Multi-objective power allocation and heat recovery optimization of solid oxide fuel cells (SOFC) including real time health management for ship power systems

Abstract

System for controlling a marine vessel comprising a number of devices, the devices including a number of Solid Oxide Fuel Cells (SOFC), the system including at least one processor calculating device score values of the devices, the system also including a control unit connected to said devices and being configured to calculate an efficiency score value based on said device score values, the efficiency score value defining the efficiency of the system The system also is configured to calculate the capacity of the system based on the status of the devices, and to calculate a capacity margin index indicating the capacity of the system devices, according to predetermined scores for each operation mode and computing the efficiency and capacity values, the status of the fuel cells including a health indicator index.

Description

[0001] MULTI-OBJECTIVE POWER ALLOCATION AND HEAT RECOVERY OPTIMIZATION OF SOLID OXIDE FUEL CELLS (SOFC) INCLUDING REAL TIME HEALTH MANAGEMENT FOR SHIP POWER SYSTEMS

[0002] Over the last decade, the regulations on emissions of pollutants have become strict. This has led to a new thinking across various sectors of maritime industries. Many transportation and shipping line companies are looking for an energy efficient system with improved fuel consumption with less carbon emissions. This shall be achieved by minimizing the utilization of diesel or gas driven engines or generators referred here as “gensets” and compensating them with renewable energy sources. There are various sources of renewable energies like batteries, photo-voltaic solar cells, wind, fuel cells etc. EP3865335 describes a control system for a hybrid energy system including energy storage system (ESS) generators. A generator system including fuel cells is described in US5532573, However, the challenge lies in integration of these energy sources with traditional ship power management system.

[0003] Among the different types of fuel cells, Proton Exchange Membrane Fuel Cell (PEMFC) and SOFCs are considered as the most promising technologies. The use of SOFC in generator systems is mentioned in CN110077221 and CN1 13346117. Even though, PEMFC have higher power densities compared to SOFC and have better response for dynamic operating ship load profiles, the requirement for purified hydrogen fueling makes it less favorable for maritime industries since considerable storage volumes are required.

[0004] On contrary, high temperature SOFC can produce electric power efficiently from various fuels including Methanol, Ammonia and LNG. Consequently, SOFC are a potential fuel cell technology for medium to long distance shipping. The integration of SOFC into the ship power systems requires the energy management system to be more intelligent considering the inherent limitations of the SOFC such as efficiency, slow transient capabilities, low specific power, slow in capacity change and limited power density at same time as balancing the Energy Storage System (ESS).

[0005] The ship power systems often have high power requirements from the propulsion, mission equipment and other auxiliary consumers and complex energy management system have been provided, for example the intelligent energy management system described in Norwegian patent application N020220197. In such systems multi-stack architecture of SOFC, i.e., considering an assembly of SOFCs in series and / or parallel, are used in order to achieve the desired voltage and current requirements and also to bring modularity and fault tolerance. The association of SOFCs in series and / or parallel requires an appropriate control and energy management system to ensure the reliable and cost-effective solution. The power drawn from these different stack of SOFCs should consider the state of health, operating conditions, heat recovery ability.

[0006] To achieve future green and efficient shipping, several technologies have already been implemented and / or are about to be implemented in the maritime industry. These technologies include technical designs for energy efficiency improvements, light construction material, ship-hull optimization, propulsionimprovement devices, air lubrication systems, energy efficiency measures, all electric ship, multi-energy management and alternative fuel or energy sources namely, energy storage, hydrogen fuel cells, ammonia fuel cells, synthetic methane, biofuels, synthetic diesels, and integration to renewable power.

[0007] According to this invention, these objects are obtained using a multi-objective optimization problem which is proposed for power allocation in a multi-stack SOFC consisting of the minimization of fuel consumption and emissions, extension of state of health including the real time health management, maximization of heat recovery capability and meeting the power demand and surges. Specifically, the objects are achieved as presented in the accompanying claims.

[0008] The solution according to the invention thus provides a system and mechanism that benefits the ship performance both in terms of efficiency, environment impact and safety (redundancy or capacity margin) such as providing recommendations for optimizing power generation using fuel cells. The efficiency margins are related to the performance of each unit while the redundancy or capacity in this case relates to the ability of the system to replace or compensate for an at least partial loss of a unit. In addition to SOFC and ESS and / or generator sets (Genset) a typical power system according to the invention may include power electronics, main engine and shaft generators.

[0009] The system is designed to be autonomous in decision making capability and therefore the balance profile feature is introduced to process the results of cost optimizer before sending the suggested configurations to the control system. The balance profile computes a combined index between the energy efficiency and vessel redundancy / capacity margin. This may be performed according to a predetermined set of rules or be based on analysis of previous incidents and stored events, possibly in combination with user feedback related to the situations.

[0010] The dynamic robustness is provided by a balance profile in terms of energy efficiency and redundancy margin index. An optimal decision is sent from the IT layer to the OT layer when the balance profile feature gives the best combined index, being based on the weighted scores calculated from the energy efficiency components and redundancy margin components. In this manner, an optimal power plant including setup of generator sets such as fuel cells and other renewable energy sources such as waste heat recovery as well as thrusters, propellers can be autonomously configured in the real time by calculating the redundancy and / or load on the system and provide an index indicating the optimal operation of the system.

[0011] The invention will be described below with reference to the accompanying drawings, illustrating the invention by way of examples: Figure 1 illustrates a 2-Split DC bus power system according to the prior art.

[0012] Figure 2 illustrates the system according to the invention.

[0013] A typical power system setup is shown in Figure 1 , where it is possible to connect a number of multistack SOFCs 2-5 to a DC bus 22 and 23 using DC- DC converters 8-11 and breaker 14-17. An AC producer 6 may typically be a variable speed four-stroke diesel engine and / or a two-stroke main engine with a shaft generator is connected to DC bus system 23 using converter 12 and breaker 18. An Energy Storage System (ESS) 1 is connected to DC bus system 22 using DC-DC converter 7 and breaker 13. The drawing shows a number of SOFC stacks 2-5, representing a redundancy if one of the stacks are lost. Similar may preferably be the case for loads and consumers as well as the AC producer 6 and the Energy Storage System, adding safety and redundancy to the system.

[0014] The two DC bus systems 22,23 can be connected together using breaker 20. Each bus systems have its own load from consumers 19, 21 .

[0015] In Figure 2, the connection between different sub-systems of the system under consideration is shown. A fuel supply system 31 delivers gaseous or liquid fuel to the Hotboxes (SOFC stacks) 33, e.g. including the SOFC stacks 2-5 illustrated in figure 1. The term “HotBox” may also be referred to as Balance of Plant (BOP) within the technical field. The fuel pressure- and temperature are controlled and regulated by the system to meet the supply requirements. The fuel supply system may use voyage predictions from an intelligent energy management system (iEMS) 36, e.g. as described in Norwegian patent application N020220197, to prepare and condition fuel delivery (i.e. Boil-Off gas management).

[0016] The SOFC stacks 33 including the SOFC stacks 2-5, comprise the high temperature components of the SOFC system. The SOFC stacks include fuel pre-crackers if required, high temperature heat exchangers for fuel and air and the fuel cell stacks. The performance is monitored by evaluating air and fuel flows, pressures, fuel, air and stack temperatures, voltage and current. As the anode of the fuel cells gets degraded over the lifespan of the fuel cells, the heat production from the system increases and electricity production is lowered for a given fuel supply.

[0017] The heat recovery system 34 produces electricity based on the excess heat available from the SOFC stacks 33. The system can be either thermo-electrical or based on a turbine system. Heat is primarily taken from the process air exhaust, but possibly also from the fuel exhaust side of the SOFC stacks 33. Excess heat from the system can be supplied to onboard thermal consumers if available. The SOFC stacks 33,2-5 are connected to the power system as illustrated in Figure 1.

[0018] The iEMS 36 receives information from the SOFC health model 32 and monitoring of process flows and temperatures in the SOFC stack 33 and heat recovery system 34 to make a prognostic of heat- and heat-to-power production available. Actual production in the heat recovery system 34 is logged by iEMS 36 and used to verify and trim the digital twin model.

[0019] In figure 2 the iEMS 36 receives information from heat recovery system 34, fuel accounting 37, emissions account and planner 39, load power prediction 38, SOFC health model 32 and the power system 35. From fuel accounting 37 the information on fuel cost and operator priority choice is received. The emissions account and planner continuously calculates the emissions profile of the ship based on actual and predicted planned load profile. The load power prediction 38 is connected to a planning system which provides details on the energy demand for a particular voyage or operation. Based on the predicted energy demand the load of the consumers are determined which is transferred to iEMS 36.

[0020] A SOFC health model 32, deploys a digital twin which performs real time monitoring and health prognostics of the fuel cell. It monitors the subsystems such as the fuel supply system 31 , SOFC stack 33 and heat recovery system 34. For the subsystem defining the SOFC stacks 33, the air and fuel flows, cell and stack voltage, current, temperatures and power production are monitored. In the fuel supply subsystem 31 , the temperatures and pressures of air supply, mass flow of air, fuel and fuel recirculation flow are monitored. In heat recovery system 34 the process and exhaust exchangers with heat to power system, if installed, are monitored.

[0021] In addition, the SOFC health model subsystem 32 utilizes relationship between,

[0022] 1 . The fuel cell stack voltage, current, temperatures as a function of power output, fuel and air supply,

[0023] 2. Pressure drops over the stacks,

[0024] 3. Process air exhaust temperatures as a function of power generation,

[0025] 4. Available heat production and possible waste heat to power production.

[0026] These relationships are used to determine a dynamic efficiency value that changes according to the fuel cell health. The running hours of SOFC stacks 33 are used to adjust this efficiency as per the baseline efficiency curve.

[0027] In iEMS 36, a multi objective optimization function is implemented. The objective is to meet the power requirements from load power prediction 38 and power system 35, minimizing the total operative cost by regulating the stack degradation and replacement costs using the SOFC health model 32, emissions costs provided by the emission account / planner 39, heat recovery benefits provided by the heat recovery system 34, fuel costs provided by the fuel accounting system 37, fuel availability provided by the fuel supply system 31 , and the availability of stacks from in the SOFC stack system 33. This results in score values relating to efficiency and redundancy / capacity defining constraints which are weighted according to operational mode and operator priorities.

[0028] The weighted data provided by the units discussed above may be fed to the iEMS 36 which, provide a score value for efficiency and redundancy or capacity based on the data received from the different parts of the system, where the score values may be calculated based on the received data or chosen or corrected by an operator. The score values represent the performance of the devices based on the sensor measurements and known characteristics of each device. The information from the SOFC health model may be especially important for calculating the weighed efficiency score values for the SOFC stacks. This way the score value regarding an inefficient fuel cell may be matched with the redundancy or capacity of the fuel cell so that for example a weak cell getting a low score from the health model may through the iEMS process be compensated for if the number of operating additional fuel cells is sufficient and / or the capacity of the remaining fuel cells can be increased accordingly.

[0029] According to the invention as presented in figure 2 the system is capable of handling limitations with respect to SOFCs where the ESS and / or Genset is a vital part of the system. For an optimal use in terms of health management and power system levels the following should be taken into account: a. The load shift capability of SOFCs should be significantly limited in order to prevent the thermomechanical stress affecting the cells and to account for fuel reforming and the thermal inertia of the system. b. There is no benefit to operate SOFCs at very low load periods, c. SOFCs in standby (idling) that are kept at high temperatures also have a long start-up procedure (several minutes).

[0030] Based on the above mentioned limitations and with inputs from other systems referred to in figure 2, the iEMS will perform the necessary computations to decide on minimum load limit for SOFCs, load shift capability of SFOCs and number of SOFCs to be in idling based on the present and predicted power demand.

[0031] The system according to the invention will have different operation modes including different setups and combinations of different devices in the system, thus having different efficiency and redundancy. The scores for the individual energy efficiency and redundancy / capacity margins may be mean values of the individual scores for each element, but may be weighted differently, either automatically through analysis of previous situations or by an operator. As discussed in N020220197, the score values for the devices or parameters of the system related to efficiency are used to calculate a weighted score related to efficiency and the score values related to the redundancy or capacity of the system are used to provide a weighed score related to the redundancy / capacity, where the weighted scores may be calculated based on the mean value of the related scores. A combined index value is determined by setting a base and target value for energy efficiency and redundancy / capacity margin scores. Then the two scores are mapped to a common index range and dynamic weights are assigned for them according to operation mode. Finally, the weighted score values for each score are summed up. The weights may be predetermined or dynamically adjusted based on machine learning algorithms and continuously samples information related to the performance of the system.

[0032] The iEMS will receive a number of different setups or operation modes with corresponding score values for each setup. These score values may be compared for selecting the optimal setup depending on the preferences in the specific situation and the health, efficiency and redundance / capacity of the system. If the weighted score values in one operation mode exceeds a predefined or calculated limit, the iEMS may calculate corresponding values for other operation modes and chose a new operation mode showing better efficiency, health, capacity and / or redundancy.

[0033] Based in this a preferred load reference may be calculated for each of the connected SOFC stack 2-5 and the information is transmitted to the Power System 35.

[0034] In addition to the efficiency and redundancy / capacity of the fuel cells the iEMS may be set to handle relevant score values from other parts of the system such as consumers and additional generators or energy storage devices.

[0035] In general, the present invention achieves a multi objective optimization to operate systems including multi-stack SOFC for optimally meeting the power requirements from real time and prediction system minimizing the maintenance, fuel and emission costs and maximizing the health of SOFC. This is based on calculation of the real time dynamic efficiency of a set of SOFC stacks 33 which is calculated based on health model and relationship from baseline efficiency. A causal relationship is established between the heat production and the heat recovery system which is used for providing score values adjusting the weight factor of the constraint in the optimization function.

[0036] A digital twin model for the SOFC health and condition may be used for monitoring with real time prediction on health score. This may be used for computation of efficiency score value of the SOFC stacks as well as the redundancy and / or capacity score values indicating how vulnerable the system is to failures of a SOFC stack or other devices in the system. This may be obtained by comparing the different operation modes as described above using the iEMS 36 comparing the operation modes and choosing a mode being within the chosen range of the combined index value. Based on this the system is able to control the risk of failure based on the efficiency or risk of failure of each component based on the health and efficiency of the devices as well as the number of alternatives in case of failure.

[0037] An emission account planner may be used which computes the accounted emissions and emissions based on the prediction. This information may also be used to adjust the scheduling and managing of power requests. A heat recovery system may also be added which predicts heat generation and available heat for recovery to electrical power using a ThermoElectric Power unit.

[0038] To summarize the present invention relates to a system for controlling a marine vessel. The system comprising a number of devices, the devices including a number of Solid Oxide Fuel Cells (SOFC) where the fuel cells may be organized in stacks each including a number of individual cells.

[0039] The system also includes at least one processor calculating device score values of the SOFCs and other devices where the score values are related to the redundancy of each of the devices, indicating the vulnerability if one device fails as well as the reliability and / or efficiency of each device or fuel cell / fuel cell stack. The system also includes a control unit connected to said devices and being configured to calculate an efficiency score value based on said device score values, the efficiency score value defining the efficiency of the system. The system is configured to calculate the capacity of the system based on information about the status of the devices being relevant to the operation of the system, and to calculate a capacity margin index indicating the capacity of the system devices using the predetermined scores for each operation mode and computing the efficiency and capacity values. The status information about the fuel cells includes a health indicator index. In addition, a combined index is calculated based on the efficiency score value and capacity margin index.

[0040] The system includes a set of setup configurations or modes for different configurations and combinations of the devices included in the system and is configured to, at the instance of the combined index being outside a chosen predetermined range, calculate the combined index for at least one alternate device setup and to indicate the preferred setup based on a comparison between the combined index values.

Claims

Claims1 . System for controlling a marine vessel, the system comprising a number of devices including a number of Solid Oxide Fuel Cells (SOFC), the system also comprising at least one processor calculating device score values of the devices, the system also comprising a control unit connected to said devices and being configured to calculate an efficiency score value based on said device score values, the efficiency score value defining the efficiency of the system, wherein the system also is configured to calculate the capacity of the system based on information from the devices being relevant to the operation of the system, and to calculate a capacity margin index indicating the capacity of the system devices, according to predetermined scores for each operation mode, and computing the efficiency and capacity values, and wherein the information from the fuel cells includes a health indicator index, the system also being configured to calculate a combined index based on the efficiency score value and capacity margin index, and the system includes a set of setup configurations for the devices included in the system, the system being configured to, at the instance of the combined index being outside a predetermined range, calculate the combined index for at least one alternate device setup and indicate the preferred setup based on a comparison between the combined index values.

2. System according to claim 1 , wherein the health indicator index of each fuel cell is calculated based on real time measurements of selected parameters, including at least one of heat recovery ability, fuel consumption, emissions, operating temperature and / or power output.

3. System according to claim 2, wherein the health indicator index calculation for each fuel cell is based on comparing said real time measurements with a predefined model specifying a preferred operationmode.

4. System according to claim 1 , wherein the health indicator index is based on the relationship between:- The fuel cell stack voltage, current, temperatures as a function of power output, fuel and air supply,- Pressure drops over the stacks,- Process air exhaust temperatures as a function of power generation,- Available heat production and possible waste heat to power production.

5. System according to claim 1 , wherein the SOFC power cells constitute at least one multistack SOFC, and wherein the capacity of the multistack power cell represents the redundancy of the SOFCs in the stack, the system, being configured to allocate power to the SOFC cells in the stack based on the dynamic efficiency of the cells.

6. System according to claim 1 , where in the calculating device score values are based on at least one of real time health / availability, dynamic efficiency and / or fuel cost / emission cost.

Images