A method for controlling a green hydrogen production system

Abstract

A method for controlling a green hydrogen production system (100; 100'), comprising geographically distributed power generating nodes (10, 300; 300') each having at least one node center (320; 320.1, 320.2, 320.3, 320.4) and at least one electrolyzer (13) for generating green hydrogen within the system from the produced electrical energy, wherein each the power generating node (10, 300; 300') comprises multiple PV units (12; 312) and multiple wind turbine generators (WTG) (11; 301...316) as power generating units and wherein the multiple wind turbine generators units (WTG) (11; 301...316) are located in geographically dispersed sites surrounding the node center(s) (320; 320.1, 320.2, 320.3, 320.4), wherein the installed capacity (IC) of the electrolyzer (13) and all other energy consuming devices in the system is smaller than the sum of maximum capacities (MG) of all PV units (12; 312) and wind turbine generators (11; 301...316) available for operation together, wherein the method comprises at least the following steps: a) an energy demand value (EDV) of electrical power required for constantly operating the electrolyzer and other consumers is defined wherein EDV < IC; b) weather conditions in proximity of the power generating units and in windward direction of the PV units (312) are constantly monitored; c) based on weather conditions acquired from monitoring, an expected energy yield value (EEY) is calculated separately for each type of power generating unit and / or for each individual power generating unit; d) an individual workload is assigned to both the PV units and the WTG units (301...316) wherein the WTG units (301...316) designated to be operated in generator mode are selected to build a cluster (361, 362) of WTG units (301...316) which are set operative to compensate shortage of electrical energy when at least one PV unit is expected to be at least partially shaded.

Description

[0001] A METHOD FOR CONTROLLING

[0002] A GREEN HYDROGEN PRODUCTION SYSTEM

[0003] The invention relates to a method for controlling a green hydrogen production system.

[0004] When using renewable energy sources in the form of solar and wind energy, there is always the problem that the energy yield is dependent on the current position of the sun and the local weather conditions like cloud coverage, wind speed etc. The coupling of photovoltaic (PV units) power systems and wind turbine generators (WTG) to form hybrid power plants can reduce these dependencies, and the construction of the plants can be targeted at locations where fewer weather-related losses are to be expected. Thus, even those kind of processes where short-term fluctuations in the supplied electrical power are disadvantageous can be based on renewable energy supply. One such vulnerable process is the electrolysis to produce green hydrogen gas. Here, shortterm power drops can disrupt the process and might damage the electrolyzer or lead to hydrogen cross-over and in worst case, if safety measures are not followed, explosion risk due a too high concentration of hydrogen at the electrolyzer anode and mix with oxygen. Electrolyzer can be shut down in a controlled process e. g. before dawn within less then 2 minutes to a hot-standby mode.

[0005] Feeding such electrolyzer with hybrid energy generation systems, which include photovoltaic power systems and wind turbine generators simultaneously, reduces the risk of short-term power drop. For this purpose, it is necessary to install wind turbine generators with such a capacity that losses in energy supply due to temporary shading of the PV units by clouds can be substantially compensated by wind energy. If even a complete failure of the photovoltaic system must be considered, the same capacity needs to be installed in form of wind turbine generators. Conversely, the capacity of the installed photovoltaic systems must be sufficient to be able to substantially feed the electrolysis process alone when there is no wind at all. Providing the full capacity with both WTG and PV units results in high investment costs.

[0006] To be able to produce green hydrogen economically, a widely dispersed installation of several photovoltaic power systems and wind turbine generators is necessary in a favorable geographical location where there is no public AC grid with further power plants as a backup to bridge short-term demand periods, and also no electric energy consumers are available to buffer production peaks.

[0007] It is an object of the present invention to provide a method for controlling a hybrid renewable energy power plant in which the economic use of the two energy sources is controlled in a targeted manner to ensure a constant current supply to the electrolysis systems.

[0008] This object is achieved by a method for controlling a green hydrogen production system with the features of claim 1 .

[0009] A prerequisite for carrying out the method according to the invention is that a hybrid energy generation plant is used in which the two different energy sources, namely photovoltaic power systems and wind energy plants, are present in sufficient numbers and are arranged in such a widely distributed manner in a wide area that they can be selected zone by zone and / or can be combined into groups. All photovoltaic power systems and wind turbine generators are interconnected with power lines to form a local area power network.

[0010] The method of the invention is based on the following assumptions:

[0011] - All cases for controlling the power sources in the hybrid power plant consider times of daylight to allow for PV units electricity generation. - All cases consider wind speeds above the wind turbine generator ’s cut in wind speed so that electricity can be produced from wind energy.

[0012] - The installed capacity (IC) of the electrolyzer and all other potential electrical consumers mid- and downstream in the local area power network together is smaller than the capacity of the sum of WTG- and PV units so that either PV units and / or wind turbine generators would need to be reduced / curtailed during times of rated wind speed and full PV units’ production.

[0013] - The PV unit’s production in different nodes does not impact the other nodes in vicinity whereas wind turbine generators create wake at their leeward side so that wind turbine generator located leeward of another wind turbine generator in operation will be operating less efficient even if minimum distances of e.g., 10, preferably of 5 rotor diameters among them are kept.

[0014] - Wind turbine generators’ wear and tear and aging is depending, among others, on the hours of operation. A wind turbine generator that is curtailed or shut down in idle mode shows less wear and tear than a wind turbine generator in permanent operation. Solar PV units aging and wear and tear is less dependent on the level of utilization.

[0015] - Wind turbine generators can easily be curtailed or shut down by pitching the blades whereas solar PV units cannot easily be curtailed. When exposed to sun and not producing electricity the panels get even warmer. Single axis tracking could contribute to reduction of electricity generation. The PV can intentionally be turned out of the sun if no electricity production is wanted, and heating of PV shall be reduced.

[0016] Taking into account the above considerations, solar PV units should be primarily used in case of too high upstream power and wind turbine generator should be curtailed first.

[0017] In the method of the invention an energy demand value (EDV) is defined with which the necessary electrical input power for the electrolysis process is defined. The energy demand value includes necessary electrical input power for any other electrical auxiliary devices on the site if any.

[0018] In particular, this energy demand value is not kept static over the entire operating period but is adjusted continuously or at regular intervals of between 10 seconds to 1 minute, e.g., depending on the position of the sun, because as mentioned before long-term fluctuations in the input power for the electrolysis process are still possible whereas short-term fluctuations disturb the electrolysis process or cause damage to the electrolysis system. For example, two hours before dawn the EDV can be decreased continuously during the shutdown process of the electrolyzer. The energy demand value (EDV) is always smaller than the installed capacity (IC) so that there are reserve units available to bridge short-term power drops.

[0019] The weather conditions in the vicinity of the power generating units, i.e., the photovoltaic power systems and the wind turbine generators, are always observed. Especially the weather conditions in windward direction (luv) of the PV units are constantly monitored and forecast data like cloud coverage, cloud height and wind speed on ground and wind speed in height of the clouds are gathered.

[0020] Based on the forecast data obtained through the weather observation, an expected energy yield value (EEY) is calculated for each type of power source for a forecast period starting multiple days ahead with a limited probability and being constantly optimized to periods of seconds at very high probability.

[0021] Digital twins of the PV units and the WTG are used to perform this calculation. So, the energy yield value (EEY) is the result of a simulation of operating the power generating units for a future period by using the probable weather data for this period.

[0022] The calculation of the energy yield value (EEY) starts multiple days ahead and it is based on weather and wind forecasts only at the beginning. The energy yield value is then continuously adjusted with increasing accuracy by more precise forecasts and on-site measurements incl. all-sky cameras.

[0023] Besides the probable difference between the EEY and the EDV is determined.

[0024] In terms of both investment costs and maintenance costs relative to the installed nominal power, photovoltaic systems are significantly cheaper than wind turbine generators. As long as the EDV of the electrolysis process can be covered by solar energy alone, all PV units are operated at full load and all wind turbine generators are kept in a stand-by mode to reduce mechanical wear and to extend lifetime.

[0025] Conversely, wind turbine generators are used only when the EEY for the photovoltaic plants falls significantly below the energy demand value EDV or even drops to zero. A significant shortage of electrical energy which needs to be compensated by measures actively taken in the control method is given if:

[0026] - the EEY is expected to fall below EDV for at least 1 min and

[0027] - the difference between EDV and EEY is greater than 1 % of EDV; typically, this is equal to about one half of the rated power of a WTG.

[0028] If it there is an expectation based on previously made simulation and calculation that the EEY of the photovoltaic systems will be reduced soon and their power will no longer be sufficient to fully meet the EDV, then wind turbine generators are additionally used to generate additional power but are curtailed to minimum load such that only the difference between the EDV and the EEY of the PV units is provided by WTG units.

[0029] If, based on the weather observation, it is expected that the EEY will only be reduced for individual photovoltaic systems due to passing cloud fields, the known wind direction and wind speed on the one hand and the equally known topography of the photovoltaic systems on the other hand can be used to precisely calculate which of the photovoltaic systems within the energy generation system will probably be shaded at which time. Accordingly, several wind turbine generators are ramped up in good time before the occurrence of the shading, with which the reduced expected energy yield of PV units can be compensated. By regulating the blade position of the commissioned wind turbine generators, a fine adjustment of generated electrical power to the temporary changes in energy demand is possible.

[0030] A special feature of the invention is not simply to use all wind turbine generators available in the hybrid power plant at once, but to select individual wind turbine generators from the full set of installed wind turbine generators for generator operation. This reduces the mechanical wear of the wind turbine generators as a whole.

[0031] Furthermore, the selected wind turbine generators are operated in a cluster, preferably in a migrating cluster. This means that all the wind turbine generators present at the site are grouped together in temporary clusters that extend across the expected wind direction. The same wind turbine generators are not permanently assigned to a certain cluster but are constantly reorganized in new clusters depending on the weather observation data. On the one hand the selection is made such that a sufficient number of WTGs in generator mode is provided to compensate power loss of PV units. On the other hand it is preferred that only those wind turbine generators that maintain a sufficient distance of at least 5 times the rotor diameter from each other when viewed in the wind direction are selected for generator operation from the full set of WTGs available.

[0032] If, for example, there are 16 wind turbine generators in an approximately regular arrangement of four rows and four columns and the wind direction is perpendicular to one side of this arrangement, it is possible to use only the first and third rows or only the second and fourth rows for power generation or to shift from operation with first and third row to operation with second and fourth row. This reduces mutual interference between the wind turbine generators on the one hand and increases the service life of the wind turbine generator on the other, since only half of the wind turbine generators are used at once. The arrangement in linear rows and columns is exemplary only. In a real large scale hybrid power plant, which extends over an area of at least 500 km2or even about 10.000 km2the positions of wind turbine generators are determined individually according to the geographical conditions. Nevertheless, such arrangements can be divided into rows and columns even if they do not extend linear in a geometrical sense.

[0033] If a wind field passes through the area of the hybrid power plant within a short period of time, it can be predictively planned to operate the rows of wind turbine generators like a travelling wave, i.e., all wind turbine generators located in the direction of passage of the wind field are switched from idle mode to power generating operation as soon as cut-in wind speed is given but before the main wind field arrives. Blade control then assigns the highest load to the first row of WTGs first, then to the second row, while the load of the first row is already reduced again, then to the third row and so on.

[0034] If the wind direction changes, the wind turbine generators are re-clustered, i.e., combined into other groups that are aligned across or transverse to the wind direction.

[0035] Since the direction of cloud movement usually corresponds to the wind direction at a height relevant for wind turbine generators, it is always possible, if photovoltaic power systems and wind turbine generators are installed next to each other, to form several clusters of wind turbine generators in a shape that corresponds to the shape of the moving cloud fields. The clusters of wind turbine generators are then operated one after the other in the expected direction of cloud migration, so that in a first cluster the wind turbine generators are already in generator operation even before the shading of the photovoltaic systems occurs. The next clusters of wind turbine generators , viewed in the wind and cloud moving direction, are then put into operation one after the other before the respective neighboring photovoltaic systems are going to be shaded, and the wind turbine generator clusters are taken out of operation again as soon as the shading of the PV units has ended. The cluster built by the method of the invention can sometimes be just a static group of individually assigned WTGs but more often is to be understood as a migrating cluster i.e. a sequence of multiple different groups of WTGs to be put in generator mode or released from that state, wherein the multiple groups are activated and inactivated one after the other.

[0036] Another aspect of the invention relates to periods in which WTG units are used only or additionally for power generation i.e. for cases (ii) and (iii) of the method of the invention some WTG units located at an windward direction are selected for the cluster to operate with higher priority or have a higher individual workload assigned than other WTG units from the whole set of WTG units available for operation located at a leeward direction. Besides

[0037] According to another aspect of the method of the invention in cases (ii) and (iii) some WTG units located closer to the electrolyzer are selected to operate in generator mode with higher priority or have a higher individual workload assigned than other WTG units that are farther away from the electrolyzer.

[0038] According to an aspect of the method of the invention a computerized service life log is kept so that in similar wind conditions, not always the same wind turbine generators are used but rather those are elected for operation e.g., that have had the fewest operating hours so far. In the example described above, it could be planned to prioritize rows 1 and 3 during a few days of operation - assuming constant wind direction - and then rows 2 and 4.

[0039] Alternatively, to preferably shutting down the WTG which are most fatigue loaded to increase their lifetime, those WTG can be selected which have already endured high load and will be subject for replacement next.

[0040] In the service life log various parameters can be recorded like overall service hours, hours in idle mode, hours in generator mode. Besides it is possible to use digital twins of the WTG to calculate fatigue of certain parts like shafts, clutches etc. based on recorded wind data. According to another aspect of the invention the selection of WTGs for the cluster in which the WTGs are run in generator mode a fatigue and wear indicator (FWI) is calculated for each individual WTG unit based on weather conditions acquired from monitoring. An individual workload assigned to each WTG unit is set such that the fatigue and wear indicators (FWI) for each WTG unit in the green hydrogen production system (100) remain below a general or below an individual threshold. The FWI can be defined as a dimensionless value referring to the quality of load like light, normal or heavy load.

[0041] An individual threshold can also be set in relation to the consumption of machine hours over the total service life of the machine based on the expected load.

[0042] Furthermore, an individual threshold can be set as a maximum operational torque to avoid damage of mechanical parts due to high load operation.

[0043] Such individual thresholds are stored in the service life log of each WTG and retrieved during the simulation using a digital twin of each WTG available in the green hydrogen production system.

[0044] An example for carrying out the method of the invention is described with reference to the flow charts in figures 16 to 17.

[0045] In a start phase illustrated in the upper half of figure 16 a start parameter is required on which the further steps will be based. The amount of electrical energy required for constantly operating the electrolyzer or a set of multiple electrolyzer is defined as the energy demand value (EDV). Basically, this value is constant for the electrolyzer(s) installed in the monitored power plant or the monitored sector thereof. However, it may vary due to e.g. change in ambient temperature. The start parameter can be set by user interface input or can by using previous data as long as the hardware scheme has remained unchanged. Having the method started, the weather conditions in the monitored area of the power plant and its vicinity are constantly monitored, at least in an windward area of the WTGs locations.

[0046] Based on the weather conditions gathered from weather monitoring and / or forecast the expected energy yield value (EEY) is calculated for each type of power generating unit within the area controlled and monitored by the method. The EEY is the forecast value for electrical energy expected to be produced.

[0047] From a comparison between the energy demand value (EDV) on the one hand and the summed up EEYs of all PV units it becomes evident whether the yield of solar anergy in the plant will be sufficient or not:

[0048] - If the solar energy yield is expected to be sufficient all PV units will be operated at full load whereas the WTG units will remain idle; running WTGs are stopped so they are not part of the energy supply scheme during the calculated forecast period anymore.

[0049] - If the solar energy yield is zero like at night or very small in comparison to the expected energy yield value of the WTGs then the energy supply for the electrolyzer(s) will be based on the WTGs only.

[0050] - In the third alternative neither the expected energy yield of the PVs is sufficient to run the electrolyzer(s) alone nor is the expected energy yield of the WTGs.

[0051] The last alternative is more complex but will occur more often in reality than the other cases. Pursuant to the invention the use of PVs is prioritized over the use of WTGs as a matter of principle so if additional WTGs are required to support the supply of electrical energy to the electrolyzer then their number is kept low and / or the WTGs running are curtailed to minimum load.

[0052] For the most effective use of WTGs in terms of maintenance costs and machine lifetime a WTG cluster is built according to specified criteria.

[0053] To find the best WTG cluster for operation in the next operation period a cost- benefit-ratio is calculated for at least two different cluster models whereby digital twins of each WTG are used so that the optimization considers individual properties of each WTG. In the example illustrated in figure 17 four cluster models are defined and compared:

[0054] - According to a first potential cluster model those WTGs are selected from the plurality of WTGs installed in the power plant which are arranged in rows across or transverse to the wind direction.

[0055] - According to a second cluster model WTGs are selected from a node area which has a similar size and shape as the moving cloud field discovered in previous whether monitoring.

[0056] - In another cluster model those WTGs are selected which have had the least operating hours since their installation.

[0057] - In still another cluster model such WTGs are selected which have the least operating hours remaining before end of life or before the end of the maintenance interval.

[0058] As a result of the calculation the cluster model with the best cost-benefit-ratio is selected and the corresponding WTGs in the field are set to generator mode.

[0059] For building cluster models to be compared to each other and then used for operation if suitable the following considerations are made:

[0060] - Wind conditions like wind speed, wake, turbulence, wind shear etc. at each WTG location.

[0061] - Temperature and solar irradiation can impact ambient temperature at the WTG site. A WTG in the shadow of a cloud can work more efficiently in hot climate and vice versa in cold climate where the wind / solar hybrid may not often be chosen. WTG in the sun and very high temperatures or WTG outside of the sun and very low temperatures might need to be curtailed or shut down.

[0062] - Operational lifetime of each WTG: which is the number of operating hours of the WTG? This could lead to the conclusion to level out the operating hours per WTG or to run some WTG harder for earlier replacement than newer WTGs. - When is next planned maintenance due? Depending on the service staff schedule and spare parts available the date can be extended or accelerated.

[0063] - Which is the current health status of each WTG and which might be the impact on it of the next operation period? Is there a component, which requires the WTG to be operated in a curtailed mode, due to a damage or wear and tear?

[0064] - Are there any known issues with a particular WTG so that not all installed WTGs in a certain area will be available?

[0065] - In densely populated areas further considerations might refer to the location of the unit regarding noise receptors / shadow flicker or other environmental constraints. Can disturbances of the environment be avoided if specific WTG are kept shut down longer or more often.

[0066] - Which is the distance of WTGs to the consumer i.e. the distance of the electrolyzer to the node center with which the WTGs are connected. The shorter the distances are the lower the electrical transmission losses will be.

[0067] - Consideration of Balance of plant (BOP): for building a WTG cluster it is important to know if there are any maintenance or repair activity is planned at a certain part of the wind farm. In case for example an overhead line to certain WTGs needs maintenance, then these WTG should not be considered to be included in any cluster or at least not be considered with priority.

[0068] Based on these considerations and input data the cost of electricity can be calculated for each WTG via a digital model. The outcome of this calculation will create a ranking of the WTGs with the lowest electricity costs in the present conditions.

[0069] The controller would then select the WTGs with the lowest electricity generating costs until the required capacity as defined by the energy demand value EDV is reached. If the wind and / or solar conditions are fluctuating a lot or are highly intermittent, then excessive energy can be stored in a battery energy storage system (BESS) or vice versa. Missing energy can be quickly discharged from a BESS at any time. Thus, the BESS balances the changes in EYV and EDV by using a battery instead of a constant turning WTGs on and off.

[0070] Alternatively, WTGs can be kept in operating but run in a curtailed mode e.g. by intentionally increased pitch angle.

[0071] In the aforementioned example of the control method as illustrated in figure 17 a simple form is used as many cluster models are compared at once and one of them is selected to become operative.

[0072] In an alternative method which is illustrated in figure 18 the definition and selection of the cluster model is more complex: first a subset of WTGs is selected in that area which is affected by shading or other weather impacts and in a second step some WTGs are selected from this subset according to any other criteria as mentioned above like lifetime or wear etc. This multi-level approach is favorably applied in large plants where the number of available WTGs is greater than the number of WTGs required.

[0073] Whereas in the example illustrated in figure 17 four cluster models are defined and compared in parallel, an alternate subroutine illustrated in figure 18 of the method provides a first set of criteria according to which a benefit-ratio is calculated. This first set of criteria is weather based. The result is a large subset of WTGs suitable to be clustered to generate auxiliary power.

[0074] In a second stage another set of criteria is used to make the final selection of WTGs to be put into operative mode from the first subset. The second set of criteria relates to lifetime, maintenance periods or wear and tear.

[0075] The two-step approach illustrated in figure 18 is favorable for large power plants with a large number of available WTGs. By this approach the built cluster is the result of both weather-based aspects and lifetime aspects. The invention will be explained in more detail below with reference to the exemplary embodiment shown in the drawings. The figures show:

[0076] Fig. 1 a first example of a topography of a green hydrogen production system;

[0077] Fig. 2 a second example of a topography of a green hydrogen production system;

[0078] Fig. 3 an exemplary power generation node in top view;

[0079] Fig. 4 an exemplary energy utilization node in top view;

[0080] Fig. 5 a functional diagram of a green hydrogen production system according to a first embodiment;

[0081] Fig. 6 a functional diagram of a green hydrogen production system according to a second embodiment;

[0082] Fig. 7 to 15 a schematic representation of the topography of a node of a hybrid power plant with wind and moving clouds;

[0083] Fig. 16 to 18 a flow chart of a method for controlling a green hydrogen production system;

[0084] Fig. 19 a cluster of wind turbine generators used in fig. 7 and

[0085] Fig. 20 a cluster of wind turbine generators used in fig. 9.

[0086] Figure 1 shows a first example of a green hydrogen production system 100, which is distributed over a coastal region 1 of a country. In the example, the maximum extension in the north-south direction is about 50 km and in the eastwest direction about 250 km. The green hydrogen production system 100 comprises a total of 24 power generation nodes 10.1 , ..., 10.24 located upstream and one power utilization node 30 located downstream, at a sea 2 or lake. The lines between the power generation nodes 10.1 , ..., 10.24 indicate area boundaries of each node. The dots within the boundaries each represent a wind turbine generator 11 , each of which is electrically connected to the respective adjacent power generation node 10.1 , ..., 10.24 in the area.

[0087] The power generation nodes 10.1 10.24 and the power utilization node 30 are interconnected by a network of an interconnection line 20 comprising at least one hydrogen pipeline. At the same time, especially in previously undeveloped parts of the region 1 , the route built for the interconnection line 20 can be used to construct a parallel roadway.

[0088] A main line of an interconnection line 20 extends from the far north-east power generation node 10.1 , ...,10.24 to the power utilization node 30. The connection of the individual power generation nodes 10.1 , ...,10.24 to the interconnection line 20 can be made in various ways:

[0089] - The easternmost power generation nodes 10.1 , 10.2, 10.3, 10.4 are connected directly to a main line of the interconnection line 20.

[0090] - Some power generation nodes, such as the westernmost power generation nodes 10.21 , 10.22, 10.23, 10.24, are interconnected in a group of four by a branch interconnection line 29.1 running in a north-south direction. At a crossing point, the branch line is connected to the main line. In addition, another group of four nodes is connected by a branch interconnection line 29.2 running in a north-south direction.

[0091] - At the power generation nodes 10.9, ...,10.12 a Y-shaped branch interconnection line is provided. The power generation nodes 10.9, 10.10 are connected to the power generation node 10.11 , from which the branch interconnection line extends to the power generation node 10.12. At the intersection of the branch interconnection line 29.4 with the main line of the interconnection line 20, the connection is made.

[0092] Figure 2 shows a second, smaller and simplified example of a green hydrogen production system 100' which is also established in the coastal region 1 of the country located near the sea 2. It consists of ten power generation nodes 10.T, 10.10’. Each power generation node 10.1 ’, ..., 10.10’ is associated with an area of about 20 km x 20 km. A plurality of wind turbine generators 11 , each represented by a dot, are arranged in the area of the nodes 10.1 ’, ..., 10.10’ and are electrically connected to their associated node. In this example, each power generation node 10.1 ', ...,10.10' is connected via its own spur line to a main line of an interconnection line 20' extending from east to west to a power utilization node 30 at sea 2.

[0093] Figure 3 shows an example of a single power generating node 10 in plan view. This is designed so that several central units are placed within an arrangement of four rectangular arrays of PV units 12. The central units include an electrolyzer 13, a battery module 16, and a central station 19 in which controls and accommodations for personnel, among other things, are located. The PV units 12 are electrically connected to each other and to the central units. The wind turbine generators 11 associated with the power generation node 10 are also electrically connected thereto; the electrical connection lines extending from the wind turbine generators 11 to the central units are shown as dashed lines in Fig. 3.

[0094] The central units are positioned along an interconnection line 20 that includes a hydrogen pipeline 22 shown as a dotted line, a freshwater pipeline 21 shown as a dash-dotted line, and a high voltage electrical power line 23 shown as a solid line. A roadway 24 shown as a double line is located along the interconnection line 20.

[0095] Figure 4 shows an example of a downstream energy utilization node 30 in top view. It is a complex of facilities built near the sea 2. The energy utilization node 30 includes a port 31 , a seawater desalination plant 32, and an ammonia plant 33. In addition, other facilities are located in a central station 35. The energy utilization node 30 is connected to the network of interconnection lines 20 and parallel roads 24. Hydrogen produced in the power generation nodes 10 is pumped through the hydrogen pipeline 22 to the ammonia plant 33. The liquid ammonia produced there is loaded onto ships at the port 31. The seawater desalination unit 32 obtains seawater from the sea 2 via a tap line 36. The fresh water obtained from this is, on the one hand, supplied to the ammonia plant 33 as cooling water and, on the other hand, pumped upstream via the freshwater pipeline 22 within the connecting line 20 to supply the electrolysis plants in the nodes 10.

[0096] Without reference to any possible topography, Figure 5 shows a functional diagram of a simple form of power generation plant 100', such as that shown in Figure 2. Two functionally identical power generating nodes 10' are shown in the left panel, each comprising:

[0097] - a plurality of wind turbine generators 11 ;

[0098] - a plurality of PV units 12;

[0099] - an electrolyzer 13; and

[0100] - an on-site electrical distribution network 14 through which the wind turbine generators 11 , PV units 12, and electrolyzer 13 are connected within the node.

[0101] For example, each node 10 is configured such that the connected wind turbine generators 11 and PV units 12 generate 1 GW of power at peak, with most of the power being used in the electrolysis process. In an exemplary configuration, the total renewable power generation is divided into similar nodes 10 of 1 GW each. Each node 10 consists of the following elements:

[0102] - 500 MW wind

[0103] - 500 MW solar

[0104] - 650 MW of water electrolysis co-located with the arrays of PV units 12.

[0105] The power generating nodes 10' are each connected to a connecting pipeline 20', which includes a hydrogen pipeline 22'. Through this pipeline, the hydrogen extracted from the electrolyzer 13 of the node 10 is delivered downstream. It can be used:

[0106] - to produce hot briquetted iron (HBI) from iron ore 40 in an HBI plant 34,

[0107] - to produce ammonia in an ammonia plant 33, which can be liquefied and easily transported by ship through a port 31 . - to be stored in tanks 37 or

[0108] - to be directly liquefied and transported by ship.

[0109] Also associated with the energy utilization node 30 is the seawater desalination unit 32, through which fresh water is pumped upstream through the freshwater pipeline 21 to the node 10.

[0110] Hydrogen consuming units may be installed along the interconnection line 20 or are centralized in the downstream node 30. For example, to operate a seawater desalination unit 32 operating on the principle of reverse osmosis, engines fed with hydrogen or ammonia can be used.

[0111] In the embodiment of a simple green hydrogen production system 100' shown in Fig. 5, the interconnection line 20’ does not include a high voltage electrical line. Thus, the energy produced in the nodes 10’ is only conducted through the hydrogen pipeline 22’ in the form of the locally produced hydrogen. The only electrical connections in this green hydrogen production system are provided by the on-site electrical distribution network 14 within each node 10.

[0112] Figure 6 is a functional diagram of a more complex green hydrogen production system 100 with a higher level of equipment. Two functionally identical power generating nodes 10 are shown in the left panel, each of which includes:

[0113] - multiple wind turbine generators 11 ;

[0114] - multiple arrays of PV units 12;

[0115] - an electrolyzer 13;

[0116] - an on-site electrical distribution network 14 connected via a transformer 15 to a high-voltage electrical power line 23 as part of the wide area network provided by the interconnection line 20;

[0117] - a flow battery module 16 for long-term buffering of any of the units connected to the internal electrical distribution network 14;

[0118] - a lithium-ion battery 17 for short-term buffering and / or stabilization of the on-site electrical distribution network 14; - a rotor inertia storage unit 18 for stabilizing the frequency of the on-site electrical distribution network 14.

[0119] With such a configuration the operation of the power generating node 10 can be extended to periods of both low wind speed and low solar energy and is more failure safe hence. Reserve capacities of electrical power and hydrogen gas are stored with in the node and can be used to continuously operate all control systems, the substation, the compressors, and the cooling equipment for the electrolyzer. Eventually the stored power can also be used to maintain the electrolysis process running at low level, too.

[0120] A rotor inertia storage unit 18 provides stability within the on-site electrical distribution network 14 during short periods of approximately 1 to 3 minutes of interruption of power generation.

[0121] The power generating nodes 10 are each connected to the interconnection line 20, which includes the freshwater pipeline 21 , the hydrogen pipeline 22, and a high-voltage electric line 23. Via the hydrogen pipeline 22, the hydrogen obtained from the electrolysis unit 13 of the node is supplied to the energy utilization node 30, where it can be utilized in the same manner as previously described with reference to Fig. 5.

[0122] An excess of electrical power which cannot be processed on-site at the electrolyzer 13 of the power generating node 10 can be supplied to the electrical high voltage network to be consumed at any energy consuming downstream node 30. Still, the on-site generation of hydrogen and the control of the system of the green hydrogen production system 10 via hydrogen gas delivery retains priority over electrical power generation to be distributed over long distance lines.

[0123] Figure 7 is a schematic representation of the topography of one node 300 of a green hydrogen production system that includes sixteen wind turbine generators 301...316 arranged in four columns 321...324 and four rows 331...334. Many PV units are grouped and arranged in a node center 320. The illustrations in Figure 7 and the subsequent illustrations in Fig. 8 to 11 are schematic only and not to scale. Rather, the distance between the adjacent rows 331 ...334 and between adjacent columns 321 ...324 of wind turbine generators 301...316 is at least 5 times the rotor diameter, preferably at least 10 times. In some appliances even a minimum distance of 3 times the rotor diameter is sufficient e.g. if WTGs are installed in sloped areas and rotors are arranged in different heights

[0124] A wind field 350 moving across the area of the node 300 from North to South is symbolized by three block arrows. With the moving wind field 350, a cluster of clouds 340 moves across the node 300 and towards the node center 320 with the PV units. Active wind turbine generators 301...304, 309...312 in rows 331 , 333 are in generator mode and build a cluster of operative WTG units. They are symbolized by the dashed circle around the propeller blades whereas wind turbine generators in rows 322, 324 without such a circle symbol are in idle mode.

[0125] Fig. 19 shows the cluster 361 of operative WTG units used in the situation illustrated in figure 7. It consists of wind turbine generators 301 , 302, 303, 304 in row 331 and wind turbine generators 309, 310, 311 , 312 in row 333.

[0126] In a following state of the same node 300 shown in Figure 8, the clouds 340 are now above the PV units in the node center 320, resulting in significant power losses of solar energy there. These can be compensated by the fact that the wind turbine generators in rows 331 , 333 have been switched to generator mode in time.

[0127] Figure 9 shows the situation with the wind field 350 departing from the node 300, which also drives the clouds 340 away from the node center 320 with the PV units.

[0128] Whereas in the state of the node 300 in figures 7 and 8 wind turbine generators in rows 331 , 333 were in generator mode, generator mode now passes to rows 332, 334 in a wave-like manner, while rows 331 , 333 return to idle mode. In this way, the wind energy is optimally utilized in the wind field 350 passing over the node 300, since, on the one hand, the wind turbine generators produce electricity where the wind speed is highest and, on the other hand, since the rows 331 ...334 maintain sufficiently large distances between each other, so that mutual interference by wake effects is significantly reduced or avoided at all.

[0129] Figure 10 shows a state of the node 300 after the direction of the wind 350 has changed and drives a field of clouds 340 over the node from West to East. The wind turbine generators 301 ...316 of the node 300 have been re-grouped such that now WTGs in columns 321 , 323 are in power generating mode first and will be switched to idle mode after columns 322, 324 have been activated.

[0130] Fig. 20 shows a cluster 362 of operative WTG units used in the situation illustrated in figure 10. It consists of wind turbine generators 305, 306, 307, 308 in row 332 and wind turbine generators 305, 306, 307, 308 in row 334.

[0131] Another situation is shown in Figure 11 to illustrate the flexibility of the control method. In this node 300’, PV units are installed at a total of four node centers 320.1 , 320.2, 320.3, 320.4. With the given direction of the wind 350 from the North and the L-shape of the field of clouds 340 determined from the weather observation, it is to be expected that the PV units in the two western node centers 320.1 , 320.3 will both have a loss of power when the clouds 340 move over, but in the two eastern node centers 320.2, 320.4 a power loss will occur only one by one, since the cloud field 340 there is not so extensive in the direction of migration.

[0132] In this example the cluster of those wind turbine generators, which are marked by the dashed circle in figure 11 and which are intended to be put in generator mode for compensation of shaded PV units, is chosen in such a way that the cloud shape is reflected in the cluster arrangement of selected WTGs. This cluster arrangement has the advantageous side effect, that the WTGs elected for generator operation are located close to the PV units in the node centers 320.1 , 320.2, 320.3, 320.4 thus minimizing the length of transmission lines for electric energy between the selected WTGs and the node centers and minimizing transmission losses consequently.

[0133] After having explained the generic concept of the control method of the invention with view to the schemes illustrated in figures 7 to 11 , the method is explained with reference to figures 12 to 15 each of which shows a Western area being part of the green hydrogen production system 100 of figure 1.

[0134] With weather conditions illustrated in figure 12 the whole Western area is free of clouds. Wind direction of approaching wind field 350 is North. All PV units 12 in all power generating nodes 10.17...10.24 shall run un-curtailed. Northmost nodes 10.17 and 10.21 can be primarily used to power downstream facilities since their WTGs will produce more power as they are located in windward direction. WTGs in the subsequent nodes 10.18, 10.22 are curtailed first so that wake is reduced. Nodes 10.19, 10.23 can also utilized in wind generation and WTG in nodes 10.20, 10.24 are curtailed, too.

[0135] A control strategy can be applied to assess where and by how much the individual WTG can be curtailed or shut down so that the overall production is optimized.

[0136] The weather conditions, especially in windward direction of the PV units, are constantly monitored and forecast data like cloud coverage, cloud height and wind speed on ground and in height of the clouds are gathered.

[0137] Based on the forecast data obtained through the weather observation, the method of the invention is carried out as explained above with reference to figures 16 to 18. The result obtained by carrying out the steps of the method is at least one cluster of WTGs which are designated to be operative in generator mode temporarily until shading of PV units is over. Referring to figure 13 now, the wind field 350 is moving a large field of clouds 340 which are impacting PV units in windward nodes 10.17. 10.21. The power generation by WTG is maximized in those nodes to compensate for losses in PV units’ energy yield. Other nodes 10.18, 10.19. 10.20, 10.22, 10.23. 10.24 continue to run on maximal PV power and curtailed wind energy production. These nodes will see lower wind energy production as windward nodes 10.17. 10.21 will create more wake.

[0138] In the situation illustrated in figure 14, the clouds 340 have passed over windward nodes 10.17. 10.21 and now cover nodes 10.18, 10.22. In nodes 10.18, 10.22 under clouds the WTG production is maximized to compensate for loss in PV units. In nodes 10.17, 10.21 the PV units are ramping up again after clouds 340 have passed by. Wind energy production in nodes 10.17, 10.21 can be curtailed again resulting in reduced wake impact so that downstream nodes 10.18, 10.22 can increase wind production. All other nodes 10.19, 10.20, 10.23, 10.24 located leeward of nodes 10.18, 10.22 will see lower wind energy production as nodes 10.18, 10.22 will create more wake.

[0139] In the state illustrated in figure 15 the clouds 340 have already passed over most part of the site and now cover PV units in nodes 10.20, 10.24 only. In cloud covered nodes 10.20, 10.24 the WTG production is maximized to compensate for loss in PV units. In all other nodes located windward of nodes 10.20, 10.24 PV units ramp up after the clouds 340 have passed by. In all windward nodes wind energy production is low thereby inducing less wake effects so that nodes 10.20, 10.24 can increase wind energy production.

[0140] Reference si ns: coastal region

[0141] 2 sea

[0142] 100 green hydrogen production system

[0143] 10; 10.1 , 10.24; 10.1’, ..., 10.10’ power generation nodes

[0144] 11 wind turbine generators

[0145] 12 PV units

[0146] 13 electrolyzer

[0147] 14 on-site electrical distribution networks

[0148] 15 transformer

[0149] 16 battery module

[0150] 17 lithium-ion battery

[0151] 18 rotor inertia storage unit

[0152] 19 central station

[0153] 20 interconnection line

[0154] 21 fresh water line

[0155] 22 hydrogen pipeline

[0156] 23 electrical high voltage power line

[0157] 24 roadway

[0158] 29.1 , 29.2, 29.4 branch interconnection line

[0159] 30 power utilization node

[0160] 31 port

[0161] 32 water desalination unit

[0162] 33 ammonia processing unit

[0163] 34 HBI plant

[0164] 35 substation

[0165] 36 tap line

[0166] 37 tank

[0167] 40 iron ore

[0168] 300; 300’ node

[0169] 301...304; 313...316 wind turbine generators

[0170] 320; 320.1 , 320.2, 320.3, 320.4 node center

[0171] 321...324 columns

[0172] 331...334 rows

[0173] 340 clouds

[0174] 350 wind field

[0175] 361 , 362 cluster

Claims

Claims:1 . A method for controlling a green hydrogen production system (100; 100’), comprising geographically distributed power generating nodes (10, 300; 300’) each having at least one node center (320; 320.1 , 320.2, 320.3,320.4) and at least one electrolyzer (13) for generating green hydrogen within the system from the produced electrical energy, wherein each the power generating node (10, 300; 300’) comprises multiple PV units (12; 312) and multiple wind turbine generators (WTG) (11 ;301 ...316) as power generating units and wherein the multiple wind turbine generators units (WTG) (11 ; 301...316) are located in geographically dispersed sites surrounding the node center(s) (320; 320.1 , 320.2, 320.3,320.4), wherein the installed capacity (IC) of the electrolyzer (13) and all other energy consuming devices in the system is smaller than the sum of maximum capacities (MC) of all PV units (12; 312) and wind turbine generators (11 ; 301 ...316) available for operation together, wherein the method comprises at least the following steps: a) an energy demand value (EDV) of electrical power required for constantly operating the electrolyzer (13) and other consumers is defined wherein EDV < IC; b) weather conditions in proximity of the power generating units and in windward direction of the PV units (12; 312) are constantly monitored; c) based on weather conditions acquired from monitoring, an expected energy yield value (EEY) is calculated separately for each type of power generating unit and / or for each individual power generating unit; d) an individual workload is assigned to both the PV units (12; 312) and the WTG units (11 ; 301...316) wherein the WTG units (11 ; 301...316)designated to be operated in generator mode are selected from the set of all WTG units (11 ; 301 ...316) available for operation in the system to build a cluster (361 , 362) of WTG units (11 ; 301 ...316) which are set operative to compensate shortage of electrical energy when at least one PV unit (12; 312) is expected to be at least partially shaded.

2. The method of claim 1 , wherein in step d) an individual workload is assigned to both the PV units (12; 312) and the WTG units (11 ; 301 ...316) which is selected according to the following prioritization scheme: i. if the EEY(PV) of the PV units (12; 312) alone will be sufficient to meet the EDV, all PV units (12; 312) are operated at full load and all WTG units (11 ; 301 ...316) are operated in idle mode or curtailed or stopped. ii. If the EEY(PV) of PV units (12; 312) is zero or below threshold but EEY(WTG) of WTG units (11 ; 301...316) is high, the WTG units (11 ; 301 ...316) are used only. iii. if moving clouds (340) are discovered and due to partial shading of some PV units the EEY of the PV is expected not be sufficient to meet the EDV alone, the WTG units (11 ; 301 ...316) are used in addition to PV units but are curtailed to minimum load such that only the difference between the EDV and the EEY of the PV units (12; 312) is provided by WTG units (11 ; 301 ...316) in generator mode, wherein the WTG units (11 ; 301 ...316) designated to be operated in generator mode are selected from the set of all WTG units (11 ; 301 ...316) available for operation in the system to build a cluster (361 , 362) of WTG units (11 ; 301 ...316).

3. The method of claim 1 or 2, wherein the selected WTG units (11 ;301 ...316) for a cluster (361 , 362) are arranged in at least one row extending across or transverse to the current wind direction (350).

4. The method of claim 1 or 2, wherein some WTG units (11 ; 301 ...316) are selected such that the cluster (361 , 362) corresponds in size and shape to the field of moving cloud (340).

5. The method of any of claims 1 to 4, wherein in cases (ii) and (iii) some WTG units (11 ; 301 ...316) located at an windward direction are selected to operate with higher priority or have a higher individual workload assigned than other WTG units (11 ; 301 ...316) located at a leeward direction.

6. The method of any of claims 1 to 5, wherein in cases (ii) and (iii) some WTG units located closer to the electrolyzer (13) are selected to operate in generator mode with higher priority or have a higher individual workload assigned than other WTG units (11 ; 301 ...316) that are farther away from the electrolyzer (13).

7. The method of any of claims 1 to 6, wherein in case (iii) single WTG units of a group of adjacent WTG units (11 ; 301 ...316) are selected to operate with higher priority or have a higher individual workload assigned such that the distance between the selected operative WTG units (11 ;301 ...316) is at least 5 rotor diameters.

8. The method of any of claims 1 to 7, wherein the method is halted in or after step (c) until an expected significant shortage of electrical energy which needs to be compensated is stated, wherein a significant shortage is given if- the EEY is expected to fall below EDV for at least 1 min and- the difference between EDV and EEY is greater than 1 % of EDV.

9. The method of any of claim 8, wherein a significant shortage is given if the difference between EDV and EEY is greater than one half of the rated power of one WTG of the multiple wind turbine generators (WTG) units (11 ; 301 ...316) surrounding the node center (320; 320.1 , 320.2, 320.3, 320.4) in which at least one PV unit (12; 312) is expected to be at least partially shaded.

10. The method of any of the preceding claims, wherein based on weather conditions acquired from monitoring a fatigue and wear indicator (FWI) is calculated for each individual WTG unit and the individual workload assigned to each WTG unit is set such that the fatigue and wear indicators (FWI) for all WTG units in the green hydrogen production system (100; 100’) remain below a threshold.

Images