Wobben v Vestas-Celtic Wind Technology Ltd

This is an action for infringement of four patents relating to various aspects of wind turbine technology. The claimant and patentee, Mr Aloys Wobben, is the owner and Managing Director of Enercon GmbH (Enercon). Mr Wobben founded Enercon in 1984 and it is now a major manufacturer of wind turbines and the market leader in Germany. The defendant (Vestas) is the UK arm of a Danish company, Vestas Wind Energy Systems A/S, which is the largest manufacturer of wind turbines in the world.

Originally four patents were in issue but one, European Patent (UK) No 1,282,774, is now accepted to be invalid. The remaining three are:

There are two other matters I should mention at the outset. First, the applications to amend are opposed on substantive and discretionary grounds. It has been agreed that all issues arising from the discretionary grounds are to be dealt with at a later date, if necessary. Second, Vestas has raised various competition law defences. These have been stayed pending resolution of the infringement and validity issues.

Each side called one expert witness. Mr Wobben’s expert was Dr Taylor. In 1993, Dr Taylor was awarded a first class degree in Electrical and Electronic Engineering at Liverpool John Moores University. Thereafter he was employed by GEC Alsthom (now AREVA) in the transmission and distribution projects team working on what are described as “High Voltage Direct Current” projects and subsequently by E.P.S. UK, a power system switchgear consultancy firm in Nottingham. In 1995, he moved away from the power systems industry for approximately two years. He was drawn back to the field of power systems in a research capacity in 1997 by his interest in renewable energy. That year, he began an industrially focused engineering PhD programme with the University of Manchester Institute of Science and Technology (UMIST) and Econnect, a consultancy firm specializing in the grid integration of renewable energy, who sponsored him. His supervisor at UMIST was Professor Nick Jenkins, an acknowledged leader in the field of wind turbines. The subject of his PhD concerned the integration and control of renewable energy in electrical networks. In 1997, and as part of his research, he carried out a thorough literature survey in relation to renewable energy and electrical networks, including wind power.

Dr Taylor completed his PhD in 2001 and thereafter worked full time for Econnect, specializing in the network integration and control of wind energy systems and managing the research and development team. His research was funded in part by Econnect and in part by the DTI and the European Commission. In March 2004, Dr Taylor left Econnect and was appointed a lecturer in the New and Renewable Energy Group in the School of Engineering at the University of Durham. He has remained in contact with industry and the research projects which he directs are carried out in collaboration with, and are part funded by, a number of international energy companies.

Vestas contended Dr Taylor was at times partisan and had difficulty maintaining his independence, tending to act as advocate for his client. Criticism was made of the way he approached aspects of the cited prior art in his reports and cross examination. He was accused of adopting an approach to the concept of “spinning reserve” which was inconsistent and perplexing and being unduly reluctant to move position on the issue of “mindset”. I address the prior art, the concept of spinning reserve and the issue of mindset later in this judgment, but I think it is fair to say that Dr Taylor did indeed qualify his evidence on occasions. However, I do not believe this can be attributed to a failure to maintain independence or a tendency to act as advocate. The case raised many issues, a good deal of them involving complicated technical matters and a consideration of their bearing on the scope of the technology embraced by the claims. It was entirely appropriate for Dr Taylor to qualify his views as the issues unfolded. I formed the view that he endeavoured to answer the questions put to him fairly and honestly throughout his cross examination.

Vestas also criticised Dr Taylor as not being a person skilled in the art at the priority date. In 1997 he was just starting his PhD. Further, in his first report he expressed the belief that the use of power electronics was extremely rare – so implying that it would not have been generally known. However, as he later accepted, in 1997 the skilled person would have known of both the full converter turbines, such as the Enercon E40, and the DFIG machines and how they worked. Specifically as to DFIGs, Dr Taylor first heard of them in about 1999 and described them as being an emerging technology in 2001 and something to which his supervisor, Professor Jenkins, drew his attention at that time. Yet they comprised over a third of the turbines installed in that year. All these matters confirm to my mind that Dr Taylor was not a person skilled in the art in 1997 and, although I have no doubt he has done his best to consider the issues arising in this case from the perspective of such a person, this is a matter I must take into account in drawing my conclusions. Nevertheless I should emphasise that overall I found Dr Taylor’s evidence of very great assistance.

Vestas called Professor Green. In 1986, he was awarded a first class degree in electrical engineering from Imperial College, London. In 1989, he gained his PhD in electrical engineering from Heriot-Watt University, Edinburgh. From 1990 to 1994 he was a lecturer in the Department of Computing and Electrical Engineering at Heriot-Watt, with a particular interest in power electronics. In 1994, he returned to Imperial, initially as Lecturer in the Department of Electrical and Electronic Engineering and subsequently as Senior Lecturer and then Reader. In 2005, he was appointed Professor of Power Engineering.

Since 1994, Professor Green has engaged in various collaborative industrial projects. Of particular relevance to these proceedings are projects with National Grid Company in 1994 and 1997 in relation to the use of power converters (such as those found in certain types of wind turbine) for reactive power compensation, with Turbo Genset Limited in 1998 in relation to reactive power and voltage control in distributed generation (such as wind turbines) and with ABB Corporate Research in 2001 in relation to the use of power electronic devices (such as the components of control systems used in wind turbines) to damp power system oscillations. He was co-author of two reports: “Impacts and Cost of Intermittency” published by the UK Energy Research Centre in 2006 and “Modelling renewable energy and the distribution system – eastern region”, for the DTI under contract to Future Energy Solutions, published in 2003. Each of these had a particular focus on the use of wind turbines in the electrical power supply system.

Professor Green was criticised for lacking objectivity and looking for difficulties where none existed. I reject these criticisms. I found his evidence to be carefully considered and clearly expressed. It was said his approach was overly academic but I did not find it so. As I later explain, he had real difficulties making sense of aspects of the teaching of the patents in issue.

Nevertheless I do accept that Professor Green has never carried out any project or work for a wind turbine manufacturer and he has never had any involvement in the practical side of the design or operation of wind turbines. So he has never had the practical experience of Dr Taylor. Further, in 1997 he did not have contacts with wind turbine manufacturers or operators. His experience of the industry at that time was based upon his growing interest in distributed generation and possible applications for his power electronics expertise and the knowledge he gained through teaching an undergraduate course in power engineering. In 1997, he was certainly no closer to the person skilled in the art than Dr Taylor. These are all matters which I have had fully in mind when assessing his evidence, particularly on issues concerning the state of mind of the person skilled in the art. Overall, however, I found the evidence of Professor Green, like that of Dr Taylor, to be of very great assistance.

A patent must be considered through the eyes of the notional person skilled in the art. He is a person who is likely to have a practical interest in the subject matter of the invention. In some cases he will be a team of people because the specification may call for a set of skills unlikely to be found in any one person: Technip France SA’s Patent [2004] RPC 46.

It was agreed the skilled addressee is the same for each of the three patents which remain in issue but there was some dispute as to whether he was a single person or a team. Professor Green suggested the patents are addressed to a person interested in the design, operation and control of electrical generators connected to electrical supply networks. Such a person would typically be a graduate in electrical engineering with at least five years’ experience in industry working for a wind turbine manufacturer (or the manufacturer of a major electrical sub-assembly) or engaged in power system analysis for a consultancy company or network operator.

Dr Taylor considered the skilled addressee would necessarily be a multi disciplinary team comprising a wind turbine design and development engineer (from a wind turbine design company); an electrical machines design and development engineer (from an electrical machines company); a power electronics and real time control design and development engineer (from a variable speed drives company); and a power systems operation and control engineer (from a distribution network or a transmissions systems operator).

In the end I think this difference of opinion was more apparent than real. Professor Green’s skilled person would have the ability to understand the teaching of the patents whilst Dr Taylor’s skilled team would be needed to implement that teaching. In these circumstances I think it right to adopt the team proposed by Dr Taylor. It is this team which would have a real practical interest in the inventions and the skills needed to attempt to put them into effect.

Although the patents have different priority dates, the key date for the consideration of nearly all the issues in dispute is December 1997. Accordingly, my discussion of the common general knowledge is as of that date.

The common general knowledge is the common knowledge in the field to which an invention relates. It is that knowledge which is generally known and accepted as a good basis for further action by the bulk of those engaged in that field: Beloit v Valmet [1997] RPC 489 at 494. I begin by summarising the background common general knowledge necessary for an understanding of the teaching of the patents. For this purpose I was provided with two primers, one prepared by each expert. They therefore inevitably covered a good deal of the same ground. This is clearly an undesirable course. A primer should, by definition, be uncontroversial. Paragraphs [18]-[51] of this section of my judgment are drawn in large measure from the primer and first report prepared by Professor Green, for which I am most grateful. I indicate those aspects on which the parties did not agree.

Electrical power is the product of two elements, the current (I) and the voltage (V). As electrical power flows through a power line, some of that power is lost from the system, either because it is used up in overcoming the resistance and reactance of the power lines themselves, or because it is drawn from the network by loads. In using some power to overcome the resistance and reactance of the line, a voltage drop is created such that the magnitude of the voltage at the consumption end of the line is lower than the magnitude of the voltage at the production end. The more power that flows through the line, the greater the difference between these two voltages. At least in the steady state a good approximation of the magnitude of the difference in voltage between two points in a network can be determined from the equation:

Professor Green characterised this equation as “the fundamental equation” whereas Dr Taylor was concerned to emphasise that it applies only in normal (steady state) situations, a proposition with which Professor Green agreed. I will therefore refer to it as “the steady state equation”. The various quantities represented in the equation are as follows:

Resistance (R) and reactance (X) both impede the flow of electrical current (and therefore electrical power). The power lines of an electrical network have both resistance and reactance. Loads (a generic term used to refer to equipment connected to the network in order to consume electrical power) also have resistive and reactive characteristics. As power flows through a network, it must overcome this resistance and reactance. The magnitude of the voltage at any given point in a network is affected by the interaction of real power P with resistance and by the interaction of reactive power Q with reactance. The steady state equation expresses (to a good approximation) how these interactions affect the magnitude of the voltage at a given point on a network.

Although reactive power merely “shuffles” in and out of inductors and capacitors, it does have an impact on network conditions. To understand how reactive power will affect the network voltage, it is important to know whether the reactive power itself is inductive (i.e. shuffling in and out of an inductor) or capacitive (i.e. shuffling in and out of a capacitor). From the perspective of the network, inductive reactive power appears to be drawn from the network (and consumed), while capacitive reactive power appears to be supplied to the network.

Consuming real power and inductive reactive power each contributes to a fall in voltage. Voltages therefore tend to be lower at points in the network to which loads are connected. Conversely, emitting real power and capacitive reactive power contributes to a rise in voltage. Accordingly, voltages tend to be higher at points in the network to which power plants are connected. If a power plant emits real power and consumes some inductive reactive power at the same time, the two actions counteract each other and the voltage will not rise as much at the point in the network to which that power plant connects as it otherwise would.

In 1997 the skilled person knew that network voltages could be controlled by regulating the flow of real power or the flow of reactive power through the lines of that network.

The presence of reactive power in a network is a reflection of the fact that the voltage and the current have become out of phase with each other. If they are not exactly 90° out of phase then there will be both real power and reactive power in the system. The relative amount of each will depend upon the phase angle (conventionally given the symbol Φ) between the angle of the voltage and the angle of the current.

An electrical network is simply a shared environment in which the actions of one party will, to some extent, affect the actions of other parties. It may be extremely simple with a single power source and a number of relatively small consumers, each with its own load. But in the context of this case much attention has been focused on national electricity supply. In this context networks are frequently characterised as transmission networks or distribution networks. Transmission networks are built for the bulk transmission of electrical power at extra high voltage (EHV). In Great Britain the transmission network operates at 400,000V. The distribution network is built for the delivery of electrical power to domestic, industrial and commercial loads. These networks operate at high voltage (HV) of 33,000V, medium voltage (MV) of 11,000V or low voltage (LV) of 230V.

The overwhelming proportion of the power used by loads connected to distribution networks is supplied from the transmission network through grid sub-stations which comprise transformers which step down the voltage (and increase the current) of the power supply. But, as I elaborate below, the 1990s saw an increasing tendency for smaller power plants, particularly wind turbines, to be connected to (or embedded in) the distribution networks.

For the purposes of these proceedings the following differences between transmission and distribution networks are particularly relevant:

There is a considerable overlap between the concepts of transmission and distribution networks and strong and weak grids. A strong grid is one in which the loss of a generator, particularly a small one, or a sudden increase in a load is unlikely to have much if any effect on the quality of the power delivered by the system. Small generators are unlikely to be required to contribute to measures intended to control network frequency or voltages. A substantial transmission network operating at EHV is the paradigm strong grid. Conversely a weak grid is fed by relatively few power plants and is more likely to be affected by the behaviour of a particular generator or load. A distribution network is much more likely to be a weak grid than a transmission network.

I should also refer at this point to the concept of an electrical island. This is a network that is electrically isolated from any other network. It must contain at least one power plant and at least one consumer. It can form within any part of a network if the connections to the rest of the network are cut, as may happen in the case of a severe localised fault. An island is characteristically weak and so the individual participants are likely to have a greater influence over the quality of the power. Correspondingly, smaller plants may be able more effectively to contribute to the control of frequency and voltage.

Those operating national networks are required to keep the voltage and frequency of the electrical power the network carries within acceptable tolerance limits by the terms of their licences. This is important. Loads connected to the network may be damaged or malfunction if the voltage or frequency deviates significantly from those limits.

The transmission network in Great Britain is operated by the transmission system operator (TSO) and each distribution grid is operated by a distribution network operator (DNO). In 1997, these operators were required by law to maintain network voltages within tolerance bands. So those operating the higher voltage networks were required to operate within tolerances of +10%/-10% in the case of voltages above 132,000V and +6%/-6 % in the case of voltages below 132,000V. In order to ensure these requirements were satisfied the operators and electricity suppliers entered into a series of agreements and grid codes. All significant power plants connected to the networks were required to enter into an agreement with one of the operators. Conventional power plants connected to the transmission network entered into a connection agreement with the TSO and were required to comply with the transmission grid code. Smaller power plants, such as wind farms, embedded in a distribution network entered into a connection agreement with the local DNO and were required to comply with the distribution grid code. However, larger embedded power plants - those producing more than 50MW of power – were also required to comply with the transmission grid code.

In the long term, voltage control can be improved by reinforcing the network by, for example, installing new equipment to raise the operating voltage or using thicker power lines. But short term fluctuations of voltage require a much quicker response which was provided by a variety of means, as I shall explain.

In 1997 the majority of power in the UK was generated by fossil fuelled and nuclear powered generators. The position remains the same today. These conventional power plants are principally connected to the transmission network and are of a very substantial size. They use large synchronous generators (typically producing in the region of 500 MW in 1997) and they form the “backbone” of the grid. They set and maintain the AC frequency of the grid at 50 Hz, at least in the UK. The power produced by these conventional plants is said to be dispatchable in that it can be turned on or off or supplied in a stable manner as required, without concerns as to the fuel source. The machines have an operating envelope which means that they can remain connected to the network and, depending upon the degree of excitation of the generator, supply almost any combination of real power P or reactive power Q in support of the network at the network operator’s command. This ability to vary the real and reactive power output of the major generators is the principal means of controlling the voltage on the transmission networks.

For points on the network some distance away from the power plant it was also conventional to control the voltage by the use of capacitors or inductors which could be switched into the network to raise or lower the voltage as desired by varying the reactive power. More sophisticated devices known as SVCs and STATCOMS were also used for the same purpose.

A further means of controlling the voltage, particularly in the case of distribution networks, was the variable ratio or tap change transformer. This is a transformer in which the number of windings on each side of the transformer can be adjusted either manually or automatically in response to conditions on the network measured by sensors. When placed at various strategic points in the distribution network they have the ability to regulate the voltage on the secondary side to a value within the desired limits. As a practical matter, the tap change transformers in a distribution network were under the control of the network operator.

The power of the wind has been used for thousands of years and wind turbines as a source of electrical energy were first developed at the beginning of the twentieth century. But wind turbines became of particular interest as a source of electrical power in the 1970s as the price of oil began to rise.

There are three main types of wind turbine of relevance to these proceedings. In the early to mid 1990s the most common and, so to speak, work horse of the industry was the directly coupled asynchronous generator turbine (also known as an induction, squirrel cage or Danish generator). It is depicted schematically below:

The asynchronous generator uses coils of wire as both rotor and stator, rather than using a permanent magnet as a rotor. The stator of the generator is connected to the existing AC network, which allows current to flow in the stator. The flow of current creates a magnetic field which rotates as the current oscillates through each cycle. This is known as magnetising the generator. The rotor, turned by the rotation of the blades, cuts across this rotating magnetic field. Provided the rotor is not turning at exactly the same speed at which the stator’s magnetic field rotates, this induces a voltage on the rotor which in turn causes current to flow in the rotor coils.

Once an electric current has been induced in the rotor, this creates a second magnetic field. This second magnetic field plays the same role as that of a permanent magnet in a conventional generator. It cuts across the stator as the rotor rotates. The key feature of the induction generator is that the alternating currents in the rotor produce a magnetic field that rotates with respect to the rotor. When this rotation is added to that of the rotor itself, the total rotational speed of the second magnetic field is exactly right for generating power in the stator at the frequency of the voltage originally presented at the stator by the existing AC network. The generated power can then be drawn off the stator and fed to that same AC network.

This wind turbine has a number of limitations. First, it is only able to generate power when connected to an existing AC voltage. There must be power in the network before it can be used to provide more power to that network. Second, the speed of the rotor is, for practical purposes, fixed and determined by the frequency of the supply grid, the gear ratio and the generator design. Third, fluctuations in wind speed are transmitted as fluctuations in mechanical torque which can yield voltage fluctuations at the point of connection to the network and require a mechanical construction which is able to tolerate high mechanical stress. Fourth, it is limited to producing electricity at the frequency of the AC network that originally magnetised the stator.

In the mid 1990s, Vestas introduced a variant of the basic induction generator which increased the range of wind speeds at which the turbine could operate by introducing a variable additional rotor resistance. Typically the speed range of this variant is 0-10% above the synchronous speed. But, as in the case of the conventional induction generator, it draws reactive power from the grid and a capacitor bank is used to provide reactive power compensation.

The 1990s also saw the establishment of, another, but very different, turbine called the full converter turbine. It was typified by the Enercon E40 and by 1997 had some 16% of the world market. It is depicted schematically below:

This machine generates electricity at whatever frequency is best suited to capturing the wind power at any given time. It then converts the frequency of that electricity to the frequency of the network to which it is connected.

The rotor has associated with it a magnetic field created, for example, by the presence of a permanent magnet. The rotation of the rotor and its associated magnetic field induces an AC electric current in the stator with a frequency directly related to the rotational speed of that magnetic field. The frequency of the power generated in this way will therefore vary according to the wind speed (and any steps taken by the turbine to adjust its capture of wind energy). The AC power generated at this undetermined frequency is first converted to DC, a process commonly known as rectification. This is achieved by an electronic component consisting of diodes or transistors and known as a rectifier (if it can only convert from AC to DC) or a converter (if it can work in either direction). This DC power then passes through the DC link to a component which converts it back into AC at the desired frequency (usually that of the network) by a process commonly known as invertion. The component consists of diodes or transistors and is known as an inverter (if it can only convert DC into AC) or a converter.

This full converter system is very flexible. It allows the generator to operate at any frequency. However, it does require all of the power generated to be processed through two converters that use semiconductor technology to process power in so-called switch-mode circuits. This is a relatively expensive technology and the cost of using it increases as the power to be processed in this way increases.

The last of the turbines is the doubly-fed induction generator (“DFIG”) turbine. This uses the same power based conversion technology as the full converter turbine but not for all of the power it generates. The first commercial machine of this type was introduced in 1996 by Tacke and was very large for its time. From 1997 other manufacturers started to produce DFIG machines but of a more appropriate size and with great success. All of the Vestas machines alleged to infringe are of this type. It is depicted schematically below:

The DFIG is a form of induction generator and has a rotor made up of a winding rather than a permanent magnet. However, in this case a connection to the existing AC network is made both to the winding on the stator and to the winding on the rotor. The stator is directly connected to the electrical network and has a voltage at the network frequency imposed on it. The rotor is connected to the network via converter circuits.

In order to generate power in the stator at the correct network frequency, the converters control the currents in the rotor so that the rotor has a magnetic field that rotates at the correct speed for the given rotational speed of the turbine shaft. Depending upon that rotational speed, power can be extracted from both the rotor and the stator or just from the stator. Power extracted from the stator is at the network frequency. Power extracted from the rotor has its frequency adjusted to the network frequency by the converter circuits.

The advantage of this system is that the converters need only process a fraction of the generated power and it is consequently cheaper than the full converter turbine. The disadvantage is that the speed range over which it can operate is not as great.

There are limits to the amount of power a plant can produce at any point in time. It follows from the steady state equation that in order for a power plant to increase the voltage at the point of the network to which it is connected it must increase the real or reactive power it supplies to that point. But in the case of a wind turbine, limits to the apparent power it can produce are set by the maximum current and voltage the electrical circuits of the turbine can tolerate. Power is the product of current and voltage. Hence if the voltage at the connection point of the turbine to the network falls, so too does the maximum power the turbine can generate without causing a damaging increase in the current. If the current was already at its rated capacity before the voltage drop, any increase in that current would be likely to cause damage and so the power of the turbine must be reduced or “capped” for so long as the voltage drop remains.

This is known as de-rating and has been a well understood concept in power generation for very many years. However, the extent to which it was used or it was obvious to use it in 1997 specifically in connection with wind turbines is a matter I consider later in this judgment.

Dr Taylor suggested in his first report that in 1997 the use of power electronics was extremely rare and that the average engineer would have had concerns about the use of power electronics in a fully rated converter because it would lead to an increase in cost and complexity. I have no doubt that the costs associated with DFIGs and full converter turbines, such as the E40, were something of which those in the field were very conscious. But I also have no doubt that the skilled person interested in making an improved wind turbine in 1997 would have been fully aware of the DFIG and full converter machines and their general principles of operation. Both were well established in the market by the end of 1997 and were described in a standard textbook Generation of Electricity by L.L. Freris (1990). Indeed, in the end Dr Taylor accepted as much in the course of cross examination.

As I shall explain, the 1990s saw the number and size of wind farms increase but by 1997 it was still the case that only a very small percentage of the power in the UK was generated from renewable sources. Wind farms characteristically comprised between five and ten turbines with an average capacity of about 600KW. So the wind farm as a whole might typically have a capacity of about 5MW. By 1997, 36 wind farms had entered into service in the UK with a total capacity of about 290MW. To put these wind farms in context, the electricity system in Great Britain has a peak demand of about 55,000MW.

A feature of wind farms is that they have to be built in windy locations, usually a considerable distance away from the transmission networks. It is also important to note that a wind turbine can only generate as much electricity as the wind available at the time allows. So wind turbines operate for much of the time at below their maximum capacities. Indeed, wind turbines in Great Britain have an average production over a year in the region of only 20-35% of their capacity. Their power is therefore not dispatchable.

This combination of features: limited capacity, physical location and dependence on wind resulted in wind plants generally being connected to distribution networks. Wind turbines were traditionally not called upon to participate in efforts to stabilise network conditions. The maximum output from wind farms did not exceed the amount of electricity required at times of least demand, so all available wind power could be taken.

Conventional (asynchronous) wind turbines also suffered from the limitation that they had only a narrow operating window and imported inductive reactive power which further limited their ability to provide active network support. They were therefore regarded as negative load, which is to say they exported power but in a relatively inflexible way. They did not contribute to the control of the power system and were seen as surplus to requirements under fault conditions.

Nevertheless, the 1990s were years of dramatic change. First, the new turbines, namely the full converter and DFIG turbines, were much more flexible, operated over a wider range of wind speeds and permitted their operators to vary the real and reactive power they produced.

Second, the size of turbines and the capacity of wind farms worldwide was increasing rapidly. In 1997, the average capacity of newly installed wind turbines in Germany was about 600KW. By 1999, it had risen to over 900KW and by 2002, to nearly 1400KW. Over the 16 years from 1988 to 2003, the typical wind turbine capacity in Germany increased by approximately 25 times.

The growth of overall operational wind power capacity in Germany, Denmark and Spain is particularly striking. By the end of 1995, the installed capacity in Germany was 1,136MW, in Denmark 619MW and in Spain 145MW. By the end of 2003 it had grown to 14,609MW in Germany, 3,110MW in Denmark and 6,202MW in Spain. As for the size of the wind farms themselves, these too increased. In 1991, the first offshore wind farms were commissioned in Norway and Sweden. The larger of the two, that in Sweden, had a capacity of 4.5MW. In 1997 the largest wind farm commissioned in the UK had a capacity of 20MW – a significant increase but still some way short of the capacity that would have required compliance with the transmission grid code. In 2001 the largest wind farm commissioned in the UK had a capacity of 30MW and it used Vestas turbines. Denmark was one step ahead: the Middlegrun offshore wind farm had a capacity of 40MW. By this time it was clear that wind farms could not simply be regarded as negative loads. They were becoming connected to the higher voltage sections of the network and were expected to provide support. Indeed, Dr Taylor explained in cross examination that as early as about 2000 Denmark began the process of changing its grid code to require renewable energy producers to support the network under fault conditions.

It was well understood in 1997 that variations in a network voltage can occur for a multitude of reasons. The voltage in a network can rise because of a reduction in load and an increase in power supply, and it can fall because of an increase in load and a reduction in power supply.

It was also well understood that faults can occur on the network which can result in a sharp drop in the voltage. Indeed the transmission grid code in Great Britain recognised the possibility that the voltage might collapse transiently even to zero until the fault was cleared. In the event of such a sharp drop in voltage major suppliers connected to the transmission network were required to remain connected for a maximum of 120ms – a feature known as “low voltage ride through” or “LVRT”. The reason for this requirement was straightforward. If a fault cleared then it was essential that power be restored to the network as soon as possible. If the fault did not clear in that time then suppliers could disconnect.

On the other hand, wind turbines and other smaller embedded generators were required to disconnect in the event of a voltage deviation of +10%/-10%. The reasons, at least in the case of voltage drops, were associated with islanding. Opening a circuit breaker in the event of a fault might isolate a distal part of a distribution network from the rest of the network. If this isolated part of the network contained both loads and embedded generation, it might continue to operate as an electrical island with a number of undesirable consequences. Chief amongst those canvassed in evidence were hazards to personnel carrying out repairs on that part of the network and the risk that the frequency and voltage in the electrical island could easily deviate from those in the rest of the network. When the relevant section was then reconnected to the network the two systems might well not be in synchronism with each other and connecting two systems which are not in synchronism is potentially damaging.

It is important to note that the requirement to disconnect was imposed upon the wind turbine operators by the network operators. They could have managed the practical problems of islanding by, for example, managing the control of the isolated network voltage and frequency and ensuring that they were re-synchronised with the rest of the network before reconnection. But, for cost reasons, they chose not to do so.

It is also important to note that the requirement to disconnect could be imposed because of the relatively small power contribution embedded wind farms were making to the network. Disconnection would have only an insignificant impact on the transmission system. However, it was, at the very least, a real possibility that had a wind farm of sufficient capacity been installed (that is to say, greater than 50 MW and with voltage and frequency control) then it would have been required to comply with the grid code, remain connected through a fault and contribute to the control of network frequency and voltage in just the same way as the large conventional synchronised power plants. As Professor Green explained, no TSO could ever tolerate allowing significant amounts of a particular type of generation to be connected to the transmission grid without requiring it to contribute to frequency and voltage control. That would jeopardise stable operation of the grid or place a disproportionate burden on other forms of generation.

Dr Taylor maintained that the mindset of the skilled person in 1997 was to connect a small number of modest capacity wind turbines to the distribution network. He would then seek to disconnect those turbines from that network at the first sign of trouble, that is to say if the network voltage went outside the tolerance limits of +10%/-10% . He was thinking how he could export as much power as possible from the turbine without infringing voltage or thermal limits and wanted to ensure that under fault conditions disconnection was achieved quickly in order to protect the turbines. Dr Taylor also considered it would have required “visionary thinking” to have the idea of remaining connected under fault conditions to assist the network operator to control network frequency and voltage. It was his view that the average wind turbine engineer was by nature conservative and unwilling to question or challenge accepted standards or perceived wisdom. I deal with the issue of mindset and its impact on the obviousness attacks against the patents in issue later in this judgment. However at this stage I would express the following preliminary conclusions.

First, I am unable to accept Dr Taylor’s characterisation of the average engineer in the context of the wind turbine industry. The 1990s were undoubtedly years of considerable change. The rated capacity of wind turbines was increasing rapidly, as was the size of wind farms. The well known textbook Wind Power in Power Systems by Ackermann (2005) described wind energy as the fastest growing energy technology in the 1990s, in terms of percentage of yearly growth of installed capacity per energy source. Moreover, the market profile of the various different wind turbines was also changing. In 1997 sales of DFIGs accounted for only 3% of the world market, but in 1998 this figure had jumped to over 26%. In that same year DFIGs and full converter turbines together accounted for over 40% of the market. This is not a picture of wind turbine engineers being conservative or resisting change. To the contrary, they were clearly adapting to and engaging with changing circumstances and new technology.

Second, it was nevertheless undoubtedly the case that wind turbine operators were generally required to disconnect if the network voltage strayed outside the +10%/-10% tolerance limits imposed upon them by the network operators. This was so whether the variation was caused by conditions of unusually high or low demand or by a fault. A real practical benefit of disconnecting in the case of a fault was the avoidance of islanding and the problems associated with it. As a result, wind turbine operators would generally disconnect at the first sign of trouble.

Third, I also accept that in 1997 wind turbine engineers were not in general taking practical steps in anticipation that there would be such an increase in the size of wind farms that they would necessarily have to comply with the transmission grid code and, in particular, remain connected under severe voltage disturbances and then emit power as soon as the voltage recovered. Indeed, it was not until 2000-2002 that a change was proposed to the grid code for Great Britain in respect of renewable energy to require wind farms to remain connected under severe fault conditions. It was met with considerable resistance from the wind turbine manufacturing industry which was concerned at the practical problems of implementation. This was the consistent evidence of Dr Taylor under cross examination and is supported by a paper by Antony Johnson and Nasser Tleis of the National Grid Company in 2005 (Wind Engineering Vol 29, No 3, pp.201-215) in which they noted the reluctance of wind farm developers to accept any requirements to enable wind generation to connect to the transmission system on the basis of additional cost and that the technology was still at the proving stage.

Fourth, manufacturers were, however, seeking to avoid the need for disconnection, as will be apparent from my consideration of the cited art. For example, the introduction of the full converter turbines, such as the E40, allowed the wind turbine operators to vary the real and reactive power output of the wind turbine with the object of maintaining the network voltage at its desired reference value.

It is convenient to consider the 324 application in some detail for it is here that the basis for the existing and proposed amended claims of the 564 and 691 patents must be found and the disclosure of the document was the subject of considerable dispute between the parties. In doing so I have numbered the paragraphs sequentially and will refer to them accordingly.

At the outset the application describes an invention relating to a method of operating a wind energy system and to a wind energy system itself. Paragraph 3 explains that in the known systems the generator is operated in parallel with the electrical load, frequently an electrical network and that both the electrical power produced by the generator and the electrical generator voltage vary as a function of the wind power. This appears to be a description of a conventional asynchronous generator.

The application then explains that such an arrangement can lead to problems. In particular, severe changes in the generator voltage can lead to severe unwanted changes in the network voltage at the connection point. As elaborated in paragraph 5, the network voltage might rise to an undesirably high value if the load on the network is low and, at the same time, a high level of power is fed into the network. Such a situation might occur, for example, on a windy night when the electrical consumption in households is low but the wind energy system is providing a high level of power. If the voltage in the supply network rises above a predetermined value, the wind energy system would then have to disconnect from the network. Such disconnection, the application explains, leads to the supply of electrical power being interrupted in a manner which is equally undesirable for the operator of the wind energy system and for the operator of the network. It is undesirable for the operator of the wind energy system because he loses revenue and for the operator of the network because of the sudden interruption in the feed of electrical power.

No details are provided as to what the predetermined value might be but, as Dr Taylor accepted, if this teaching is applied to Great Britain then the skilled person would understand it to be the normal voltage limit of +10%.

Paragraph 7, on page 2, then explains the object of the invention is to provide a system which avoids voltage over fluctuations (that is to say, fluctuations over a certain size) at the point of the consumer, in particular in a network, and undesirable disconnection of the wind energy system.

This object is achieved by a method and apparatus of the invention as described in paragraphs 8 and 9, on page 2, in which the power supplied to the network by the wind energy generator is regulated as a function of the network voltage. The invention is to be contrasted with the prior art system in which the power supplied by the generator varied as a function of the wind power and it is clearly contemplating a weak grid in which the network voltage will be affected by the power supplied by the wind energy system.

The invention is elaborated in paragraphs 10 to 14. Paragraph 10 is a paragraph on which both parties focused. It explains that the basic starting point of the invention is not fluctuations which occur in the capacity to generate power as a result of changes in wind strength but rather that fluctuations in power consumption also occur on the consumer side and these can be reflected in critical voltage fluctuations which damage the electrical equipment of consumers. Importantly, the invention therefore takes into consideration both fluctuations in energy generation on the generator side and fluctuations on the consumer side in determining the feed of energy into the system so that the voltage is regulated at the feed in point to the desired reference value. I see this as a clear teaching that the aim of the invention is to hold the network voltage to the reference value and so avoid fluctuations which would otherwise require disconnection because they would take the network voltage outside the normal tolerance limits.

There was some considerable debate between the parties and a good deal of evidence from the experts as to whether the fluctuations contemplated are only those caused by large loads or whether they also include fluctuations caused by faults. Professor Green considered that the teaching is not concerned with faults whereas Dr Taylor maintained that it was. I have reached the conclusion that the skilled person would understand the teaching to include fluctuations caused by faults of other than a severe nature. In the event of such a fault, all the wind energy system sees is a voltage fall. It has no idea whether that fall is the result of a large load or a fault. The teaching of the application is to safeguard against fluctuations whatever the cause and to do so by feeding in more power in the case of a voltage drop or feeding in less power in the case of a voltage rise so as to regulate the network voltage to the reference value. However, that is as far as the teaching in this paragraph goes and, in this respect, I agree with Professor Green. In particular, there is no teaching that in the case of a severe fault which reduces the network voltage to zero or close to zero then the wind energy system should remain connected but emit no power. This is an altogether different situation in which there is no hope of regulating the network voltage to a desired reference value.

Paragraphs 11 and 12, on page 3, reiterate that the invention avoids undesirable fluctuations in the network voltage by regulating the power delivered by the generator as a function of the network voltage – including fluctuations which can arise from changes in wind power. As a result, there is now no need to disconnect.

Paragraphs 13 and 14 return to the theme of the consumer. They explain once again that the power drawn by consumers can vary but the supplied power is controlled so as to regulate the network voltage to a desired reference value. This is to be achieved in the case of an increase in the load by supplying more power as explained in paragraph 13:

“In addition the invention makes it possible to compensate for network voltage fluctuations as regularly occur in electrical networks for supply of electricity even with a constant wind power, as some consumers connected to the network from time to time draw large amounts of power from the network, possibly resulting in a reduction in voltage. In the event of such a reduction in voltage, the wind energy system according to the invention can feed an increased amount of electrical power into the network, thus compensating for voltage fluctuations. For this purpose, the feed-in voltage is raised at the interface between the wind energy system and the network, for example on the basis of the network voltage value which is sensed in accordance with the invention.”

The application then proceeds to describe the invention by way of an example which is illustrated in figures 1 to 5. This appears to be a description of a full converter wind turbine. Figure 1 is a diagrammatic view of a wind energy system feeding into a simple network. Figure 2 shows a control device according to the invention. It also shows a rotor coupled to a generator which will produce electrical power as a function of the wind speed and thus of wind power. An optimum generator voltage Unom (as shown in the figure but referred to in the text as Uref) is calculated as a function of the determined network voltage and the control device is then used to regulate the actual generator value Uact to the desired voltage value Unom. The text explains that by feeding the power supplied by the wind energy system regulated in this way, fluctuations in the network voltage can be avoided or considerably reduced.

This brings me to figure 3 which was the subject of considerable dispute. It is shown below:

The explanation of this figure is given in paragraph 22, on page 5:

“Referring now to FIG. 3, there is illustrated in the graph how the power entered on the ordinate and delivered by the wind energy system relates to the network voltage entered on the abscissa. If the network voltage differs only slightly from its reference value, which is between the voltage values U_min and U_max, then a uniform level of power is delivered to the network by the generator, corresponding to the upper straight portion of the curve (straight line parallel to the abscissa). If the network voltage rises further and exceeds a value defined by point P1, the power fed into the network is reduced. When the value U_max is reached, then the power fed into the network is equal to zero (point P2). Even in the case where there is a high level of wind power, no power is fed into the network at point P2. If the wind power falls sharply, then only a reduced amount of power can still be fed into the network. Even if no further power is delivered by the wind energy converter, it continues to be operated – although without delivering power – so that power delivery can always be effected as soon as the network voltage has again assumed a value between U_min and U_max.”

At the outset it is necessary to clarify one aspect of the figure. Although the graph purports to be a plot of wind power (ordinate or y axis) against network voltage (abscissa or x axis) it was agreed that the skilled person would understand it to be a plot of power output from the wind energy system against network voltage.

As for the substantive disclosure, the paragraph begins by describing that portion of the graph which runs parallel to the x axis. It explains that somewhere along that portion is the voltage reference value and that at voltages which differ from it “only slightly” (that is to say at all points between P1 and the equivalent voltage at the left hand end of that parallel portion) a uniform level of power is delivered to the network by the generator.

It then proceeds to describe the right hand slope of the graph. If the network voltage value exceeds P1 then the amount of power delivered by the generator is gradually reduced until, at point P2, corresponding to the network voltage value Umax, no power is fed into the network. This is so even if there is a high level of wind power available. Then the last sentence of the paragraph explains that even if no power is delivered by the turbine, the converter continues to be operated, though without delivering power, so that delivery can be effected as soon as the network voltage has dropped below Umax.

This is entirely consistent with the rest of the description of the application. It might typically occur on the windy night described in paragraph 5. As the network voltage rises the power delivered by the generator is reduced so as to bring the network voltage back towards the desired reference value. The aim is to avoid a voltage rise which would result in the need for disconnection. Umax is not defined but the teaching is consistent with it representing the upper limit set by the network operator. Once the voltage reaches this upper limit the generator ceases delivery of any power but continues to operate. As the application explains in the last sentence of paragraph 22, this means that power delivery can always be effected “as soon” as the network voltage has again assumed a value between Umin and Umax. The application does not explain what is to happen to the generator in the event the network voltage rises above Umax, save that I think it is clear from figure 3 that the generator will not deliver any power to the network. This is an aspect of the disclosure to which I must return.

In my judgment the left hand side of figure 3 is considerably less clear. On the assumption that the network voltage reference value lies somewhere along the upper straight portion of the graph, it can be seen that initially the power emitted by the generator remains constant as the network voltage falls. However a point is reached at the end of the straight portion of the graph parallel to the x axis and opposite to P1 at which the power emitted by the generator begins to fall and thereafter it does so in a linear manner with any further reduction in the network voltage. Eventually the power emitted by the generator falls to zero at the point where the network voltage corresponds to Umin.

Umin is also not defined but once again I think the skilled person would consider a reasonable starting point is that it represents the lower voltage limit set by the network operator. However this interpretation has a major difficulty, namely there would seem to be no sensible reason to reduce the power emitted by the generator before Umin is reached. Indeed, to do so would be flat contrary to the teaching of the application that the invention is concerned with regulating the network voltage to the desired reference value. Specifically paragraphs 13 and 14 of the application explain that the generator of the invention can feed an increased amount of power into the network, and so compensate for a voltage reduction caused by customers drawing large amounts of power.

Another possible explanation of the left hand side of figure 3 was provided by Professor Green, namely that it represents the effect of a fall in wind speed. This would reduce the power the generator is able to produce. This explanation has the attraction that it finds some support in the statement in paragraph 22 of the application that:

“If the wind power falls sharply, then only a reduced amount of power can be fed into the network.”

However this explanation also faces difficulties. At the outset it is hard to see why a reduction in the power produced by the generator should result in a fall in the network voltage. It is possible that the left hand side of figure 3 represents a weak network fed by wind turbines which all experience the same fall in wind speed. But then one would expect to see Umin much closer to the origin of the graph. Instead, the x axis is broken which indicates that Umin is not close to zero volts. Moreover, if the depicted reduction in power is attributable to a drop in wind speed then the two slopes of figure 3 are the result of completely different phenomena and there was no reason to draw them in a symmetrical fashion. It is also difficult to reconcile this explanation with the closing words of paragraph 22 of the application:

“so that power delivery can always be effected as soon as the network voltage has again assumed a value between Umin and Umax. ”

In this scenario this would not be the case – the ability to deliver power (and so cause a rise in the network voltage of this weak network) would depend upon the wind speed increasing.

A yet further possible explanation, and one put to Professor Green in the course of cross examination, is that the voltage limits imposed by the network operator are at P1 for the upper limit and its equivalent point at the opposite end of the straight portion of the graph parallel to the x axis for the lower limit. Outside these limits, so it was suggested, the prior art generators would have disconnected whereas the generators of the invention continue to provide power and assist the network. Focusing on the left hand side of figure 3, it was further suggested that this illustrates de-rating. As the voltage on the network falls, so too does the maximum power the generator can produce without the current exceeding the rated capacity of that generator.

Once again, this explanation has its difficulties. As Professor Green pointed out, P1 (and so also its equivalent point at the opposite end of the straight portion of the graph parallel to the x axis) cannot represent the upper and lower voltage limits imposed by the network operator - because the graph would then depict the generator continuing to supply power into the network above the network limit, in which circumstance it would be much better for the generator to cease supplying power altogether. Secondly, if the left hand side of figure 3 was intended to represent a generator maintaining the maximum possible amount of power as the voltage falls (that is to say de-rating), the skilled person would expect to see Umin at the origin. This follows because power is proportional to the product of voltage and current (IV) and when a generator is de-rated, the current (I) is held at the maximum permissible level. Instead, however, the graph drops sharply down to zero power delivery at Umin, a point evidently some distance away from the origin.

It was in these circumstances that Professor Green described figure 3 and paragraph 22 of the application as perplexing. I have to say that I agree. The general teaching of the application is that the invention provides a generator which will avoid over-fluctuations in the network voltage and avoid undesirable disconnection of the wind turbine. It contemplates providing more power in the event of a fall in the network voltage and providing less power in the event of a rise in the network voltage, in each case with the aim of regulating the network voltage to the desired reference value. In this context the right hand side of figure 3 makes sense if Umax is considered to be the upper limit of the network voltage. The object of the invention is to ensure that the limit is not exceeded. To this end the turbine emits less power as the network voltage approaches Umax. However the left hand side of figure 3 presents the skilled person with a puzzle. Quite what it is intended to represent is something of a mystery. It appears to depict the exact opposite of what is taught elsewhere in the application, that is to say a reduction rather than an increase in the power emitted by the wind turbine as the network voltage falls. Nor do I consider there is any clear description of what the turbine is to do in the event the invention fails to achieve its object and the network voltage fluctuates outside the voltage limits. There is no general teaching that this forms part of the invention. All the application teaches is to be found in the description of figure 3 in paragraph 22 that

“Even if no further power is delivered by the wind energy converter, it continues to be operated – although without delivering power – so that power delivery can always be effected as soon as the network voltage has again assumed a value between Umin and Umax”

Quite apart from the difficulties over figure 3 itself, there is no explanation as to whether this teaching applies to values above Umax and below Umin and, even if it does, whether it applies to all values above Umax and all values below Umin and for how long the wind turbine must continue to operate. As will be seen, this assumes considerable importance in considering the questions of validity and infringement. In that context Mr Wobben contended the application discloses an invention which encompasses the ability of a turbine to stay connected but emit no power for a period of 100-200 ms in the event of a serious network fault which results in the network voltage dropping to zero or close to zero – that is to say, LVRT. But there is no teaching of LVRT or how it is to be achieved and, importantly, it is a situation in which there is no prospect of regulating the network voltage to a desired reference value, as both Dr Taylor and Professor Green explained in the course of their evidence.

Figure 4 is described as showing the essential elements of the control and regulation arrangement from figure 1. Interestingly, this same figure was published in Enercon’s brochure for the E-40 full converter wind turbine in 1995, which is one of the citations relied upon by Vestas in support of its attacks on the validity of the 564 and 691 patents (“D8”) and was, as I have found, something of which the skilled person would have been aware in 1997.

Finally I should refer to the claims. These are consistent with the general teaching to which I have referred. They are directed to a wind energy system and to a method of operating such a system in which the power supplied by the generator is regulated as a function of the network voltage.

The specification and drawings of 564 are identical to the specification and drawings in 324 save in the following respects:

564 has three claims which are set out below with the words and expressions in issue underlined

It can be seen immediately that these claims are concerned with the operation of a wind energy system in circumstances where the network voltage is greater or less than a predetermined network voltage value (Umin and Umax). They call for a system which continues to operate or remain connected but without emitting power when the network voltage falls outside predetermined limits, whether caused by low loads, high generation output, high loads or faults. Indeed, Mr Wobben contended that the patent is primarily concerned with major faults and provides a system which continues to operate and remains connected during severe voltage depression, that is to say one which has LVRT capability. Specifically he contended, and was supported in this contention by Dr Taylor, that the patent claims a concept akin to “spinning reserve”. This is a term taken from conventional power stations where it is used to describe a synchronous generator which is maintained in a condition in which it is spinning, synchronised and ready to take up load and so provides a source of reserve capacity which can be brought on stream the moment it is required. With that introduction I will deal with the disputed words and expressions in turn.

These words appear in claim 1 and claim 3 respectively. There is no doubt that claim 3 describes a system for carrying out the method of claim 1 and which emits no real power while remaining “connected”. There was however a substantial dispute between the parties as to whether the word “operated” in claim 1 embraces a different concept. In his first report Professor Green identified six ways in which a full converter turbine might be prevented from emitting power:

As Professor Green explained, options (1) and (2) involve preventing the wind turbine from generating electrical power from the wind. Options (3) to (6) involve preventing the emission of electrical power by the wind turbine even if wind energy is captured and electrical power is generated. They are not mutually exclusive. In order to use any of options (3) to (6) it would be necessary either to dump real power or to combine that option with either option (1) or option (2). Further, to maintain a wind turbine in a condition in which it was spinning but not emitting power to the network would require the use of one of options (3) to (6) in combination with dumping real power or option (2).

Professor Green also expressed the opinion that “operated” means that the turbine’s generator should be spinning and regulated by the control system. Hence a wind turbine might be said to be “operated” in any of the states identified in (2) to (6). By contrast, he would understand “connected” to mean that there is no physical break in the electrical circuit between the turbine and the network. Accordingly (1) to (5) could be said to be connected, although (3) and (4) were only arguably so.

I think it is far from easy to determine what the patentee intended by the use of the terms “operated” and “connected” in the context of the claims of 564 because the body of the specification does not explain how the system should behave once the network voltage has fallen below Umin or risen above Umax. Nevertheless I have reached the conclusion that the skilled person would understand the terms to embody the same concept for essentially the reasons advanced by Mr Wobben. First, and as I have explained, the body of the specification is concerned with a description of an invention which provides a method of operating the system in such a way that voltage over fluctuations and the need to disconnect are avoided. Hence the patent appears to contemplate that operation and connection are concerned with the same state of the system.

Secondly, claim 1 is a method claim. It is concerned with a way of operating the system such that no power is emitted to the network when the network voltage is outside the predetermined range. Claim 3, on the other hand, is a product claim directed to the wind energy system itself. This calls for a system which does not emit any power whilst remaining connected. I think the skilled person would be surprised if the patentee intended the method and product claims to be directed to different concepts. He would think it more likely the patentee intended the two expressions to convey the same concept, namely that the system remains connected and so in operation despite emitting no power. I therefore conclude that “operated” in claim 1 means “connected”.

As to the meaning of “connected”, I think this would be given its ordinary meaning in an electrical context, namely that the wind energy system must be electrically connected to the network. It would not be so connected if the circuit breaker is opened. Nor would it be connected in the event that either the inverter or the rectifier is turned off.

I begin with a point of agreement. Both experts expressed the view, which I accept, that the skilled person would understand the specification and the claims to be concerned with real, as opposed to reactive, power.

As to the expression “without any power”, Mr Wobben contended it means without any significant power relative to the output power of the turbine. Vestas contended it means essentially no power. In my judgment the skilled person would not understand the expression to mean the system must emit precisely no power. It is unlikely the control systems would be sufficiently sensitive to ensure that no power at all would be emitted. However, subject to this small and for practical purposes de minimis exception, I see no reason to give the words anything other than their ordinary meaning. I do not accept that “without any power” means any significant power relative to the power of the turbine. There is no basis in the specification for introducing any such gloss. To the contrary, in so far as there is any discussion of this feature in the specification it explains that when the value Umax is reached the power fed into the network is equal to zero. I therefore accept the submission advanced by Vestas.

The first question is whether the requirement is conjunctive or disjunctive. Must the system be such that it will remain in operation and connected both when the network voltage is greater and less than the predetermined network voltage values or is it enough that it will do either? Both sides submitted that the requirement is disjunctive and I agree. It is enough that the system will do either.

The second question is whether the claim requires the wind energy system to be operated without any power being emitted when the network voltage is greater than Umin or less than Umax or the other way round. Vestas contended that it means the former and relied in support of this argument on the fact that the claim is written in that way. I have no hesitation in rejecting this contention. On any basis the specification contemplates power being emitted when the network voltage is between Umin and Umax. The skilled person would appreciate the claims are concerned with the area of operation above Umax and below Umin.

A predetermined network voltage value is a voltage value which is determined in advance of the method being performed or the system operated. In the system of the patent it is the value above which (in the case of Umax) or below which (in the case of Umin) the system will remain connected and operating but emit no power. The skilled person would understand that these may be the voltage limits imposed by the network operator or, at least in the case of the upper limit, the upper voltage limit of the equipment.

Vestas originally contended this element required a measurement to be taken on the network and nowhere else. I reject that contention. In order for the control system to take account of the network voltage it is necessary to input a figure which corresponds to what is happening to the network voltage. But that figure can be derived from a measurement taken on the turbine side of the grid transformer. Such a measurement is equivalent to a measurement of the actual voltage on the network.

This feature of claim 2 seems to me to be a consequence of the fact that the wind energy system continues to operate. It means the power supply can be resumed once the network voltage value has dropped below Umax or risen above Umin. If the claims encompass major fault conditions which, in my judgment they do, then, as Dr Taylor explained, one is concerned with remaining connected for only a period of 100-200 ms before the fault is cleared. This would require a combination of Professor Green’s option 5 (control of the voltage at the inverter) with only a small degree of option 2 (pitching of the blades). If the fault did not clear then the wind energy system would disconnect. If the fault did clear then emission of power could be resumed in about 0.5 seconds – that is to say, almost instantaneously.

In the case of the network voltage exceeding Umax, for example because of a low load on a windy night, then the position is very different. Here, as Dr Taylor accepted, the concept of spinning reserve and the production of near instantaneous power does not apply because the blades of the turbine would have to be pitched to prevent the turbine generating unmanageable amounts of power and it would take seconds if not minutes to resume power production.

There can be no doubt claim 2 is concerned with fluctuations both above Umax and below Umin. Accordingly, I can see no basis upon which the claims should be limited to systems which can emit power the instant the network voltage is restored to a value within the predetermined network voltage tolerance limits or otherwise between Umin and Umax.

The validity of 564 is attacked for:

The allegation of added matter can be stated quite shortly. Vestas contended that the specification of 564 discloses that the wind energy system is operated or connected to the electrical network without any power being emitted when the network voltage is less than a predetermined network voltage value Umin and that there is no such disclosure in the 324 application.

As I have explained, the body of the specification of each of the two documents is for practical purposes the same. The allegation is founded on the claims of 564 which, it is said, are framed in such a way that they extend to and describe the concept of the wind energy system remaining connected but emitting no power during a major voltage depression caused by a severe fault (LVRT) when this is simply not taught in the application.

The test for added matter is well known and was explained by Aldous J in Bonzel v Intervention Ltd [1991] R.P.C. 553 at 574:

"The decision as to whether there was an extension of disclosure must be made on a comparison of the two documents read through the eyes of a skilled addressee. The task of the Court is threefold

(a)

To ascertain through the eyes of the skilled addressee what is disclosed, both explicitly and implicitly in the application.

(b)

To do the same in respect of the patent as granted.

(c)

To compare the two disclosures and decide whether any subject matter relevant to the invention has been added whether by deletion or addition. The comparison is strict in the sense that subject matter will be added unless such matter is clearly and unambiguously disclosed in the application either explicitly or implicitly."

In European Central Bank v Document Security Services Inc [2007] EWHC 600 (Pat) I elaborated on those principles in paragraphs [97]-[102] of my judgment in the following terms:

“97.

A number of points emerge from this formulation which have a particular bearing on the present case and merit a little elaboration. First, it requires the court to construe both the original application and specification to determine what they disclose. For this purpose the claims form part of the disclosure (s.130(3) of the Act), though clearly not everything which falls within the scope of the claims is necessarily disclosed.

98.

Second, it is the court which must carry out the exercise and it must do so through the eyes of the skilled addressee. Such a person will approach the documents with the benefit of the common general knowledge.

99.

Third, the two disclosures must be compared to see whether any subject matter relevant to the invention has been added. This comparison is a strict one. Subject matter will be added unless it is clearly and unambiguously disclosed in the application as filed.

100.

Fourth, it is appropriate to consider what has been disclosed both expressly and implicitly. Thus the addition of a reference to that which the skilled person would take for granted does not matter: DSM NV's Patent [2001] R.P.C. 25 at [195]-[202]. On the other hand, it is to be emphasised that this is not an obviousness test. A patentee is not permitted to add matter by amendment which would have been obvious to the skilled person from the application.

101.

Fifth, the issue is whether subject matter relevant to the invention has been added. In case G1/93, Advanced Semiconductor Products, the Enlarged Board of Appeal of the EPO stated (at paragraph [9] of its reasons) that the idea underlying Art. 123(2) is that that an applicant should not be allowed to improve his position by adding subject matter not disclosed in the application as filed, which would give him an unwarranted advantage and could be damaging to the legal security of third parties relying on the content of the original application. At paragraph [16] it explained that whether an added feature which limits the scope of protection is contrary to Art 123(2) must be determined from all the circumstances. If it provides a technical contribution to the subject matter of the claimed invention then it would give an unwarranted advantage to the patentee. If, on the other hand, the feature merely excludes protection for part of the subject matter of the claimed invention as covered by the application as filed, the adding of such a feature cannot reasonably be considered to give any unwarranted advantage to the applicant. Nor does it adversely affect the interests of third parties.

102.

Sixth, it is important to avoid hindsight. Care must be taken to consider the disclosure of the application through the eyes of a skilled person who has not seen the amended specification and consequently does not know what he is looking for. This is particularly important where the subject matter is said to be implicitly disclosed in the original specification.”

In the present proceedings both parties accepted this to be an accurate statement of the law.

The allegation of added matter is founded on the following key passages which appear in claims 1 and 3:

In neither case do these words appear in the application. Yet Mr Wobben contended that they are disclosed implicitly, if not explicitly, in the text and in figure 3. I have discussed the disclosure of the 324 application in detail earlier in this judgment and accordingly now simply highlight those aspects upon which Mr Wobben placed particular reliance. At the outset he emphasised the fact that paragraphs 4 to 7 of the application contain a description of the problems facing those operating the prior art systems and these include severe unwanted changes in the network voltage which might lead to disconnection. I accept that this is so but, as the application explains, it is the object of the invention to avoid such over fluctuations and this is to be achieved by providing a system in which the power supplied by the generator is regulated as a function of the network voltage.

Paragraph 10 was also the subject of considerable attention. It is undoubtedly an important paragraph which explains that the invention is also concerned with voltage fluctuations which occur on the consumer side and draws particular attention to critical voltage fluctuations – which I accept the skilled person would understand to include faults for the reasons I have given earlier in this judgment. However, the teaching of the application is that the invention provides a way of regulating the network voltage to the desired reference value and so once again avoiding such undesirable fluctuations and thus the need to disconnect. There is no teaching here of a system which will deal with severe voltage depression by remaining connected for a period of 100-200 ms (or even longer) whilst delivering no power. LVRT does not avoid the undesirable fluctuation at all; rather, it provides a means of surviving the critical event and then resuming delivery of power as soon as possible thereafter.

Paragraph 13 is also an important paragraph which I have set out in paragraph [79] of this judgment. It discusses supplying an increased amount of power to the network to compensate for voltage reductions caused by consumers drawing large amounts of power. Mr Wobben suggested that this is not part of the invention. I disagree. I have no doubt the skilled person would regard it as a description of how the invention makes it possible to compensate for voltage fluctuations on the consumer side.

Finally, there is figure 3. I would emphasise that it is extremely important to consider this figure in the absence of any understanding of the claims of 564. Absent those claims I think the disclosure is extremely confusing for the reasons I have given in paragraphs [81] – [94] of this judgment. It certainly discloses that no power is emitted when the network voltage drops below Umin but neither the figure nor the associated description provides any clear and unambiguous disclosure that the wind energy system must remain operating and connected for 100-200 ms in the case of a severe collapse of the network voltage, that is to say that it has a LVRT capability. It seems to me that further support for this conclusion is gained from the following matters. First, and considering the right hand side of figure 3, a skilled person would understand this to be representative of conditions involving low power consumption and high power availability. When the network voltage has risen above Umax, it would not be practicable to simply “dump” the generated power for other than a very short period of time and it would be necessary to allow the turbine to “spin light”, as in option 2 referred to in paragraph [100] above. However, as Dr Taylor explained, it is not possible to come back and emit power quickly from this position. Hence I believe the skilled person would look at figure 3 and its description, including specifically the last sentence of paragraph 22 of the application, in this light. I do not believe he would understand the application to be teaching clearly and unambiguously that at all values above Umax the wind energy system must necessarily remain operating and connected. The skilled person might choose to remain connected for a while or he might choose to disconnect. So also, there is nothing in the application to suggest he should understand the teaching in respect of the left hand side any differently. There is nothing to tell him the wind energy system must remain operating and connected in the face of the network voltage dropping below (and possibly considerably below) Umin.

Second, I believe the skilled person would be very puzzled as to what the left hand side of figure 3 was intended to represent. However, and as I elaborate in considering the allegation of insufficiency, he would know that it would be extremely challenging to implement LVRT in a wind turbine. He would recognise it as an onerous task. Yet the application does not give any explanation as to how it might be achieved. Indeed, it does not mention it expressly at all. In these circumstances I think he would be very surprised to be told that the invention described in the application was precisely that, namely the provision of LVRT in a wind turbine.

For all these reasons I have come to the conclusion that the application does not contain a clear and unambiguous disclosure of a wind energy system which is operated and remains connected without any power being emitted to the electrical network when the network voltage is less than a predetermined network voltage value (U_min). It follows the patent is invalid for added matter.

Before leaving this topic I should briefly mention one further allegation. It was said that in so far as figure 3 is relevant, it only discloses a method and system in which no power is emitted at both Umin and Umax. In short, the figure cannot provide support for a selection of one side of the figure only. Had I decided the primary allegation in favour of Mr Wobben I do not believe this would have constituted added matter. There is nothing in the application to lead the skilled person to believe that the system must necessarily incorporate both sides of the figure.

Overtoner was published in 1996. It is part of a Danish study called Harmonics and Operating Conditions of Variable Speed Wind Turbines and forms part of a larger report entitled Power Quality and Grid Connection of Wind Turbines. The purpose of the report was to explore the effect of full converter turbines on the quality of power in a network. It points out that the use of frequency converters between the network and wind turbines is a new technique and it was therefore considered necessary to carry out a renewed evaluation of the equipment required to ensure a proper operation and a safe de-coupling in the event of failures. It then proceeds to describe the demands for protection of wind turbine installations.

Specifically Overtoner describes the de-coupling of a wind turbine in circumstances where

Overtoner explains the meaning of de-coupling on page 3:

“The disconnection from the network can be effected either by opening the main switch between the inverter and the network, or by stopping the control signals to the inverter stitches”

Overtoner therefore describes stopping the emission of power using either Professor Green’s option 4 (stopping the control signals to the inverter) or his option 6 (opening the main circuit breaker).

As I have construed the 564 patent, it is clear that Overtoner does not deprive it of novelty. It does not disclose a wind energy system which is operated (in the case of claim 1) or connected (in the case of claim 3) without any power being emitted to the electrical network. To the contrary, Overtoner teaches disconnection once the network voltage has risen above or fallen below specified network voltages for a particular period of time.

Heier is entitled Wind Power Plants in Grid Operation and was published in 1996. Two sections are relied upon, the first headed “Protective Measures” and the second headed “Management System”.

“Protective Measures” explains that faults may occur in the grid or the turbine which cause voltages and currents to be so high that functional groups in the frequency converters of the wind energy systems can be destroyed. This requires the constant monitoring of all relevant variables. Moreover, the control devices of rectifiers and converters must recognise anomalous plant conditions and initiate suitable protective measures. Having described problems that can occur on the generator side or in the converter, it continues in respect of the grid side:

“According to these regulations [of the power supply companies], frequency converters on the grid side must recognise voltage and frequency changes in order to prevent unintentional isolated operation. Wind power plants and thus also their frequency converters must disconnect from the grid immediately in the case of over- and undervoltages outside the stipulated limits, or rapid auto-reclosing in the grid. Systems for the recognition of these faults can be integrated into the plant control system or management system or designed as external units.”

In this passage Heier recognises the problem of islanding which I have discussed in paragraph [62] of this judgment. He also discusses the operation of auto-reclosing. As Professor Green explained, the network protection system which detects, locates and isolates faulty equipment uses circuit-breakers that can automatically re-close after opening to test whether a fault is transitory or permanent. However, if the initial opening of the circuit-breaker caused an island to form, and if that island adopted a slightly different frequency from the main grid, then the recloser would seek to join two systems that are out of synchronism. This is a dangerous event that must be avoided by using circuit breakers that check for synchronism before reclosing or by requiring embedded generation to be disconnected after the circuit breaker has opened.

It was in these circumstances that Professor Green accepted that it was usual for a network operator to require “embedded” generation to disconnect in the event of a fault that isolated the part of the distribution network to which that plant was connected – so as to avoid unintentional isolated operation.

Turning to the “Management System” section, Heier explains that the management system must monitor the relevant system components and variables, recognise faults or emergency situations and bring about various operating states. In a passage upon which both sides relied, Heier states in relation to grid failures and rapid auto-reclosure:

“…grid failures can only be recognised by the frequency converter. It must therefore shut down immediately and send a message to the management system. As the turbine’s generator is no longer opposed by a load moment, the speed increases. Using blade angle adjustment and, if necessary, the brake (in the upper speed range), the speed must be run down into the waiting state. As soon as all conditions (grid OK, among others) are again fulfilled, running up can be automatically initiated once again.”

Heier therefore describes a system in which once a grid failure is detected the inverter or rectifier is switched off immediately. In this condition it is electrically disconnected. Then, using blade angle and, if necessary, a brake, the system will be run down into the waiting state and the circuit breaker may be opened. From this state the rotor speed can either be run up to a value at which it is possible once again to connect the wind energy system to the network or it can be shut down.

It is apparent from this description that Heier does not deprive 564 of novelty. It teaches disconnection in the event of a fault. It does not disclose a wind energy system which is operated (in the case of claim 1) or connected (in the case of claim 3) without any power being emitted to the electrical network.

The correct approach to the issue of obviousness has recently been re-stated by the Court of Appeal in Pozzoli v BDMO SA [2007] EWCA Civ 588:

I have identified the person skilled in the art and the common general knowledge earlier in this judgment. As to step ii), this is a case where I think it is convenient to take the inventive concept as being what is claimed. Steps iii) and iv) must be considered in the light of each of the specific allegations.

The differences between the common general knowledge and what is claimed can be shortly stated. It was not common general knowledge for a wind energy system to be operated or connected without any power being emitted to the electrical network when the network voltage was greater or less than a predetermined network voltage value (Umin, Umax), or to operate such a system in which power could always be emitted when the network voltage value had once again assumed a value between those predetermined values.

There were two attacks that it was obvious to take those steps. The first, and major attack, was at the heart of the claim. It was said it was conceptually obvious to design a wind turbine with LVRT capability. The second was that it was obvious to continue to operate a wind turbine at or above Umax. I will address them in turn.

There can be no dispute that it was part of the common general knowledge that voltages on transmission and distribution networks must be controlled and that this was achieved by reference to minimum and maximum tolerance limits. From this point, the argument that it was conceptually obvious to design a wind turbine with LVRT capability ran as follows. It was well known that that the average capacity of installed wind turbines and the rotor diameter and rated capacity of new wind turbines were increasing year on year. Network operators had a duty to ensure that their networks were run safely, securely and in accordance with statutory requirements and the relevant grid codes. In 1997 the size of wind energy installations in the UK was such that they were only subject to a distribution grid code. However, the transmission grid code applied not only to power plants that connected to the transmission network but also to any embedded power plants connected to the distribution network which had a capacity of greater than 50MW. Hence if any wind farms of a capacity in excess of 50MW had been connected to a distribution network or if any wind farms had been connected to a transmission network then they would likely have been required to comply with the transmission grid code and remain connected during a fault. The skilled person therefore knew that once wind farms reached a certain size, LVRT was not only desirable but potentially mandatory. It was therefore conceptually obvious in 1997.

Mr Wobben responded that this is a classic hindsight step by step attack on the patent, and one which is not supported by the evidence. In 1997 the mindset of the skilled person was to disconnect at the first sign of trouble. Further, the wind turbine industry was taken aback and dismayed when, much later in 2002, it was suggested that LVRT would be required. Plainly this was not obvious to the skilled person some five years before.

In assessing any obviousness attack it is clearly important to avoid hindsight: Ferag v Muller Martini [2007] EWCA Civ 15 at [13]. This is particularly so in a case where the attack is based upon the common general knowledge: Coflexip v Stolt Connex Seaway (CA), [2000] IP & T 1332, at [45], for the reason that such an attack may be unencumbered by any detail which might point to non obviousness.

I also recognise that it may be inventive to break out of a mindset that imprisons those skilled in the art. Dyson v Hoover [2002] RPC 22 is a good example. There the vacuum industry was described as functionally deaf and blind to any technology which did not involve a replaceable bag. So also there may be invention in overcoming a prejudice held in a particular industry. As Jacob J explained in Union Carbide v BP Chemicals [1998] RPC 1 at 13, invention can lie in finding out that which those in the art thought ought not to be done, ought to be done. However, it is also important in assessing any such argument to consider all the circumstances. It may well be that the explanation for those in the industry not taking a step at a particular moment in time is that the technology was not yet available to permit it to be taken or that the industry was held back by other limiting factors which meant that the necessary investment to bring it about could not be justified. All must depend upon the facts of each individual case.

It is convenient to begin by considering the issue of the mindset of the skilled person in 1997. I have addressed aspects of this issue and set out a number of preliminary conclusions in paragraphs [65] to [69] of this judgment. I have rejected the submission that wind turbine engineers were conservative or resisted change. To the contrary, the rated capacity of turbines and wind farms was increasing rapidly and wind turbine technology was itself becoming increasingly sophisticated with those in the art fully aware of the existence and general mode of operation of the DFIG and full converter turbines. Machines such as the E40 monitored the grid frequency and voltage and incorporated power electronics. These features permitted the wind turbine operators to vary the real and reactive power output of the turbines and so respond to changing conditions on the network and allow active voltage regulation with a view to maintaining and stabilising the network voltage. Moreover, the market profile of the various wind turbines was changing, with the years from 1997 seeing a rapid growth of sales of DFIGs and full converter turbines.

Nevertheless, I have accepted it was the case that wind turbines did disconnect in the event that the network voltage strayed outside the specified limits. However, it is important to note that this was not a reflection of a blinkered approach to design by wind turbine manufacturers as in Dyson. Nor was it the result of a prejudice of the kind discussed in Union Carbide. Rather, it was a requirement imposed by the network operators who treated the wind turbines as a negative load and were concerned about islanding. Consequently I do not accept that the skilled person did not look ahead or that he had a mindset which was to connect a small number of modest capacity wind turbines to the distribution network, as suggested by Mr Wobben. The reality was quite the opposite. The industry was developing larger, more powerful and more sophisticated turbines which used power electronics to support the distribution grid. Moreover, the power contribution made by these turbines was increasing year on year and the wind farms themselves were growing ever larger. The time would come when the network operators could no longer treat wind farms as they had done in the past.

Against this background the question that arises is whether it was technically obvious to implement LVRT capability in a wind turbine: Hallen v Brabantia [1989] RPC 307; [1991] RPC 195 (CA). In considering this question the evidence of the experts is of particular importance. Professor Green’s position was straightforward. He maintained that the skilled person would have appreciated in 1997 that if any wind farms exceeded 50MW capacity or were connected to the transmission grid then they would have been required to remain connected during a fault. It was therefore conceptually obvious to make a wind turbine which had that capability.

Dr Taylor accepted that by 2002 wind farms were expected to support the network and were of a sufficient size that if they were disconnected during fault conditions they could exacerbate the power system’s problems. Nevertheless he maintained that it was not obvious in 1997 to design a wind turbine with LVRT capability because the mindset of the industry was to disconnect at the first sign of trouble and because the necessary technology had not been developed. But as he was questioned further I believe it became apparent that his real concern was the lack of the necessary technology. So, for example, the following interchange took place on day 7 at 776-778:

“Q. What actually happened was that they said we will make wind farms behave like conventional power stations -- is that not right -- and make them provide low voltage ride-through?

A. That is what happened, yes.

Q. That is innovative as well, in your view, to say behave like conventional power stations?

A. It was not innovative, particularly, it was a requirement. It was a performance requirement they decided they needed. It is not innovative to ask for something but it can be innovative to come up with a good way of delivering it.

Q. Let me go back to my original question. There was nothing innovative from a technical perspective in realising in December 1997 that once wind farms got to a certain size the issues which you mention in the paragraph of your expert's report that we are looking at would need to be addressed. There is nothing innovative from a technical perspective about that.

A. There is nothing innovative about understanding that issues would be raised as wind power grew. There is something innovative about what is described in the 564 patent, which is to use a wind energy system to remain connected under severe voltage disturbances and then emit power again as soon as the voltage recovers.

Q. What the 564 patent says is that wind farms should do exactly what conventional power stations had been doing hitherto. Is that not right?

A. But it is much harder for a wind energy system to do that than it is for a conventional power station to do it. Technically, you have to use different techniques and, technically, it is a harder job to do it with a wind energy system than it is with a conventional power plant.

Q. Please assume we are not concerned with how you are going to do it because there is nothing in the 564 patent which applies any significant guidance on how. We are just at this conceptual level as to whether it would be an obvious idea to apply to wind farms what has been required of conventional power stations before December 1997?

A. The reason I was answering like that is because you asked me if it was technically innovative, and you said it was not; so I was trying to show you that it is. It is technically innovative to deliver this with a wind energy system, because it has not been done before and it is difficult to do because of the constraints of the technology.”

The cross examination continued and Dr Taylor maintained that there were other ways the issue could be addressed, for example by providing more spinning reserve, putting a cap on the size of the biggest wind farm or taking care to locate wind farms in less sensitive locations. However he accepted that one obvious way of dealing with the growth in capacity was for wind farms to behave in the same way as conventional generators, as the following further interchange on day 7 at 784-787 demonstrates:

“Q. You say: "From my recollection of the history and development of wind energy production (and as confirmed in the textbooks I referred to in section 5), it is clear that wind turbines and therefore wind farm capacities have grown significantly since 1980. In 1980, a typical turbine rating was 50kW. It is now around 3MW. This means that wind farms can no longer be treated as a negative load. They are now expected to contribute to the operation of transmission networks or power systems." That is the position, is it not?

A. That is the position now, yes.

Q. No. That is the position when they get to 3 megawatts. It would have been ----

A. It is not that the ---

Q. Sorry, let me finish. That is the position when they get to 3 megawatts. If they got to 3 megawatts that much sooner, then they would have been expected to contribute to the operation of the transmission networks or power system that much sooner?

A. There are two things here. First of all, it is not that a turbine is 3 megawatts that matters. It is the size of the wind farm. It does not matter whether the turbine is 3 megawatts or whether the turbine is 50 kilowatts. It just means you need more turbines to produce a big wind farm. It is the size of the wind farm in total that matters. So it does not matter that we were at turbines of 3 megawatts particularly. What it means is if you put a wind farm of 20, 3 megawatt wind turbines up, that is the first thing to point out.

The second thing is it is not obvious in '97, even if you can look forward and realise that it is going to cause issues, that the way you are going to try to solve it is by asking them to behave exactly the same way as conventional generators. You could just provide more spinning reserve, instead, and not ask the wind farms to do it. You could put a cap on the size of the biggest wind farm. You could locate the wind turbines in places that were less sensitive, for example. So there are a number of different things that you could have done. It is not obvious in '97 that you would definitely ask them to do it, and it is definitely not obvious with most of the technology that they could do this, either.

And, do not forget, the Grid Code was written around the capabilities of large synchronous plant. So then just to take those codes that were written around the capabilities of large synchronous plant and just apply them to wind farms was not obvious and was actually met with dismay by the wind industry when it looked like that was what they were going to try to do.

Q. But one obvious way of dealing with it, apart from what you say, spinning reserve, a cap on wind farms, is simply to say, "Well, we will have the wind farms behave in the same way as conventional power generators." At its conceptual level, that is one obvious approach to the problem, is it not?

A. That is one way you could do it.

Q. One obvious way you could do it?

A. Yes.

Q. Yes?

A. Yes.

Q. That would involve, if possible -- and you have to look at the possibilities -- staying connected to the network in the case of faults?

A. Yes.”

The thrust of Dr Taylor’s evidence was that in 1997 it would have been apparent to the skilled person that the growth in wind farms would require management, and that one of the obvious ways to manage them would be to require them to behave in the same way as conventional generators, including the capability to remain connected in fault situations. On the other hand, implementation of this capability would have been an altogether different matter, and one which presented a major challenge. Dr Taylor elaborated this on day 7 at 778-779:

“Q. The patent, the 564 patent, does not help you with the question of how does one do this with a wind farm.

A. The 564 patent, first of all, its main teaching is that you can and should think about doing this, which, as I say, in 1997 is innovative. And then it also shows you the kind of measurements you would need, the fact that you would use a microprocessor and the fact that you would use power electronics to do it.

Q. The fact that you could use one of the wind turbines which was on the market was common general knowledge, yes?

A. The fact that you could use a turbine that was out on the market -- I think we have already said that you would need to do further development to the E40, if that is what you are referring to, to make it be able to behave in the way described in the patent.

Q. Is that development work? That would have to be innovative, would it, to be able to function in accordance with the 564 patent claims?

A. You would have to come up with new ideas, some new control algorithms.

Q. Innovative or not?

A. Broadly speaking, yes.”

At this point it is important to remember that the 564 patent gives no details as to how the invention is to be implemented beyond what was common general knowledge in 1997. Figure 4 of the patent shows the major components of the control and regulation arrangement but, as I have mentioned, this is to be found in the promotional material for the E40. The 564 patent is therefore directed to a concept rather than a method of implementation and in my judgment the expert evidence taken as a whole indicates that the concept was obvious.

I must also take into account the reaction of the industry in 2000-2002 to the suggestion that turbines should remain connected. As Dr Taylor explained, the suggestion met with considerable resistance. For example, the British Wind Energy Association (the BWEA) recruited consultants to explain that the impact on the industry would be severe, indeed it would kill the industry because LVRT would be very difficult to implement. The paper by Johnson and Tleis to which I have referred in paragraph [68] is to the same effect. Mr Wobben submitted this demonstrated not only the mindset for which he contended but also that no-one had taken steps to implement the concept of LVRT in wind turbines before that time. Hence it could hardly have been obvious. Similarly, a 2006 report entitled DISPOWER, prepared by a broadly based European consortium of research institutes, universities, manufacturers, the power industry and distribution and transmission system operators, noted that “latest developments” showed “a change of paradigms” and renewable energy sources and distributed generation not only being perceived as a burden but as integral parts of electricity networks and able to contribute to at least some system services. It also explained the “disconnect at the first sign of trouble” approach was no longer acceptable. Once again it was submitted this showed the persistence of the mindset long after 1997.

These are powerful points but on close consideration I do not believe they are inconsistent with the conclusions I have drawn from the expert evidence. I have accepted that in 1997 the skilled person considered that he had to disconnect if the network voltage strayed outside the predetermined limits. This was more than a paradigm or approach in that the requirement was imposed on him and would likely remain so until the wind farms approached 50 MW capacity. At that point it was plain the position might change. I also accept that the industry resisted the notion of having to remain connected in severe fault situations. But that does not mean that the concept of remaining connected in severe fault situations was innovative or that the industry was stuck in a mindset. Indeed I think it equally consistent with the industry anticipating it would be very difficult to implement LVRT on wind turbines. I think this is also an explanation as to why manufacturers were not ready with developed systems in the period 2000-2002 when the question became actively debated. Consultations followed. But manufacturers resisted the imposition of a LVRT requirement. Had it been routine to implement it is hard to see what the basis for that resistance could have been.

For all these reasons I have come to the conclusion that the technical contribution contained in the 564 patent is no more than a concept and that the concept was obvious. However, as I explain in considering the allegation of insufficiency, I do not accept that the implementation of LVRT could be implemented without undue difficulty and this necessarily affects whether the claims are invalid for obviousness in the light of this particular allegation.

The second attack can addressed much more shortly. It was submitted it was obvious to stay connected and operating as the network voltage approaches Umax and as Umax is exceeded.

It is conceded that it was obvious in the light of, inter alia, the promotional material for the E40 to reduce the real power output of the turbine as Umax is approached, as, for example, in the case of conditions of high power output and low load on a weak grid. This is a very different situation to LVRT, taking place over a longer time scale and necessarily involving pitching the blades of the turbine because of the impracticality of dumping power for any length of time. For these reasons Dr Taylor considered there is little point in remaining connected once Umax is approached, let alone exceeded, and he so explained in relation to the Tande publications which I address later in this judgment. However, that is exactly what the claim covers. Professor Green considered that there was nothing conceptually different about using a wind turbine rather than any other type of generation to control network voltages. He also considered that in the event that Umax is reached it was obvious to continue to operate for a while in case the load increased.

In the light of all this evidence I have reached the conclusion that it was technically obvious not only to reduce the power emitted as Umax is approached but to remain connected and emit no power when it is exceeded, at least for a while and so that the power supply can be resumed in the event of an increase in the load. In my judgment that is within the scope of the claims. There is no suggestion that there would be any undue difficulty implementing such a system in the light of the availability of the E40. All the claims are therefore invalid for obviousness.

It is convenient to deal with this citation next because the allegation of obviousness based upon it is closely related to that based upon the common general knowledge.

This document was written by Mr Wobben and presented at a conference in 1996. As Dr Taylor explained in his first report, it discusses the E40, 500KW, wind turbine and how its characteristics and controllability make it suitable for power system integration. It expressly looks to the future and states that if a significant amount of wind power is to be used, wind turbines will have to be able to behave in a similar fashion to traditional large scale power plant. Specifically it says at the outset:

“Wind energy converters have increased exponentially over the last few years with regard to numbers as well as power installed. Therefore in the interest of the electric utility and the customer it is absolutely necessary to have a grid-compatible power output.”

The document considers the need to reduce the power output in off-peak periods and then describes the characteristics of 44 installed turbines at the Fehmarn wind farm and how they can be actively managed. In doing so it expressly anticipates wind farms in the future with a capacity of more than 50MW.

On page 3, the document notes, albeit in somewhat optimistic terms:

“If we succeed in actively operating wind parks on the grid like a power plant, it will be possible to increase today’s power of passively operating machines from about 10% of the supply power up to well over 50% ”

And it concludes:

“The bigger the size of a wind energy converter the more they will be put under the criteria of conventional power stations .”

It is important to note this document does not refer to remaining connected under fault conditions; indeed the E40 did not have that capability. However it anticipates the growth of wind farms and the imposition upon them of the criteria applying to conventional power stations. As Dr Taylor accepted in the evidence cited above, one obvious aspect of that was remaining connected under fault conditions and hence having the LVRT capability. In my judgment this publication confirms the conclusion that the concepts claimed in the 564 patent were technically obvious.

It also confirms my conclusion in relation to the secondary case. Dr Taylor accepted in his first report that D5 discloses reducing power emission after a specific network voltage has been exceeded. For like reasons to those given in relation to the attack over the common general knowledge, it was, in my judgment, obvious to gradually reduce that power to zero as Umax was approached and to remain connected, at least for a period, if it was exceeded

This is another Enercon publication and it promotes the E40. It describes the grid management system incorporated in the turbine and states, on page 11, that the system can “adjust output as a function of the existing grid voltage” and so help to stabilise grid voltage and frequency. The system has the ability to adjust the power output, voltage, power factor and frequency. It depicts a figure of the control system and this is the same as that appearing in the 564 patent as figure 4.

In my judgment the contents of this publication were common general knowledge. But in any event it confirms the case of obviousness over the common general knowledge. It reveals that the 564 patent contains nothing new other than the concept of remaining connected once the network voltage has strayed outside the predetermined limits.

This publication also confirms my conclusion in relation to the secondary case. Dr Taylor accepted that it renders obvious the concept of power reduction in situations of over voltage. It was equally obvious to reduce the power to zero as the network voltage approached Umax and to remain connected, at least for a period, if it exceeded Umax.

I have discussed the disclosure of Overtoner in paragraphs [129]-[133] of this judgment. It describes the conventional approach of disconnecting the wind turbine in the event the network voltage strays outside the predetermined limits. It achieves this by stopping the control signals to the inverter (Professor Green’s option (4)) or opening the main circuit breaker (option (6)). It does not disclose a wind energy system which is operated or connected without any power being emitted to the electrical network when the network voltage is greater or less than a predetermined network voltage value.

On the face of it Overtoner is therefore not a promising starting point for an allegation of obviousness and it appears to add nothing to the case based upon the common general knowledge. Nevertheless, it was argued by Vestas that Professor Green’s option (5) was part of the common general knowledge and an obvious way of preventing the emission of power as an alternative to disconnection. It based this argument upon the following evidence of Professor Green given in paragraph 1.0.11 of his third report:

“As regards option (5) in particular, it was well-known that an inverter controls the export of power by varying the voltage at its output terminals relative to the voltage at the connection point. In 1997, I understood that a wind turbine with a power electronic interface (such as the Enercon E40) would have made the transition between exporting power and not exporting power at low wind speeds by implementing option (5). I would expect the skilled person in 1997 to have assumed that option (5) is how this transition is achieved if he did not already know that to be the case.”

Mr Wobben did not accept that the E40 had such a capability at all. But in any event I do not accept that the low wind speeds referred to by Professor Green would necessarily involve anything other than normal network voltages and the evidence does not establish that the E40 could implement option (5) if the network voltage collapsed or otherwise dropped below Umin.

This position is, I believe, confirmed by the evidence of Dr Taylor in relation to Heier on day 6 at 685-686:

“Q. Sorry, let me get to the end of this -- would there have been anything clever about using option 5 with D2 as an alternative to the frequency converter, which is option 4.

A. I think what I meant, I hope what I have tried to say, was that under normal conditions it is not so difficult to match the volts to the network and therefore emit no power. But if we have a fault, then the voltage is heavily depressed and also it is likely to be unbalanced and it is transiently changing. So you are not just matching it once and leaving it there, you are having to track that voltage around as it changes. You are having to balance it up across the three phases, so you may have to be having one amplitude and one phase on one -- one amplitude and one phase angle on one phase, a different one on the second phase, a different one on the third. That is dynamically moving, it is transient. We are talking microseconds. That is quite difficult -- it is very difficult.

Q. That is power electronics, though, is it not? That is par for the course for power electronics

A. No, that is difficult.

Q. It is difficult but I mean the skilled person knows how to do it.

A. No.

Q. They cannot do it?

A. I think that is a very challenging thing for somebody to do in 1997.

Q. Let us just get clear what they cannot do with option 5 because I thought you were relying on option five for various things. Option 5, when can it not be used?

….

A. I think what I am saying is that in 1997 nobody was remaining connected because during fault conditions and matching the volts on the output of the inverter to the volts on the network under fault conditions. Nobody was doing that in 1997.”

It is apparent from this evidence that it was simply not part of the skilled person’s common general knowledge that he could implement option (5) in place of option (4) in any conditions. I therefore reject this further attack of obviousness.

I have discussed the disclosure of Heier in paragraphs [134] - [140] of this judgment. Once again it does not disclose a wind energy system which is operated or connected without any power being emitted to the electrical network when the network voltage is greater or less than a predetermined network voltage value. The allegation of obviousness based upon Heier was the same as that based upon Overtoner and I reject it for like reasons.

Tande 96 describes the modelling of a situation in which an additional turbine is added to a conventional wind farm comprising Danish type fixed speed 500KW machines. The wind farm is connected to a weak grid and the authors consider the impact of the wind farm on the network voltage in circumstances where the wind farm is able to feed a high level of power but the consumption is low – the typical windy night scenario – so tending to make the network voltage rise.

The paper then describes methods for assessing and minimising the impact of wind turbines on voltage quality. Specifically it identifies four options:

In the context of these proceedings attention was focused particularly on option ii). As to this, the authors explain on page 5 that an alternative to grid reinforcement is to disconnect the wind turbines if their production causes any probability of an over voltage. However, they continue:

“The production losses due to the voltage dependent disconnection of the wind turbine is estimated at 158MWh/year. It should be noted that if the wind turbine was reduced in small steps rather than disconnecting the wind turbine for avoiding overvoltage, less loss of energy would be obtained.”

Accordingly, the primary teaching of the paper is that the power fed into the network is reduced by disconnecting the extra turbine. But it also teaches that disconnection can advantageously be avoided if the output of the turbine is reduced in small steps – through a straightforward application of the steady state equation. However, there is no express teaching that the output of the turbine can be reduced to zero, nor that the turbine can remain connected whilst emitting no power if the network voltage continues to rise.

Professor Green explained in his first report that the skilled person reading the paper would understand that the emission of no real power is intended to be the endpoint on a continuum that begins with a small reduction in the emitted real power and that the wind turbine might remain connected throughout, rather than being disconnected. Dr Taylor was reluctant to accept this but I have no doubt that Professor Green was right. I believe that it would be plain to the skilled person that the authors believe disconnection is undesirable and that the system they describe seeks to avoid such disconnection by regulating the power supplied as a function of the network voltage. As the network voltage rises the power supplied by the wind turbine is gradually reduced by pitching the blades and with the aim of restoring the network voltage to the reference value. However, if the network voltage continues to rise the skilled person would realise that he must continue to reduce the power supplied. Dr Taylor maintained that a point would be reached at which the skilled person would simply disconnect the wind turbine. I accept that would be one of his options. But another option, which would be equally obvious, would be to remain connected and gradually reduce the power supplied to zero. If the network voltage begins to fall the power supply can then be resumed.

In my judgment this is precisely the concept sought to be claimed in the 564 patent in the case of network voltages rising above a predetermined network voltage value. All the claims are therefore obvious in the light of Tande 96.

Tande 97 is again concerned with power control for wind turbines in the case of weak grids and specifically with options for overcoming voltage quality constraints.

In section 1 the authors explain that there exist a number of ways of overcoming such constraints:

Section 2 describes the County Donegal power system. Section 3 explains how the concept of dissipation of wind power has been implemented at the wind farm at Cronalaght. Specifically it explains that the wind farm consists of five 600KW wind turbines, a wind farm control unit and a voltage control unit (VCU) and that:

“The VCU facilitates voltage dependant reduction of the output power of the wind farm. This means that in case there is a risk for unacceptable high voltage at the grid due to the wind farm output power, the VCU gives the wind farm control system a signal to reduce the output power of the wind farm.”

It continues that load flow analyses have shown that in the case of minimum load and maximum wind power output, there is a risk of voltages exceeding the level accepted by the utility. Hence the VCU must limit the power output from the wind farm so that the voltage level does not exceed the accepted level. It then explains:

“It is evident from [Figure 3] that in case the load is less than 40% of its assumed maximum level, the voltage level at the PCC [the point of common coupling] may get critically high. In that case the VCU must limit the wind farm output power as to maintain an acceptable voltage.”

Professor Green acknowledged that the figures in the paper do not show the output power of the wind farm being reduced to zero, but maintained they do show a steady reduction from the 3MW maximum of the wind farm to less than 1MW. He expressed the opinion that it was obvious that in other circumstances (such as if the critical voltage were reached when the load was at 60% of its assumed maximum level rather than 40%), it might be necessary to reduce the output further, down to and including zero.

Dr Taylor, on the other hand, maintained that the VCU works on a wind farm basis rather than at each individual wind turbine and it is unlikely that the wind farm capacity would be such that significant amounts of power are dumped for significant periods of time. Therefore he believed the VCU operated to dump a fraction of the wind farm output infrequently at times of high wind farm output and low local loading. By contrast, the 564 patent operates on an individual turbine and reduces the power to zero.

It is clear that both experts accepted that, as in the case of Tande 96, there is no express teaching that the power output of a wind turbine can be reduced to zero, nor that the turbine can remain connected whilst emitting no power if the network voltage continues to rise. The question is whether or not that is obvious in the light of this paper.

Once again I am satisfied that step was indeed obvious. Dr Taylor maintained under cross examination that it was very hard to envisage any normal situation in which it would be appropriate to reduce the output of the whole wind farm to zero and nevertheless remain connected. That I accept. However he did agree that it was possible to envisage a normal situation involving minimum load and maximum wind power output, where there was a risk of the network voltage exceeding the level accepted by the utility. In these circumstances the skilled person would appreciate he had two alternatives. One would be to limit the power output from the wind farm as a whole. The other would be to disconnect one single turbine or reduce the power output from a single turbine. Reconnection might require anything from one to 15 minutes. Accordingly there was an advantage in remaining connected and adjusting the power output with a view to maintaining the network voltage within the permitted upper limit. As in the case of Tande 96 this would be a continuum down to zero. All the claims are obvious over Tande 97.

During the course of the trial Vestas advanced a case that the E40 anticipated the claims of the 564 patent. The argument ran as follows. In the period between a fault causing a severe drop in the network voltage and the circuit breaker being opened, the E40 remains connected and in operation. During this time the power emitted by the turbine would be zero. Accordingly the E40 falls within claims 1 and 3.

When originally asked the time it would take the circuit breaker to open, Dr Taylor indicated he thought it would take about 100ms. The following day he corrected this to about 40ms. But, as he later explained, in the meantime the E40 has already disconnected by stopping the control signals to the power electronics. This happens almost immediately. I accept the submission advanced by Mr Wobben that the skilled person would not understand the E40 to be operating or connected when going through this procedure. I therefore reject the allegation that the E40 renders the 564 patent invalid for lack of novelty.

Vestas alleged the claims of the 564 patent are insufficient because:

In the end Vestas focused on the second of these allegations. It was right to do so. In general terms it was known how to control the output of wind turbines. For example, the E40, which was common general knowledge, had a control system which enabled the turbine to vary its output of real and reactive power in response to changing conditions on the network.

The law in relation to insufficiency is well established. The key principles are these:

The question whether a burden is undue must be sensitive to the nature of the invention, the abilities of the skilled person and the art in which the invention has been made. At the end of the day the court must consider whether the effort required of the skilled person is undue having regard to the fact that the specification should explain to him how the invention can be performed. Nevertheless a number of cases have sought to provide guidance and a helpful exposition is to be found in the decision of the Court of Appeal in Mentor v Hollister [1993] RPC 7. Lloyd LJ explained at 13-14:

“…..if a working definition is required then one cannot do better than that proposed by Buckley LJ in giving the judgment of the Court of Appeal in Valensi v. British Radio Corporation [1973] R.P.C. 337. After referring to a number of earlier authorities, including Edison & Swan v. Holland, he said:

"We think that the effect of these cases as a whole is to show that the hypothetical addressee is not a person of exceptional skill and knowledge, that he is not to be expected to exercise any invention nor any prolonged research, inquiry or experiment. He must, however, be prepared to display a reasonable degree of skill and common knowledge of the art in making trials and to correct obvious errors in the specification if a means of correcting them can readily be found."

Then a little later:

"Further, we are of the opinion that it is not only inventive steps that cannot be required of the addressee. While the addressee must be taken as a person with a will to make the instructions work, he is not to be called upon to make a prolonged study of matters which present some initial difficulty: and, in particular, if there are actual errors in the specification -- if the apparatus really will not work without departing from what is described -- then, unless both the existence of the error and the way to correct it can quickly be discovered by an addressee of the degree of skill and knowledge which we envisage, the description is insufficient."

In that case there was a mistake in the specification. But Buckley LJ's language is equally apt to cover an omission.

…….

Before leaving the authorities, I should mention No-Fume Ltd. v. Frank Pitchford & Co. Ltd. (1935) 52 R.P.C. 231. Quoting from the judgment of Romer L.J. in that case, Buckley LJ in Valensi is reported as saying:

"The test to be applied for the purpose of ascertaining whether a man skilled in the art can readily correct the mistakes or readily supply the omissions, has been stated to be this: Can he rectify the mistakes and supply the omissions with the exercise of any inventive faculty? If he can, then the description of the specification is sufficient. If he cannot, the patent will be void for insufficiency."'

"With" in that quotation must be a misprint for "without".
I turn to the judge's conclusion on the law. What he said was this:

"The section requires the skilled man to be able to perform the invention, but does not lay down the limits as to the time and energy that the skilled man must spend seeking to perform the invention before it is insufficient. Clearly there must be a limit. The sub-section, by using the words, clearly enough and completely enough, contemplates that patent specifications need not set out every detail necessary for performance, but can leave the skilled man to use his skill to perform the invention. In so doing he must seek success. He should not be required to carry out any prolonged research, enquiry or experiment. He may need to carry out the ordinary methods of trial and error, which involve no inventive step and generally are necessary in applying the particular discovery to produce a practical result. In each case, it is a question of fact, depending on the nature of the invention, as to whether the steps needed to perform the invention are ordinary steps of trial and error which a skilled man would realise would be necessary and normal to produce a practical result."

I have already quoted the remainder of that paragraph.

I can find no vestige of error in that statement of the law. It was at first argued that the skilled man should not have to carry out any research, enquiry or experiment at all, whether prolonged or otherwise. But Mr. Thorley subsequently retreated from that extreme position. There is no support for setting so high a standard of disclosure, whether in Valensi itself or in any of the previous authorities, save possibly the judgment of Lindley LJ in Edison & Swan v. Holland. When, a little later, Aldous J came to apply the law to the facts of this case, he refers to "routine trials" and "normal routine matters that the skilled man would seek to do and would be able to do". Mr. Thorley criticises the use of the word "routine". To require the performance of routine trials is, he said, to ask too much of the addressee. I do not agree. "Routine" is just the word I would have chosen myself to describe the sort of trial and error which has always been regarded as acceptable; and "routine trials" has the further advantage that it is a positive concept, which is easily understood and applied. In practice, therefore, it may provide a surer test of what is meant by "clearly enough and completely enough" in section 72(1) of the Act than the negative test proposed in Valensi, namely the absence of prolonged research, enquiry and experiment. If the trials are unusually arduous or prolonged, they would hardly be described as routine.”

In this case Dr Taylor emphasised in the course of his evidence that he considered the essence of the invention to be the provision of a wind turbine which could deal with severe voltage disturbances associated with fault conditions. He arrived at that conclusion because the claims require the turbine to remain connected whilst emitting no power which, to him, meant LVRT. He also described the wind turbine as entering a “spinning reserve” state because the turbine is in a position to resume the emission of power as soon as the fault is cleared and the network voltage restored.

As I have indicated, I agree that the claims of the 564 patent cover (but in my view are not limited to) this condition. The question I must therefore decide is whether the disclosure of the specification was sufficient to allow LVRT to be implemented without undue burden in 1997.

Professor Green gave rather ambiguous evidence in his first report. He expressed the belief that although the skilled person would be able to cause a wind turbine to operate and remain connected while emitting no power, he would find no clear instructions in the patent as to how to achieve this and would have to rely solely on his common general knowledge.

In normal conditions involving a weak grid and a period of minimum load and maximum wind power output the turbine could remain connected and reduce its power output, just as Tande described. However remaining connected in a fault condition involving a network voltage collapse is an altogetherdifferent matter. As I have explained in considering the issue of obviousness, this is technically very difficult to achieve in a wind turbine, even if the starting point is a turbine equipped with the power electronics of the E40. In the passages of Dr Taylor’s evidence I have cited in paragraphs [152] - [154] above, he variously described it as technically innovative and difficult to do. At other points in his evidence he described it as a development task and more than routine but that a research group would succeed in producing a functioning prototype. In the passage of his evidence cited in paragraph [175] above, he explained why it was so difficult. Emitting no power involves matching the output voltage to that of the network. Under normal conditions that is not so difficult. But when the network voltage is heavily depressed it is likely to be unbalanced and transiently changing. So the control system must track it around and do so across three phases in a period of microseconds. This is very difficult and very challenging. Yet the patent provides no assistance as to how it is to be achieved.

The practical problems this posed to those in the art were graphically illustrated by their reaction to the suggestion they might be required to introduce this feature into their machines. As Dr Taylor emphasised, the BWEA thought it would kill the industry.

I have been left with the overwhelming impression in the light of all the evidence that implementing LVRT on wind turbines in 1997 was not routine, would have been regarded by the skilled person as a very difficult and very challenging task and would have required innovation. But the 564 patent provides no practical guidance as to how it is to be achieved. All it discloses (beyond what was known from the E40) is the concept of remaining connected to the network whilst emitting no power in circumstances where the network voltage has fluctuated outside predetermined limits. In my judgment the 564 patent does not disclose the invention in a manner sufficiently clear and complete for it to be carried out by a person skilled in the art. Accordingly, the 564 patent is invalid for insufficiency.

It was confirmed by the House of Lords in Synthon v SmithKlineBeecham [2005] UKHL 59; [2006] RPC 10 that an earlier publication will only deprive a patented invention of novelty if it discloses the invention and that the skilled person would be able to perform the disclosed invention if he attempted to do so by using the disclosed matter and his common general knowledge. In my judgment the same approach must apply to obviousness. There can be no justification for finding that a “near miss” which does not enable the claimed invention nevertheless renders it obvious.

It follows that although I have concluded the concept of implementing LVRT in wind turbines in 1997 was obvious in the light of the common general knowledge, D5 and D8, that conclusion does not render the claims invalid. D5 and D8 are no more enabling disclosures than the 564 patent itself.

The allegation of infringement is maintained in relation to the Vestas VCS wind turbines which have the so called AGO2 feature. It is said that these wind turbines have a LVRT capability in which they can remain connected for up to 200ms whilst emitting no real power and consequently they infringe each of the claims.

The AGO2 is used to ensure the VCS turbines are able to comply with the requirements of certain grid operators that generators remain connected and inject reactive current into the grid during low voltage faults.

The way the AGO2 works is explained in the confidential product and process description and a series of graphs (X4) handed in during the course of the trial. It can be summarised as follows. In the event of a low voltage fault during which the voltage on the grids falls below 85% of its nominal value, the AGO2 puts the VCS turbine into what is described as rotor current control mode. In this mode the behaviour of the wind turbine is no longer controlled with regard to reference values for real and reactive power. Nor does the turbine attempt to maintain power at the level it was immediately prior to the fault. Instead, the AGO2 controls the production of reactive current so as to follow a 2% slope as a function of the voltage reduction on the grid. It is so configured that at a 50% drop in grid voltage, the turbine is able to deliver 100% of its rated current in the form of reactive current. However, the wind turbine can operate above rated current for a period of time and a certain amount of real current is also supplied. If the wind turbine was operating at rated capacity immediately before the fault, at least 37% real current will be maintained during the fault.

The real power emitted by the wind turbine in circumstances in which the voltage falls to 50% or less of its nominal value is then determined by the mathematical relationship between power, current and voltage. So, for example, and as is apparent from X4, when the voltage is at 20% of the regulated voltage the real power emitted will be about 10% of the rated power output. The power emitted decreases in a linear fashion to the origin. It is only when the network voltage has collapsed to zero that no real power is emitted.

If the network voltage does indeed collapse to zero then the wind turbine will remain connected for 200ms or thereabouts. If the voltage is restored in that time, the power recovers first to 40% of the rated output and then after a period of time to the maximum power output. If it is not, the turbine will disconnect.

Various non infringement points were taken but I need only deal with three. The first relates to claim 3 and is that the wind turbine measures the voltage at a point inside the wind turbine itself and not at any point on the network. This is a bad point. The claim does not require the sensor to be on the network. It only requires the sensor to sense the voltage on the network, and that is exactly what it does because the turbine voltage is directly proportional to the network voltage.

Second, Vestas contended that the prospect of the network voltage falling to zero is vanishingly small. It would require a 3-phase short circuit right up against the turbine. I accept that may be so, but the turbine is nevertheless built and configured to operate in the way I have described.

The third point has more substance. Claims 1 and 3 require the system to remain connected and operational without any power being emitted to the network when the network voltage is less than a predetermined network voltage value (Umin). But there is no such predetermined value in the VCS wind turbine because at no stage is a network voltage determined below which the wind turbine will emit no power. Instead, the wind turbine continues to emit power until the network voltage drops to zero. This is vividly illustrated by a comparison of figure 3 of the patent with X4. Umin is readily identifiable in figure 3. By contrast, no such value can be identified in X4.

Mr Wobben contended that at a value of Umin of one or two volts, power is zero or insignificant. Hence, it was argued, Umin is one or two volts and is predetermined in the sense that it is fixed in advance in the design of the control system.

I am unable to accept this contention. As I have construed the claims, the predetermined network voltage value is a voltage value which is determined in advance of the method being performed or the system operated. It must therefore be ascertainable. But, as became apparent during the cross examination of Dr Taylor, it is simply not possible to identify any particular value as being Umin on X4. As he accepted, it depends what the reader considers to be an insignificant amount of power. Different but reasonable engineers might well come up with different numbers. In my judgment this is fatal to the allegation of infringement. The VCS turbines with the AGO2 feature do not have a predetermined network voltage value below which no power is emitted.

The 691 patent also stems from the 324 application. It had nine claims. By an application dated 28 March 2007, Mr Wobben sought to amend the patent essentially to replace original claim 8 with three separate claims to clarify that the word “or” in the phrase “greater or less than a predefined network voltage” refers to a wind energy system which reduces the amount of power emitted when the voltage is “greater than a predefined network voltage”, “less than a predefined network voltage” and “both when the voltage is greater than a predefined network voltage and when the voltage is less than a predefined network voltage”. On 9 May 2007, and in the light of the evidence of Dr Taylor, Mr Wobben made a further application to amend 691 to delete claim 1 and those parts of old claim 8 which mirrored claim 1. During the course of the trial and as a result of submissions made by Vestas it became apparent to Mr Wobben that his second application did not go far enough and he made a third application to clarify new claim 5 and to delete new claim 6.

The final version of the proposed claims 1, 5, 6 and 7 are set out for convenience below with the words and expressions in issue underlined:

In the remainder of this judgment I will adopt this claim numbering. As I have indicated, claims 1 and 5 are said to be independently valid. Claims 1, 5, 6 and 7 are alleged to be infringed.

The body of the specification is in the same terms as that of the 564 patent, save that the description of the prior art is different, as is the consistory clause. Accordingly, the 691 patent once again teaches the principle of controlling the real power emitted by the wind energy system in response to changes in the network voltage. However, claim 1 is now directed to a method of operation in which the power emitted is reduced as the network voltage falls below a specific network voltage value. It is conveniently considered by reference to figure 3. The claim focuses on the left hand slope of the graph and claims a system in which the power emitted is reduced after the network voltage has fallen below P3 (a specific network voltage value) and before it has reached Umin (a defined minimum network voltage value). The emitted power is said to be less than theavailable generator power and is emitted for network protection.

Claim 5 is directed to a method of operation in accordance with claim 1 in which the system is operated without any power being emitted once the network voltage falls below Umin. In short, it seeks to incorporate the claimed invention of the 564 patent into the claims of the 691 patent.

Mr Wobben contends that whilst the 564 patent is most beneficial in dealing with major fault conditions, the 691 patent is primarily concerned with smaller power output reductions, and less serious voltage excursions such as those that might be caused by heavy loading or minor faults. I should observe that 691 originally also contained claims to a system which responded to an undesirable increase in the network voltage by reducing the power emitted to the network. Such claims were accordingly directed to the right hand slope of the graph in figure 3. However, as I have indicated, Dr Taylor concluded that it was obvious to the skilled person that if the voltage on the network was too high then the emission of more power would exacerbate the problem. Correspondingly, it was obvious to reduce the power, which would tend to contribute to the lowering of the network voltage and so help to deal with the problem. These original claims are therefore accepted to be invalid.

Mr Wobben maintained that the under voltage situation is, however, quite different. Here a reduction in the power emitted will tend to contribute to a further fall in the network voltage. It was therefore counter-intuitive to take such a step. For this reason claim 1 is inventive.

Vestas, on the other hand, contended that in so far as the claim makes sense, it is directed to the simple and routine step of de-rating and as such was entirely obvious in the light of the common general knowledge.

With that introduction I now address the disputed words and expressions in turn, save in so far as they arise and have already been dealt with in relation to the 564 patent.

There was some dispute about the expression network protection. Professor Green was uncertain what it meant. Dr Taylor considered it meant the provision of support to the network, protection of the network from loss of stability and protection of network components, including the wind turbine itself. I agree with Professor Green that the expression is somewhat ambiguous and I think it requires undue strain to suggest that it includes the circuitry of the turbine. However, I think the skilled person would understand that the intention of the patentee was that if the wind energy system remains connected and continues to emit power then it can continue to provide a measure of support to the network voltage. In so doing it assists in maintaining the stability of the network and, in that sense, provides some network protection.

This is an important expression. It makes clear, in a way the 324 application did not, that the invention is concerned with a method of operating a wind energy system in which the actual power emitted is a fraction of the maximum available generator power for the system in a particular set of conditions. In short, it includes de-rating and so allows the wind turbine to be protected against the risk of excess current whilst at the same time avoiding a loss of generation from the network.

For much the same reasons as discussed in relation to the 564 patent, this is a very confusing limitation. It introduces into the claim the requirement of a defined minimum network voltage without specifying what it is supposed to represent. A reasonable starting assumption is that represents the lower limit of the permissible voltage variation set by the network operator. But, as I have already discussed, the skilled person would normally expect the turbine to emit as much power as possible down to this lower limit. An alternative possibility is that it is the limit imposed to protect the turbine from damage caused by the flow of excess current through its power electronics. However, this makes no technical sense. For the reasons I have given, the skilled person would expect Umin to be at the origin. However, it is clear from figure 3 that it is not.

Overall I think the skilled person would be left with the conclusion that the claim is concerned with a rather inefficient and sub-optimal form of de-rating which involves a reduction in power which is not necessary to protect the turbine. He would understand that the method must involve the use of two predetermined or defined values P3 and Umin. The power emitted by the turbine must be reduced at a point below P3 and above Umin.

At one time there appeared to be an issue as to whether the power emitted had to be “directly” controlled. I see no basis for such a limitation. All that is required is that the power emitted is controlled as a function of the electrical voltage.

This element of claim 5 introduces the features of the 564 patent. It explains that the system must continue to be operated without the emission of power once the network voltage has dropped below Umin – the predetermined network voltage.

The validity of 691 is attacked for:

I have discussed the relevant legal principles in considering the validity of the 564 patent and do not repeat them here.

The allegation of added matter falls into two parts. The first, directed at claim 5 and dependent claims, is the same as that advanced in relation to the 564 patent. It succeeds for the same reasons and I need say no more about it. The second, directed at claim 1, is that the specification discloses that an amount of power which is less than the available power is emitted for network protection and that the amount of power which is emitted is reduced before reaching a defined minimum network voltage value (Umin) after falling below a specific network voltage value (P3).

Once again the allegation is founded upon words of the claim which do not appear in the 324 application. Specifically, reliance is placed upon the requirement in claim 1 that:

“an amount of power which is less than the available generator power from the wind power system is emitted for network protection”.

For the reasons I have given, I believe these words make it clear that a fraction of the maximum available generator power is emitted. The question that now falls to be considered is whether this was clearly and unambiguously disclosed in the application. Mr Wobben turned again to figure 3. I have discussed the difficulties of interpretation posed by this figure and its associated description in paragraphs [81] – [94] of this judgment and I do not repeat that discussion here. But to summarise, there can be no doubt the figure shows the power emitted by the turbine is reduced as the network voltage falls. However, the cause of that reduction is not explained. There are a number of possible explanations but none is satisfactory. The skilled person is presented with a puzzle and I accept Professor Green’s description of the figure and its associated text as perplexing. Not only is it perplexing but it seems to be flat contrary to the description of the invention in the body of the specification. In these circumstances I do not believe it can be said that the application clearly and unambiguously discloses that “an amount of power which is less than the available generator power from the wind power system is emitted for network protection”. The patent is therefore invalid for added matter.

Finally, I should mention that Vestas again contended that in so far as figure 3 can be relied upon as a relevant disclosure, it only discloses a system in which reduced power is emitted below a specific network voltage and reduced power is emitted above a specific network voltage. I reject this contention for the same reason I rejected the equivalent contention in relation to the 564 patent. There is nothing in the application to lead the skilled person to believe that the system must necessarily incorporate both sides of the figure.

I have discussed the disclosure of each of these publications in considering the validity of the 564 patent. They were primarily directed to original claim 1 of 691 which is now accepted to be invalid. None of them discloses the emission of an amount of power which is less than the available generator power from the wind power system after the network voltage has fallen below a specific network voltage value (P3), or the emission of such reduced power for network protection. The allegation of lack of novelty over each of these publications therefore fails.

This is another publication describing certain features of the E40 turbine with particular reference to weak grids. Specifically it describes a reduction in the emission of power to avoid the voltage on the grid rising too high. It does not disclose the emission of an amount of power which is less than the available generator power from the wind power system after the network voltage has fallen below a specific network voltage value (P3), or the emission of such reduced power for network protection. The allegation of lack of novelty over this publication also fails.

This was a major attack on new claim 1. As I have construed the claim, it covers sub-optimal de-rating, that is to say a greater reduction in the power emitted by the wind turbine than is strictly necessary to prevent damage to its electrical circuits.

There can be no doubt that de-rating was common general knowledge. Professor Green described it as part of the most basic knowledge of any electrical engineer and that it was obvious and technically trivial to implement it on wind turbines. Dr Taylor accepted that the skilled person would have known of de-rating as part of his basic training and would have understood that it applied to a wind turbine as much as to any other kind of generator. The issue between the parties was whether it was obvious to incorporate this feature in a wind turbine in 1997.

Mr Wobben contended it was not. As I have indicated, he argued that it was counter-intuitive to reduce the power emitted by a wind turbine in situations of under-voltage. However, I am unable to accept that this was so. The skilled person would have appreciated that in principle it was desirable to maintain the power output in the case of an under-voltage, or even increase it if possible and so try to restore the network voltage to a desired reference value. However, he would also have known that if the voltage at the connection point of the turbine to the network falls too far then the maximum power the turbine can generate must also be reduced to avoid a damaging increase in the current flowing through its circuits. So, if the turbine remains connected as the network voltage falls, it was certainly not counter-intuitive to reduce the power. Indeed, it might be a technical necessity.

Mr Wobben’s second argument was that in order for the skilled person to appreciate he needed to de-rate, he must first have taken the decision to remain connected. And in accordance with the mindset the wind turbine would have long since been disconnected.

I am unable to accept this argument either. I have considered the issue of mindset in addressing the obviousness of the 564 patent. For the reasons I have given I believe it was conceptually obvious to the skilled person in 1997 to remain connected in fault conditions even though in practice wind turbines were required to disconnect to avoid the risk of islanding. So also, it would have been perfectly clear to the skilled person that if the turbine continued to emit power despite a severe drop in the network voltage then it would be necessary to de-rate.

Furthermore, even in the circumstances pertaining in 1997, de-rating was an obvious thing to do. Dr Taylor accepted as much in the course of cross examination, as shown by his evidence on day 6 at 652-654:

“Q. …. Anyway, this concept of de-rating, what I want to suggest to you is the concept of de-rating made it obvious for the skilled person -- again considering the position prior to December 1997 -- to consider limiting the current in his wind turbine and thus the power emitted by the wind turbine in an under-voltage situation?

A. That method, the de-rating, holds for cases where the voltage depression is not particularly severe and that you can protect your power electronics by de-rating up to maybe 20% or something like that. But if the voltage dips very, very severely and in order to protect your equipment you have to reduce the power output significantly, then in 1997 you would just disconnect because there is no technical, commercial regulatory driver to tell you to stay on.

Q. I see. I think you may be running ahead of me. I was only interested in the slightly higher level of generality there. What I was saying to you was if there was an under-voltage situation prior to December 1997, a skilled person would, just as a matter of basic training, would be alive to the concept of de-rating and the fact that the under-voltage situation might lead to a need to de-rate. That was as far as I was going in the question.

A. I think what I was saying is I agree with you on that but only for small deviations from nominal.

Q. I see. To 20%, I think you mentioned.

A. Yes. Maybe that is a bit strong.

Q. All right.

A. Maybe more like 10%.”

The use of power electronics in, for example, the E40, was common general knowledge. The skilled person was well aware this permitted the operator to vary the real and reactive power output of the turbine. As Dr Taylor accepted, the skilled person also knew of the need to de-rate in the event of relatively minor voltage deviations which did not require disconnection. This was made clear by his evidence a little later on day 6, at 656-657:

“MR. MILLER: You have said de-rating can protect, I think you said, 20% or 10%. Were you just thinking of typical uses of de-rating or were you suggesting that de-rating ----

A. The thing is ----

Q. Sorry, let me jut get the question out.

A. Sorry.

Q. Was that just typical cases of de-rating or are you saying de-rating cannot really be of any use above the 10% or 20% figure

A. De-rating can be of use over a wide range, but you will only de-rate until you hit the protection setting. When you hit the protection setting, you will disconnect and then there is no point doing any de-rating.

Q. I follow.

A. Yes?

Q. Yes, I understand.

A.

If that protection setting is 20%, you might have to de-rate down to that; if it is 10%, you will de-rate down to that. So you are protecting yourself while you are remaining connected.”

Claim 1 does not require the wind turbine to de-rate to zero. It is enough that the amount of power which is emitted is reduced even before reaching a defined minimum network voltage value U(min). Dr Taylor and Professor Green both considered that to be obvious. Further, there was no suggestion it would pose any technical difficulty.

Claims 1, 6 and 7 are therefore invalid in the light of the common general knowledge. Claim 5 adds the obvious concept of the 564 patent but is insufficient.

These add nothing to the case of obviousness save that, as in the case of the 564 patent, D5 looks to the future and anticipates the imposition on wind farms of the criteria applying to conventional power stations and D8 expressly discloses the control system of the E40 and explains its functionality.

The allegation of insufficiency against claim 1 was only weakly argued. It was suggested the specification is insufficient because it provides no teaching of how to put the invention of claim 1into effect to achieve “network protection”. I reject this argument. I have construed this requirement at paragraph [225] above and it could be implemented without undue effort.

The allegation of insufficiency against claim 5 succeeds for the reasons discussed in relation to the 564 patent.

Vestas objected to the proposed amendments on the basis they extend the protection conferred by the patent. As granted, claim 8, the first apparatus claim, called for a wind energy system in which:

“the amount of power which is emitted to the network is reduced when the voltage is greater or less than a predetermined network value”

It was argued by Vestas that this required the power emitted to be reduced both when the voltage was greater than the predefined network voltage and less then the same predefined network voltage value. By contrast, so it was said, the proposed claims include neither aspect of this limitation.

I reject both aspects of this argument. It is apparent from figure 3 of the specification that the inventors did not contemplate a single predefined network voltage value both above and below which the system reduced the emission of power. Moreover, the skilled person would have understood the requirement of claim 8 as granted to be disjunctive – for like reasons to those given in paragraph [236] of this judgment.

All of the claims in issue require there to be a defined or predetermined network voltage value (U min). Claim 1 requires the power emitted by the system to be reduced before this value is reached. Claim 5 incorporates the additional requirement that the system must remain in operation but emit no power once the network voltage falls below this value. For the reasons I have elaborated in relation to the 564 patent, I do not believe the VCS turbines equipped with the AGO2 feature have such a value. The claim for infringement therefore fails.

The claim for infringement of claim 7 fails for the further reason that the VCS turbines do not operate without emitting power when the network voltage is greater than a predetermined network value.

I should mention two further non infringement arguments for completeness. It was submitted that claim 1 is not infringed because the reduction in power illustrated in X4 is not for network protection. I reject this argument. By remaining connected the VCS turbine continues to provide a measure of support to the network voltage and assists in maintaining the stability of the network. That is enough to satisfy this requirement of the claim.

It was also suggested that claim 1 is not infringed because the VCS turbines do not directly control the emission of real power. However, power is controlled as a function of network voltage. So this requirement of the claim is also satisfied.

The 078 patent claims an earliest priority date of 24 April 2001. It describes a method of operation of a wind energy system (and the wind energy system itself) in which the network voltage is controlled by regulation of the reactive power Q as opposed to the real power P.

Claim 1 of the patent is directed to the basic concept of varying the amount of reactive power which is emitted or absorbed by a wind energy system when the network voltage is greater or less than a predetermined range of values and to do this by controlling the phase angle Φ.

Both experts agreed that all the claims of the 078 patent were obvious, with the exception of claims 6 and 7, where they disagreed. Professor Green considered all the claims were obvious in the light of the common general knowledge because the skilled person was well aware that the starting point for the control of network voltage is the steady state equation: ΔV=(PR+QX)/V. The skilled person also knew that in the short term the only two variables which could be used to control the network voltage were real and reactive power. As I have explained, in the case of transmission networks, regulating the supply of reactive power is a significantly more effective means of voltage control than regulating the supply of real power because the reactance of the power lines is generally greater than their resistance. But in distribution networks, reactance and resistance are more evenly balanced, so real and reactive power each have an effect on network voltage. However, it was Professor Green’s view that, in such networks, operators are paid primarily for providing real power. So they prefer, so far as possible, to control the network voltage by varying reactive power rather than by sacrificing the supply of real power.

Dr Taylor, on the other hand, considered the patent showed foresight because at the priority date the majority of wind farms were being connected to weak distribution networks with a relatively low reactance and accordingly the most powerful way to affect the network voltage was with real power. Nevertheless he accepted that all the original claims, save for claims 6 and 7, were rendered obvious by the citations Improvement of the Grid Compatibility of Wind Energy Converters (D5) and Benchmark (D8) which each disclose how the E40 was a full converter turbine with the ability to vary the phase angle Φ of the power supplied to the network and so consume inductive or emit capacitive reactive power.

In the light of Dr Taylor’s evidence that all the claims except for claims 6 and 7 were obvious, Mr Wobben sought to make various amendments in an attempt to validate the patent. A considerable number of points, some small and some of more substance, have been taken in relation to the proposed amendments but before addressing them it is convenient first to set out claims 6 and 7 in their original form:

Claim 6:

“Method according to one of the preceding claims, characterized in that the control system can directly or indirectly control the operation of a switching device in the network.”

Claim 7:

“Method according to one of the preceding claims, characterized in that appropriate voltage detection operations and control processes are carried out separately by means of the phase angle Φ for sub-areas of the network.”

The proposed amendments involve first, collapsing old claim 6 into new claim 1, which I set out with the words sought to be added italicised:

Then Mr Wobben seeks to introduce a new claim 2 which combines old claims 1 and 7. To highlight the combination I have italicised the features of old claim 7:

Old claim 6 is then deleted and old claim 7 made dependent only on claim 1. Finally, I was referred to old claim 8, the first apparatus claim, which is amended to introduce the features of old claim 6, which I have again italicised:

Two points of construction arise on new claim 1. Vestas pointed out that the additional feature is not precisely in the form of old claim 6. In particular, old claim 6 refers to “the control system” whereas new claim 1 refers to “a control system”.

I do not think there is anything in this point. The claim refers to a control system only because it contains no antecedent reference to a control system.

Vestas also submitted there is nothing in the claim which requires that the control system is controlling (1) the operation of a switching device in the network and (2) the phase angle mentioned in the earlier part of the claim. In other words, the claim will be satisfied by the (obvious) method of old claim 1 together with the use of any control system to control the operation of a switching device in the network.

I reject this submission too. I consider the skilled person would understand the claim to require that the method of operating the wind energy installation must itself include the use of a control system to operate a switching device in the network. The skilled person would not understand the claim to be directed to the use of any control system. The same applies in relation to new claim 8. The skilled person would understand this to be directed to a wind energy installation which itself comprises a control system which can operate a switching device in the network.

The experts were, however, agreed that the reference to a switching device is to a tap change transformer. New claim 1 is therefore directed to a method of operation of a wind energy installation in which the control system also controls the operation of a tap change transformer in the network. New claim 8 is directed to the installation itself.

The experts were also agreed that the claim offers a solution to a condition which they described as “hunting” or “ping pong” in which the action of the wind turbine to control network voltage can cause the tap change transformer to take its own action to counter that of the wind turbine. So also the activity of the tap change transformer can confuse the wind turbine with the result that they each successively take action to counter the other.

As to new claim 2, this is unhappily worded. Mr Wobben contended that the concept of the claim is to break the wind farm down into parts with each part supporting different parts of the network with different reactive power requirements. However, the problem with the claim is that it refers only to a method of operating a wind energy installation and it is clear from the body of the specification that a wind energy installation is a wind turbine and not a wind farm. Hence the claim is not limited to the concept of breaking a wind farm down into parts at all. Rather it simply requires a wind turbine to be separately controlled (to another wind turbine) for sub-areas of the network. There is no requirement that such wind turbines are part of the same wind farm. This point is highlighted by reference to old claim 9 which does call for the wind turbines to be part of the same wind park, as its wording makes clear:

“Wind park having at least two wind energy installations according to claim 8, characterized by an apparatus (10) for carrying out the method according to one of the preceding claims, and in each case one voltage detection device (22), 27) for each separately controllable part of the wind park.”

These claims were attacked on the grounds of obviousness over the common general knowledge and insufficiency. In considering these allegations I again have well in mind the legal principles discussed in relation to the 564 patent.

Professor Green considered that old claim 6 (and hence new claim 1) was obvious because the problem had a readily apparent solution. Where there are two controllers trying to act on the same variable with perhaps slightly different target values then it is obvious that one must be subordinate to the other and it is also obvious that this can be done either way around. He accepted that the commercial and regulatory framework rendered it radical to give control over the tap change transformer to the wind turbine operator but considered that did not affect his conclusion it was a conceptually and technically obvious thing to do.

Dr Taylor accepted that the skilled person would appreciate that the facility of a wind turbine to provide variable reactive power would produce a risk of ping pong and that one of the ways of avoiding that risk would be to put the wind turbine and the tap change transformer under common control. However, he disputed that it would occur to the skilled person to give that common control to the wind turbine operator. He considered that was a radical step which was not obvious, essentially because the tap change transformer was already part of a more complex arrangement with an associated control system.

I have reached the conclusion that new claim 1 (old claim 6) was not obvious. Attractively though the case for Vestas was put, I do not believe it takes sufficient account of the practical reality described by Dr Taylor. I accept that the skilled person would appreciate the need for common control but I am not satisfied it would have occurred to him to give that control to the wind turbine or wind farm operator. That is not something he would have had any reason to think of doing. I believe he would have approached the issue on the basis that the network operator would retain control of the network assets. That would solve the problem which faced him.

As to the feature introduced by old claim 7 (new claim 2), neither expert suggested it was inventive for wind turbines in different wind farms to support different sub-areas of the network. Hence claim 2 is obvious, as I have construed it. However, I am not satisfied it was obvious to break a single wind farm down into parts with each part supporting different sub-areas of the network with different reactive power requirements. I had very little evidence on the topic but Dr Taylor thought this a clever idea and one which was out of the ordinary, albeit not ground-breaking. It would not have been a routine development, would have required the implementation of appropriate switching arrangements and is not suggested in the prior art. Had the claim been appropriately limited I would therefore have considered it was inventive.

Neither side addressed any specific arguments to any of the other claims and at this stage I say no more about them.

It was alleged the patent was insufficient because the specification fails to teach how to make or implement the necessary control systems to implement the invention. I reject this allegation. It was not supported by Professor Green and Dr Taylor considered it could be implemented, albeit with some engineering effort.

This allegation was not pursued in the light of the evidence.

My conclusions are as follows:

As to validity:

The allegation of infringement fails.

As to validity:

The allegation of infringement fails.

As to validity:

The allegation of infringement was not pursued in the light of the evidence.

I will hear argument as to the form of order if it cannot be agreed.