Energy Management Planning is described in [1] as follows :
In all electrical systems the production and the consumption should always be in close conformity as it would otherwise be physically impossible to keep the system frequency within the allowable tolerances. This balance is the primary reason why continuous production planning is a necessary activity for all utilities. In addition, production planning is of course contributory to the economic and environmental optimal operation of the electrical network.
Focus on the economic optimisation including risk management is increasing with the liberalisation of the electricity markets in the EU and the subsequent competition between the utilities.
It is a precondition that the overall long-term planning ensures the availability at all times of the necessary production and transmission capacity.
Energy Management Planning can be split up into the following activities:
Load Forecast / Unit Commitment :
Primary and Secondary Control
Security Analysis
Training
Emergency Control.
Generic and national information regarding diurnal and seasonal variations of consumption patterns is systematically collected by national grid operators. Well-proven load forecast tools are available. These tools are used on a daily basis for generating unit commitment. In most countries generic information on consumption patterns is publicly available.
Generic information on diurnal and seasonal variation of wind energy production can readily be derived from measured wind data eg. [30,31]) Also, production statistics are available in several countries from utility companies or other organisations.
Short-term variability (i.e. variability on a time-scale of 10 minutes) and long-term variability (i.e. variability on a time-scale of 12 to 24h) of the power production of the existing onshore wind farms has been analysed for Germany [2], Denmark and the Nordic Countries [21]. The applied methodology can easily be extended to LSOWE applications.
Worst-case power gradients typically occurring in large wind farms in normal operation are of the order of 10 to 15% of rated power in 15 minutes. It is however expected that much larger variations in the power production of LSOWE can be caused by passing weather fronts and thunderstorms, which could possibly cause a nearly simultaneous shut-down of all wind turbines in the wind farm (though not in several wind farms at once). Quantitative investigations into these power gradients are not publicly available.
An important question to address is this: is it better to cope with variability of output by:
curtailing wind production during critical periods (in the worst case, ceasing production entirely during storms or other critical periods);
technical measures to reduce the variability (storage, power gradient limits, VAR control etc.);
utilising (and paying for) the ability of conventional thermal and hydro plant to compensate for variability of wind production.
In principle the existing information should be sufficient to assess the unbalance between consumption patterns and consumption patterns. The result will however depend largely on local circumstances. Generic assessments (e.g. On a European scale) are only meaningful if cross-border transmission capacity limitations are taken into account.
Since the development of LSOWE has a concentrating effect on wind energy production (i.e. a larger proportion of the total wind energy production is concentrated in some specific areas), it may be expected that the development of LSOWE increases the production/consumption unbalance. Part of this unbalance may be compensated by a geographical smoothing effect (if cross-border transmission capacity is sufficient). For the remaining part however, additional energy management measures are required.
A study of equalising effects from wind energy in Northern Europe [6] has shown that wind power from sites with a separation of more than 1500 km is nearly uncorrelated. This leads to a smoothing of the wind power production. Change on a time scale of 12 hours can reach ca +/- 30% about once a year. The existing study only considers existing onshore wind farms. An extension of the study based on LSOWE plans in the different EU countries, would yield valuable information regarding the impact of LSOWE development on Energy Management .
By means of currently available information technology, existing demand side management systems (e.g. Double or dynamic tariff systems during night hours and switching of interruptible loads) could be extended to take into account the availability of power from wind energy.
Such measures, although not specifically taking into account wind energy production, are now standard practice with most electricity suppliers for managing peak demands.
RISOE has published a study on the possibility to recharge electric cars with LSOWE by using a tariff signal. The electric cars are in standby in the garage and when the signal is present (thus when there is offshore energy) they charge. RISOE indicates that the cars are often in the garage and that it would be the best way to use clean energy in transport (technology H2 is in gestation).
The increased unbalance between production and consumption created by the development of LSOWE can to some extent be covered by increasing the flexibility of conventional power plants. Conventional power plants have some flexibility allowing them to participate in secondary control and to compensate for load variations by modulating their power output. The extent to which conventional power plants can be modulated depends on the type of plant. Hydro-power plants are among the most flexible, even if pumped hydro-power plants often can only be modulated in turbinating (power producing) mode. Fossil-fuel fired plants also have some modulating potential, although modulating is significantly more difficult for coal-fired plant which have a relatively high start-up lag time, and a much higher thermal inertia than for instance gas-fired plants.
In all cases, increasing the flexibility of conventional power plants inevitably decreases their efficiency (due to operation at a sub-optimal working point, and due to increased start-up and cool-down losses). Therefore the extent to which conventional plants can be modulated, may in practice be largely determined by economical constraints. It should also be noted that the modulating potential of co-generation plants is often very small due to constraints imposed by the heat demand. The increased penetration of co-generation plants effectively reduces the modulating potential of the existing conventional power plants.
Few studies are known in which a detailed assessment is performed of the potential and the cost of operating conventional power plants in a modulating mode to compensate for the variability of large-scale non-dispatchable wind energy [27]. However, pilot projects (e.g. Hybrid generation consisting of wind power in parallel to a gas-fired power plants) are underway in some countries. There has been various studies on this subject in Ireland
Power output from LSOWE plants may rapidly decrease when storm fronts pass, and all wind turbines in the wind farm are nearly simultaneously shut down (when the wind speed exceeds the cut-out speed).
In some countries, it is a requirement from the TSO that the (negative) power gradient caused by the LSOWE plant should never exceed a reduction from 100% of rated power to below 20% in 2 seconds.
Even so, with a large penetration of LSOWE plants, a significant amount of rapidly dispatchable power may be required to compensate for the LSOWE power gradients.
However, in countries where the power system comprises large units (such as 1000 MW nuclear power plants), rapidly dispatchable power of some kind should already exist to cover an unexpected shut-down of one of these large units. Mostly, (pumped) hydropower plants will be used for this purpose. For many countries the existing capacity of rapidly dispatchable plants will be sufficient to cover for the power gradients occurring in their offshore wind farms.
Notice that in [28], it is stated that the Danish requirements for transmission-connected wind farms will make them the fastest-responding generating plants on the system (power ramp rates of 5 MW/minute in controlling mode).
Technically, storage is a perfectly viable solution for the unbalance between production and consumption. The cost of energy storage may however often be prohibitive, either due to low efficiency (eg. Pumped hydropower) or due to high investment cost.
Although energy storage should be investigated at the level of the entire electricity system (and not only in relation to offshore wind energy), research efforts aiming to decrease the cost of energy storage will definitely be beneficial to the development of LSOWE.
Pumped Hydropower Technology is technically mature, and can in many cases be implemented without unacceptable environmental consequences. However the cost of pumped hydropower is high, mainly due to low efficiency and high investment costs. Typically, the COE [1] for electricity from pumped hydropower ranges between 0.02 and 0.1 EUR/kWh.
Hydrogen technology is often considered promising for use as an energy vector and for energy storage. Research activities in this area are ongoing. The technical maturity of hydrogen technology however remains very uncertain.
According to [1], a technology which looks promising for maturity by 2010 is the so called Regenerative Fuel Cell(=RFC) under development by Innogy (UK) [2] .
The principles of the technology are verified on small-scale laboratory tests and are based on the electrochemical conversion process such as is used in the fuel cell technology. The RFC comprises two compartments separated by an Ion-selective membrane, separating the electrolytes flowing through each compartment. Electrodes are immersed in the electrolyte as electron transfer surfaces, but do not take part in the electrochemical process and so do not limit the energy storage capacity. The energy storage capacity (MWh) is determined by the size of two external reservoirs containing the two different electrolytes which are pumped through the fuel cell, while the capacity (MW) is determined by the surface area of the ion-selective membrane.
Plans are to build a demonstration plant to provide 120 MWh of energy and up to 15 MWe power rating. The authors claim a full conversion cycle (Electricity-chemical-electricity) of about 70%, with interesting dynamic operating capability including quick start (response time of the order of 0.02 s), quick switching from charge to discharge mode, and can be modulated by a control system. Such system would be able to store the total electrical energy from a 5 MWe wind farm during 24 hours. Such storage system combined with adequate wind prediction models could to a large extend compensate for the supply/demand forecast errors.
This technology promises to be a good alternative to pumped storage power plants, because they have a short response time, are flexible in MW and MWh and independent from dedicated sites. Furthermore, this new technology could become interesting in the future for the delivery of ancillary services, in general.
The Regenesys system should be tested in Laeso (Denmark).
Smaller scale techniques are available such as batteries, flywheels, superconducting magnetic storage and super-capacitors, but do not have the power and energy storage capacity needed for storing large quantities of intermittent energy such as produced by LSOWE feeding into the main transport grid.
Long distance storage (using for instance the large Scandinavian hydropower capacity) could in principle be a viable energy management option. Currently however, long distance storage on a European scale is not feasible due to technical restrictions on cross-border electricity transmission, reinforcement costs, and due to the organisation of the electricity markets in Europe.
According to [1], energy management of electrical systems implies that the production is adjusted to the inevitable load variations throughout the year, all the way down to day level even down to hourly or quarter of an hour level - by following the prognoses for the load. The load forecast is at first based on statistical data for consumption during previous years adjusted to the expected development in consumption and on a day-to-day basis corrected for temperature conditions, other climatic factors and the actual consumption on the previous days or weeks. On basis of these prognoses a Unit Commitment Plan is prepared showing which production plants should be in operation, and the output which they are to feed to the network within given intervals. In this planning process considerable consideration is paid to the operating costs of the various plants as the base load should be covered by those plants which have the lowest production costs inclusive of net losses. Consideration should also be made to start-up costs, which may vary greatly.
Pumped-hydro plants are very suitable to cover peak loads, which normally occur a few times a day. If such plants are not available other production plants must be scheduled for these situations.
The producers on the power market must also consider the expected price level on the Power Exchange and any constraints in the network when their Unit Commitment Plan for the following day is prepared.
Obviously any electrical system can absorb a certain amount of unregulated and stochastic production from LSOWE as the system must be designed and operated in order to accommodate the changes in the consumption, a trip of a conventional production unit or a fault on a transmission line. However, the exact amount of LSOWE that can be accepted without any modifications of the procedures or tools for system operation must be carefully analysed in each case.
For systems with a high penetration of LSOWE, the most significant difference is that in addition to prognoses of the consumption, prognoses are also to be prepared of the unregulated wind power production. Such prognoses are necessary both for the TSO and for the players on the power market that own significant wind power production sites as well.
In recent years wind energy forecasting tools have been developed in countries such as Denmark or Germany (e.g. [12],[13],[14],[24]). These tools are being applied by utilities for trading and unit commitment purposes. Also, some forecasting tools are available on a commercial basis. Projects are currently underway to merge some of the existing forecasting tools.
Currently available forecasting tools are in general not considered sufficiently reliable for assessing balancing requirements, i.e. to dispatch other power plants to compensate for the short-term variability of wind farm output. It is not expected that this will change with the development of new models. Other short-term energy management solutions (e.g. Energy Storage or Modulation of Conventional Power Plants) remain necessary to compensate the short-term variability of wind power.
Currently available forecasting tools are to some extent considered sufficiently reliable for use in electricity trading (i.e. for day-ahead planning of purchase and sales bids on the power exchange). This means that, with careful risk management methods, the average accuracy of forecasts is sufficient to limit financial losses due to erroneous forecasts to an acceptable level. However, trading based on forecasting tools is currently applied to the average wind power from many different wind farms dispersed in a certain geographical region. With LSOWE it may be expected that wind power will be geographically more concentrated, and more advanced forecasting tools may be required.
Overviews of currently existing electrical connection technology for LSOWE technology haven been presented in e.g. [18], [25]. Special workshops dedicated in particular to HVDC transmission for offshore wind farms have also provided a useful discussion forum on this subject.
For large distances to shore, cable losses become important, and cable cost may become excessive. The maximum distance from shore for which grid connection is technically and economically feasible depends on the choice between AC and DC transmission. With state-of-the-art AC technology the maximum distance is in the order of a few hundred km [29]. With DC technology the maximum distance has no practical upper limit.
Using state-of-the-art technology, the maximum (active) power which can be transmitted by a single three-phase AC cable between wind farm and shore is about 30-40 MVA for a transmission voltage of 30 kV, and about 140-150 MVA for a transmission voltage of 150 kV. This implies that very large offshore wind farms (with an active power larger than 100 MW) would require more than one cable at a 150kV transmission voltage. Cables may run in parallel tracks, but safety distances are required between the cables, which causes an increased use of sea surface often in vulnerable areas close to the cost. [Due to electric interference or to installation requirements??]
It should also be noted that advanced power cable technology would allow transmission up to 450kV for single-phase cables, and that DC transmission of about 600 MW should be possible through a single cable operating at 500 kV.
Very important developments have been made recently in forced commutation AC/DC conversion technology (Voltage Source Converter based HVDC), which could possibly make DC grid connection of offshore wind farms economically attractive in the near future (eg. [26]).
Whereas conventional (natural commutation) has been in use for more than 50 years, it is still unclear whether the new forced commutation systems are sufficiently mature for large-scale application in offshore conditions.
Even if the economical analysis of different grid connection options is very much project-specific, it is estimated that HVDC-based links may become more attractive than AC links starting from distances of about 50km from the shore.
It should also be noted that the HVDC-technology offers additional benefits regarding transmission network and power quality management (e.g. STATCOM (abbreviation?? functionality) (cfr. 2.2.3).
With adequate design measures (protection levels, sheltering, use of gas-insulated closed switchgear, ) state-of-the-art electrical equipment is considered sufficiently reliable for offshore installation.
Maintenance requirements for offshore electrical equipment are expected to be similar to requirements for land-based installations (about one service visit per year). Maintenance costs are however very difficult to estimate, since they highly depend on accessibility and work conditions. This uncertainty can only be reduced by feed-back from test and demonstration projects.
Whether or not a redundant grid connection lay-out (e.g. Two separate cables between wind farm and shore, or an interconnected internal wind farm grid) is necessary to assure a sufficiently reliable grid connection, is project-specific. The question can be answered by straightforward risk analysis, provided that adequate cable failure data and cable repair times are available.
The development in recent years of power electronics equipment to the point at which it can be used in electricity transmission is a major development in electrical engineering. According to [1], the most important features of this development are, firstly, reliable application of thyristor equipment in High Voltage Direct Current (HVDC) equipment and in Static Var Compensators (SVC) and, secondly, the more recent advent of IGBT or GTO devices with a controlled on/off capability at power levels compatible with the necessary rating for transmission. This technological advance opens new possibilities for power equipment permitting better management of transmission networks through rapid, continuous and flexible control of reactive and active power flows. Such techniques are of interest to those faced with new challenges in their transmission activities, such as increasing environmental pressure and deregulation, both resulting in a less predictable future. In fact, this electronic equipment enables more extensive use of the thermal capacity of power lines, without decreasing the stability margin.
Electricity is supplied at a specified quality level, expressed (as regards voltage characteristics) in terms of standard thresholds for the following (ref. [1]) :
voltage imbalance;
slow voltage fluctuations;
rapid voltage fluctuations and flicker;
harmonics.
The voltage quality rules governing network access for generating plant operators define minimum network characteristics at the connection point [3] , plus minimum technical conditions for the plant (for existing producers or new incomers). These rules are determined to ensure that consumers enjoy supply quality within applicable standards.
Various factors contribute to voltage fluctuations at the terminals of a wind turbine : aerodynamic phenomena (wind turbulence, tower shadow effect, etc), short circuit power at the grid connection point. Under particular connection conditions, this may result in a significant flicker level. As a consequence, some limitations for installed power could be recommended in case of weak network or particularly polluting devices.
Special attention should be paid for LSOWE since the grid connection point may be a weak point and correlatively the building of a new onshore transmission line, in opposition with environmental politics, may be required.
Experience from onshore wind farms shows a noticeable impact on flicker levels on small island grids (e.g. [19]). Similar effects might be expected from large offshore wind farms.
However, in general flicker emission is not considered to be a limiting factor for the development of LSOWE for the following reasons :
Flicker emission of wind turbines is highly dependent on the wind turbine technology. Modern variable or semi-variable speed wind turbines show relatively low flicker emission levels.
High frequency fluctuations of power output from wind turbines in a wind farm are not correlated. Increasing wind farm size does not cause an equally important increase of the flicker level.
Large offshore wind farms need to be connected at a sufficiently high voltage level to limit transmission losses. Typically at high voltage level (e.g. 150 kV) the short-circuit power of the grid at the point of common coupling is sufficiently high to limit flicker to an acceptable level.
Harmonics and interharmonics are not considered to be a limiting factor for the development of LSOWE. Even if inverters used in modern variable speed wind turbines or HVDC systems generate harmonics and interharmonics, standard filter techniques can be applied to limit the emission of harmonics and interharmonics to an acceptable level.
A particular feature of offshore wind farms, compared to onshore wind farms, is the influence of the power cable between the wind farm and grid connection point on power quality. The power cable generates reactive power which may need to be compensated. Some voltage control problems may be expected, e.g. due to severe inrush transients when switching on transformers.
It should also be noted that the power cable may have a beneficial impact on power quality by acting as a harmonics filter.
Current assessment methods for flicker, harmonics and interharmonics (e.g. Draft IEC 61400-21, which is expected to be issued as an international standard in 2001) are considered suitable for application to large scale offshore wind farms.
It is expected that variable speed wind turbines will in general augment the power quality of large scale offshore wind farms. Additional measures which may be taken are : installation of filters for harmonics, VCS-controlled switching to reduce switching transients.
In severe cases STATCOMs or SVCs may be installed to reduce or eliminate flicker. The same functionality is obtained from state-of-the-art HVDC-links.
There are no generic or firm requirements regarding grid characteristics at the grid connection point for connecting large scale offshore wind farms. The required voltage level will be dependent on the wind farm active power output. Typically wind farm sizes in the range of 100 MW would require connection at the 150kV level. The required Short-Circuit Power Level also depends on the wind farm nominal power. Some countries and grid operators limit the allowable ratio of wind farm nominal power to short-circuit power at the grid connection point. However the limiting ratio appropriate for small wind farms is likely to be inappropriate (conservative) for large wind farms.
In general the transmission grid is not very strong in coastal areas, except for areas with significant industrial or harbour activity, or in cases where nuclear power plants have been built close to the seafront. Therefore onshore grid connection (i.e. covering the distance from the shore to the nearest substation where the grid is sufficiently strong) may represent a significant part of the total offshore wind farm cost. This is especially true when the use of overhead lines is not acceptable for environmental protection reasons, and cables must be used instead.
.
The strength of the currently existing grid in coastal areas may become a limiting factor to the large-scale development of offshore wind energy. The impact of this limiting factor has until now not been taken into account in offshore wind energy potential studies.
Connecting large-scale offshore wind farms will require some grid reinforcements in coastal areas (cables and switchgear) . There seems to be a need for studies into the relationship between the technical-economical offshore wind energy potential, and the investment cost required for grid reinforcements. There is also no clear view on whether grid reinforcements should be born by the project developers, or by the grid operators (see 2.4.1.7).
Slow voltage variations can be eliminated by means of tap-changing transformers and/or reactive power control, and should therefore not become a limiting factor for the grid connection of large scale offshore wind farms.
Set points of grid protection equipment should in principle not be affected by the presence of offshore wind farms. Neither will the control system set points of other power plants be directly affected (although the introduction of large amounts of wind power will increase the control effort to be performed by these control systems).
Load flow analysis is required to assess the feasibility of connecting large-scale offshore wind farms to the grid. Traditional load flow analysis methods are suitable, although some stochastic features need to be taken into account due to the variability of wind power production. Load flow analysis should cover all realistic scenarios. As a minimum, four scenarios must be analysed, comprising minimum and maximum wind farm output, combined with minimum and maximum system demand, taking into account possible shut-down of conventional plants.
Dynamic grid stability issues may be a limiting factor to the grid connection of large offshore wind farms. In particular, the dynamic impact on the transmission grid of a nearly simultaneous shut-down of all wind turbines due to high wind speeds (above cut-out wind speed) should be taken into account. The importance of dynamic grid stability issues depends to a large extent on the imposed specifications for connecting wind farms to the grid. Most specifications imposed until now (mainly intended for dispersed wind power generation connected to the medium voltage grid), simply allow wind farms to disconnect in case of grid faults (and to reconnect when normal grid conditions are restored). A similar requirement cannot be imposed in case of large-scale, geographically concentrated wind power generation. For example, the Eltra specifications imposed in Denmark, require the wind farm to withstand three-phase grid faults without attempt to reclose, and two-phase faults with an unsuccessful attempt to reclose [28]. These requirements may lead to modifications in the wind turbine control systems. Dynamic grid analyses are required to demonstrate that the specifications can be met.
State-of-the-Art Dynamic Grid Simulation Codes in use in many countries (such as EUROSTAG, NTUA, INETI, ) are suitable for the analysis of dynamic grid stability of offshore wind farms. In general these codes have standard models for synchronous generators, and even for standard asynchronous generators. However, further development of wind turbine/generator models is required to take into account the particular dynamic behaviour of advanced (variable speed) wind turbine generators and their power and speed control systems.
Primary and Secondary Control
According to [1], by primary and secondary control is meant the production control which is necessary to handle the inevitable difference between prognosis and the actual load including of losses. The turbine controllers in the controlled units are to engage automatically and increase the production in case of frequency drop or reduce the production in case of frequency increase. This is called primary control and is as such part of the Automatic Generator Control (AGC), acting with a time constant in the range of 1 sec.
The primary control is defined by the following two parameters :
Spinning Reserve (MW);
Droop (%).
Sufficient and fast primary control ability is necessary in all electrical systems and must be very well co-ordinated to avoid instability. The co-ordination embraces both the geographical dispersion and the above mentioned primary control parameters.
The control possibilities including response time are highly dependent on the type of plant as hydro plants have far better control possibilities than thermal plants. Within the framework of NORDEL, among others, so-called Power Station Specifications have been prepared which set up specific requirements to the utilities capability of participating in the primary control.(see also chapter 3.2.2.1)
The primary control ability is at the present not commercially available on the deregulated power market.
Secondary control is the subsequent system-wide adjustment of the production to a new operating situation in order to maintain the frequency to its nominal value, e.g. 50 Hz. There is a desire in order to keep the exchange of power across borders within the planned programmes. The time constant of the secondary control is roughly 10-15 minutes from the time that the intervention of the primary control is achieved.
In some cases, e.g. in large nuclear power plants, the secondary control is automatic. In other cases, the secondary control is performed manually at the power plants on request from the regional dispatch centre.
LSOWE plants at least with todays technology do not constitute reserves of primary or secondary control.
Therefore, a high share of unregulated production causes a markedly higher need for secondary control on conventional units in order to compensate for the stochastic production from, e.g. wind power. A high production from LSOWE plants often means that the primary utilities are forced down in the load area and as such away from the economic and technical optimum operating load.
Additionally, the ability to perform primary and secondary control may be expected to be a valuable product in a competitive market, as regulating power could be sold to the TSO for a price above the spot price. A large LSOWE production on European level would then require a higher amount of regulating power thus increasing the value/costs for this product.
Even using advanced variable speed technology, wind farms can only generate reactive power to a limited extent. Unless the wind farms are grid connected through innovative HVDC-links with STATCOM-functionality, the development of LSOWE, partially replacing conventional power plants, will increase the reactive power generation requirements for the remaining conventional plants.
Due to the short-term variability of wind power, the development of LSOWE will impose an increase of the primary regulation effort on all non-wind generators operating in the interconnected grid system, unless primary control requirements are equally imposed to the wind farms (i.e. unless wind farms are treated on an equal basis compared to conventional power plants).
At present, there are no studies showing quantitatively how much additional primary regulation will be required from conventional generators, if (offshore) wind energy develops according to national plans without primary control requirements.
Since some countries have already decided that wind farms should participate to primary control, there is manifest a need for the development of improved methods allowing offshore wind farms to participate to primary control. [Is this achievable?]
Due to the short-term variability of wind power, the development of LSOWE will impose increased secondary (frequency) control requirements to the generators in the national systems to which the wind farms are connected (due to operating practices and international rules), unless secondary control requirements are equally imposed to wind farms.
Whether or not the development of LSOWE will create unit commitment problems will largely depend on the availability of reliable wind power forecasting tools.
Spinning reserve requirements will probably need to be adapted, taking into account the relative unpredictability of wind power generation, but also the possibility of losing an entire offshore wind farm due to grid faults, unless redundant grid connection is provided.
The main operating requirement of an electrical power system is that voltage and frequency should be kept within permissible ranges. This determines both service quality and safety factors (ref. [1]):
power system dependability directly determines the quality and continuity of service offered to consumers, especially since large-scale electricity storage is not yet a viable option for palliating the wide variations in consumer demand with season and time of day;
protection of persons and property (under normal and incident conditions) is an increasingly critical issue given the growing dependence on electrical power in many aspects of everyday life.
To ensure dependable power system operation, the operator must substantiate satisfactory performance of system components, both in the network (line protection efficacy, waveform quality, etc.) and in the generating plant (stability, supply capacity, impact on regulation, etc.). The term performance is used here, to refer indistinctly to functionalitys, technical characteristics and capacities.
Voltage regulation
To maintain contractual voltage quality and ensure system safety (i.e. avoiding voltage collapse), the system must be capable of keeping voltage within the permissible range.
As far as the power system is concerned, the crucial factors are voltage regulation and reactive power capabilities at the supply terminals, i.e. the busbar of the generating plant. Generating plant performance requirements are therefore specified at this location.
Again at the supply terminals, generating plants must be able to withstand voltage variation within normal and exceptional ranges, the former being specified with no duration or performance limitation.
Frequency regulation
The supply frequency reflects the power systems generation/load balance, and should be maintained close to rated value. This requires a certain spinning reserve, to compensate against incidents, natural demand fluctuations, and generating plant tripping . The primary frequency regulation system is responsible for automatically adjusting generating plant output to accommodate changes in load.
UCPTE (Union for Co-ordination of Production and Transmission of Electricity) rules on primary frequency regulation [4] for member organisations state that above a power level determined by the connection voltage, generating plant operators must afford a constructive primary reserve capacity of at least 2.5% of the total plant power output.
Other recommendations specify the frequency ranges within which equipment is required to operate (normal frequency range of 49.5 to 50.5 Hz, plus exceptional frequency range).
Currently, in some countries voltage regulation and frequency regulation demands for LSOWE are identical to those of other power plants (ref. Denmark Eltra). In other countries no specific requirements are imposed. Nevertheless, when LSOWE forms a significant part of the total generating capacity, voltage and frequency regulation requirements can only be fulfilled if :
the LSOWE take part in frequency and voltage regulation, which requires development of improved wind turbine control systems ;
the burden of frequency and voltage regulation is put on the traditional non-renewable generating capacity, and the cost of frequency and voltage regulation is accounted for in the transmission fees.
The development of LSOWE may therefore require :
to re-examine the UCPTE (supranational TSO) rules in respect with regulating capacity and margins at individual TSO level;
to provide voltage and frequency regulation on wind turbines by using electronic power converters in combination with variable speed and pitch control;
to implement non regulated RES-E with associated Static Var Compensator (SVC) for voltage control and/or energy storage (SMES, Pumped storage, Reversible Fuel Cells) for both frequency and voltage control.
Like the network itself, a generating plant (with its regulation system) should possess electromechanical characteristics capable of ensuring uninterrupted operation of the power system under incident conditions.
Very broadly, we distinguish between two types of incident situation: those which involve separation of the generating plant from the power system (to form an island or separate network), and those which do not.
Incident conditions without separation from power system
Because of the difference between protection plans for interconnection networks (400 and 225 kV) and distribution networks (less than 100 kV), we must specify the generating plant connection point. Specific incident conditions for examination could include the following :
static stability or stability-related performance under minor incidents
stability under heavy transients: response to single-phase/three-phase short-circuit or voltage dip
behaviour under slow voltage collapse
On the opposite, voltage stability is a major issue. For instance, wind plants using induction generators are very sensitive to voltage drops which leads to loss of the wind turbine generation for the grid. As for the power electronic systems it is possible to operate, to a certain limit, at less than full voltage output capability, these wind turbines should not lose synchronism and thus stay stable after slight voltage drops. According to the French Technical Requirements for Power Plants Connection 17 , it is required to IPPs plants to be stable and to stay connected to the grid after a defined voltage drop pattern. As a conclusion, it appears that requirements into generator manufacturer design for better stability to voltage variations, are still needed in order to avoid transfer limits on power from remote LSOWE plants.
An important design and operation requirement for safe utility and TSO operation is overall system stability on typical faults.
In a conventional system, this is essentially an electromechanical problem, whereby the behaviour of the (synchronous) production units is essentially represented by adequate models.
LSOWE plants with voltage sensitivity (inverters) or "induction generator" behaviour may well drop out for faults far earlier than the conventional generators in the grid.
In case of massive LSOWE injections, this may have unexpected adverse effects on the stability of the remaining system.
On the side of LSOWE technology, developments should be made to allow "reconnection in flight" of production units after a transient fault in the network. This supposes correct discrimination of the "external" fault by the wind turbine generator, safe disconnection without irreversible effects (blown fuses ) and smooth reconnection as soon as the normal grid conditions are re-established.
On the side of conventional network techniques one should consider network operation strategies, in which a substantial LSOWE production may disappear without resulting in system instability or voltage collapse.
Incident conditions with separation from power system
Following a serious incident, an isolation of generating plants from the power system may occur, in which case they may be isolated altogether or remain connected to an isolated sub-network. Generating plants must be able to withstand the resulting transient and sustain a satisfactory situation under these conditions (islanding or isolated sub-network) until restoration of normal conditions as part of the main network
All grid operators apply technical requirements (e.g. Related to reactive power, flicker and harmonics,), but these requirements vary greatly from country to country.
Some of the existing regulations are specifically adapted to wind power generation, but in most cases these special regulations are limited to connection at the MV-level. As such they may not be applicable to large offshore wind farms which will require connection at the HV-level.
Regulations specifically developed for large wind farms connected to the HV grid have already been developed in Denmark If LSOWE develops significantly it may be expected that technical requirements for grid connection of offshore wind farms will become very similar to requirements imposed to other kind of power plant.
Among the non-technical requirements imposed by most grid operators in a liberalised market is the requirement for a power purchase agreement (PPA) between the project developer (producer) and a client (consumer).
Regarding grid support requirements, two very different approaches are currently in use :
1. In some countries (eg. DK) grid support requirements (primary and secondary control, reactive power production) are shared between all generators on a pro rata basis. In this case LSOWE plants must participate to grid support ;
2. In other countries (eg. UK, B), grid support is provided by some generators only, in return for an economical incentive. In this case a free market for so-called ancillary services exists, alongside the free market for the physical electricity. This means that LSOWE plants have no obligation to participate to grid support, but they may choose to do so for economical reasons. It is important that the markets in ancillary services are fairly set up and regulated.
Requirements imposed by national grid codes are in general not considered to be a limiting factor for the development of LSOWE. From a technical point of view, even the most stringent current requirements can be fulfilled by using state-of-the technology. Nevertheless compliance with grid code requirements may cause a significant cost increase for some LSOWE projects.
In general the requirements imposed by national grid codes are considered as being not particularly suitable for non-predictable, highly variable energy sources, with the exception of the Danish requirements which have been specifically developed for use in offshore wind energy.
In all cases, project developers may have to take additional measures to comply with the grid codes, such as : use of variable speed wind turbines, special purpose remote control systems (with individual power set points for the wind turbines, )
In the long term, HVDC transmission and/or on-site storage facilities with controllable reactive power output, might present interesting opportunities allowing LSOWE plants to meet grid access requirements more easily.
Priority access can be relatively easily accomplished for a small number of small-scale renewable energy projects (e.g. < 10MW). For large-scale offshore wind farms however, unrestricted priority access would create significant additional costs for the transmission system and/or the other generators. Then it has to be decided how these costs are calculated and distributed amongst the stake-holders. The question is : Who pays for priority access ? The project, the other generators, the customers, or the tax-payers ?.
In some countries (e.g. UK, B, ) sea cables will be owned by the LSOWE project, and ownership limits will be decided on a project by project basis. In some other countries however (e.g. DK) the sea cables are owned by the national grid operator, who effectively extends his grid to the location of the LSOWE plant. This causes a significant cost reduction for the LSOWE project (as the grid connection cost is largely incurred by the grid operator who transfers this cost to all customers).
According to [1], ancillary services are services needed to transmit the energy from generation plants to end users with guaranties concerning power system dependability. The main ancillary services concern active power and frequency regulation, reactive power and voltage regulation and system restoration after collapse.
Primary regulations : voltage and frequency
LSOWE plants using induction generators without any power electronic interface to the system (cf. 5.2.1.2) can not ensure satisfactory performances, as far as primary regulations (voltage as well as frequency) are concerned, particularly when no wind, the basic power source, is blowing.
Concerning voltage control or VAR control (reactive power), the two other groups of generation systems (cf. 5.2.1.1 and 5.2.1.3) can provide good performances and hold a unity power factor. This is a well established result for generation systems using conventional alternator. For generation systems using power electronic interfaces, it is possible to operate less than full voltage and thus to hold a various range of power factor.
According to
the French Technical Requirements for Power Plants Connection
[5]
, IPPs are required to maintain their output voltage to the nominal
value plus/minus 5%, and thus to be able to regulate the voltage in one of the
three different following options decided with the System Operator : at
constant voltage, at constant reactive power output, or according to the linear
relation U = Uo + k.Q (Uo nominal voltage, Q constant reactive power
output and k constant reactive droop).
Contribution of stochastic RES generation (e.g. wind turbines or PV) to frequency control is possible only if they use power electronic interface. But in such cases, an efficient frequency control would lead to decrease the energy output and consequently decrease their economical profitability.
Moreover in case of high penetration of stochastic RES generation in a network, power fluctuations from stochastic RES generation increase the need of spinning reserve [6] and frequency control includes an extra burden on the conventional power plants.
According to the French Technical Requirements for Power Plants Connection to, for example HTB23, it is required to IPPs to provide spinning reserve equal to 2.5% of their nominal power. For the time being, nothing is specified for RES generation but should be in the future as far as RES penetration is expected to increase.
Technical solutions (blades and speed control, electronic devices, etc.) are possible for RES generation plants so that they can participate to frequency control. The funding of these technical solutions may be balanced in the new energy market by the ancillary services payment and less power fluctuation from stochastic RES may imply lower additional spinning reserve for the system.
Black start
Black
start ancillary service is the ability for a plant to generate power after a
complete collapse of the network (no more voltage and other generation).
Even if black start capability is not frequently required for traditional power
plants, except for hydro units, it may be profitable for biopower and
geothermal plants to propose this ancillary service particularly if they are
well geographically located (e.g. close to a large power plant).
If required, LSOWE plants can contribute to reactive power generation. Controllable reactive power generation (i.e. with set points for reactive power determined by the grid operator) is however most easily accomplished by means of variable speed wind turbines and/or VSC HVDC transmission.
If required, LSOWE plants can contribute to primary control (i.e. production/consumption balancing on a 10sec time scale), although this is associated with a significant energy penalty.
If required, LSOWE plants can contribute to secondary control (i.e. production/consumption balancing on a 10min. time scale), although as before this would be associated with a significant energy penalty.
In general wind turbines require a pre-existing voltage source with stabilised frequency and can therefore not provide black-start capability. From a purely technical point of view, by means of advanced technology, black-start capability could be realised, but only at a very significant cost.
In principle, delivery of ancillary services, could be seen as an new economical opportunity for LSOWE plants. Nevertheless, wind farms can only provide ancillary services at a very significant cost. Therefore it is unlikely that LSOWE plants could become competitive to conventional plants on the ancillary services market, and it is equally unlikely that ancillary services would become a major source of revenue for LSOWE projects.
Offshore extension of the transmission grid (i.e. the creation of offshore substations owned and operated by the TSO) may have an important cost reducing effect since this would allow to share grid connection resources between different wind energy projects in the same area. Shared grid connection has the additional benefit of reducing the environmental impact of the sea cable and its landfall. A single 100 MW cable is indeed expected to have lesser environmental impact than 5 separate 20 MW cables. Nevertheless, taking into account the current technological limits for 3-phase AC connection, large clusters of offshore wind farms (many times 100 MW) will still need several cables, unless HVDC links are used.
In principle it would also be feasible to interconnect several offshore substations by means of sea cables to create an interconnected offshore grid. This would increase the grid connection reliability of offshore wind farms. Nevertheless, it should be noted that onshore grid reinforcements are far less expensive and much easier to maintain than an offshore grid.
International Area Control Error (ACE) requirements force national TSOs to keep the imbalance between production and consumption on a 15min time-scale within pre-set limits, either by secondary control of production units, or by cross-border traffic. If these ACE requirements are not changed to reflect the increased production variability induced by large concentrations of LSOWE plants in some areas, this would severely penalise the countries having installed LSOWE. This would also effectively prevent achieving a truly international market for electricity from wind.
Both for long term transmission system planning as for daily operation of the European interconnected system a faithful representation is needed with sufficient (but not excessive) detail. These power system models, including data relating to production as well as equivalent network representations, should be developed and publicly made available or commercialised to all operators and players on the liberalised market at European scale.
An overview of exiting cross-border connections is presented in [1]. There is at present no comprehensive assessment the impact of cross-border transmission capacity on geographical smoothing of wind power variability , nor of the feasibility of long-distance energy storage.
Long term planning has been carried out for long by utilities as they had to prepare generation and transmission investment plans to be submitted, amongst others, to national authorities and their management boards. Although it is expected that this practice will evolve in the near future according to the introduction of the market liberalisation, it is worthwhile examining how such planning practices were undertaken until now. Indeed, indicative planning will remain a necessity for national authorities in order to monitor their electrical system and for the utilities, in order for them to comply with their own market share of the electrical power demand.
Long term planning covers time spans from one to 20 years. Each time-span can be divided in several sub-periods for analysis purposes. Several categories of analysis are currently undertaken in the framework of long term planning :
- investment planning, either by use of a pluri-annual optimising model or by simulating scenarios of investments on an annual basis using techniques of probabilistic costing;
- probabilistic generation costing, for the simulation of the generation system on an annual basis;
- reliability analysis in generation and transmission, usually by using Monte Carlo simulation techniques.
Currently available methods and tools for long term power system planning can already cope with centralised large-scale renewable energy sources, connected to the transmission grid. (ref. [1]) As such these methods should be suitable to deal with LSOWE.
Generic information regarding investment budgets is available from many feasibility studies and to some extent from existing offshore wind energy projects. The available data however do not cover the entire range of cost influencing parameters (such as distance from shore, water depth, wind and wave climate, soil conditions, ) which may be encountered in case of a significant Europe-wide development of LSOWE.
It should be noted that some of the above conditions, in particular soil conditions, may show significant variations even within a single country. For some projects, the differences may be large enough to require the use of different foundation solutions.
It should also be noted that investment costs related to the compensatory measures are difficult to evaluate by the lack of experience : example for the radars or the UHF emissions, for protections anti-erosion, for the compensations of the fishermen. The costs of dismantling can be integrated in the initial invest (2% of the total budget as on the Danish projects, which appears weak) or in exploitation by constituting a reserve (as what is required in Belgium). Their evaluation is very difficult today. In addition, the regulation will be able to evolve (and will evolve): will it be possible in 20 years simply to explode the monopiles ?
Investment risk for onshore wind energy projects is well known and has been described to some extent [EWEC99, Raftery et al].
For offshore wind energy projects, additional risk arises due to :
The possibility of major transport and installation delays due to bad weather,
Large uncertainties on accessibility and availability
Large uncertainties on O&M-costs (incl. Eg. Taxes and royalties)
In addition, even more far reaching risk may be caused by the application of relatively new technologies in environmental conditions that are badly defined. Questions remain related to eg. wake effects in large wind farms, lifetime of offshore wind turbines, etc.
Industry takes uncertainties into account by applying proper margins to budget estimates. However, for a better understanding and quantification of investment risk, it is required that operational feed-back from test and demonstration projects around Europe is analysed in a coherent way.
!
In the long term, a cost reduction could be possible by the introduction of a cost sharing system between different operators. This requires cross holdings between operators limiting the risks and reducing the costs. This organisation is different from those of the terrestrial projects, and will be easier to realise with operators of big size that with smaller size wind turbine developers. Thus these operators could share a data base on the available resources and on the maintenance costs like it is done in offshore oil Companies. This would create a professional organisation gathering these data for the benefit of all. It is necessary to note that on sea the actors are naturally more "united" by the difficult conditions and of the significant costs. A mutation of wind energy seems necessary to this stage. We could then propose to provide the foundations of such exchange structure between operators (with various under-topics : technology, impacts, compensation, dismantlement).
From the current developments of demonstration LSOWE projects of various sizes , it would appear that sufficient equity capital is available for financing LSOWE projects. Some major oil & gas companies and utilities have announced LSOWE projects which could be financed by company equity. Nevertheless, many other projects have apparently been announced well before financing was secured. It still remains to be determined under which conditions (due diligence, certification, insurance,) bank loans will be granted for LSOWE projects. Only test and demonstration projects will allow to establish an answer to this question.
Even in a country such as Germany where financing is easily achieved for onshore wind energy, it is not certain that off-shore wind energy projects could be financed in the same way. Obviously the financial risk involved in a large-scale off-shore wind energy investment is much larger than the risk involved in a series of smaller on-shore wind energy investments.
Whereas insurance conditions for onshore wind energy are well established (and typically amount to about 2.5% ot the annual 0&M costs), it remains to be determined at which costs machinery breakdown and/or production loss insurance will be available for LSOWE projects. Only test and demonstration projects will allow to establish an answer to this question. The evolution of safety regulations may have an important impact on the evolution of insurance costs.
Support mechanisms applicable to LSOWE projects vary from country to country. Some of the existing support mechanisms are not applicable to large scale projects (e.g. 100 MW) connected to the HV-grid. There is no consensus regarding the suitability of the different existing support mechanisms for LSOWE projects.