Improving the effect of demand - Side managment programs by control methods for distributed generations

nstall the power system at each load node (can be regarded as each power consumption location in the EPS) at the low voltage level is being raised, making the power flow throughout the EPS not. as expected and also makes it difficult for the dispatching of the whole system. The reason for this is that the power emitted from PVG is only available during the time there is solar radiation and the power emitted from WG is always abnormal. This has made it extremely difficult to mobilize additional resources for the power shortages relative to load. At the same time, keeping power balance on the grid is also difficult, making the power at the nodes and frequencies in the whole system easy to lose control. This shows the role of the energy management program implementation at each load node with the participation of the source system. At the same time, studies in the world and in Vietnam on the source system DSM have not considered the application of a large capacity ES capable of discharging/charging to meet the unstable characteristics of nature and love. load. • Source control problem: With PVG, the commonly used control method is to combine a MPP max power point monitoring technique in the MPPT max power point monitor with a -5- control technique such as voltage control technique. AVC average, SMC slip control technique, FL fuzzy control technique, control technique using ANN neural network. The techniques of finding MPP are also quite many, diverse and according to many different goals such as CV constant voltage, P&O disturbance and observation, INC inductance increase, Temp temperature, OG slope optimization, detection ESC extreme, .... Each technique has its advantages and disadvantages and can be assessed in terms of cost of investment, ease of use/ease of use, accuracy, and energy loss. Most recently, IB detection and splitting technique has been proposed to be used based on PVG's full mathematical model but only used as a single source for the 3-phase grid; Only the IB-AVC method can be built with taking the signal of the current on the inductor and the voltage at the input of the DC/DC converter to act as control signals. The implementation of this control structure is complicated because two measurement variables have to be used. WG can be controlled via an active rectifier or through two converters, in which the active rectifier or DC/DC converter acts as a regulator of the power consumption corresponding to the maximum amount of capacity at a time. With WG, the extraction method is quite similar to PVG when there is a combination of MPP searching technique with control technique. However, there are only a few techniques for finding MPP commonly used with WG such as HCS hill climb, TSR pitch ratio, PSF power signal feedback. Of these techniques, the HCS technique is most commonly used due to its ease of implementation under practical operating conditions. At the node with the participation of the power system, the DC/AC converter is controlled to perform the role of power flow regulation and grid coupling. Single-phase DC/AC converter control is often studied in two main directions. The first direction studies the grid-connected current control to regulate the DC voltage maintained at a fixed value when there is no energy- balancing element such as the power warehouse on the DC side. The second direction studies the power current control on the AC side in systems that have a constant voltage holding element on the DC side. In particular, the control of the power current flowing through the DC/AC voltage fluctuation according to the preset values has been researched recently. The determination of the parameters of controllers should be made clear when applicable to DSM programs. • Determining the optimal ES capacity: currently there are several methods of determining the optimal capacity for the ES to be able to respond to different problems. The research direction to evaluate the ratio of lack of power (RLP) is based on the assessment of the ratio of the total capacity shortage due to the source not meeting the load and the total load capacity. The optimum ES power is then determined as the value that ES can compensate for the load so that RLP reaches the required RLP value. It can be seen that this method can use past or future data on the load and source system parameters to give the optimal value of the ES capacity. When RLP = 0, the load is fully powered by the ES. However, this method cannot evaluate the time of purchasing power, that is, it cannot -6- evaluate the effect of electricity price on the determination of ES capacity. The research direction uses the quantity of the probability of a power failure at any one time in the whole considering cycle. The value of this probability varies from time to time and is applied to achieve the minimum cost function goal caused by a power outage or to reach LSLP. It can be seen that this research also gives an assessment of the power supply capacity and has not yet evaluated the impact of the electricity price on the capacity of the ES. Direction of research and assessment of required energy of the load or requirement of energy storage of the source system. By specifying the storage time taking into account redundancy over a certain period of time, the capacity of the ES is considered to be able to guarantee the power supply to the load throughout the consideration cycle. However, the capacity determined by this method is quite large and cannot evaluate the interaction between the source system and the power system as well as the effect of the electricity price on the capacity of the ES. 1.4.2. Propose a solution Most studies on PVG and WG are still single or combined studies based on traditional methods with low energy efficiency. Therefore, the thesis proposes the following solutions: • Develop a DSM program at the load node with the participation of the source system in actual conditions in Vietnam and with the participation of a large-capacity ES: this program will be based on the regulation of 3 electricity prices for secondary households. load and regulations on electricity prices of households to load for EPS when exploited from source to EPS in Vietnam. The thesis will propose operational strategies for the DSM program to plan the whole system's operation on the basis of requirements management of loads and generation capacity from sources, storage capacity of ES. The DSM program helps to regulate the flow of capacity throughout the system and determine the optimal capacity for the ES that meets the EPS requirements in Vietnam. • Construction of control structure according to IB-AVC method for PVG object using PID controller with the use of a single measuring signal which is the voltage at the input of the DC/DC converter. This will also be the method of maximizing power from PVG under all operating conditions because it is always possible to determine the parameters that need to be controlled before controlling. Combined with HCS method and on-demand power control for single-phase DC/AC converter, the thesis will implement the control according to the requirements of the DSM program while ensuring maximum exploitation of the capacity of the source system. • Construction of a control structure for single-phase DC/AC converter, which uses a resonant controller to overcome the disadvantages of a conventional PI controller. 1.5. Conclusions of chapter 1 -7- Chapter 2 DESCRIPTION OF RESOURCES MATHEMATICS AND DSM MATHEMATICS 2.1. Solar battery source 2.2. Wind power 2.3. Develop a DSM program at the node to exploit the source system in specific conditions of the Vietnamese electricity system 2.3.1. The strategy of power flow regulation according to DSM model 2.3.2. Some constraints and limitations • Predictive curves are reallocated to rectangular plots. • The ith step of the divided variables corresponds to the timelines. • All power quantities in the entire system are converted to DCbus. The total power of the predictive power system obtained at DCbus at step i is determined by (2.29): Gconv PVGconv WGconvP (i) P (i) P (i)= + (2.29) The total amount of power that can be obtained from the DCbus on the DCbus at H, M, L hours in the cycle is determined by (2.30), (2.31), (2.32):     GconvH1 GconvH 2 14 23 GconvH Gconv i Gconv i i 12 i 21 E E E P (i) P (i) = = =  +   (2.30)       GconvM3GconvM1 GconvM 2 11 20 25 GconvM Gconv i Gconv i Gconv i i 6 i 15 i 24 EE E E P (i) P (i) P (i) = = = =  +  +    (2.31)     GconvL1 GconvL 2 5 27 GconvL Gconv i Gconv i i 1 i 26 E E E P (i) P (i) = = =  +   (2.32) Total required energy of the load in hours H, L, M in cycle  is determined by (2.33), (2.34), (2.35):     loadH1 GH 2 14 23 loadH load i load i i 12 i 21 E E E P (i) P (i) = = =  +   (2.33)       loadM3loadM1 loadM 2 11 20 25 loadM load i load i load i i 6 i 15 i 24 EE E E P (i) P (i) P (i) = = = =  +  +    (2.34) -8-     loadL1 loadL 2 5 27 loadL load i load i i 1 i 26 E E E P (i) P (i) = = =  +   (2.35) The total energy generated of the EGconv source system and the required power of the load Eload over the duration of the consideration cycle  are determined by (2.36) and (2.37): Gconv GconvH GconvM GconvLE E E E= + + (2.36) load loadH loadM loadLE E E E= + + (2.37) 2.3.3. Proposing DSM algorithm to operate at nodes with the participation of the source system in specific conditions of the Vietnamese power system Figure 2.8. DSM algorithm for systems that exploit source systems The following algorithms will not repeat the input data blocks because they are both serialized algorithm in Figure 2.8 or serial of each other. -9- Figure 2.9. System-wide operating algorithm for DSM mode 1 -10- Figure 2.10. Algorithm for the whole system of hours L1 of the scheme DSM1 With: Ede is the lack of electricity, EsL1 is the excess amount of electrical energy in the source system in the hour L1, Es(i) is instantaneous excess power. -11- One thing to note here is that the quantities related to the capacity in the ith step represent the ES capacity at the end of the calculation in order to be ready for the next calculation. Figure 2.11. Algorithm operating the whole system of hours L2 For option DSM 2, the release time of ES is determined by the algorithm described in Figure 2.12 and Figure 2.13. In this case, the algorithm proposes to evenly distribute the missing electricity of the hours L1, H and M must buy from the EPS to accumulate energy into the ES. This algorithm has been published in Article 7. -12- Figure 2.12. The whole system operation algorithm for DSM plan 2 -13- Figure 2.13. Program of phase L1 of the DSM 2 option 2.3.4. Proposing a method to evaluate the effectiveness of the DSM program and the optimal ES capacity for the DSM problem The optimal capacity of the proposed ES is determined based on the power purchase cost function Zrb, profit obtained from the sale of electricity Zas and the economic function Z. -14- Figure 2.15. Algorithm to determine the optimal ES capacity 2.4. Results of simulation of the DSM program operating the source system operation applied to the Vietnamese electricity system 2.4.1. Input parameters 2.4.1.1. DSM plan 1 Total required amount of electrical energy of the load: less than the total amount of electrical energy generated from the source system in hours H and M; is greater than the total amount of electricity generated from the source system -15- in hour L1; is greater than the total amount of electricity generated from the source system in the whole consideration period. Figure 2.17. Capacity obtained from source system and required capacity of load of DSM plan 1 2.4.1.2. DSM plan 2 Figure 2.19. Power obtained from the source system and the required capacity of the load DSM plan 1 Total required amount of electrical energy of the load: greater than the total amount of electrical energy generated from the source system in hours H, M, L1 and in the whole consideration period. 2.4.1.3. Data related to converter DC / DCES inequality efficiency is 2 = 0.95; DC / AC inequality efficiency is  = 0.95. 2.4.2. Determine the optimal size of the ES Select the starting value of the rated capacity Cr = 30 kWh, the step capacity of capacity C = 5 kWh. Simulation results show that the minimum value of Cr must be between 365 kWh and 370 kWh. For redundant capacity (5  10)%, the optimal rated capacity value of the ES is selected as Cropt = 400 kWh. With the selected Cropt value, the next content will be simulated to verify the economic efficiency of implementing the DSM program with no DSM implementation in options 1 and 2. These simulation contents all assign the capacity of the ES before entering the calculation cycle: Cins(0) = 0.2Cr. This value is significant to evaluate the purchase and sale of electricity to serve the load or ES is completely in cycle , thereby seeing the significance of the DSM problem. -16- 2.4.3. Results of simulation to evaluate the effectiveness of the DSM 1 Corresponding to Cr = 400 kWh, the graph showing the instantaneous capacity of Cins of ES when applying DSM and when not using DSM is shown in Figure 2.20. a. Applying DSM b. Not applicable DSM Figure 2.20. Instantaneous capacity graph of ES option 1 Figure 20.a shows that the Cins capacity has reached the rated capacity Cr before the end of phase L1 and launched to the minimum Cmin capacity according to the DSM requirement. With the result of Figure 20.b, ES has passive operation because it only charges when the power system has excess energy and discharges electricity when the power system is not able to meet the load. Electricity Erb to be purchased from EPS and electricity from Eas to be sold to EPS using DSM and when not using DSM is shown in Figure 2.21. a. Applying DSM b. Not applicable DSM Figure 2.21. Erb and Eas graph of DSM1 Figure 2.21a shows that the DSM program purchased electricity from EPS in phase L1 to both meet load and ES charging. Also, do not buy electricity from EPS at H and M hours under the DSM program. With the result of Figure 2.21b, the system buys electricity at times with L, H and M hours to meet the load. Zrb costs to buy electricity from EPS, profits Zas from selling electricity to EPS EPS when applying DSM and when not using DSM is shown in Figure 2.22. a. Applying DSM b. Not applicable DSM Figure 2.22. Zrb and Zas graph of DSM1 -17- Figure 2.22a shows that DSM program bought electricity from EPS in period L1 and sold electricity at H, M hours to limit cost of electricity purchase. As shown in Figure 2.22b, the system has lost the cost of purchasing electricity without collecting money for selling electricity. These simulation results demonstrate the strength of the DSM program. The above results were published in article 5. 2.4.4. Results of simulation to evaluate the effectiveness of the DSM 2 Corresponding to Cr = 400 kWh, Figure 2.23a shows that the Cins capacity has reached the rated capacity of Cr before the end of phase L1 and launched to the minimum capacity Cmin according to the requirements of the DSM program. Figure 2.23b, ES has passive operation because it only charges when the power system has excess energy and discharges electricity when the power system is not able to meet the load. With the problem 2, the ES had to continuously discharge the load in response to the load during the H and M hours to avoid buying high-priced electricity. a. Applying DSM b. Not applicable DSM Figure 2.23. Instantaneous capacity graph of ES problem DSM 2 Graphs showing the amount of electricity Erb needs to buy from EPS and how much electricity Eas will sell to EPS using DSM and when not using DSM is shown in Figure 2.24. a. Applying DSM b. Not applicable DSM Figure 2.24. Erb and Eas graph of DSM2 Figure 2.24a shows that the DSM program purchased electricity from EPS in phase L1 to both supply load and load ES. Also, the system does not buy electricity from EPS during the period H and M under the DSM program. Figure 2.24b, the system buys electricity in times of L, H and M to supply the load. With the problem 2, the load is mainly increased during H and M hours, causing the system to continuously buy electricity from EPS if DSM is not performed. Zrb costs to buy electricity from EPS, profits Zas from selling electricity to EPS when applying DSM and when not using DSM shown in Figure 2.25. -18- a. Applying DSM b. Not applicable DSM Figure 2.25. Zrb and Zas graph of DSM2 Figure 2.25a shows that DSM program purchased electricity from EPS in phase L1 and sold electricity at H, M hours to limit the purchase cost as required. Figure 2.25b, the system has lost the cost of buying electricity without collecting money for selling electricity. These results demonstrate the strength of the DSM program in the resource system exploitation system, especially with the help of ES and DSM. The above results have been published in the article No.7. DSM efficiency is assessed through the difference between the cost of purchasing electricity and the profit earned from selling electricity. The results show that: • With program 1: the difference between the cost of purchasing electricity and the profit when implementing DSM is negative while this difference when not implementing DSM is positive. This shows that the DSM program was very effective. • With program 2: the difference between the cost of purchasing electricity and profit when implementing DSM or not implementing DSM is all positive. However, it has made a profit by selling electricity and reducing the cost of buying electricity from the grid to meet the load (nearly 50%). 2.5. Conclusions of chapter 2 • Synthesize the mathematical descriptions of the main objects in the source system, that is PVG and WG. Factors affecting the parameters on the mathematical model such as G, T, ... have been evaluated in detail. Develop a strategy to regulate the capacity flows throughout the system according to the requirements set out by the DSM program in the actual conditions in Vietnam. • The simulation results show that corresponding to the installed capacity of PVG and WG, power generation graphs under operating conditions show the correctness of the proposed DSM strategy. The capacity of the selected ES can help absorb all the power from the mains or the mains at hour L to accommodate the load at H hour and M hour in both cases where the required energy of the load is greater. generation capacity of the source system. These results can be used to plan the operation of each element throughout the system through the design of the controllers for the converters and provide set values for those controllers. -19- Chapter 3 Figure 3.1. Structure diagram of the control system of the source system 3.2. Base control theory and mathematical description of converter power electronics Figure 3.15. Loop control structure inverter DC / DC boost method IB-AVC PID controller has the form: I cv P D K G K K s s = + + in it: dc pv eq dc P dc eq S L C R R K 2V R T + = , dcI dc eq S R K 2V R T = , dc pvD dc S L C K 2V T = 3.4. Construction of wind power controllers: controlled under the HCS method. 3.5. Construction pairing grid controller as required DSM 3.5.1. Control structure: Control structure applicable to 1-phase DC / AC converter is shown in Figure 3.20. -20- Figure 3.20. Control structures inverter DC / AC 1 phase 3.5.2. Current controller: iici pi 2 2 0 K s G (s) K s = + + (3.21) 2 2 pi bwK R (L ) 2R ,= +  + 2 2 2 2 2bw 0 ii p bw p bw bw ( ) K (R K ) 2(L ) 2K L  −  = + +  − +    3.5.3. Power controller Loop control structure power control for the inverter DC / AC is described in Figure 3.23. Figure 3.23. Model of the power loop structure of a DC / AC converter ip cp pp K G (s) K s = + (3.39) in it: 2 2 tdc dc 0 L pi pp 2 pi g i 0 R C (1 2T )(K R) K K U (K L)  + + = + 2 2 0 L pi ip 2 pi g i 0 2 (1 2T )(K R) K K U (K L)  + + = + -21- 3.6. Simulation results Figure 3.34. Pmpp features and PPV Figure 3.35. Curve of the power generated by WG Figures 3.36 and 3.37. Power characteristics obtained on the branch DCbus PVG and WG Figure 3.38. ES power characteristic Figure 3.39. Pgref characteristics, Pg inverter DC / AC a. Time interval from 1.4s to 1.7s b. Time interval from 3.9s to 4.2s Figure 3.41. Features output current at the inverter DC / AC Comment on simulation results: - The controller IB-AVC helped exploit the full capacity of the MPP from PVG. Power generated from PVG fluctuates instantly according to the fluctuation in input. - The HCS controller helped exploit the capacity at the WG's MPP. The power characteristics obtained from WG always vary in the same pattern as the wind speed, even when the wind speed is completely lost. -22- - The net-side controller controls the power running through the DC / AC converter, precisely adhering to the set capacity, only a very small fluctuation in a short time when there is a change in the set power. Demonstrates the precision of the design of the net controller, achieving the quality required by the DSM program. - The simulation results show that ES plays a very good energy balance role. The whole system had a very strong variation of input parameters, making the exploited capacity from each source always fluctuate while the required power from the grid side also increased or decreased abnormally. 3.7. Conclusions of chapter 3 • Building a hybrid system structure between PVG and WG operating under the DSM model. • Construction of controllers: PID according to IB-AVC method applied to PVG objects; HCS applies to WG subjects; mesh operation under the DSM program. • The simulation results have solved the set requirements: Utilizing maximum power from each source under all operating conditions of the input parameters, controlling the power set to the transducer on the side. grid and perform power balance for the whole system. The simulation results show the accuracy of the controller design as well as the feasibility of the DSM program when all units are closely linked together. Chapter 4 EXPERIMENTAL POWER LINE CONTROL PROBLEM AT BUTTON WITH SOURCE PARTICIPATION 4.1. Build real equipment structure models Figure 4.1. Structural model real devices 4.2. The method of operating the real equipment model 4.3. Main equipment 4.4. Experimental results • To verify the ability to maximize PVG capacity, three sampling times were carried out under real conditions in Thai Nguyen province on June 20, 2018. -23- Figure 4.13. The first sampling results tested the ability to exploit MPP Figure 4.14. The second sampling results tested the ability to exploit MPP Figure 4.15. Results third sample tested ability to exploit MPP The experimental results show that G always varies in a very wide range while T varies very slowly. The voltage controller along with the variation of (G, T) created the lines ipv, iDCbus, VDCbus have the same profile as the variation of (G, T). This is similar to the graphs representing ppv, Pmpp, pDCbus, in which the ppv line always sticks exactly to the Pmpp line and the pDCbus line is always a little lower than the other two lines. This demonstrates the accuracy of the technique for determining MPP, Voltage Controller and hardware options. • Experimental capacity control as required: Sampling results on June 20, 2018 in Thai Nguyen. The first and second results show that corresponding to the change of (G, T), the power emitted from PVG always pursued exactly MPP. In both of these sampling, the controller for the DC / DCs2 converters did exactly that the voltage on the DCbus to the value of the applied voltage and the power current from source 2 made up only a small fraction of even a small percentage. when the AC load increases. -24- VDCbusref (V) POPref vDCbus (V) vs1 and vs2 (V) is1 and is2 (A) G (W/m2) Pmpp, ppv, pDCbus Graphs of power across DC / AC converters and ps1, ps2 a. First sampling b. Second sampling Figure 4.16. Test the ability to distribute natural power stream The results