NZSM Online

Get TurboNote+ desktop sticky notes

Interclue makes your browsing smarter, faster, more informative

SciTech Daily Review

Webcentre Ltd: Web solutions, Smart software, Quality graphics

Feature

Blowin' in the Wind

How strong is the wind on top of a hill?

Tony Bowen

Everyone knows that it is usually more windy on top of a hill than on the valley floor below but who would like to put a figure on it? This problem has held a special fascination for me ever since I was first confronted by it at the Mechanical Engineering Department, University of Canterbury, over two decades ago.

At that time the question was being asked by the New Zealand Electricity Department because they were specifying the design of transmission towers to be sited in the foothills of the Southern Alps for their DC link. They wanted to know more accurately what maximum wind speed these structures should be designed for if they are to give, say, 50 years service without failure. The New Zealand Standards Association was also becoming interested in similar problems while developing codified prediction methods for the wind loading of buildings and other structures, some of which would be sited on exposed hills around the country.

Long-term wind statistics have been gathered around New Zealand over many years, and the statistics for the occurrence of maximum mean wind speeds over a long period can be estimated for these largely flat sites. However, the problem is to know how these mean wind speed data should be modified to represent the conditions on a hill site some distance away. A related question is how the gustiness of the wind varies with position over the hill.

Testing the Winds

We first turned our attention to full-scale measurements of the wind flow over simple long escarpments and compared the results with those obtained from wind-tunnel modelling. The results provided some of the first reliable hill data to be incorporated into the National Building Codes of New Zealand and the UK. Further work with simple ridges modelled in the wind tunnel showed that speed-up factors of close to 2:1 could be expected over smooth, gentle hills with long ridges perpendicular to the wind flow. However everyone knows that it is very difficult to find such simple hill shapes, and most building sites in complex terrain are invariably surrounded by a unique three-dimensional pattern of ridges and valleys.

The interpretation of test data taken from simple hills requires a lot of imagination if it is to be used to predict conditions at sites in real hilly terrain. Some help has been provided by outstanding analytical work carried out elsewhere. This has improved the understanding of the main parameters involved and has allowed various empirical rules of thumb to be developed for use by the building industry and others. In the meantime, other disciplines have begun to show an interest in the wind flow over complex terrain. This attention has come from a wide range of activities which includes forestry land management, the spreading of forest fires, atmospheric pollution dispersal and the prediction of the local wind environments near the ground.

Because of the paucity of guidance for estimating maximum design wind speeds, the use of the wind tunnel to model strong wind flows over complex hill sites has made an important contribution. Canterbury University's Department of Mechanical Engineering has developed techniques to ensure accurate measurements using scale-model hills down to 1:2500.

One particular highlight was the international field experiment in the Outer Hebrides. Apart from participating in the field tests, we also tested a model of the actual hill in our boundary-layer wind tunnel. Successful comparisons of the results with two other model tests at different geometric scales and with the field data confirmed the integrity of the wind-tunnel techniques used. The data sets created by this experiment are now used widely throughout the world as benchmark data for the development of empirical code rules, numerical solutions and in the calibration of wind tunnel hill-modelling techniques.

Wind Energy

The recent surge of popularity in wind energy has intensified the interest in the wind flow over hills but has also added another layer of complication. Instead of only extreme winds being of interest, a wide range of commonly occurring and more gentle wind speeds must now be considered because substantial amounts of wind energy are still available from these lower speeds. This means that the effects of solar radiation, thermal and density gradients, inversion layers etc, which could be ignored under strong wind conditions, must now be included, and these make the problem of predicting wind energy levels even more difficult.

I was fortunate to work for ten months at the Ris National Laboratory in Denmark recently, where a program has been developed to predict wind energy levels at hill sites using wind speed statistics taken from nearby measurement sites. It is used widely around the world for wind energy predictions of proposed wind farms in order to estimate their potential for commercial viability and for the optimum siting of individual turbines. The problem is that the program cannot cope with steep hills such as we are used to in New Zealand, nor can it cope with thermal effects.

Leaving aside the question of thermal effects for the time being, attention was focussed on the prediction errors from this program when it is used outside its operational envelope; that is, for steep hills which force the separation of the wind flow away from a substantial area of the hill surface. An effective method was developed which can predict the error based on the ruggedness of the sites in question so that the user can now compensate for the expected error. It is hoped to confirm these methods soon with more field data taken from New Zealand sites.

So after all this time and effort, the question of how windy it is on top of a real hill still cannot be answered accurately, except for the rare case of a smooth, low hill of simple shape in strong winds. Practitioners of wind power and structural building design who need a better estimate than that given by the existing crude rules of thumb must still resort to actual measurements of wind speeds at the site over a substantial period of time. Short-term measurements of a few months can also help but the data must first be modified using sophisticated Measure-Correlate-Predict techniques developed by the wind energy fraternity for that purpose.

It is frustrating to admit that although the understanding of the wind flow over hills has come a long way since the 1970s, it is still not possible to put an accurate figure on the speed-up over many of our rugged New Zealand hills.

Dr Tony Bowen is a Senior Lecturer in Mechanical Engineering at the University of Canterbury.