Lessons from Installing an Eoltec Scirrocco [Archive]

View Full Version : Lessons from Installing an Eoltec Scirrocco

Frank Janssens

27th November 2007, 09:38

Hi All,

I 'm a newcomer and have no experience with building wind power installations. I want to share the story of my son-in-law Lambert when he decided to install an Eoltec Scirrocco (=SC), 6kW. Lambert ( a handy-man)grows flowers and wanted to cut back on the electricity bill for warming up his green-houses. So after some browsing on the internet he ordered a SC. It is a professional quality turbine and there was an agent in Holland. These arguments made up for the high price. He ordered the SC in July06 (announced delivery time 6 weeks, downpayment 2/3) to be ready for the winter. The monopole tower came in Oct.The agent ,who shall remain nameless, had not a clue about the concrete foundation.The turbine arrived in Jan 07. Also there, the agent had no idea how to go about. Lambert asked around and started the installation. The part shown on the Solacity website (demo Woodstock) was all right but the electrical part was a real problem.Only the third wiring scheme received , for connecting the wind interface (PVI-7200) and 2 units PVI-3600+ transfo, was usable in Holland. The connection to the Energy counter, after the circuit breakers is still doubtful. In Feb 07, the SC was running fine at 3 Beaufort. After 3 weeks, the SC developped a frightening howling at 8 Beaufort . (See www.jajaja.info/3scirocco.MOV) So, panic all around but the agent said "no problem". 1 week later, the turbine jammed. Contacting Eoltec via the telephone number on their website, you get an answering machine telling you to sent an e-mail. After several panic e-mails and letters in solid french, we NEVER got any reaction from Eoltec. At this point the Green Power community (Matt Tritt, Rob Beckers, Logan Bryce) came to our rescue. Via Rob Beckers we got some reply from Eoltec and Logan Bryce with his experience found the problem while in Montana !. He advised us to check the Bussman 15 Amp fuzes in the wind interface (The thermal breakers on the wild current hadn't tripped). Indeed, one of the fuzes was open. The turbine was shipped back for repair. I understand that in such situation, the mechanical load on the shaft is not symmetrical as 2 phases give a reaction force to the wind and the third is not. So, when current is drawn, their is a torque on the turbine that wants to go to somehere not compatible with his axis or a re-orientation movement. (Yaw shudder is also an example of this, although of different origin). In July the turbine was back and repaired (so again, order a crane for the tower etc.,...). All our questions, what exactly had happened remained unanswered. In August, the SC was running again. The thermal breaker is, by trial and error, now set to 14 Amps instead of the 17 Amps on the diagram to make sure that all 3 phases of wild current are in the same state ( A replacement fuze in the PVI-7200 showed the same problem !). Now the SC is performing up to its specifications. Only, at low speed, when picking up wind, their is a strange noise that disappears at higher RPM. My wild guess is that the first bending frequency of the tower( ~1000 kg steel for 12 m) is too close to 40 RPM where the SC starts turning.

In short:
a) For such a project, make sure you have an EXPERIENCED dealer/"installateur" or you risk to have collected a heap of expensive junk. Solacity is compiling a list of thrustworthy dealers. Unfortunately for us ,Canada only. For us, Logan Bryce (pineridge products - Montana) has the highest possible ranking.
b) Eoltec is not such a big company that it has an independent PR unit. It's the kind of company that you expect to care about their clients and help them out with problems. In fact, the opposite is true, all letters, e-mails remained unanswered. Only via Rob Beckers (Solacity) did we get some feedback. The shortest route from France to Holland was via Canada ! Matt Tritt , (wind department for DC Power Systems of Healdsburg, Calif) had a similar experience with Eoltec. Eoltec should either give their nameless agent in Holland some training before accepting him as their representative or get rid of him.
c) Make absolutely sure that the 3 phases of wild current to the interface are always in the same state.
d) Noise when the turbine picks up speed (low RPM) ??
see: www.jajaja.info/2scirrocco19AUG2007.MOV

Thanks to the GREEN POWER community , for their disinterested help and making it possible that Lambert's project ended well.
Frank Janssens

Rob Beckers

27th November 2007, 15:05

Hi Frank,

A couple of notes: The rectifier box, PVI-7200-WIND-INTERFACE from Power-One/Magnetek, should have come with 20A fuses in it. As your story shows, and Eoltec has told us, there are boxes that erroneously shipped with 15A fuses in them. I've been checking every box we received to make sure it has 20A fuses. So far they all have. If you have not done so yet, you can replace all 3 of the 15A fuses with 20A ones, the type is KLKD 20 (fast) from LittelFuse. I'm sure Bussman makes equivalent fuses too.

Did you have a thermal (motor) breaker in line with the 3 phases coming from the wind turbine when it was first installed? I'm no specialist on thermal breakers, but according to my information an open phase will automatically trip the breaker (and disconnect all 3 phases) at a low current. At 14A I would expect that it will trip regularly during high winds. If Mark Parsons reads this, I would like to hear his comments (he IS a specialist on thermal breakers). For my part, I have tried to collect all the information that is important to a successful install in an electrical diagram, see http://www.solacity.com/Docs/Scirocco%20Hookup.pdf. This is of course for North America, though the European setup is largely identical. The diagram is to aid installers in getting it right the first time.

One of the reasons that 3-phase power is so popular for alternators is because it presents a constant mechanical load as the alternator turns. There aren't any spots in the circle with a smaller or larger load, making for a smooth operation. When one phase is dropped, there will be a number of places during the alternator's turn with a smaller load (causing it to speed up). To make matters worse, the inverter for the Scirocco tries to draw a certain amount of power based on either voltage or frequency (both largely depend on rotor RPM), causing it to put a larger load on the two remaining phases. It seems this sequence of speeding up and slowing down sets up a rather nasty vibration, leading to mechanical failure of the alternator. I suppose this is true not just for the alternator in the Scirocco, but other 3-phase wind turbines have a similar weakness. As it happens, the alternator in the Scirocco is made by a French company named Alxion. When one talks to those in the know, the general opinion is that Alxion makes some of the finest (and most expensive!) alternators available for wind turbines.

I realize you're stuck with it now, but still I'd like to say this so others will not be tempted: A 12 meter (40 feet) tower is very, very marginal. Especially for a wind turbine this size. It may work in a wide open field with no obstruction in sight, or at the edge of a cliff facing the wind. In general though, the rule of thumb (which in itself is an absolute minimum requirement) is that the tip of the lower blade should be 30 feet (10 meters) above any obstacle in a 300 feet (100 meter) radius. Even without significant obstacles, a wind turbine at 40 feet is going to have a hard life in rather large wind shear between the upper and lower blade. Add to that any turbulence created by obstacles and you get the idea. Personally I would like to put the minimum tower height at 60 feet, and even that only in open terrain without obstacles. I'm a big fan of tilt-up towers, for their ease of installation and the ability to tilt the tower down for turbine maintenance. Renting a crane is not cheap!

I certainly don't want to defend Eoltec here. In fact I hope they read this and draw their conclusions from it. I've been in business (not the renewable energy business, but where I had to interact with customers) for 20-some years now. It is my sincere believe that being responsive to customers is probably the most important thing if you want to continue doing business with people. The other side of the coin is that Eoltec is a relatively small company, and they seem to be doing very well. Meaning that they're all working their behinds off in trying to keep up. My guess is that upper management there thinks it's better to let their distributors talk to customers and personally stay out of it, to avoid opening a potential can of worms that can cause endless additional work. I believe they're wrong, but it's not my business...

To end with; I hope your story has a happy ending and your son will enjoy his Scirocco for a long time to come! May it produce many kilo-Watt-hours!

-RoB-

P.S. If you have a moment, post one or more pictures of the wind turbine. It's easy to do, just hit the 'attachment' button and browse for the picture(s). The forum will automagically resize them.

Paul Bailey

27th November 2007, 20:25

Yes thermal overloads on one or more of the three phases trips the breaker and disconnects all three legs. Paul:)

Frank Janssens

28th November 2007, 03:38

Hi Frank,

...erroneously shipped with 15A fuses in them.

The attachment shows the 15Amp fuzes in our PVI-7200...

Did you have a thermal (motor) breaker in line with the 3 phases coming from the wind turbine when it was first installed?

Initially, no thermal breaker was installed as it was not delivered with
> the system. This illustrates the kind of problems you have when the
> installer doesn't know the first thing about his product. You expect the
> delivery to be the complete package.

...3-phase power presents a constant mechanical load as the alternator turns.... When one phase is dropped, there will be a number of places during the alternator's turn with a smaller load (causing it to speed up)... the inverter for the Scirocco put[s] a larger load on the two remaining phases... this sequence of speeding up and slowing down sets up a rather nasty vibration, leading to mechanical failure of the alternator. ......

Thanks for the clarifications. As you can hear from the 5 sec. movie in my post, the noise is really frightening

A 12 meter (40 feet) tower is very, very marginal.

I agree with your comments on the installation. Our Scirrocco operates in, above average, turbulent conditions. Lambert had no choice, he pleaded
with the community that in their regulations, 15m height MAX, had to be read as to the center of the blades but the burocrats insisted that it was the highest point of a blade. At the same time they claim to promote the use of wind energy......After looking at Eoltec's installation in Nice (see attachment), I thought , well it's certainly not optimal but will probably only reduce the lifetime from , say 20 y to 15y.

... It is my sincere believe that being responsive to customers is probably the most important thing.......

I wish (sigh) more managers shared your view about interacting with the
> customers !!

....I hope your story has a happy ending ...
So do I, TX :)
-RoB-

P.S. If you have a moment, post one or more pictures of the wind turbine..

Hereby some pictures of the first Dutch Scirrocco

Rob Beckers

28th November 2007, 07:05

Yes thermal overloads on one or more of the three phases trips the breaker and disconnects all three legs. Paul:)

Actually, what I was trying to get at is that no load (the opposite of an overload) on one phase causes the breaker to trip at a current well below the set trip current. At least, that's what I was told by a breaker guy...

Regardless of that, the way the inverter works is that it will try to draw a set amount of power based on either the AC frequency (directly proportional to turbine rotor RPM, the the way the PVI-6000 inverter is programmed), or the DC voltage (somewhat proportional to rotor RPM, and used for the PVI-3600 inverter). When one phase drops out the inverter will therefore try to draw more current from the remaining phases, making it more likely that the thermal breaker will trip that way.

-RoB-

Frank Janssens

28th November 2007, 11:37

Yes thermal overloads on one or more of the three phases trips the breaker and disconnects all three legs. Paul:)

;) I agree fully that only an overload (=too much heat) trips the thermal breaker directly. If one phase is open, the power-point-tracking system in the PVI-7200 tries to draw more current from the 2 remaining phases , hence the thermal breaker is more likely to trip. So, this is a kind of indirect activation of the thermal breaker. Did I get it right, Rob ?

As there are many types of thermal breakers with different sensitivities, it would be nice to have either some brand names that are recommended or the specifications that must be met.

As I mentioned, lambert is now using a setting of 14 Amps that was determined by trial and error to make sure that this is compatible with the fyzes he has in the inverter.

Mark Parsons

29th November 2007, 08:57

Greetings Frank, Rob and Paul,

I am wading in on this thread. I have had communications with Rob in the past regarding the wild AC side of the Scirroco. The Telemecanique breaker specified has both thermal and magnetic trip properties. I also specified to Rob an equivalent Allen Bradley product for him that would have better availability in North America.

I would source the AB 140M-H8P-C15 motor circuit breaker as an alternate to the Telemecanique.
Specs can be found at http://www.ab.com/en/epub/catalogs/12768/229240/229254/229469/229517/tab7.html

Fuses should not be used in place of a 3 phase thermal / magnetic breaker in the feed side from the turbine to the rectifier. It could allow a phase loss situation to continue causing potential alternator and / or mechanical damage. As apparently happened to Lambert. However, fuses are required to meet branch circuit wire protection and fault current protection by code. Trip characteristics for fuses are very different from breakers.

The above breaker trip characteristic charts are at: http://www.ab.com/en/epub/catalogs/12768/229240/229254/229469/229517/tab10.html

Bussman class CC fuses trip characteristic curves are at: http://pdf.directindustry.com/pdf/cooper-bussmann/full-line-catalog/12341-260-_233.html

It is important to coordinate trip characteristics in any power circuit to ensure that the proper protection element is working first.

IMO, the breaker should be monitored and a shunt trip stop switch automatically engaged if the breaker trips. This would prevent overspeed damage from the open circuit. I understand the Scirroco has RPM limited auto pitching blades so maybe this is not a significant problem.

Regards,
Mark

Mark Parsons

30th November 2007, 09:01

Frank Janssens

1st December 2007, 07:54

Gentlemen,

Further analysis of trip characteristics between circuit breaker and fuses in the Power One rectifier shows issues to protect against phase loss situations.

Rob indicates replacing the Bussman KLM 15A fuses with Littelfuse KLKD 20. Datasheet showing curves: http://www.littelfuse.com/data/en/Data_Sheets/KLKD.pdf

Compare with the AB 140M trip characteristics near bottom of page: http://www.ab.com/en/epub/catalogs/12768/229240/229254/229469/229517/tab10.html

It looks to me like the fuse will clear a substantial overload of durations 0.02 to 80 seconds. This will continue the failure mode of fuse related phase loss causing alternator damage.

Can the fuses in the Power One device be sized up and to different motor protection type fuses like Littlefuse CCMR 35A? Characteristics: http://www.littelfuse.com/data/en/Time_Curve/CCMR.tc.pdf

This CCMR 35 fuse would provide a smaller window of failure characteristic. The midget fuse holder in the Power One would accept a class CC 30 amp fuse but not the 35A. Would require a different fuseholder. Someone from Eoltec should approach Power One to resolve this conflict and failure mode.

Perhaps Power One could integrate the circuit breaker instead of fuses?
My $0.02 worth...

Regards,
Mark

Hi Mark,
Thanks for your valuable comments and information. I like your idea
Perhaps Power One could integrate the circuit breaker instead of fuses?

It would make life easier for the installer and customer....

some questions:
a) you quote the AB 140M-H8P-C15 as an alternative to the telemecanique,
but on the url you give, tabblad PRODUCTS, it says under the 3-phase kW ratings : 230V => 3 kW ; 400V => 5.5 kW
and the Eoltec scirrocco gives 6 kW at 240V (website Solacity)
Do I misunderstand this or do we need the 140M-H8P-C30 that has 5.5kW at 230V ?
b) the fuse will clear a substantial overload of durations 0.02 to 80 seconds.
I do not understand what you mean by this or how you get these numbers from the characteristics on the page "Littelfuse :Class CC Fast-acting...."
(excuse my ignorance)
From the next sentence I see that the fuse does not trip under the conditions given by this sentence and that the failure mode continues

Best Regards, Frank

PS What are you holding on your avatar picture :confused: ?

Mark Parsons

1st December 2007, 10:02

Greetings Frank,

Sorry for any confusion with some electrical power fault clearing technical details.

a) you quote the AB 140M-H8P-C15 as an alternative to the telemecanique,
but on the url you give, tabblad PRODUCTS, it says under the 3-phase kW ratings : 230V => 3 kW ; 400V => 5.5 kW
and the Eoltec scirrocco gives 6 kW at 240V (website Solacity)
Do I misunderstand this or do we need the 140M-H8P-C30 that has 5.5kW at 230V ?
These products are designed to protect motors from mains supplied power. The 3kW @ 230V 3 phase is the motor protection rating. A 3kW induction motor can easily draw 6kW of power under overload conditions. In your case you are protecting from an overcurrent and phase loss situation for a PMA. 15A * SQR(3) * 230VAC = 6kW. On the 140M trip characteristic curve you can see it actually requires about 1.3x of setting current to trip after 1000 seconds. In a continuous static overload for a 15A breaker setting this would be 15A * 1.3 * SQR(3) * 230VAC = 7.8kW

b) the fuse will clear a substantial overload of durations 0.02 to 80 seconds.
I do not understand what you mean by this or how you get these numbers from the characteristics on the page "Littelfuse :Class CC Fast-acting...."
Trip characteristic graphs show the amperage vs time characteristics for the protection element (breaker or fuse) to clear the fault condition. By overlaying the 2 graphs you can see which device will trip first under specific amperage and time conditions. My analysis of this overlay indicates the KLKD 20A fuse tripping before the breaker and continuing to potentially cause the phase loss failure mode situation in certain short duration substantial overcurrent conditions. A powerful wind gust perhaps? The CCMR 35A fuse would provide a smaller window of failure opportunity when coordinated with the AB 140M breaker.

Does this explanation help any?

Regards,
Mark

PS. Answer. My avatar is me holding my spinning bladeless turbine being powered by the output from my paddlewheel sling pump during a river site test. More info on the paddlewheel can be found under a Microhydro thread on this site and also on Dennis Buller's site at www.wildwaterpower.com.

Rob Beckers

4th December 2007, 09:12

Paul, from talking to Mark it seems thermal breakers don't trip on differential current. My information was wrong. So, at least one of the phases needs to go beyond the set current for it to trip.

Looking at the fuse curves, this particular 20A fuse won't trip for currents below approx. 25A. What will happen when a phase gets disconnected from the alternator is that the inverter will try to draw the same power from the remaining two phases (this is true for the PVI-6000 inverter, but not entirely for the PVI-3600 setup used by Frank's son-in-law). For full power, when there's lots of wind, this means 6500 Watt and a voltage between the two remaining phases of 240V, so around 27A. This is a bit simplified, in that the voltage will likely drop a little or a lot (depending on how the alternator handles the increased phase load), and I don't know where the alternator tops out in terms of current. Still, in view of those numbers it looks like it will take some time, if at all, to blow a 20A fuse on two phases. At 27A it would be around 500 seconds judging from the curve.

The thermal breaker curve shows nothing much happening until 1.3x the set current, or 22A for a breaker set at 17A. If the alternator would indeed draw 27A in case of a phase loss, it would take around 250 seconds by the looks of it for the breaker to trip. That would still be faster than the 20A fuses would blow. Just in case a 20A fuse would blow before the thermal breaker, this means there's only one phase remaining and no current will be drawn from the alternator. That's OK, the Scirocco can run unloaded for any amount of time (it gets noisy though when running unloaded).

There are lots of "ifs" in this analysis, lots of uncertainly. This little exercise is interesting though. We want the thermal breaker to trip in case of a disconnected phase. We want this to happen reliably and preferably quickly, so the alternator does not get loaded asymmetrically for too much time. We would like for this to happen at wind speeds short of "full power" as well. Since the thermal breaker's curve shows it won't trip for anything short of 1.3x the set current, and a worst-case full power current for the Scirocco is just short of 17A, it would seem that setting the thermal breaker to 14A or 15A makes more sense.

Frank mentions that the breaker for their turbine is set to 14A. Maybe he can comment how often, if at all, the breaker trips in high winds. Meanwhile, I think I'm going to change my recommendation on the Scirocco hookup diagram to set the breaker at 15A (instead of 17A) and increase it in 1/2A increments in case of nuisance tripping.

Comments welcome!

-RoB-

Frank Janssens

4th December 2007, 09:34

Greetings Frank,

On the 140M trip characteristic curve you can see it actually requires about 1.3x of setting current to trip after 1000 seconds. In a continuous static overload for a 15A breaker setting this would be 15A * 1.3 * SQR(3) * 230VAC = 7.8kW

This part is now clear..so, with this setting (15A) the thermal braker trips after 16min 40 sec if the generator delivers 7.8 kW

Trip characteristic graphs show the amperage vs time characteristics for the protection element (breaker or fuse) to clear the fault condition. By overlaying the 2 graphs you can see which device will trip first under specific amperage and time conditions.
Does this explanation help any?

For this part it helps to some extent. I do understand that we want to make sure that the thermal breaker goes first (while the 3 phases deliver current) before one of the fuses blows. Once a fuse in the PVI7200 is open only 2 phases deliver current and the generator operates in bad conditions and, eventually, breaks down.
I'm happy to leave it to specialists like you (thanks for your patience) to get that info from the graphs to find combinations motor/fuses that assure that the thermal breaker goes first.

At this moment there are "Ferraz Shawmut slow fuses 15A/ 600V AC-DC IR 110kA" in the PVI and the thermal breaker is set to 13 A. Two day ago their was a heavy storm in Holland (Beaufort 9) and the thermal breaker tripped , so the scirrocco was save...

Do you have an idea what kind of damage is done to the generator when 1 phase is open ? In our case is got stuck, Logan apparently , experienced that condition also, but did halt the system before any real damage was done. The noise is really frightening as you can see/hear from the little movie on my initial message.
I was told that the only effect would be that the rotor spins up and down inside one revolution (so much faster than changing speed due a variable wind) as the mechanical load on the axle turns around. A priori, I would expect also some side torque ??
If this is a silly question, please disregard it.

Many thanks for your help,
Frank Janssens
PS my guess for the low resolution avator picture was that were holding a plate with fine liquors and were starting a party....;)

Mark Parsons

4th December 2007, 15:17

Greetings Frank and Rob,

I believe you guys got this power coordination thing in hand.:)

Good to hear that the breaker tripped and not the fuses in a recent situation. Frank - If you can get the Shawmut Fuse catalog number, (should be printed on the fuse) we can look up the curves to ensure coordination is expected in most (all) conditions. I assume you are using the Telemecanique circuit breaker? The characteristic curves are also published for this device.

Single phasing a PMA under heavy load can do very significant shock torque damage from the high asymetric loads. Damages bearings, bolts, shafts and perhaps magnet degauss. Unlike an induction motor (generator) that will induce a 3rd phase in the open windings to protect considerably against single phase shock loads, a PMA is especially vulnerable.

Attached is the photo my avatar was extracted from to hopefully reduce any further speculation. Fine liquors and a party eh? :D

Regards,
Mark

Frank Janssens

7th December 2007, 12:39

Greetings Frank and Rob,

Good to hear that the breaker tripped and not the fuses in a recent situation.

The thermal breaker (=tb) Lambert has installed is MOELLER PKZ MO16

the Shawmut Fuse catalog number,
On its way.....

Results from experience:

tb set at 13.2A => Beaufort 8 , after 1 hour the the tb goes "open" OK
tb set at 13.8A => Beaufort 8, after about 1 hour, a 15A Shawmut fuse blows
starting the now well known growling of the Scirrocco. (Supplementary cooling on the PVI7200 installed !)

While functioning at 8 Beaufort, 6kW are delivered to the grid.
Lambert is still hesitant to replace the 15A fuses by 20A as Rob proposes, out of fear for internal overheating/damage in the PVI7200. May Rob can reassure him ?

Single phasing a PMA under heavy load can do very significant shock torque damage from the high asymetric loads.

Glad to read this ! Nice to have a confirmation from a specialist on my line of thinking on this one.

Unlike an induction motor (generator) that will induce a 3rd phase in the open windings to protect considerably against single phase shock loads, a PMA is especially vulnerable.

This is a difficult one for me, but...maybe if I scratch my head enough, the light will come...

the photo my avatar was extracted

This photo does justice to you and your project. Sorry for my speculations :)

All the best, Frank

Rob Beckers

7th December 2007, 13:18

...
While functioning at 8 Beaufort, 6kW are delivered to the grid.
Lambert is still hesitant to replace the 15A fuses by 20A as Rob proposes, out of fear for internal overheating/damage in the PVI7200. May Rob can reassure him? ...

As I said before in a private conversation with you: The PVI-7200 comes from the factory with 20A fuses installed. In other words, this is not my idea, but Power-One/Magnetek approves the use of 20A fuses and that is how the unit is UL/cUL listed here. If you look at Power-One's own brochure (see http://www.solacity.com/docs/PVI-WIND-INTERFACE.pdf), it says in there "Input overcurrent (fuse protected): 20A". It also says that the maximum approved operational current is 16.6A, ie. that's enough to blow any 15A fuse. That some units were shipped with 15A fuses seems to have been a mistake on Magnetek's part.

Thank you for the thermal breaker setting information. Based on the curves for the recommended Télémécanique I have changed my recommendation. Instead of 17A I recommend installers to set it to 15A (see http://www.solacity.com/Docs/Scirocco%20Hookup.pdf). The idea is that this makes it more likely for the thermal breaker to trip in case a phase is disconnected (especially if the wind is not enough to produce full power). This looks to be a safe value, not causing nuisance tripping of the breaker.

Hope this helps...

-RoB-

Frank Janssens

17th December 2007, 11:37

As I said before in a private conversation with you: The PVI-7200 comes from the factory with 20A fuses installed. In other words, this is not my idea, but Power-One/Magnetek approves the use of 20A fuses and that is how the unit is UL/cUL listed here. If you look at Power-One's own brochure (see http://www.solacity.com/docs/PVI-WIND-INTERFACE.pdf), it says in there "Input overcurrent (fuse protected): 20A". It also says that the maximum approved operational current is 16.6A, ie. that's enough to blow any 15A fuse. That some units were shipped with 15A fuses seems to have been a mistake on Magnetek's part.

Thank you for the thermal breaker setting information. Based on the curves for the recommended Télémécanique I have changed my recommendation. Instead of 17A I recommend installers to set it to 15A (see http://www.solacity.com/Docs/Scirocco%20Hookup.pdf). The idea is that this makes it more likely for the thermal breaker to trip in case a phase is disconnected (especially if the wind is not enough to produce full power). This looks to be a safe value, not causing nuisance tripping of the breaker.

Hope this helps...

-RoB-

hi Rob and Mark,

I hope I'm not boring you by continuiing on this subject ....
With thermal breaker settings of the MOELLER PK-ZMO-10 (see attachment) of 13.7 and 13.5 A, a 15A Shawmut Ferraz in the PVI goes first. The next experiment Lambert did was to take the fuses OUT of the PVI and replace them with standard 16A domestic fuses (attachment fuze) as :
- outside they are better cooled
- they cost about 30 Eurocent a piece while the Shawmut fuzes are close 10 Euro's a piece (and he has blown already several !)
Since then, the thermal breaker , at 13.5A has not tripped and none of the domestic 16A fuzes has blown. At the same time the funny noise at the starting (low) RPM has gone. This funny noise can be heard on the url
www.jajaja.info/4scirocco.mov (the url in my starting message had a typo, this 5 sec movie should be OK). I have no clue for an explanation ...maybe, it's just a coiincidence. The fact that 2 things happen at the same time is not a proof that one causes the other.
Now the scirrocco is fine up to beaufort 8 and there is an evolution towards the 20 A that Rob/Magnetek advise. Maybe the 15A fuzes were a leftover of their previous model the PVI-5200 ?.
The argument that we proceed with care are is that we don't want the PVI to overheat. Also Logan installed a supplementary cooling ( I didn't hear from Logan any more ?).
From Beaufort 6, the scirrocco delivers his 6kW. Noordwijk is just in the green region of attachment "wind nederland" (average 6.5-7 m/s or beaufort 4). So more wind is quite common and occasionally there are peaks that it produces more (theoretically the power delivered remains constant). So the PVI-7200 has then to dissipate more and becomes quite warm.
With the current settings, everything is fine, it's just that the return is not optimal as in strong winds the thermal breaker is activated. The procedure of halting the scirrocco with the mechanical brake to 40 RPM and then shorting the thermal breaker works to great satisfaction.

Many greetings and thanks for all the help, Frank

Bryan Rendall

3rd December 2008, 12:05

Hi All,

Its a long time since this thread has been added to, but heres something new and worthwhile I hope!

For all the UK turbines, we have now re-defined the protection scheme to greatly reduce the likelihood of single phasing, yet, provide a similar level of protection to the Wind Interface Box, and the generator for overload / short circuit.

To achieve this, we now replace all Wind Interface Box fuses with solid links, and achieve protection using a Z characteristic 20 amp 3 pole circuit breaker on the 3 phase AC between the generator and the Wind Interface Box. An extensive study into protection grading has shown us that this gives us a better level of protection, as well as 3 pole tripping. Z characteristic breakers are designated for semiconductor protection, and have the fastest / lowest multiple of rated current tripping curves we can attain. They are however, not the easiest to find, and probably not in small multiples either.

We are almost complete in retrofitting all UK turbines to this spec - This is a UK initiative we are carrying out as the UK/Ireland distributor, with the approval of Eoltec.

Result is an elimination of the risk of single phasing due to a fuse rupturing in the Wind Interface Box for whatever reason, and a greatly reduced risk of single phasing full stop!

Cheers

Bryan Rendall

Rob Beckers

3rd December 2008, 13:08

Thanks for the info Bryan!
Would you have a link to a spec sheet for those breakers? I would like to take a closer look at them.

For the Sciroccos here in Canada we provide a Télémécanique GV2ME20 thermal/magnetic breaker (they are made by Schneider/Square-D). The breaker is then set to 15A. It'll handle 125% on 3 phases indefinitely without tripping, so it works for the 16.5A that a Scirocco expects to see on each phase at full power. When there is a phase failure, these breakers will trip after approx. 250 seconds (4 minutes) at the set value (15A), faster for higher currents.

-RoB-

Frank Janssens

5th December 2008, 20:05

Bryan Rendall's reaction confirms that it was not a bad idea to start this thread.
As a newcomer to this field I worried that my question's were stupid. I understand now that in the design of the Scirrocco, the problem of single phasing was complicated due to the evolution in the different models of wind interface boxes. It is clear that single phasing is dangerous and must be excluded. Now, different valid solutions were given in Canada (Solacity) and the UK (BJRE) which shows again the importance of the technical competence of the installer. The situation in Holland, the country by excellence for wind-energy devices, is still bad. No recognized dealer on EOLTEC's list although it's possible to order their products.
I like very much the Bryan's idea to rely solely on an appropriate type of thermal breaker (if I understand correctly his modification).

The only dutch scirrocco is now running flawless for over a year up to Beaufort 8.
Lambert still shuts it down at Beaufort 9 but that that has more to do with the vibrations of the tower which could only be plaved in very poor environment conditions (turbulence).
If anyone on this forum is interested in tower vibrations, I would be happy to start a thread on this one .

Many greetings to all of you,
Frank Janssens

Bryan Rendall

3rd September 2009, 06:49

All,

Sorry I have been away a long time from this forum!

We now have almost 1 year of experience using the Z20 breakers, and they are perfect for the job. Coupled with solid links in place of fuses in the Wind Interface box, they provide ideal protection for the generator - Even a short circuit in a rectifier diode will take out the breaker instantly, dropping off all 3 phases and protecting the generator.

We now fit these in all UK turbines as a matter of course, and I would encourage anyone else to contact their installer to enquire about upgrading to this higher level of protection.

Cheers

Bryan

Rob Beckers

3rd September 2009, 07:21

Bryan, could you please post a link to a manufacturer/type of those breakers. I'd like to take a closer look at them. Still using the thermal breakers here, we now set them to 18A though to avoid nuisance tripping (had that with some installs during gusty high wind conditions).

-RoB-

David Simms

6th March 2012, 12:27

Actually, what I was trying to get at is that no load (the opposite of an overload) on one phase causes the breaker to trip at a current well below the set trip current. At least, that's what I was told by a breaker guy...

-RoB-

Unfortunately, I'm a bit late with a response to this interesting thread. My take on this situation -one phase goes open and causes the fuses to blow in the other two - is as follows;
wind power in = electrical power out, all things being equal.

Hence, when one phase in a three-phase alternator, goes open and conditions remain constant, the power out must adjust to equal the power in.
So, if the voltage remains constrained then the current in the two operational phases will rise to 3/2 of the value that was flowing before one phase went open. if the turbine is operating at fairly low wind speeds it should be able to lump along, quite literally, for a while in this state. However, if the wind picks up to moderate values, the rating of the remaining fuses will be exceeded and they will blow.
All this will happen well below wind speeds that would cause the pitch mechanism to limit the turbine speed.
Throwing in a few convenient numbers...suppose that the fuses are rated at 15 A.
If there is just 10A flowing in each phase, before one fuse blows, (30A total) then that 30A would now be spread over two phases, instead of three,a and the fuse rating would be reached.

Rob Beckers

6th March 2012, 14:08

Hi David,

That's a blast from the past (the thread)...
That quote of mine is a bit unfortunate, what it should say is that the thermal breaker will blow at a lower current than the set value, rather than "no load" (it will still take a load to reach that current).

Your are largely correct regarding the regular fuses, with one blowing it would cause the others to carry more current. Unfortunately we've found in every instance where the Scirocco alternator broke due to single-phase running that one fuse had blown, while the others were still intact. So, it seems entirely possible to run for a long time at partial output power, without blowing another fuse, while still blowing the alternator.

With that in mind we replace the 20A fuses in the rectifier box now with 30A ones (so they never blow), and use the thermal breaker as the overcurrent protection device. It will switch off all phases at the same time when it pops, and pop at a lower current when it is going single-phase. So far that has worked well.

-RoB-

David Simms

6th March 2012, 18:01

This is quite interesting. Of course, knowing Southern Ontario as i do, higher wind speeds could be quite rare...maybe rare enough to allow the turbine to function for a fairly long time.
Next hypothesis...as pointed out, when one phase is disconnected, the alternator will run in a pretty rough fashion. The counter torque, applied by the dead phase, will be absent and yet, the two other phases will apply their counter torques. So, in each revolution, the braking action of the alternator will be applied and released. i wonder whether, if this scenario persists for a while, could that lurching action cause more "micro" movement in the live coils because they'll have regular surges, for short periods of time. Then, the coils move a bit, one coil rubs on its neighbor, or on the core laminations, eventually causing a short or a ground.

Was it possible to do a post-mortem on the alternator to pinpoint the failure location ? I wonder whether an improvement in coil insulation might help.

Certainly, what you've done to disconnect all three phases at the same time, is the way to go.
I used to run a Jacobs, rebuild same and sell them. I still have two but I'm impressed by this Scirocco.

Rob Beckers

7th March 2012, 07:55

David, when one phase wire is disconnected (blown fuse or otherwise) the alternator will run on a single coil (out of 3 phases). Loosing one wire knocks out two of the three phases, leaving just one. And, yes, it runs very rough. The latter is the issue as it sets up a vibration in a spring inside the alternator (large machined aluminum piece that is there to dampen the rectifier noise that otherwise feeds back to the rotor, without it the turbine would be 'humming' loudly, something you may have observed from other turbines). The spring fatigues after a few weeks and breaks. Usually the rest of the alternator is fine. The coils are entirely encapsulated in epoxy or some other substance, nothing happens to them. The spring gets replaced and it's back up and running again. It's just a pain to do since it means taking the turbine down, and taking the alternator apart.

-RoB-

Rob Littiken

7th March 2012, 21:30

Rob Beckers

8th March 2012, 07:29

Rob, no, this is a spider-like flat structure that connects the rotor (with the magnets) to the shaft of the alternator. It's machined out of a thick sheet of aluminum, and the arms have been dimensioned just right so that they form a low-pass filter for the frequencies that the rectifier injects back into the alternator, so the rectifier hum doesn't feed back to the rotor/blades. It works very well (running without it results in a much noisier turbine), but adds vulnerability.

-RoB-