固態氧化物燃料電池系統建模與控制

Modeling and Control of Solid Oxide Fuel Cell Systems

吳健彰(Chien-Chang Wu)

指導教授 ： 陳宗麟

國立交通大學/工學院/機械工程系所/博士(2020年)

摘要

本論文針對SOFC系統建立數值模型。所建立的模型包括：SOFC電化學模型、反應氣體動力學模型、燃燒器燃燒反應模型、重組器重組反應模型與熱交換器熱量交換模型。根據這些模型設計設計一套以CH_4為燃料的SOFC-CHP系統。該系統採取較複雜的廢熱回收機制，讓燃燒器所產生的廢熱大部分用於協助重組器重組反應與SOFC電化學反應，藉此獲得較多的電能輸出。此外，還在SOFC之前配置了一組額外的熱交換器，用於降低進入SOFC的重組氣體與空氣之間的溫度差異。在所提出的廢熱回收架構下，還設計了二組溫度控制器，用以維持SOFC的工作溫度與調整燃燒器所排出的廢氣溫度。從模擬結果可以知道，所設計的SOFC-CHP系統能輸出4.35 kW的電功率與6.96 kW的熱功率。與現有的SOFC-CHP系統相比，本論文所設計的SOFC-CHP系統具有較多的電能輸出。此外，從另一模擬結果還可以知道，在所設計的SOFC-CHP系統中，重組器、SOFC與燃燒器的工作溫度分別是830.8 K、973.4 K與1124 K。與現有的熱回收架構相比，本論文所設計的SOFC-CHP系統能在較低的工作溫度下進行發電任務。
由於所設計的SOFC-CHP系統排出的廢氣仍有6.96 kW的熱功率可以使用，因此本論文接著提出一套SOFC Tri-generation系統來使用該熱量。該系統是由SOFC-CHP系統、吸附式冷水機與配有溫度控制器的水箱組合而成。為了深入探討SOFC Tri-generation系統的性能，本論文整合了前述SOFC-CHP系統、吸附式冷水機元件與水箱的動態數值模型。基於這些模型，本論文還提出了一種從工作溫度的響應時間常數來決定水箱容量大小的方法。從模擬結果可以知道，所提出的SOFC Tri-generation系統能輸出4.35 kW的電功率、2.448 kW的熱功率與1.348 kW的冷功率。整體SOFC Tri-generation系統的能量轉換效率是64.9 %、製冷系統的COP是0.32。此外，從另一模擬結果還可以知道，當改變SOFC電功率輸出時，雖然會造成SOFC Tri-generation系統廢氣排出時的熱量變化，但是不會影響到吸附式冷水機的冷功率輸出；相對地，改變吸附式冷水機的冷功率輸出時，雖然也會導致SOFC Tri-generation系統廢氣排出時的熱量變化，但是不會影響到SOFC的電功率輸出。這些理想的工作性能，可歸因於本論文所提出的溫度控制策略。
本論文的另一重點在於發展大型發電系統中的燃料電池並聯策略與控制電路。大型燃料電池系統會透過數組SOFC電堆而獲得千瓦或兆瓦等級的電力輸出。然而，鮮少研究報告對SOFC電堆並聯之後的特性進行深入地探討。為此，本論文提出了一種用以模擬SOFC電堆並聯系統的動態數值模型。該模型整合了反應氣體供應架構、電性並聯特徵與熱傳遞機制。所建立的數值模型解決了傳統SOFC電化學模型無法用於讓並聯的SOFC電堆們具有相同輸出電壓、不同輸出電流的問題。這是因為傳統的SOFC電化學模型必須先制定反應氣體與輸出電流這兩個輸入變數，而這兩個輸入變數在並聯系統中，卻是無法先被預知得到的。為此，本論文採用控制理論的方法，將這兩個變數整合在多個回授路徑上，以進行SOFC電堆並聯系統的數值模擬。從模擬結果可以知道，並聯電堆的輸出電流即使是不一致或是發散的狀況下，每一顆電堆的輸出電壓仍會是相同的。另外，本論文還探討了造成並聯系統熱失控的可能原因。從模擬結果可以知道，當並聯的SOFC電堆的特性不匹配時，並聯系統可能會因為較慢的溫度暫態響應時間而變得較不穩定，甚至會引起熱失控的現象發生。另外，從模擬結果還可以知道，當並聯系統工作在較低的工作溫度條件下，並聯系統間會存在著一種不穩定的正回授機制，這個正回授機制將導致熱失控現象發生；然而，當並聯系統工作在較高的工作溫度條件下時，並聯系統間則是具有一種穩定的負回授機制，這個負回授機制能夠消弭SOFC電堆間的性能差異。
數個SOFC電堆透過並聯形成電堆模組，數個電堆模組透過並聯控制電路同時滿足負載電壓準位與大功率輸出的要求，常用的控制電路是直流/直流轉換器。然而，SOFC的發電特性與傳統的電源系統不同，因此傳統的直流/直流轉換器較不適用。為此，本論文提出了一套適用於SOFC發電特性的並聯系統控制方法。該控制方法除了能夠滿足負載的電壓準位要求之外，還能夠控制每一顆SOFC電堆的輸出功率，並且能在負載功率改變時，按照比例將額外的負載功率分配至每一顆SOFC電堆之中。也因如此，所提出的控制方法解決了傳統直流/直流轉換器無法在並聯系統中控制每一顆SOFC電堆輸出功率的問題，同時也放寬了並聯系統對於並聯SOFC電堆們的特性一致性要求。為了證明提出的控制方法的性能與有效性，本論文還進行了模擬驗證與實驗驗證。其中，在進行實驗驗證時，還設計了一種能呈現SOFC發電特性的電子電路，該電路能夠增加提出的控制方法的可信度。從模擬結果可以知道，在額定的工作條件下，控制SOFC模組輸出功率的誤差小於0.79%；在負載功率改變之後，額外增加的負載功率分配比誤差小於2.56%。另外，從實驗結果可以知道，在額定的工作條件下，控制SOFC模組輸出功率的誤差小於6.82%；在負載功率改變之後，額外增加的負載功率分配比誤差小於7.43%。由此可知，本論文提出的控制方法除了能夠控制每一顆SOFC電堆在額定條件下的輸出功率之外，還能夠規劃額外負載功率分配至每一顆SOFC電堆的大小。

關鍵字

固態氧化物燃料電池 ； 熱電聯產系統 ； 冷熱電聯產系統 ； 電堆並聯系統 ； 負載功率分配系統

並列摘要

Several numerical models are developed for the solid oxide fuel cell (SOFC) system in this study, including SOFC electrochemical, reactive aerodynamic, burner combustion reaction, reforming reaction, and heat exchanger models. An SOFC–combined heat and power (SOFC–CHP) system using CH_4 as fuel is designed on the basis of the aforementioned models. The system adopts a complex waste-heat recovery mechanism that enables the use of most of the waste heat produced by the burner to facilitate the reforming reaction of the reformer and the electrochemical reaction of the SOFC, thereby increasing electricity power output. A set of heat exchanger is installed in front of the SOFC to reduce the difference between the temperature of the reformed gas entering the SOFC and that of the air. Within the proposed waste-heat recovery architecture, two sets of temperature controllers are used to maintain the SOFC operating temperature and to adjust the temperature of the exhaust gas emitted by the burner. Simulation results indicate that the designed SOFC–CHP system can output 4.35 kW of electric power and 6.96 kW of thermal power. In other words, it yields greater power output than that of current SOFC–CHP systems. In another simulation, the reformer, SOFC, and burner in the designed SOFC–CHP system have operating temperatures of 830.8, 973.4, and 1124 K, respectively. Compared to current heat-recovery architecture, the SOFC–CHP system designed in this study can generate power at lower operating temperatures.
Because the exhaust gas emitted by the designed SOFC–CHP system has 6.96 kW of available thermal power, a SOFC tri-generation system is designed to use the heat. The system combines a SOFC–CHP system, an adsorption chiller, and a water tank equipped with a temperature controller. To analyze the performance of the SOFC tri-generation system, this study incorporates the aforementioned dynamic numerical models of the SOFC–CHP system, adsorption chiller, and water tank. On the basis of these models, this study develops a method to determine the capacity of the water tank by using the response time of the operating temperature. Simulation results reveal that the proposed SOFC tri-generation system has an electric, thermal, and cooling power of 4.35, 2.448, and 1.348 kW, respectively. The entire SOFC tri-generation system has an energy-conversion efficiency of 64.9%, and the coefficient of performance of the adsorption chiller is 0.32. In another simulation, changing the SOFC electric power output results in a heat change when the SOFC tri-generation system discharges exhaust gas. However, this heat change do not affect the cooling power output of the adsorption chiller. By contrast, although changing the cooling power output of the adsorption chiller caused a heat change in the exhaust gas discharged by the SOFC tri-generation system, it do not affect the power output of the SOFC. This ideal performance can be attributed to the temperature-control strategies used in this study.
This study focuses on the development of fuel cell parallel strategies and control circuits in large-scale power-generation systems. Large fuel cell systems attain kilowatt- or megawatt- grade power output by using SOFC array stacks; however, few studies have thoroughly discussed the characteristics of parallel SOFC stacks. Accordingly, this study develops a dynamic numerical model for simulating an SOFC stack parallel system. The model integrates reactive gas-supply architecture, electrical parallel characteristics, and a heat-transfer mechanism. The established numerical model resolves the inability of conventional SOFC electrochemical models to make parallel SOFC stacks produce the same output voltage and different output currents. This is because conventional SOFC electrochemical models must first formulate two input variables, namely reaction gas and output current. In a parallel system, these variables cannot be predicted in advance. Therefore, this study uses control theory to integrate the two variables on multiple feedback paths and perform a numerical simulation of a parallel SOFC stack system. The simulation indicates that even when the output current of the parallel stacks is inconsistent or divergent, the output voltage of each stack remains constant. Furthermore, this study explores the possible causes of thermal runaway in parallel systems. In the simulation, when characteristics of the parallel SOFC stacks do not match, the parallel system become unstable because of the slow temperature-transient response, even causing thermal runaway. Moreover, when the parallel system operates at a lower operating temperature, an unstable positive feedback mechanism, which causes thermal runaway, is observed between the SOFC stacks. However, when the parallel system operates at a higher temperature, a stable negative feedback mechanism is observed between the SOFC stacks. This negative feedback mechanism can eliminate the performance difference between the SOFC stacks.
Several SOFC stacks comprise a stack module by using parallel connections. Several stack modules can meet requirements for load-voltage level and high power output by using parallel control circuits, and DC/DC converters are commonly used control circuits. However, the power-generation characteristics of the SOFC differ from those of conventional power systems. Therefore, conventional DC/DC converters are unsuitable control circuits for SOFCs. Accordingly, this study develops a set of suitable parallel-system control methods for SOFC power generation. In addition to meeting the voltage-level requirements of the load, the method can control the output power of each SOFC stack and allocate additional proportional load power to each SOFC stack when load power changes. Therefore, the proposed control method resolves the inability of conventional DC/DC converters to control the power output of each SOFC stack in a parallel system. The method also relaxes the consistency requirement of the parallel-system characteristics for the parallel SOFC stacks. To verify the performance and effectiveness of the proposed control method, simulated and experimental verification are performed. During experimental verification, an electronic circuit capable of exhibiting SOFC power-generation characteristics is designed to increase the reliability of the control method. In the simulation under the rated operating condition, the load power-distribution error is less than 0.79%, whereas the error of the additional load power distribution is less than 2.56% after the load power changed. In experimental verification under the rated operating condition, the load power-distribution error is less than 6.82%, whereas the error of the additional load power-distribution ratio is less than 7.43%. Accordingly, in addition to controlling the output power of each SOFC stack under the rated condition, the developed control method can plan the magnitude of additional load power allocated to each SOFC stack.

並列關鍵字

Solid oxide fuel cell ； SOFC combined heat and power system ； SOFC Tri-generation system ； Parallel-connected fuel cell stack system ； Power sharing control system

參考文獻

[1] Y. F. Xing, Y. H. Xu, M. H. Shi, and Y. X. Lian, ‘‘The impact of PM2.5 on the human respiratory system,’’ Journal of Thoracic Disease, vol. 8, no. 1, pp. 69-74, 2016.

[2] P. M. Fine, C. Sioutas, and P. A. Solomon, “Secondary particulate matter in the United States: insights from the particulate matter supersites program and related studies,” Journal of the Air and Waste Management Association, vol. 58, no. 2, pp. 234-253, 2008.

[3] C. Pu, “Causes of haze pollution under the regional compound environment and legal governance countermeasures,” Nature Environment and Pollution Technology, vol. 16, no. 3, pp. 917-922, 2017.

[4] R. O’hayre, S. W. Cha, W. Colella, and F. B. Prinz, Fuel Cell Fundamentals. Hoboken: John Wiley and Sons, 2016.

[5] Y. Nonobe, “Development of the fuel cell vehicle mirai,” IEEJ Transactions on Electrical and Electronic Engineering, vol. 12, no. 1, pp. 5-9, 2017.