicts of interest to declare.