Wind & Solar Data for Southwestern Ontario

[Ric Murphy]

Good Day,
Not sure who will find this data useful or interesting but assume someone will. Wish I had it before I invested in my system but it likely wouldn't have changed the end result much.
In Mar 2008 I purchased the new Flexnet DC battery monitoring system from Outback. In conjunction with Wattplot monitoring software this allowed me to finally collect “accurate” production and usage data for all the components of my system. I now have one full years worth of data that I can share with anyone interested.
A brief overview of the system and site...
Location
60 acre farm located half way between Sarnia & Chatham in Southwestern Ontario, Canada
Very flat land with unobstructed exposure to the west
System Components
8 – 150w, 24V Isofoton solar panels (wired for 48VDC)
Outback MX60 charge controller
900w (@28mph) wind turbine on 75' tilt up tower
Outback VFX3524 inverter with X240 auto transformer (120/240VAC)
1000ah FLA battery bank (single lift truck style battery – 1910 lbs)

Components are housed in an insulated (but unheated) power shed 150' from the house. The turbine is 100' from the power shed and the solar panels are mounted on the shed roof. The panel angles are changed twice yearly for summer and winter. Grid power is supplied to the power shed with an underground 6/3 conductor from the main panel in the house. A greenhouse, shed, and barn are hard wired from the power shed. A single circuit runs from the power shed back to the house. This circuit powers all my main floor lighting, basement lighting and living room receptacles (TV, satellite, stereo, etc). Grid use is controlled by Outbacks HBX feature from the inverter. Grid use is triggered by user programmable battery min/max voltage/time delay settings. Grid use setting is 23.8VDC for 1 hour. Grid drop is 28.0VDC for 1 hour. During extreme cold periods I raise the lower setting to increase battery SOC to prevent the potential for freezing.

Data Collection
The Flexnet DC allows the connection of up to 3 shunts. In my system they are connected as follows:
Shunt A is connected between the battery and inverter
Shunt B is between the MX60 charge controller and battery
Shunt C is between the wind turbine and battery
The shunts measure the current passing thru each circuit. A positive value indicates current into the battery. A negative value indicates current leaving the battery.
The Wattplot software reads this data and provides a real time graph for all 3 shunts. Each night at midnight Wattplott takes the previous 24 hours worth of data for each shunt and dumps it into a spreadsheet. At the end of each month I retrieve the data from the power shed and reformat to my preference.

Data File
The attached PDF file is an export from the master spreadsheet. You will notice a few “glitches” in the export that I could not correct. They have no impact on the data. The file contains the last 12 months of data I've collected.
The first 12 pages contain the daily data in KWH's for each shunt for each month. Also included is a graph that I create from the monthly data.
Page 13 is a summary of the previous 12 months data.
Page 14 is a graph of the wind and solar production for the entire 12 month period.
Notes
1)The system is not capable of backfeeding into the grid therefore any surplus energy is wasted under the current setup. I need to look into some method to capture and use this energy in the future. One of the options is to convert to a grid tie inverter. I suspect I may be able to increase production by as much as 20%.
2)In December there was a problem with the Wattplot software and the data is incomplete for those days. I was able to extract the solar data from the MX60
3)During extreme cold periods in the winter I raise the grid use setting to maintain a higher battery SOC. The down side of this is a reduction in wind and solar production. The battery recharges sooner in the day and energy produced is diverted after that. I need to find someway to warm the battery other than electric heaters.
4)The turbine I'm using has a problem with the furling mechanism. It is furling too soon and as a result I am missing out on a fair amount of wind energy. I cannot quantify how much but would estimate possibly 30% annually.
5)Since every site is unique, this data only represents my specific location.

Conclusions
1)Initially I thought I had a decent wind site, however there are many, many days in the summer where the turbine never turns a single revolution. There are even more where the winds are so light that I do not make enough energy to even bother recording. However as the data indicates there are many days when the sun barely shines but the wind is blowing strong. The wind and solar clearly compliment each other and without both I would have many more days of zero production.
2)The investment cost of the solar vs wind was about 4:1. The energy production from each follows that ratio closer than I would have expected.
3)Solar is much more predictable than the wind. If I could only have one source I'd have to choose solar.
Ric

633

[Brian McGowan]

Ric,
This is quite a report. I like it. As I was reading it and got to this part I had some thoughts.

3)During extreme cold periods in the winter I raise the grid use setting to maintain a higher battery SOC. The down side of this is a reduction in wind and solar production. The battery recharges sooner in the day and energy produced is diverted after that. I need to find someway to warm the battery other than electric heaters.

My thought was to build a battery box out of that pink R10 rigid insulation glued together so you can remove it in the summer if you want. Then build solar air heaters into the south side of your shed and route the air to and from the battery box only. Between the mass of that battery and the insulated box and the air heaters you should be able to maintain a higer average battery temperature. I don't know what that massive battery sits on but getting insulation under it would probably help as well. You could also make an air heater as a diversion load and add that to the battery heating system. If you are making excess power anyway no matter where it comes from, you might as well use it to warm the battery so you can store more power in it. You would have to consider proper venting of hydrogen but that shouldn't be hard to beat. After the heat seeps out of your battery box it would be going into your insulated shed anyway making perhaps a little more comfortable. That probably doesn't matter. Do you have any data about the temperature in the shed itself?

Just my thoughts.

How do you move that battery around? I have the possiblilty of getting a 420ah fork lift battery for $280.00. It is 560# though and tall and skinny which kind of scares me with that high center of gravity and weight. If I tip it over I would never be able to stand it up again by myself. I also cannot guage it's condition beyond the voltage which read about 12.45V while it was sitting there but I hear these are very robust batteries. I currently have 3 type27 deep cycle at 115ah each but they are beat and are probably very sulfated and little more than energy sinks at this point.

Brian

[Ralph Day]

Brian,
Before you jump on a fork lift battery think...why has someone scrapped it? My experience around a lift truck service centre (minimal , but some) was that batteries with problems were changed out/turned in/traded in on new ones. One salvager used cells from 2 48's and a 36v systems to get one 48v battery that worked, other cells were fubar.

Can you get specific gravity readings on it? how long has it been sitting without charge? You may have hard crystalyzed sulfation on a "new"battery...and getting more problems than you already have.

Ralph

[Brian McGowan]

Ralph,
Trust me I'm not jumping. I have known about these since the begining of November and I've still not moved on them for the reasons you mention and the weight. I don't like the fact that they have been sitting and I don't know why they are not used for fork lifts anymore. The sulfation problem is what is probably wrong with the batteries I have now and it is likely that these fork lift batteries have some also. My only alternative is to get 4 new type27s at 115ah each from Walmart for $284.00 total ($4.00 more than the forklift battery and not as robust but much easier to move and brand new) but I know they will not last either for more than maybe 3 years. If I go away for a week and get a lot of cloudy nonwindy days they will be toast by the time I get back. Any other alternative I can find costs more than twice that and would only last marginally longer.
I don't want to hijack this thread but I have concluded that FLA batteries are poorly suited for alternative energy applications due to the sulfation problems. For this application you need a battery that is happy to hang out at 75-80% charged so you always have power available and you always have room to store power since you never know when power from nature will be available and you have to get it while it's there. In order to keep FLAs from sulfating you need to keep the fully charged all the time which means running a generator or having an overly large energy collection system so the batteries will get charged even under poor conditions and this is inefficient. The only type of battery that I can find that can handle these conditions are is the Edison Nickel/iron battery. The only ones I can find are made in China and are very pricy. I am considering making 1 cell for testing and playing with but it kind of burns me up that we had this technology and now we don't. They are leakier and less efficient during charging than FLA batteries but they take a beating without degrading. We will see where my little experiment lands me but for now I still have the current actual real world dilema. Can we have a vote? Which would you choose. Unknown forklift battery or 4 new type27s from Walmart?
Sorry to rant but I'm not happy with the available choices.
Brian

[Ralph Day]

Vote
For a full time off grid home...with me there all the time...FLA's. For a cottage/cabin with use on weekends, or short stays...1 week at a time...probably vrla/agm's. They may not last as long, but there's so much less work and worry, fully charged over the week then discharged when you get there. Never without an internal combustion type charger (genset) for either though. So, that is a vote for both! How Canadian!

Ralph

[Rob Beckers]

Interesting data, thanks Ric!
I've been doing a little comparison with what models would predict for your location. The Ontario wind atlas has your area down for an annual average wind speed of 4.82 m/s at 30 meters height. For 75' that works out to 4.63 m/s annual average. If memory serves me you have a Lakota wind turbine. When I plug the wind speed into the Lakota production model (using it's published power curve), the prediction is for 872 kWh/year. That takes 10% losses due to turbulence into account. Your measurements show 438 kWh for the year. Any ideas what could account for producing just 50%(!) of predicted? Maybe it was an anomalous year? Maybe the Lakota power curve exaggerates grossly? Maybe there are obstacles around the site that cause more turbulence? Or the simple models using a Weibull distribution just aren't accurate?

For solar, your area has an annual average insolation of 4.2 hours/day for a fixed area tilted at the site's latitude. You mention changing the angle bi-annually, which should improve production. For 1200 Watt in PV, average annual production would be 1.2*4.2*365.25 = 1840 kWh/year. I like to derate PV by 30% to account for temperature, charger, batteries etc., so that would make a predicted production of 1288 kWh/year. Your measurements show 1444 kWh for the year, or 112%. That makes sense, since you're looking at what goes into the batteries (and the batteries should be about a 10% loss).

So, in short, wind is 50% less than predicted, solar is spot on.

-RoB-

[Ric Murphy]

Brian/Ralph,
Some additional info & comments...
I chose this style battery after seeing one used in another system. Cost wise it was comparable or slightly less than using good quality 2V or 6V FLA deep cycle batteries. The main advantage was no cabling to fabricate or buy and maintain. Main diadvantage is the size/weight issue. You certainly wouldn't get one in your basement. Out here on the farm I used a fork lift to get it into the power shed and a pump jack hand cart to move it into position. I had a lot of experience with electric lift trucks in a past life and found these batteries to be extremely robust. The lift truck drivers would run the truck until the battery was almost dead before putting them on the charger. Then 3 hours later some idiot would take it off just to move one pallet and forget to put it back on. Even after all the abuse we'd still get well over 5 years out of them. I figured if they could take this then they'd live far longer in my application. I bought mine new so concerns over health were a non issue. I too would be concerned about used ones unless they came with maintenence records. The only reason to split up a battery from a lift truck is usually because it's worn out. I'm sure there are exceptions to this but....
My power house is an old stake truck van body that I insulated myself (8'x8'x20'). However the floor is not insulated so after the ground finally freezes that cold permeates up thru the floor. The battery is on a 4" high pallet which is filled with rigid foam. The battery is wrapped in 2 layers of foil bubble wrap insulation. Last winter I added 2" of rigid insulation to the floor and laid a plywood sub floor over it. I also partition off the small section where the equipment is to create a smaller area to heat. I'm sure most of my heat loss is thru the vent I'm using. I made a "hood" for the battery with a 4" dryer hose coming out the top. This hose exists thru the north wall of the building. Theres a "flapper" door on the wall where the hose exits I'm sure there must be a better way to vent the enclosure. I have on occassion used an electric heater to warm the space during times of extreme cold. I try to turn it on only during the day when we have high winds or bright sun. I could wire in a relay from the aux on either the inverter or the CC. There are a number of low cost improvements I could make to improve the situation. Now that I have data the reflects what the loses are maybe that will motivate me!!!
Rob
Thanks for the comparison between my data and the predicted output...very interesting. Here's my thoughts on the differences.
1) The system overall (wind and solar) is diverting energy frequently. A windy, sunny day in mid afternoon will be generating more energy than I can use and the battery will typically be full. I can't say how often this happens or how much is being diverted but I know there are loses. High wind gusts are diverted regularly during daylight hours
2) I suspect the Lakota's power curve is exagerated as many small turbine manufacturers are.
3) The primary loss in the turbine I'm sure is the early furling issue. The turbine should furl around 38 mph I believe. I would guess that it's actually furling closer to 25 mph. The potential energy between those 2 wind speeds is significant. You'll see a few days in my data where the turbine produced 7-9 kwh's. On those days we had winds in excess of 30 mph for as long as 24 hours in a row. The turbine should have produced closer to 24 kwh's on those days.

So considering the above I would estimate theres an additional 10-20% more power available from the solar thats currently being diverted. (just a gut feel from watching the system).
I have no problem believing the turbine is only producing 50% of it's rated output. Of that 50% I'd estimate maybe 10% is being diverted due to no place for it to go. The remaining 40% is due to early furling. Next time I lower the turbine I'm going to try to alter the furling. There's no factory adjustments on this version as there was on earlier models. I can either replace the springs with stiffer ones or tie off a portion of the existing spring to limit travel. Ultimately I'd like to replace the turbine with a larger model (maybe 2kw) but it's not in todays budget (sound familiar?)
Ric

[Rob Beckers]

Ah, yes, diversion. I forgot all about that!
Somehow I had the idea in my head that you were measuring energy coming off the turbine. :nono: It's energy going into the batteries, so when they are full you're not seeing turbine production. Same goes for solar I presume. So who knows, actual production may be much closer to predicted.

-RoB-

[Gary Houle]

Ric, Love the data, I would love to do something like that myself. After reading this I looked up flexnet and found that you might be interested that they have a recall on all FLEXnet DCs within the serial number range of 00000 to 00947. This might affect some of your data. Gary

[Steven Fahey]

Ric,

Thanks for posting this information. I haven't been able to read through it all, yet, but I do appreciate the effort you took in making it available.

The weibull distribution doesn't describe the wind at my house, either.

[Wolfgang Wan]

Hi Ric:

Thanks for using our Lakota and great job.

What I'm thinking is maybe you have a Mod 3 Low Wind Version there which has got 600W output at most in instantaneous wind speed.

Wolfgang
Zytech Aerodyne (http://www.zytechaerodyne.com)

[Ric Murphy]

Hi Wolfgang,
Sorry I haven't had time to respond to your other thread. I was going to refer you to this one but looks like you found it on your own. I'm confident the MOD setting is correct for my location. When the current rebuilt unit was initially put back in service I immediately noticed a serious lack of output. No matter the wind speed it would not put out anymore than a few hundred watts. After contacting Dave Cooke at True North (the former Canadian distributer) he advised the MOD setting should be changed. He gave me the wiring details and I changed it myself. Can't recall what setting it was changed to but again I'm confident it's the best one for my location. I'm also confident my power loss is strictly related to the early furling condition. The turbine will actually produce over 1000 watts for very brief moments just as it starts to furl. After that it immediately stalls and quickly loses RPM. Once the RPM drops it slowly comes out of furl and starts to gain speed again. This cycle repeats itself over and over. Although I don't have wind speed measuring capability here's what I suspect. The turbine is rated for 900w at 28 mph. I know the turbine is capable of achieving (and slightly exceeding) that output. If the rated output is 900w at 28 mph then I assume the windspeed that it should furl at should be a higher value. For some reason I thought the furling point was around 35-38mph by design. Thru observations in all speeds I'm guessing that it is actually furling around 25-28 mph. The highest sustainable output (without furling) is around 700w.
My belief is that the springs should be slightly stiffer and the amount of travel during furl is too great. The earlier versions had spring adjustment via threaded eye bolts. The later models do not. The excessive travel during furl causes an almost complete stall. Those are my observations/conclusions after almost 5 years of experience with this turbine
Hope this additional info is helpful.
Ric

[Wolfgang Wan]

Hi Ric:

Sorry for the late answer. A bit busy on orders.

I agree with you on furling up too earlier (25-28mph) which should be designed a initiation furling speed at 28.8 mph (rated wind speed). Longbow's initiation furling up speed is 25.6mph.

I think the early furling up is from spring's fatigue which should be changed every 2 years during inspection. We are working and improving on this critical issue now undergoing data collection on our test truck.

We can send you new set of springs to change but we prefer to send you new version spring after test. Under our test, Lakota can keep producing peak power without furling over 30mph and no damage to the generator. But take consider of the generator longevity and avoid overheating at high wind, 28.8mph is still the ideal initiation furling up speed.

Thanks

Wolfgang
Zytech Aerodyne (http://www.zytechaerodyne.com)