WREC 1996

WlNDFARM PROSPECTS IN CENTRAL OTAGO, NEW ZEALAND

K. R. Dawber and G. M. Drinkwater

Department of Physics, University of Otago, PO Box 56, Dunedin, New Zealand

ABSTRACI

Detailed wind investigations have begun near the Clutha River in Central Otago, New Zealand to look for topographically augmented sites suitable for windfarm development close to large existing and potential hydro- electric power stations on the River. The main problem for immediate economic development of the sites is the competition from the very low priced electricity usually available from the national grid, so only easy access sites with the highest available wind speeds are being studied. Of the 8 sites so far investigated, all appear to be suitable for further investigation. Support through an ECNZ Research Contract is gratefully acknowledged.

KEYWORDS

Wind energy; topographic augmentation of wind; potential electric grid connection

THE ELECTRICITY MARKETING SFUATION IN SOUTHERN NZ

The hydro-electric power stations in the southern part of New Zealand, consisting of the areas known as South Canterbury, Otago and Southland, supply two fifths of the country’s electricity. However only about 6% of the country’s population lives in the area so that apart from the electricity used by one very large industrial consumer (an aluminium smelter), much of the energy is transmitted north via the national grid. There are under-sea DC cables across Cook Strait between the two islands with a power carrying capacity of 1240 MW. These cables are fed by a

wairnnpi

350 kV DC line from Benmore and the DC system terminates at llrol lllld 1

Haywards; (see Figure 1). The main grid connections in both the South i”uckland

and North Islands are 220 kV AC. Thus there is a large national grid system with very few consumers at remote sites. The potential for

“i? !4s11icsron wind generation of electricity therefore depends on the sale of the electri- city to customers throughout the country, in competition with electricity from other sources. Benm0Te

4

I! Chri.rchurch

Details of this marketing and the pricing structure are discussed fully -edin elsewhere (Dawber, 1994, Dawber, 1996). In this paper a summary is ad hM1&m sufficient. The consumers buy their electricity from a marketing utility

SM lllld p *iuai PDinr 0200 h

of which there are many in competition with each other. These utilities obtain their electricity from various generating authorities, the two main

Figure 1 Map of New Zealand

suppliers being state owned enterprises (SOE’s). A third SOE operates the national grid, but the local distribution is usually under the operation of further utilities. Local domestic marketing utilities negotiate general supply contracts for the base load but use the spot market to reduce their costs and to take advantage of the surplus electricity which sometimes exists especially from hydro sources when the rivers are in flood or the demand is light such as during summer nights. Large consumers negotiate with numerous marketing utilities in order to get the lowest priced electricity. The government requires the local distribution utilities to provide similar service at a similar price to all users irrespective of whether the marketing utility is from the same locality or not. The spot market prices for 1 kWh in southern New Zealand have ranged from (NZ)O.O 1 cents

802

WREC 1996 in the 1995-6 summer to over (NZ)lKOO cents in recent winter peak load times. However much of the night- time load is very cheaply supplied at all times and domestic consumers can expect to pay no more than about (NZ)6.00 cents including line charges per kWh from 2300 hours to 0700 hours. For the rest of the day they would pay about (NZ)lO.OO cents per kWh. [Note:- (NZ)l.OO cent is approximately (US)O.65 cent.] A further feature at present is that the transmission SOE and the distribution utilities seek full cost recovery on all operations, so for small consumers the line charges can be much higher than the energy charges.

POLITICAL INFLUENCE FOR RENEWABLES

Under Agenda 21 adopted in Rio de Janeiro, Brazil at the United Nations Conference on Environment and Development, 3-14 June, 1992, New Zealand agreed to aim for a 20% reduction in the country’s 1990 CO2 emissions by the year 2000 (Ministry of External Relations and Trade, et.., 1991). As of the beginning of 1996 the actual CO2 emissions from New Zealand are about 7% higher than in 1990. The rapid replacement of the present 25% electric generation by fossil fuels with further renewable energy sources such as wind and hydro development seems a sensible way to achieve the Agenda 21 aim, but unfortunately the only moves so far have been the erection of a demonstration 225 kW wind turbine in Wellington and the approval to erect a 3.5 MW windfarm in the Wairarapa near Wellington (at present under construction). However both internal and international pressure on the NZ Government to increase activity in this substitution of fossil fuel burning, combined with the planting of “carbon banks” in the form of permanent forest regeneration may help. If no improvement is rapidly forthcoming, the Government is expected to introduce a “carbon tax”.

A second strong political influence is the 456 page Resource Management Act (1991). This Act seeks to abide by both Agenda 21 and the Treaty of Waitangi (1840) which was the founding document for the formation of New Zealand and is supportive of the rights and privileges of the indigenous Maori population of New Zealand. The Resource Management Act is particularly protective of undeveloped coastal areas which is one of the reasons why the present Otago potential windfarm investigations have concentrated on inland rather than coastal regions, even though preliminary studies showed that in Otago, coastal winds are very strong.

THE SEARCH FOR ECONOMICALLY VIABLE WINDFARM SlTES

The research project discussed in this paper was carried out under a research contract supported by the Electricity Corporation of New Zealand (ECNZ). Using the general information obtained under a previous wind energy survey (Dawber and Edwards, 1988) it was realised that inland sites of normal exposure would not be sufficiently windy for economic development, as away from the coast Otago has the characteristic that the annual mean wind speed varies linearly with the altitude of the site, and at 1500 m which is about as high as is practicable for wind farming from the point of view of inclement weather during the winter, the expected annual mean wind speed at 10 m above ground level (AGL) is only 7.5 m/s. At present in New Zealand, only sites with a hub-height mean annual wind speed of at least 10 m/s are being considered seriously as potential windfarm sites. Thus sites with typical wind augmentation features such as exposed ridges, rounded prominences or gorge funnelling were selected for further study.

As two large hydro-electric dams have at present been constructed across the Clutha River, and the power from these is fed into the national high voltage transmission grid, it was decided to limit investigations to sites sufficiently close to strong grid connections so that there would not be large expenses involved in transmitting the power to consumers. It was realised that there would be little hope of selling much of the wind generated power locally due to the pricing structure discussed above. The variability in water storage levels in the major hydro-electric lakes operated by the SOE’s has resulted in quite a large amount of redundancy being built into the national grid in Otago, so that power can be transmitted via a number of routes. Thus the injection of an appreciable amount of wind generated electricity into the system at these strong grid connect&s would be possible without the need for further transmission wiring. It would probably also be possible to manage the flow of water through the hydro-turbines in a different pattern to that at present in order to take advantage of the wind when it was blowing.

A third criterion used in site selection was that of likely reproducibility of the site results nearby. Only a farm of at least 100 turbines would be likely to be economically viable from the point of view of keeping a service team busy full time in maintenance work associated with the generators, roading and system control. At this stage of our investigation, we had funding only for one anemometer mast per location, so the actual siting of the mast had to be representative of the entire location. Therefore uniform ridges were favoured in preference to single hills, and wide valleys in preference to narrow gorges.

803

WREC 1996

Finally, in most cases the expected ease of erection of wind turbines at the site was considered. In some situations due to the “tilt fault block” type of topography common in the area, it was possible to select sites where no roading would be required. The base rock would be suitable for firm foundations and 4WD vehicles could roam over these areas without difficulty. Thus sites where self erecting turbines might be employed were favoured. In two cases, however, advantage was taken of existing hut, tower and electricity supply facilities for recording “base station” reference data. One of these (Mt Horn) was at a very rugged site and gave interesting wind results discussed later.

SELECTION OF SITES AND TYPICAL RESULTS

Three sites at which wind recording equipment had been used for some time were selected as base stations and six further new or previously investigated sites were selected for this ECNZ Contract. The base stations were Rocklands Tower, Mt Horn and the Obelisk, at altitudes 592 m, 1136 m and 1695 m respectively. Rocklands Tower is described in a paper presented at the 1994 WREC (Dawber et., 1994). Mt Horn is a steep prominence of complex shape on the south end of the Dunstan Mountains which range up to about 2000 m. The site which overlooks the entrance to the Cromwell Gorge, through which flows the Clutha River, is used for various UHF radio transmitters and receivers by the local electricity utility, Central Power Ltd. As it has 230 V AC power and a data landhne to their headquarters‘in the town of Alexandra, they kindly offered us use of one of their wooden poles for mounting equipment, and AC power and the landline for handling data. As rime ice formation is a problem with guyed masts at this altitude in winter, we were pleased to operate our anemometers on one of their poles, although it should be recorded that we were unable to fit them at the standard 10 m AGL. However, even at the height of 7.55 m AGL the vertical component of the wind measured using an array of Gill-type proI eller anemometers was quite remarkable. Typical results are shown in Figures 2 and 3. Wind analysis results are given in Table 1. The “generating potential” is the percentage of time the anemometer at 10 m AGL was recording within the range of 5 and 25 m/s.

Site Name Altitud~m

Castle Dent 730 Mt Horn 1136 Mt Moka 1390

Mt Teviot 977 Obelisk 1695

Rocklands Tower 592 Rocklands Trig P 955

Roxburgh 910 Thomson’s Saddle 925

Mean Wind Possible Site Type Maximum Generating Speed/ms-1 Turbine Speed/ms-l Potential/ %

Number Recorded 7 200 smooth hills 31 67

12 10 sharp ridge 47 62 7 1000 slope to west 30 50

:.: 100 smooth ridge 74

1000 rough flat area :: 5:5 2000 flat land 28

:: 2000 nearly flat 1000 steep ridge 28 67

10 100 ridge across valley 42 62

Table 1 Summary of 1995 results

Figure 2 Typical Mt Horn horizontal data Figure 3 Mt Horn vertical wind speed component vs direction

804

WREC 1996

DISCUSSION OF RESULTS

The results given in Table 1 have come partly from wind run data and partly from computer logging equipment which was developed and installed during the period of the Contract. Unfortunately there was not enough funding to enable these loggers to be used at all sites. Figure 2 (Mt Horn) and Figure 4 (Mt Teviot) are typical of the computer logged outputs while Figure 3 gives an example of the sort of analysis that may be performed if Gill-type anemometers are used. The main problems at the Mt Horn site appear to be the frequent reaching of typical wind turbine shut-down wind speeds of about 25 m/s, and the large vertical component to the wind velocity. These results are interesting to compare with those of Mt Teviot, a very smooth rounded hill-top associated with a ridge of comparable height and exposure. The Mt Teviot site is about 40 km east of Mt Horn. The most promising of the other sites investigated were the ridge at Thomson’s Saddle, which is by NZ standards a broad valley, and Rocklands Trig P in the Lammermoors, an almost flat extensive elevated area. None of these locations are within 5 km of habitation, but they are all near the high voltage grid.

Due to the complexities of the terrain both at Thomson’s Saddle and at Mt Teviot, further study will be necessary, perhaps including modelling. Both Mt Moka and Castle Dent results are interesting due to the high per- centage of potential generating time available. As difficulties were experienced with ice at both sites during winter months in 1995, further study is needed in the coming year with better mast equipment. Rocklands Trig P deserves further study using computer logging of data as initial studies some years ago with analogue logging gave lower wind speed figures than were recorded during this Contract.

CONCLUSION

The results from these sites are very encouraging. The decision to put our effort into building data loggers based on the commercial Unidata units turned out to be a very useful one, as much higher quality data was obtained for little additional total cost. Operating in these areas in the winter months presents some problems with rime ice on guy wires and more substantial masts will be needed for reliable service in the future.

One of the consequences of selecting high wind sites in fairly rugged country is that it may well be more economic to choose cheaper smaller low-tech wind turbines for operation than the high tech and often much larger turbines at present favoured by many engineers. While the problem of providing attractive economics for windfarming in the area is still with us, the desire to reduce greenhouse gas emissions and the probability of considerable growth in electricity demand in New Zealand will both help considerably in the eventual establishment of this method of electricity generation, to supplement existing hydro-electric generators.

REFERENCES

Dawber, K. R. and Edwards, P. J. (1988) Wind energy resources in Otago and Southland, New Zealand E erxv R s arch and Develoument Cm Report No 160

Dazber K.\e(l994). The problems of establishing wind farming in New Zealand. Renewable Enerev, Vol. 5. Part I, pp. 658-660.

Dawber, K. R. (1996). Problems in getting further renewables started in New Zealand’s deregulated economy and fragmenting energy sector. Proceedines of the UK Solar Energv Societv Conferenc e, 22-23 February 1996, London.

Dawber, K. R., Neilson, G. D. and Fraser, D. R. A. (1994) Wind gust statistics and turbulence studies during storms at a wind generator site in Central Otago, New Zealand. Renewable Enerpv. Vol. 5, part pp. 730-732.

Ministry for the Environment and Ministry of External Relations and Trade (1991). New Zealand’s National R DO to e United Nations Conference on Environme I;B;O-4?7-05872-8.

nt and Develooment, page 42;