Adaptive Operation Strategy of a District Cooling System with Chilled Water Storage and Its Validations by OpenModelica Modeling and Simulations

Abstract

Developing operation strategies for district cooling systems with chilled water storage is challenging due to uncertain fluctuations of cooling demand in actual operations. To address this issue, this paper developed an adaptive operation strategy and performed its validations by modeling and simulating a commercial cooling system in Shanghai using OpenModelica. Firstly, the originally designed operation strategy of the cooling system was evaluated by simulation but was found unable to meet the statistically averaged ideal cooling requirements due to the early exhaustion of stored chilled water at about 5:30 PM. Then, to build foundations for adaptive operation strategy development, a newly designed operation strategy was established by increasing the operation time of base load chillers in the valley and flat electricity price periods. The new strategy proved numerically sustainable in satisfying the ideal cooling demand. Moreover, to realize the strategy’s adaptability to actual cooling load fluctuations, an adaptive operation strategy was developed by tracking the target stored chilled water mass curve that was calculated by implementing the newly designed strategy. The simulation results verify that the adaptive operation strategy enables good adaptability to representative cooling load fluctuation cases by automatically and periodically adjusting the operation status of base load chillers. The adaptive operation strategy was then further widely numerically tested in hundreds of simulation cases with different cooling load variations. The time-lagging problem resulting in strategy failures was found in numerical tests and was addressed by slightly modifying the adaptive strategy. Results indicate that the adaptive operation strategy enables adaptability to deal with cooling demand fluctuations as well as allowing low cooling supply economic costs and power grid-friendly characteristics. This study provides theoretical support to strategy design and validations for district cooling system operations.

1. Introduction

The district cooling system is the essential energy supply system for building clusters [1]. As a key component of the cooling system, chillers generate chilled water using the inputted electric power based on the reverse Carnot cycle principle, then the chilled water carries the cooling energy and continuously supplies it to users [2]. Regarded as a high electricity consumption system, the district cooling system actually has natural grid-unfriendly characteristics. Taking a day in summer as an example, the ambient temperature and cooling demand are high in the daytime, which includes most of the high electricity load periods of the power grid, thus the cooling system has to consume a large amount of electricity during those periods, further increasing the burden of the grid. Additionally, due to the widely implemented time-of-use electricity price policy [3], the cooling system’s high consumption of peak-price expensive electricity leads to the inevitable rise in cooling supply economic costs.

Chilled water storage is an effective and easily implemented approach to address the above issues in district cooling system operations [4]. During operations with chilled water storage, chillers begin to operate in advance during the valley electricity price period in the nighttime, then produce the chilled water and deliver it into storage tanks. In the daytime, the stored chilled water is prioritized for being used for cooling supply, thereby reducing the chillers’ operation time in high electricity load periods. The essence of the chilled water storage operation lies in decoupling the production and supply of the chilled water by introducing the water storage facility, thereby transferring the electric power consumption peak of the cooling system from daytime to nighttime, realizing reductions in both power grid burden and cooling supply economic costs. However, the operation strategy of the cooling system with chilled water storage becomes more complicated and difficult to design, bringing great challenges to system controls and adjustments [4,5,6,7,8]. In general, in order to obtain a detailed, optimal, and applicable chilled water storage operation strategy for someday in the future, such as the best operation time of chilled water storage chillers, the full-day cooling demand variation on the target day must be precisely predicted as it is the foundation to determine the desired amount of stored chilled water. Usually, only the long-term statistically averaged cooling load variation is derived easily and then reliable enough to be adopted. By contrast, actual cooling load fluctuations in the real system future are challenging to accurately predict in the long-term or even in the short-term [9], because actual cooling loads are highly related to the uncertainties of users’ behavior as well as ambient temperature fluctuations [10]. Therefore, the key to establishing the operation strategy of the cooling system with chilled water storage consists of developing an adaptive strategy using limited statistically averaged information so as to respond to complex actual cooling demand variations that may possibly happen in the future. However, the existing relevant studies on the adaptability of cooling system operation strategy considering chilled water storage are still lacking, unfocused, and insufficient, needing further investigations:

(1)

Previous studies [4,5] employed quite complicated mathematical optimization frameworks to solve the best operation strategies. Generally speaking, the increases in algorithm complexity can result in poorer optimizing efficiency and stability, thus making the findings of these studies difficult to apply in actual operation applications for security reasons.

(2)

Current studies have only conducted research on a few cooling load situations [4,5,6], leading to insufficient analysis and discussions on the impacts of cooling load fluctuations on strategies.

(3)

Some studies focused on special scenarios, such as a data center [5] and a university building [6], and the variation pattern of cooling load in these scenarios differs from that of general commercial building clusters.

(4)

Reference [7] considered the uncertainty associated with cooling loads; however, the model is simplified and the load variation cases are simple, so it is difficult to provide generalized support to other studies.

There are generally two methods to study the operation strategy of the district cooling system. The traditional experimental method has been widely regarded as possessing many limitations, such as high economic costs, low research and development speed, restricted parameter adjustments, and potential damage to devices; thus, the numerical method has become more important and indispensable due to its good solutions to the above limitations. Programming languages and software platforms such as MATLAB/Simulink, TRNSYS, and Modelica have been widely applied in simulating cooling systems [11,12,13]. Among those languages and platforms, the object-oriented language Modelica has attracted increasing attention for its advantages of equation-based programming, simple and intuitive modeling, and rich open-source model libraries [14]. By using graphical modeling platforms such as OpenModelica [15], the physical system models can be easily built by dragging and connecting visual model components as well as setting simulation parameters inside [15]. Then, platforms automatically create the system’s flattened and optimized differential-algebraic equations and perform simulations by solving equations using advanced solvers [15]. Modelica has been successfully applied in the modeling of energy systems such as floating offshore wind turbines, building energy, and heat recovery systems [16,17,18]. This study will adopt the open-source modeling platform OpenModelica to numerically investigate the district cooling system’s adaptive operation strategy design and validations.

According to the above research background, the current studies of the adaptive strategy of cooling systems lack practicality and generality, and they have not overcome the key issue of how to leverage limited statistical averaged cooling load to respond to complex actual cooling demand variations. Thus, the main objective of this study is to develop a practical and generalized method to design an adaptive strategy for a cooling system, enabling adaptability to unknown cooling load fluctuations. Hence, this paper takes an unfinished commercial district cooling system in Shanghai as the simulated object, and studies the system’s adaptive operation strategy under chilled water storage conditions, providing early support to future commercial operation applications. Firstly, the complete model of the district cooling system is built using OpenModelica. The model is then used for numerical validations and improvement of the system’s designed operation strategy based on monthly averaged ideal cooling demand, establishing the foundation for developing the adaptive operation strategy. Then, the adaptive strategy of the cooling system with chilled water storage is developed according to representative cases of actual cooling demand fluctuations and then validated by simulations. Moreover, the newly developed adaptive strategy is further numerically tested by simulations under more conditions so as to optimize and modify the strategy, enhancing the strategy’s adaptability and robustness.

2. District Cooling System Configuration

The district cooling system studied in this paper was designed with a chilled water storage operation strategy. As seen in Figure 1, five water-cooled chillers were configured in the system and planned to operate following the designed strategy shown below:

(1)

During the valley period with low electric power consumption (typically from 0:00 AM to 6:00 AM), chillers S1–S3 generate the chilled water using the low-price electricity from the power grid and then transport the chilled water into storage tanks. Two base load chillers B1 and B2 keep operating simultaneously to meet fundamental cooling demand.

(2)

In the other periods with high power grid electricity supply load, chillers are all closed with cooling load from buildings in the district supplied only by chilled water in storage tanks.

By operating with the chilled water storage, the district cooling system was expected to realize significant reductions in both cooling supply economic costs and electricity supply burdens on the power grid. However, the system is still under construction without any field test experience. Therefore, the adaptability and feasibility of the above-designed operation strategy need early validation by using the numerical approach. As introduced in Section 1, an adaptive operation strategy considering actual cooling demand fluctuations also needs to be developed using the basic principle of the designed chilled water storage operation strategy, so as to support future commercial applications.

3. Time Series Data of Cooling Load and Electricity Price

To perform simulations of the cooling system and then study the system’s operation strategies, the required data on the cooling load and the electricity price were prepared before the simulations.

The cooling load in the district was estimated based on the open-source chiller system energy data sets on the Kaggle online platform [19]. The data sets were collected from a cooling system in Singapore [19]. Considering that the cooling system in this paper is located in Shanghai, China, the open-source cooling load data were averaged across days in August and then processed using min–max normalization to infer the ideal 24 h time series data of cooling load in the summer season of Shanghai. The total demand of chilled water mass flow rate could then be calculated according to the basic energy balance law. As shown in Figure 2a, the cooling demand exhibits a typical characteristic with an apparent increase after 6:00 AM and a starting decrease in the evening.

The Shanghai government has implemented the time-of-use electricity price policy as shown in Figure 2b [3]. The electricity price fluctuates with the change in different types of time periods. The designed operation strategy mentioned in Section 2 intends to only use low-price electricity in the valley period for chilled water generation and storage, avoiding electricity consumption during the other periods, especially the peak and advanced peak periods.

4. System Modeling and Simulation Settings

4.1. Modeling of the Cooling System

As shown in Figure 3, the district cooling system model was established based on the open-source non-causal modeling platform OpenModelica [15]. The cooling system model with a similar appearance and same topology as the real system was built by connecting the energy equipment models and mathematical function blocks in existing open-source model libraries [20,21]. These models follow the fundamental and widely accepted theories and equations for mass and energy balance, ensuring the simulation accuracy in this study. The five chillers were modeled according to the “Carnot_TEva” example case in the open-source Buildings library [21]. The chilled water storage tanks were simplified as one integrated tank and then modeled using the “OpenTank” model in Modelica Standard Library [20]. The chilled water demand from the district and the corresponding power consumption of secondary chilled water pumps were inputted through the “CombiTimeTable” block. The power consumption of secondary chilled water pumps increases with the increasing total chilled water mass flow rate as calculated by the given fitting equation shown in Equation (1): P_CWSPumps is the total electric power consumption of all secondary chilled water pumps (unit: kW), and Q_CW,Total is the total chilled water mass flow rate (unit: kg/s).

$$P_{CWSPumps}=0.08459\times {Q_{CW,Total}}^{1.286}$$

(1)

4.2. Simulation Settings

To build foundations for adaptive operation strategy development, a simulation case was set to verify the effectiveness of the designed operation strategy with chilled water storage in Section 2. In this case, all chillers only operated from 0:00 AM to 6:00 AM to store chilled water as well as to carry the base cooling load in that period. The total simulation time was set to 86,400 s (24 h) with the time step of 1 s. The governing equations of the system model were solved by the default DASSL (differential/algebraic system solver) with a simulation tolerance of less than 1 × 10⁻⁶ [22]. Other detailed simulation settings are listed in

Table 1. The cooling system operation conditions used in simulations.

Parameters		Chiller No.
		S1	S2	S3	B1	B2
Cooling capacity (kW)		7030			3510	3510
COP (−)		5.27			5.21	5.04
Evaporator	Chilled water flow rate (kg/s)	210			116.67	115.28
	Return water temperature (℃)	15			14.3	14.3
	Chilled water temperature (℃)	7			7	7
	Pump power consumption (kW)	90			37	45
	Chilled water pressure drop (kPa)	110			82	81
Condenser	Cooling water flow rate (kg/s)	333.06			166.94	166.94
	Cooling water inlet/outlet temperature (°C)	30/36			30/36	30/36
	Pump power consumption (kW)	160			90	90
	Cooling water pressure drop (kPa)	148			97	97

. The simulation parameters in Table 1, such as chiller performances (capacity and COP), chilled water outlet/return temperatures, and water flow rates were all set according to operational parameters at the designed condition of the simulated cooling system (in Figure 1), thus maintaining consistency with the future actual operation specifications. The inlet temperature of the cooling water was set according to the average ambient temperature in the summer season of Shanghai. The pressure drops in Table 1 were estimated based on the typical chiller data from a professional book [2].

5. Results and Discussion

5.1. Validation of the Designed Operation Strategy

As shown in Figure 4, by implementing the designed operation strategy with chilled water storage, the system power consumption mainly occurs in the time period from 0:00 AM to 6:00 AM to generate chilled water using valley price electricity, leading to the increase in the total mass of the chilled water in the storage tank. The amount of chilled water in the tank then decreases after all chillers closed at 6:00 AM and the stored chilled water is exhausted at about 5:30 PM. Obviously, the simulation results indicate that the desired ideal cooling demand in the district cannot be fully satisfied by implementing the designed operation strategy. The originally designed strategy needs adjustment by additionally running the chillers in other time periods.

5.2. Newly Designed Operation Strategy and Numerical Verification

By analyzing the results of the originally designed operation strategy in Section 5.1, it was found that there were still 6538.62 tons of chilled water that needed to be generated and then supplied to the district. If the two base load chillers B1 and B2 are assigned to supplement the chilled water, an additional operation time of 7.83 h is required for the base load chillers. Thus, the newly designed operation strategy was then developed by increasing the operation time of chillers B1 and B2 during the flat and valley periods, as shown in Figure 5. The new operation strategy was also verified through the simulation using OpenModelica. The new simulation case was similar to the previous case in Section 4, but the operation time of the two base load chillers was adjusted by removing the interlock between chillers S1–S3 and chillers B1–B2 and then inputting the new strategy into the system model. The other simulation settings in the new case were consistent with the settings in the previous case.

As seen in Figure 6, the district cooling system is capable of meeting cooling load requirements in the district after using the newly designed operation strategy. As supplementing the chilled water using two base load chillers in more time periods, the electricity consumption of the system naturally increases compared to the previous simulation results using the originally designed strategy (Figure 4a). But, more importantly, the chilled water in the storage tank is consumed slower after 6:00 AM than before (Figure 4b) and then could match the cooling demand for the entire day, indicating that the new operation strategy is sustainable to satisfy the ideal cooling demand in the district.

To further evaluate the newly designed operation strategy, the above simulation results were compared with the estimation results of the traditional operation mode without chilled water storage (the chilled water is supplied directly to the district without passing through storage tanks). According to the conditions in Table 1, it was assumed that two chillers with cooling capacities of 7030 kW and 3510 kW respectively were configured in traditional mode to estimate the performance parameters. The costs in

Table 2. Comparison of performance between two operation modes.

Parameters	Operation Modes
	With Chilled Water Storage (Newly Designed Strategy)	Without Storage
Daily total electricity consumption (kWh)	54,573.6	52,867
Daily total cost of electricity consumption (RMB)	30,374.8	65,293.4

were calculated by multiplying the electricity consumption (in kWh) during each time period by the corresponding electricity price (in Figure 2b) and then summing up the resulting costs across all time periods; this method ensures that the total cost accurately reflects the varying prices at different times of the day. As shown in Table 2, although the time-domain distribution characteristics of the power consumption are quite different (Figure 6a), the daily total electric power consumption in the two operation modes is similar. However, as the power consumption during the chilled water storage operation mainly occurs in the valley and flat periods with low electricity prices, the operation with water storage could significantly reduce the economic cost of electric power consumption (calculated based on the normalized electricity price in Figure 2b) by more than 50% compared to the result without cooling energy storage, further verifying the cost reduction and grid-friendly features of the chilled water storage operation with the newly designed strategy.

5.3. Adaptive Operation Strategy for Actual Operations

Although the newly designed operation strategy proved feasible to meet the given ideal cooling demand in Figure 2a, its adaptability still needs further evaluation for commercial operation applications due to the existence of deviations between the actual real-time cooling demand and the statistically averaged ideal demand. Figure 7 shows three representative cases of the actual cooling demand fluctuations:

(1)

In the first case, the daily total cooling demand remains approximately unchanged but the temporal distribution is different. This is a very common case of fluctuation caused by the uncertain behavior of the cooling energy users.

(2)

In other cases, the overall cooling demand increases or decreases due to weather changes, such as the sudden rise or drop in atmospheric temperature.

If still applying the operation strategy in Section 5.2 to the above cases of actual cooling demand without any strategy modification, it is difficult to ensure that the chilled water in the storage tank could just meet full-day cooling requirements.

However, unlike the ideal cooling demand, the actual cooling demand is real-time information, which is hard to know in advance, as also introduced in Section 1. Without an extremely accurate prediction model of future full-day actual cooling demand, it is challenging to precisely predict the cooling system performance ahead of the day and then modify the operation strategy. With the above considerations, the determined simulation results under ideal cooling load conditions in Section 5.2 were used to develop the adaptive operation strategy for actual operations. As mentioned in Section 3, the ideal cooling load is the monthly averaged statistical information that reflects the basic cooling demand characteristics in the district, so it is reasonable to set the results in Section 5.2 as the criterion for actual real-time operation strategy adjustments. In this study, the result of chilled water mass in the storage tank (in Figure 6b) was used as the target parameter, then the corresponding adaptive operation strategy was established:

(1)

The operation strategy of the chillers S1–S3 was unchanged and consistent with the designed operation strategies in Section 5.1 and Section 5.2.

(2)

The base load chillers B1 and B2 no longer operate at fixed time periods like the previous strategies. The operation states of the base load chillers were determined according to the deviation between the actual and target stored chilled water mass. As shown in Figure 8, a day was divided into 144 time periods with each period’s duration of 10 min, which is the required restart time for the base load chillers. At the start of each period, chillers B1 and B2 automatically operate until the period ends if the actual water mass in the storage tank is lower than the target mass in Figure 6b. In summary, the essence of the adaptive operation strategy is to track the target curve of the stored chilled water mass by periodically and conditionally adjusting the operation status of the base load chillers.

The simulations were then performed to numerically test the adaptive operation strategy. The three cases of the actual cooling load variations were generated by adding random noise signals with different scales to the ideal cooling demand (in Figure 2a), just the same as the variations in Figure 7. The adaptive operation strategy was realized and coupled into simulations using the “sampler” and the “logical” blocks, as indicated by the red box in Figure 8. The other simulation settings were consistent with the settings in previous simulations.

As seen in Figure 9, by implementing the adaptive operation strategy, the actual chilled water mass in the storage tank could closely track the target curve calculated under ideal cooling demand conditions (in Figure 6b). The tracking of the target curve was realized by adaptive adjustments to the base load chillers’ operation states according to actual cooling demand characteristics:

(1)

In case 1 of the actual cooling demand (in Figure 7), the temporal distribution of the transient cooling demand changes, but the daily total demand is nearly unchanged. Thus, as shown in Figure 10a, the daily total operation time of base load chillers in case 1 is 14.2 h, which is almost the same as the 14 h by implementing the designed operation strategy in Figure 5. However, the operation time period slightly changes as an additional operation at around 12:00 AM occurs due to the cooling demand fluctuation.

(2)

In case 2 and case 3, there is an overall rise or drop in the actual cooling demand, resulting in the corresponding variations in the total operation time of base load chillers, as shown in Figure 10a. The total operation time of base load chillers increases to 15.33 h in case 2 and it decreases to 12.67 h in case 3. The cooling system electricity consumption in Figure 10b varies with the base load chillers’ operation states as the electricity consumption rises in the operation time periods of chillers B1 and B2.

More importantly, by using the adaptive operation strategy, the base load chillers’ operation and system electricity consumption peaks still mainly occur in the valley and flat periods of the electricity price (in Figure 2b). Thus, the low cooling supply economic costs and grid-friendly features of the designed strategy are still inherited by the adaptive strategy. As seen in

Table 3. System performance in actual cooling demand cases.

Parameters	Case No.
	1	2	3
Daily total electricity consumption (kWh)	54,875.6	56,952.8	52,208.5
Daily total cost of electricity consumption (RMB)	31,872.1	35,174.3	28,519.7

, the daily total costs in different cases are also maintained at low levels compared to the total cost with the previously designed operation strategy in Table 2.

5.4. Further Testing and Modification of the Adaptive Operation Strategy

To further validate the adaptive operation strategy, more simulations of operations under actual cooling load conditions were performed while implementing the adaptive strategy. An additional 100 groups of full-day actual cooling demand were generated respectively in each of the three representative cases (Figure 7) by altering the seeds in the random signal generator blocks of the cooling system models in Figure 8. The other simulation settings remained unchanged. However, the simulation results in

Table 4. Numerical test results of the adaptive operation strategy.

Test Results	Number of Simulation Cases
	Actual Cooling Demand Case 1	Case 2	Case 3
Passed	46	35	50
Failed	54	65	50

indicate that the adaptive operation strategy failed in more than half of all numerical test cases due to the too-early exhaustion of the stored chilled water at 11:00 PM–11:59 PM, as shown in Figure 11.

In order to solve the problem above, the simulation results of a failed numerical test case were used for analysis as plotted in Figure 12. As discussed before for the newly designed operation strategy in Section 5.2, the base load chillers only operate in the valley and flat periods to avoid the powder grid’s electricity load peaks. Thus, the cooling demand during the peak period of 6:00 PM–9:00 PM is just carried by the stored chilled water, and the target chilled water mass in the storage tank drops quickly. The actual stored water mass then becomes larger than the rapidly decreasing target mass, resulting in a similar decrease in actual stored chilled water mass when tracking the target curve. However, because of the delay during the parameter tracking using the adaptive strategy, there still exists a continuous rapid decrease in actual stored water mass for a short period after 9:00 PM, even though the target value has already dropped slower, as indicated by the red circle in Figure 12, thereby making the amount of the actual stored chilled water unable to meet the remaining cooling demand. In summary, the failures of the adaptive operation strategy were caused by the strategy’s time-lagging feature.

As mentioned in Section 5.3, considering that the time required for base load chillers’ restart is about 10 min, the minimal time interval between the two adjacent adjustments to chillers operation status was also 10 min in the adaptive operation strategy, thus causing the above time-lagging problem when tracking the target curve. However, the base load chillers were closed for hours before their last restart at around 9:00 PM, indicating that the restart time interval limit could be ignored here. Therefore, in order to prevent the time lag and too early exhaustion of the stored chilled water before 12:00 PM, a slight modification was applied to the adaptive operation strategy, which was to force the base load chillers to operate at once after 9:00 PM. The modified strategy then was implemented into the OpenModelica system model by using the relevant blocks shown in the blue box in Figure 13.

After the modification, the adaptive operation strategy passed numerical tests in all simulations as the early stored chilled water exhaustion problem was successfully solved, as shown in Figure 14, meaning that the adaptability to deal with actual cooling load fluctuations was effectively achieved in the modified strategy. Figure 15 exhibits the daily total performance parameters that were obtained by averaging across all 100 test cases in each cooling demand condition, and these results further richly verify the low cooling supply cost feature of the adaptive operation strategy as compared to similar to the results in Table 2 and Table 3. The adaptive operation strategy will be applied and further examined for its performance in future real-time operations of the actual commercial district cooling system.

6. Conclusions

(1)

The district cooling system with chilled water storage was modeled and simulated using OpenModelica. The simulation results indicate that, under the statistically averaged ideal cooling demand condition, the originally designed operation strategy is not feasible to meet the daily cooling requirements in the district as the stored chilled water is exhausted at nearly 5:30 PM.

(2)

The newly designed operation strategy was then established by increasing the operation time of the base load chillers. The new strategy was numerically verified to have the capability of meeting full-day cooling demand, and it reduces the economic costs of electricity consumption by over 50% compared to costs in the normal operation mode without chilled water storage.

(3)

To deal with the actual cooling demand fluctuations, the adaptive operation strategy was developed based on tracking the target stored chilled water mass by dynamically adjusting the operation status of base load chillers. After the strategy’s slight modification to address the time-lagging problem, hundreds of simulation test results sufficiently verify that the adaptive strategy has adaptability to different typical cooling load fluctuations, and it also features low cooling supply economic costs and power grid-friendly characteristics. The results of this study provide solid theoretical support for the future commercial operations of the district cooling system. In future operations, the proposed adaptive strategy will be integrated into the automatic control program of the cooling system, thereby validating if the numerically predicted performance can be achieved, ultimately leading to further refinement and improvement of the operation strategy.

(4)

In chilled water storage tank modeling, a simplified open tank model was employed in this study. However, actual storage tanks are usually in the type of naturally stratified tanks that have thin thermoclines inside; these thermoclines lead to stored cold energy loss, which can become thicker over time. Thus, to further improve simulation accuracy, it is needed to utilize a computational fluid dynamics (CFD) approach and then couple it into OpenModelica, thereby simulating the detailed temperature distributions inside the storage tank and thus more precisely estimating the available stored chilled water amount. The CFD reduced-order technique may be required so as to balance the accuracy and efficiency of CFD.

(5)

In this work, the return temperatures of chilled water were fixed to their designed values. However, chilled water return temperatures may vary with load and also other factors; thus, to improve the reliability of simulation results, a more comprehensive model of cold energy consumption by users should be developed and then coupled into the simulations in the future.

(6)

This study lays the foundations for cooling system designs and optimizations; more devices such as a cooling tower and LiBr absorption-type chiller can be introduced based on the developed simulation framework, and leading-edge optimization methods such as machine learning and digital twins can be interlinked with simulations, thus enhancing the research advancements.