24 October 2008

Delta wind trial in Netherlands; BWEA conference

During the week we exhibited at the BWEA conference and exhibition and in one of the little seminar sessions I gave the usual update on Ampair etc plus I summarised the results that have just come in from the first 6 months of the Delta wind trial in Zeeland in the Netherlands, managed by Zeeuwind.

Ampair - BWEA seminar presentation c/w Delta / Zeeland windtest results: dowload pdf file from this link:


Refer to the pdf file which can be downloaded. The first five slides are just an Ampair commercial from the seminar presentation I gave.


Ampair has been participating in the Delta wind trial in the Netherlands of small grid connected wind turbines. The first report has just been released with 6-months of data and the results are very interesting because they show clearly the state of the art in commercial small wind turbines (with Ampair as one of the leaders), and because they show the performance of several vertical axis designs.

The trial is running slightly late – this is a common feature of all such trials because in practice many products are not as available as press releases would have us believe. The trial is a collective effort of some consumer testing organisations (the equivalent of Which in the UK) and some utilities, and some provincial governments. It is taking place in the Zeeland area of southwest Holland.

Originally twelve turbines were pencilled in for the trial. Not all arrived as scheduled and so one of our Dutch distributors Eco-Energie Rietpol put forwards the Ampair 600 to fill a vacancy. The 11 turbines which did participate are summarised in slide 6. Sincere congratulations to everybody who participated for having the courage to do so – it is easy to snipe at the results but I can assure you that the participants all deserve our support. As always the people who avoid participation in these sorts of serious trials are the ones who may merit negative comment, and not those who genuinely participate and then have some public disappointment.

Eco-Energie Rietpol installed the Ampair 600 early in the year, and not long afterwards it was shut down because we needed to upgrade the blades, tail, and electronic safety circuits. The trial started whilst the Ampair was shut down and so you will notice that the first seven weeks of production data for the Ampair are essentially zero.

Each week the organisers release information to the installers about performance of their turbines but not of all the others. Last week the organisers released the summary data for the first six months of the eleven turbines that did arrive. They have given monthly production data and total consumption data. In my version which is slide 7 I have added in a row showing the net power output, i.e. production minus consumption. I have translated from the Dutch and any mistakes are my fault, sorry.

On slide 8 I have graphed the results to display installed versus net power output.

This trial includes four vertical wind turbines, two from Ropatec and one each from Turby and Windwalker. This seems to be the first time that so many VAWTs have been represented in a trial side by side with several different HAWTs and so the results would be fascinating for that reason alone.

The only two turbines in the trial which are directly comparable are the Ampair 600 and the Zephyr AirDolphin. The Zephyr is 1.8m diameter and the Ampair is 1.7m diameter but basically they are both 3-bladed upwind HAWTs. Again this is fascinating because it gives us all an opportunity to compare results from two serious manufacturers of good standing. As an industry there is a lot of consideration being given to conducting so-called round-robin tests where one turbine moves from test site to test site to compare the abilities of test organisations to reproduce each others figures. It seems quite likely that the Ampair 600 and the Zephyr Airdolphin will be used for this as they are small enough and cheap enough, yet also stable enough to survive and produce statistically meaningful results. So looking at them side-by-side is interesting.

Slide 8 is the most interesting slide. The most obvious thing is the shaded zone running crudely Ampair 600 – Fortis Passaat - Southwest Skystream – Fortis Montana. These four turbines are closest to the ideal of more power for less money and so they appear to represent the efficient frontier of the current state of the art in modern small wind turbine manufacture. All are HAWTS from manufacturers with credible histories, even if only one of them has had huge (relatively) amounts of cash to put in to product R&D.

Looking at the rest the most obvious conclusion is that there is no correlation between money invested in R&D and product performance. With one exception all of the rest are simply poor performers in terms of closeness to the efficient frontier. This does not mean that they are necessarily not worth further consideration but it does mean that from an investment perspective (whether society, shareholder, client) one should think very carefully before heading in that direction as for whatever reason they are simply expensive. I know that Ampair and Fortis are cash-poor (i.e. without VC investment to date) and we are demonstrating that we can perform and that in my opinion is the standard we should all be assessed against.

All the VAWTs are bad performers. One is so bad that it does not even make it onto the graph. In fact I am not sure it was at the test site as the Windwalker appears to have consumed 39 kWh and produced nothing. It may be that it has not been installed and the meter is reading incorrectly. I note that no price is given and so I have chosen not to graph it. The other three VAWTs do at least appear to produce but their cost/benefit is noticeably poorer than any of the HAWTs. This is in supposedly relatively clean airflow and they will argue that as the airflow becomes more turbulent they will become relatively better performers. In fact many VAWT manufacturers are claiming that they are best suited to low velocity turbulent urban airflows, but so far no good comparative data set has existed to support this claim. From the Zeeland data the best one can do is to point out that maybe the VAWTs will be better but they’ve got a large handicap to overcome.

I know the Ropatec people had some confusions in the first two months as they visited me at the BWEA conference in London. If I understood their explanation their installer put the 3kW inverter on the 6kW machine and vice versa. Towards they end of the first two months they corrected this mistake. There is some confusion as to whether the inverters have been correctly reprogrammed now. Good luck to them in the next six months. It is fascinating to observe that the Tesco supermarket chain has installed thirty plus of the 6kW Ropatec units at over £35k each in the UK as, on the basis of this data a couple of Ampair 600 or one small Fortis Passaat would have been better, and in the case of Ampair would have had the benefit of buying British.

This Turby is producing. In previous tests of the Turby this has not always been the case – note that the Turby is a serious power consumer. The Turby is spun up to speed by using the generator as a motor. It appears that whatever Turby are doing to improve their spin up algorithm is working, or that the winds are sufficiently steady on this site. But a 10:6 ratio of production : consumption does not leave much margin for error. In this respect the Ropatec layout is a clear winner in the VAWT stakes.

The Renewable Devices Swift is another non-performer. It simply is not producing. It has a 2.1m diameter and so has twice the swept area of the Zephyr Airdolphin and the Ampair 600 and so should produce approximately twice the energy. Instead it is producing less than half. Clearly their electronics design is not ideal as it is a serious power consumer, but even if we just look at power output in any given month it is producing half of the Ampair and the Zephyr. This is before one looks at the additional handicap of the huge cost of the product. Their latest claim is that they have orders for 3,000 machines and that they are silent. Well they may be silent but they also do not appear to produce power. They have not participated in any public trials before so it is not possible to comment as to whether this is an unusual result. One benefit of being a repeat participant in trials as that one can be given a bit of slack if something goes wrong in any one trial.

The Energy Ball deserves an especial good mention. The Energy Ball is one of two similar products on the market with the other being the 2kW Loopwing from Japan. Apparently Scottish & Southern Energy have just bought two Loopwings to install in the UK in Weymouth at a cost of £30-£40k or so we were told at a BWEA meeting which set us all (it was a manufacturers meeting) back on our heels with much collective gnashing of teeth. This makes the EUR 4k Energy Ball an extremely cost-effective proposition as it is the only product that is below the efficient frontier. Because of this I am not considering (for now) that the price is a true reflection of the actual commercial installed cost. I may be wrong and I will be happy to correct myself as it is a truly good result and I look forward to learning more about a design I had hitherto disregarded.

Now we get to the four machines that seem to form the efficient frontier. The Fortis units are doing exactly what I expect them to do – dependable performers that are good in low winds and not quite harvesting to the maximum in high winds. To understand this point compare April and May production for the Fortis units with that of the Zephy AirDolphin, and also look at the monthly windspeed table on the lower left of slide 7. See how the Apr/May winds obviously came through as high wind pulses contain a greater fraction of energy and how the Zephyr could more than double its performance whereas the Fortis units were held to 1-1/2 or so. This suggests that the Zephyr has over-invested in high-wind performance to the detriment of its low-wind cost-effectiveness. The Fortis designer is very focussed on cost-effectiveness and he would appear to have made a commercially sound compromise here. The Fortis installed prices are perfectly believable.

The Southwest Windpower (SWW) Skystream unit is the third good performer. For many years now Ampair has competed commercially with the SWW designs and we have been under no illusions about their competency in making cost-effective turbines (and the slickest marketing in the industry). We have been warning our bigger brethren to brace themselves and we appear to have been right to do so as the Skystream is very good – even more so if you replot this graph as a swept area graph. There are varying stories about exactly how much US government support the Skystream and SWW have received over and above the venture capital but it is paying off (I hope the UK government realise that this is a hint, but the UK government representative for microgeneration cancelled his talk at the last moment which gives you a feel for things here). I will point out that the Skystream is underpriced in this report and the true installed price is more like EUR 14k. The Belgian distributor has installed this unit and this is because the Dutch one was more expensive – someone has swallowed a chunk of cost here. Once the Skystream is upped to EUR 14k it is exactly on trend with the Fortis products.

It is time to critique the Ampair 600. Firstly our installer has overcosted this one as our boxed set only costs £2850, so an installed price of EUR 89250 is rather eyebrow raising. Reducing this to a more realistic EUR 6000 brings us on-trend, but don’t go beating Eco-Rietpol up for that price as he is well aware that fixed costs are high for small installs and he won't underprice. Secondly the energy output is low versus the Zephyr because we have effectively missed the first two months which are energy rich, because we were shutdown until week 8 whilst we waited on our safety upgrade (but that is our fault so we will take it on the chin). Comparing the next four months one sees that the Zephyr and the Ampair are trading kWh for kWh with each other which is exactly what I would expect as we are both manufacturers with an aero/mech/electrical competence and a power electronics competence and we both basically know what we are doing. In the last month the Zephyr gets slightly ahead again which is again to be expected as this was another energy rich month. I know that the inverter on this Ampair is the SMA 700W whereas the Zephyr tend to use the larger SMA units and so can better harvest the high wind pulses. Conversely however one can see that the Zephyr is a substantial energy consumer whereas the Ampair is a real miser. So I expect Ampair and Zephyr to continue to go head to head with these models for some years to come except that Ampair have a cost (or price) advantage at the momet. Once the Ampair data is corrected it is possible to draw an almost perfect line along the four efficient frontier units.


One thing I forgot to mention last night is that this Zeeland trial is supporting the general validity of the AWEA and BWEA small wind standards. In the Ampair 2008 catalogue we have tried to set out the information required in accordance with the BWEA standard - see page 5. One very important piece of information is the Annual Energy Production graph and for the the grid connected Ampair 600-230 this is the lower left hand graph.

At the test site the average windspeed was about 3.5m/s for the first six months. Drawing a vertical line on the AEP graph at 3.5 one can read off an annual production of 440 kWh which means that the Ampair 600 should have produced 220 kWh in the six month period. To compensate for the missing 7 weeks if we add approximately 60 kWh (per the Zephyr) to the measured 82 kWh we get 140 kWh.

This is actually a pretty good prediction (220kWh predicted versus 140 kWh actual) and a lot of the error is probably in the shape factor used to calculate the wind speed distribution. I am guessing from the limited data released so far but the Summer months have probably had a shape factor of 1.2 to 1.6 whereas the AWEA/BWEA standard requires the calculation to be performed with a shape factor of 2.0. I say this from observing the monthly differences, not because I have been crunching the data that the test sends out each week. During the winter months I expect that the shape factor will rise somewhat and hopefully in the next report it will be included. The shape factor is the small wind industry's equivalent of the solar PV industry's mythical 1000W/m2.

Of course the AEP is utterly dependent on the dependability of the power curve data. By now anyone reading should have figured out that this must be a power curve for net useful power to the consumer, and will now see why some folk prefer to give power curves that are somewhat ambiguous in this regard. It makes a lot of difference whether you measure power out of turbine or power into grid. It also makes a lot of difference whether you measure power out, or net power out. Ask your manufacturer !

Whether you trust your manufacturer is a separate issue. I have written previously (see blog for 29 December 2007) about Ampair practice in this regard. Our AEP figures in the 2008 catalogue come from the low tech approach. Now we have several further sets of data for the Ampair 600 which show it as anything from a 'rated' 350W machine to a 700W machine. Some of those numbers are high tech data logged outputs and some are low tech data logged outputs. Until we understand why the differences arise we will not be in a great hurry to commit further commercial suicide by publishing the lowest of these. Please excuse us but we already lose enough sales as it is. Again testing per the AWEA / BWEA standard helps but I see a nasty commercial dilemma coming where some machines test and publish independently certified results, but then go out of business whilst others merrily claim super-hero status and laugh all the way to the bank.


I expect that we will all scurry around frantically improving our designs and carry on into the next round. These wind speeds are absolutely typical of the sort of Summer windspeeds I expect at most urban fringes in reasonable wind locations. After 12-months we will know what the annual average will be.

Now could someone please send me the name of the Tesco highly paid expert consultant and of the people within Tesco who believed him. At least I hope they get their consultancy fee back.

Good luck to all manufacturers who participated. I fully expect that we will all (Ampair included) have lots of very public failures over the next six months, and I look forward to revising my opinions as new facts come to light.

For further info on the trial see:

Labels: , , , , , , , , ,


Post a Comment

<< Home