11. Recommendations for Further Work
The authors and funders of this study were aware at the start of some fundamental limitations on the accuracy of the estimates made during the study, not the least of which is the lack of readily accessible wind speed data. Other limitations have become apparent during the course of the study.
The following recommendations for further work are listed in order of priority.
11.1 Wind Dataset
The electricity industry around the world relies heavily on detailed modelling studies for a variety of purposes, and these in turn rely on having accurate input data including, for example, patterns of demand and generation down to the half hourly level. In New Zealand a database of historical hydrological inflows is also an essential input for modelling and studying security of supply and other issues.
Given the potential for wind energy integration in this country, and the many issues to work through, establishing a comprehensive wind energy dataset with which to work on subsequent studies, is considered essential. Such a dataset does not necessarily have to be unique, nor does it have to be publicly available, but it does need to be available to those responsible for overseeing the electricity industry.
There are a number of ways such a dataset could be produced in a cost-effective manner. For example, it could be based around historical wind speed data available from climate databases. This data dates back to about 1940 in some regions of the country, but is typically not in the places where wind farms would be built. For example, it might be at an airport beside the sea whereas a wind farm in that region would be more likely to be built on top of a hill where the wind speed could be significantly different on average. However there are ways in which this data could be transformed to produce a reasonable wind speed dataset with which to perform further analysis and study.
It could also be synthesised using purely theoretical statistical models or it could use whatever actual wind speed data might be made available from actual or potential wind farm sites. Or it could be put together using a hybrid approach. However it is created, it must have the following attributes:
- data available down to at least the resolution of one half hour;
- data available for at least one notional wind farm site in each major region with potential for wind farm developments in the foreseeable future;
- data relevant to wind farms that would be economic by contemporary standards, which currently means annual average wind speed in excess of 9 m/s.
11.2 Impact on IR
During the course of this study significant efforts were made to understand how modern WTGs would impact on adjustments to the basic N-1 risk calculated by SPD, and on the frequency response of the grid after an under-frequency event. With the benefit of assumptions relating to the type of WTG that would be installed, namely DFIG or synchronous (or WTGs with equivalent capabilities and inertia), it appears that IR would not limit the penetration of wind energy.
However, much uncertainty remains. It is not totally clear, for example, that the wind energy industry understands or is prepared for the importance of IR in the New Zealand context, simply because it has not been a high priority.
It is also important to understand in much more detail how the grid would respond to an under frequency event with many wind farms connected. This requires the use of the appropriate detailed models of the behaviour of system frequency.
Further work is therefore required to understand and analyse these issues, if it is not already underway.
11.3 Connection Standards
A common theme in this study is that modern WTGs are appearing which have technological solutions to issues like fault ride-through and voltage support. While this may be true, it is recommended the industry decide sooner rather than later which, if any, connection standards should be mandatory for new WTGs. What should be avoided is the need to retrospectively enforce standards which become necessary at a much later date due to high levels of wind energy penetration.
Along with this, there should be clear guidelines about the criteria under which the connection standards are mandatory or otherwise. For example, it may be that wind farms under a certain size are exempt from a range of standards, or that certain aspects of connection are determined by studies relating to the wind farm development under consideration at the time.
Connection standards for wind farms could include:
- compliance with minimum requirements for fault ride-through;
- the ability to disconnect wind farms remotely or to limit their output if for any reason there is too much wind, or if there is a line constraint, for example;
- ramp rate limiting and other features for wind farms in regions where there is a high concentration of wind power;
- a requirement for wind farms and other generation to connect to the SO via a system of automatic generation control (AGC) which would potentially have advantages in producing a more accurate dispatch.
They would also benefit from clarification of any conflicts between the EGRs and Transpower requirements, or between the requirements for grid connection versus the requirements for safe operation when embedded with a distribution network.
11.4 Wind Speed Forecasting
As a key factor in allowing the SO and the market in general to deal with uncertainty in wind farm output, the industry must continue with work on forecasting wind farm output, both on an individual basis, a regional basis and a national basis.
Wind farm forecasts need to be accurate both in the magnitude of wind speed changes and in their timing, the latter applying particularly in respect of strong wind events moving over the country.
Transpower-sponsored work has started with the efforts of the Mathematics in Industry Study Group, but progress appears to have been slow. No doubt individual wind farm owners have made progress of their own but, in anticipation of growing wind energy integration, the industry must make concerted efforts to develop methods and standards for forecasting which will ensure that the SO is provided with forecasts of consistent and well-understood accuracy.
11.5 Smaller Wind Farms
As a general rule, geographical dispersion of wind farms leads to lower variability in aggregate wind farm output and maximises the potential for wind energy integration. A number of network companies and smaller generators are investigating wind farm projects which may not be viable due to lack of scale or difficulties in obtaining hedging and other contracts. If these projects are not progressed by smaller players for one or more of these reasons, it is possible that geographical dispersion will end up being lower and, as a result, wind energy penetration may be lower than it could be.
In some European countries utilities are required by law to purchase the output of small wind farms at a guaranteed tariff rate. While this might not be an appropriate solution for New Zealand, the issue of smaller wind players is worthy of detailed and extensive consideration.
11.6 Dry Year Security Analysis
High levels of wind penetration will inevitably impact on dry year security, but as yet no one can be certain of whether the impact will be positive, negative or neutral. Dry year security analysis requires, amongst other things, access to a regional wind speed database to be used for correlation studies with hydro inflow data and for modelling studies which include the impact of hydro storage lakes.
11.7 Other Work
A number of items of work remain or will be able to be tackled once some of the higher priority work is completed, or at least underway, particularly the development of a wind speed database. These include:
- an assessment of whether reserves should be scheduled to cover large "wind events", such as storms, which could shut down a number of wind farms within a short period;
- consideration of the potential for more frequency excursions with high levels of wind penetration;
- studies of the loading on lines - around the Manawatu, for example - to determine how a higher concentration of wind farms might affect the occurrence of line constraints;
- studies of the impact of high levels of wind penetration on the ability of the grid to maintain stable operation during and after transient events (if not already addressed by other work);
- consideration of the potential to introduce rules allowing wind-hydro or wind-thermal blocks;
- consideration of market rule changes to require all plant to be offered in if it is reasonably available to run, and when such a rule might be needed, if at all.
Back to Top