Around the world on a solar ebike

Thanks! I guess I should make the video blogs public again... Were you able to view them? I think I took them down a while ago because of mean comments and they are a little embarrassing :D

I'll have to check that out!
 
grindz145 said:
Thanks! I guess I should make the video blogs public again... Were you able to view them? I think I took them down a while ago because of mean comments and they are a little embarrassing :D

All the embedded YouTube videos appear to be broken/missing. You could disable comments (under Advanced Settings) and/or set them to "Unlisted" so they show up in ES posts but not in YouTube searches?
 
I wonder if there is a way to build simple and small reflective mylar panel(s) to redirect some light from the outside of the panel onto the area with the shadow to compensate for the power loss when the shadow is present?
 
I'm changing them all back to unlisted for the benefit of ES folks. Please try to forgive my terrible monologues to the camera, it's pretty bad. Keep in mind these were mostly produced while on the road and somewhat under duress :D

Hopefully those videos will fuel your excitement for your own trip though!
 
John and Cecil said:
I wonder if there is a way to build simple and small reflective mylar panel(s) to redirect some light from the outside of the panel onto the area with the shadow to compensate for the power loss when the shadow is present?

Devil's in the details me thinks. Where are the reflectors placed? How do they react to the sun angle (or do they?) in order to get a good amount of light with enough intensity into the problem area. How do the reflectors affect aerodynamics?
 
grindz145 said:
I'm changing them all back to unlisted for the benefit of ES folks. Please try to forgive my terrible monologues to the camera, it's pretty bad. Keep in mind these were mostly produced while on the road and somewhat under duress :D

Hopefully those videos will fuel your excitement for your own trip though!
If you find the monologues embarrassing you could replace them with music. That said, I think most of us will understand the context.
 
wturber said:
grindz145 said:
I'm changing them all back to unlisted for the benefit of ES folks. Please try to forgive my terrible monologues to the camera, it's pretty bad. Keep in mind these were mostly produced while on the road and somewhat under duress :D

Hopefully those videos will fuel your excitement for your own trip though!
If you find the monologues embarrassing you could replace them with music. That said, I think most of us will understand the context.

I'm not going to go back and re-edit these 30 videos at this point, they are what they are :lol:
 
grindz145 said:
I'm changing them all back to unlisted for the benefit of ES folks. Please try to forgive my terrible monologues to the camera, it's pretty bad. Keep in mind these were mostly produced while on the road and somewhat under duress :D

Hopefully those videos will fuel your excitement for your own trip though!

Awesome. Thanks. I can see them now. I'll leave these running in the background in my workshop while I try to finish up my solar recumbent build.

It takes guts to put your face on YouTube and open yourself up to negative comments from anonymous cowards. I plan to shoot some video along the way but my plan is to edit it down to one documentary to be posted when it's over. For regular updates from the road, I figure I'll keep the more technical discussion going here and post photos and anecdotes for a more general audience on Instagram.
 
John and Cecil said:
I wonder if there is a way to build simple and small reflective mylar panel(s) to redirect some light from the outside of the panel onto the area with the shadow to compensate for the power loss when the shadow is present?
Probably not.

1) You'd have to replace the illumination, which means 2x mirrors for each cell (mirrors are not 100% efficient.)

2) You'd have to aim them somehow.

3) They tried to do that back when solar-PV was really expensive at the Carrizo solar plant for more power output. Result - overheated/damaged panels. (I know this because I bought some of the damaged ones. Only $3 a watt, and they still had 70% of their original output! A great price back then.)

All that being said, some manufacturers are using bifacial panels (transparent backsheet) to gather some additional reflected light. It can improve efficiency per square meter by 10-15%.

At some point someone is going to build a tiny (0805-sized) device that will work as a perfect diode and put one on each cell. That will solve the problem completely.
 
solarEbike said:
I plan to shoot some video along the way but my plan is to edit it down to one documentary to be posted when it's over. For regular updates from the road, I figure I'll keep the more technical discussion going here and post photos and anecdotes for a more general audience on Instagram.

Sounds awesome. Looking forward to it!
 
solarEbike said:
My custom-made PV modules arrived last week. Here's a quick peek at what they look like.
Cool panels.

Can I ask why you chose those over a more standard 4x6 panel like the Solbian SP78? Was it the extra few watts you could get, or was there some other benefit in going completely custom?
 
Hello, I've been following your project with fascination.

I have been riding and touring on a custom electric Bullitt front loader cargo bike for several years now, and just recently started charging my batteries with solar power.

I have purchased a small stack of the SunPower v3 solar cells, with the goal to create a similar foam panel to attach to the top of my cargo bin. My question is how are the solar panels face mounted to the fiberglass board? I cannot seem to locate sources for ETFE material, or the corrugated aluminum.

All the best,

Peter
 
billvon said:
solarEbike said:
My custom-made PV modules arrived last week. Here's a quick peek at what they look like.
Cool panels.

Can I ask why you chose those over a more standard 4x6 panel like the Solbian SP78? Was it the extra few watts you could get, or was there some other benefit in going completely custom?

I have a couple of Solbian panels and they're great but I really wanted to push the weight to power ratio further. By going to a custom build, I'm getting about 5% more power per cell than the SP78 you quoted. The Solbian panels take up 19% more surface area due to larger inter-cell spacing, larger border and the rather huge area for the front junction box. Finally, mine weigh 305 grams compared to Solbian's 1100 grams. That's 8 lbs (3.6 kg) lighter when we consider the additional savings from having a smaller honeycomb support panel.
 
billvon said:
At some point someone is going to build a tiny (0805-sized) device that will work as a perfect diode and put one on each cell. That will solve the problem completely.

Looking forward to it. It's going to have to be that small to be encapsulated with the cells.

Ten years ago, the first successful PV module I ever made used dual-junction gallium arsenide cells designed for satellites and each cell had it's own bypass diode. You can see them here. I think this was only possible because the per-cell voltage is much higher and the economics of launching cells into orbit are very different from terrestrial PV. If I recall correctly, micro-meteorite impact was the driver for needing cell-level bypass diodes. Each cell also has a thin cover glass bonded to it to protect it from radiation.


View attachment 1


Fast forward to today and we have "active diode" or "smart diode" devices like this Texas Instruments SM74611 (PDF) I will be using. It's not small enough to encapsulate within the cell but the cost ($3-4), forward voltage (26 mV) and reverse leakage current (3 µA) make it feasible to put one for every 4 or 8 cells.

TI SM74611.jpg
 
mkebullitt said:
I have purchased a small stack of the SunPower v3 solar cells, with the goal to create a similar foam panel to attach to the top of my cargo bin. My question is how are the solar panels face mounted to the fiberglass board? I cannot seem to locate sources for ETFE material, or the corrugated aluminum.

Hi Peter. I've switched from 1" foam to 1/4" Nomex honeycomb for my latest build (source). 1/4" (6 mm) core thickness is plenty rigid enough for a 4x6 cell panel. If you use foam, I recommend staying away from EPS/XPS insulation foam as it doesn't handle high temperatures very well. There are specialty urethane foams designed for use as core material which don't have the heat issue and should bond much better to the skins. They're not cheap, though.

Corrugated aluminum was brought up earlier in this thread by Cephalotus. I don't know if it's available outside Germany. Also, Bernard Cauquil posted photos and video of his SunPower cell soldering and encapsulation process. He used aluminum honeycomb core material (not shown).

I've never found a source of ETFE (Tefzel, fluoropolymer, etc.) film for making just a couple of panels. If you find some, make sure it's treated to be bondable otherwise nothing will stick to it. Your bottom sheet needs to be treated on both sides and your top sheet on one side.

Lightweight, robust cell encapsulation is not a trivial problem to solve, particularly if you just need one or two modules. Most solar race car teams outsource this step and having tried it myself 10 years ago, I came to the same conclusion. The commercially available semi-flex panels are a great choice for most ebike projects. Just be aware there's a lot of low-quality stuff out there with fake specs.
 
solarEbike said:
It's not small enough to encapsulate within the cell but the cost ($3-4), forward voltage (26 mV) and reverse leakage current (3 µA) make it feasible to put one for every 4 or 8 cells.

I'm probably ignorant of the potential downfalls, but I'd be inclined to tear into one of those diode packages to find out how small the active components really are to see if encapsulation along with the solar cells really is feasible. I can imagine how that might be a bad idea, but if having everything encapsulated together provides a real advantage (shock, weather?) maybe it is worth dissecting one. :^)
 
wturber said:
I'm probably ignorant of the potential downfalls, but I'd be inclined to tear into one of those diode packages to find out how small the active components really are to see if encapsulation along with the solar cells really is feasible. I can imagine how that might be a bad idea, but if having everything encapsulated together provides a real advantage (shock, weather?) maybe it is worth dissecting one. :^)

My Solbian panels have (traditional?) diodes embedded in the encapsulation but they're a much smaller package:


Solbian diode.jpg


At first I thought that's what I wanted for my custom panels but as the SunPower white paper helpfully points out:

All diodes will eventually fail; and, the life depends on temperature as well as several other factors, including module design, diode quality, junction box heat transfer, and module installation. Depending on how a diode fails, it can either permanently remove a substring from that module’s production or allow a shaded cell to run in reverse bias unmitigated, causing high heating in areas of a conventional cell which allow current to flow, generally causing backsheet damage.

I'll be putting my diodes in a junction box where I can access them for troubleshooting and replacement.
 
solarEbike said:
I'll be putting my diodes in a junction box where I can access them for troubleshooting and replacement.

Ah. So you don't want to encapsulate them. I infered from your comment that the diode you are using was too big to encapsulate that you would have preferred to do so.
 
solarEbike said:
Fast forward to today and we have "active diode" or "smart diode" devices like this Texas Instruments SM74611. It's not small enough to encapsulate within the cell but the cost ($3-4), forward voltage (26 mV) and reverse leakage current (3 µA) make it feasible to put one for every 4 or 8 cells.
Note that that one works by turning on and off rapidly, so when they are active you will see a ~30KHz voltage signal coming out of the array. It will be about .6 volts riding on top of the voltage at the output. This _probably_ won't affect the Genasun controller, but it's possible if that signal frequency lines up with the sampling period of the MPPT algorithm.

On the plus side you can use that characteristic to see how many of them are active at a given time. 30KHz signal? One is active. 60KHz signal? Two are active. Etc.
 
Those diodes are pretty cool. I haven't seen that particular one before. A higher voltage version would be nice.
 
I read through the thread like a page-turner thriller and then registered to post. I am from Australia and researching e-bike options and regulations. Once you have a motor supplying power there are legalities that come into play. Riding through Australia and Europe, you can not have >250w nominal or >25km/hr motor support. The general solution is to be inconspicuous but thats not the approach you are taking. Whats your plan?

And best of luck on your journey!
 
jalwa said:
I read through the thread like a page-turner thriller and then registered to post.

Hello jalwa and thanks for your kind words. I lurked here for 10 years before creating an active account. Sometimes it's hard to tell if I'm going too deep and then I remember this is Endless Sphere and there's no such thing as too deep. Still, the view counts don't tell you much about how many people find your post relevant or interesting. As much as I hate how much I care about the number of likes my Instagram posts do or don't get, it might be helpful to have some level of community upvoting à la Reddit or social media-style "likes" so ES readers have a way of saying "more like this, please!" without filling up the thread with too many "me too" posts.

jalwa said:
I am from Australia and researching e-bike options and regulations. Once you have a motor supplying power there are legalities that come into play. Riding through Australia and Europe, you can not have >250w nominal or >25km/hr motor support. The general solution is to be inconspicuous but thats not the approach you are taking. Whats your plan?

I figured it was only a matter of time before someone brought this up. Ok, here goes...

I was aware of these regulations when I was selecting a motor and considered several 250w options. I'm using a motor to extend daily distance traveled over pedaling alone so high power and high speed were not high priorities. Mostly, I was attracted to the lower weight of the 250w options but running a motor on the ragged edge of it's rating for tens of thousands of miles/kms is not a good recipe for long-term reliability. On flat ground with no headwind, I will typically use 250 watts or less due to energy budgeting. But for serious hill climbing with a fully loaded bike, I want a motor that can maintain 500-1000 watts until my 600 Wh battery pack cuts me off. Designing for the lowest common denominator of nominal speed/power didn't seem like the best choice.

I intend to follow the spirit, if not the letter, of the law. In populated areas when using cycling infrastructure and sharing a dedicated lane or path with other cyclists and pedestrians, I intend to travel at the speed of other cycle traffic. This is where 25 kph (15.5 mph) makes sense. The remaining 99% of the time, when I'm on open roads with little to no vehicle traffic and no other cyclists, I expect to cruise at 25-32 kph (15-20 mph) which is marginally over the limit in a few countries. On rare days when I have an energy surplus thanks to lots of sun and shorter daily distance, my 36V battery pack limits my assisted speed to about 40 kph (25 mph).

This year's The Sun Trip saw most participants going through Germany, which has arguably the strictest views on enforcement of rules. While one or two of the Belgian riders had registration plates on their bikes which allow 45 kph, the rest were technically supposed to use assist only up to 25 kph.The GPS trackers they carried showed their speeds and I was paying attention. Wanna guess how many of them actually kept it under 25? There were plenty of police stops throughout the trip but mostly those were about satisfying curiosity and taking selfies.

As for the letter of the law, I can legitimately say that the Grin 5T hub motor has a 218 watt rating at 68 RPM (motor simulator). I'm seriously considering taking my motor to a local engraver and having them carve "250 watts" in the side plate as an insurance policy against overly eager customs officers. Also, I can program my Cycle Analyst to comply with any local regulations if the need arises.

If you're a local, commuting daily along the same route every day, traveling at 2-3 times the speed of other cyclists without pedaling and generally acting like an ass in traffic you may attract some unwanted attention from law enforcement. If you're a foreigner, pedaling through on an exotic-looking solar machine you're more likely going to be stopped for a show and tell session, offered a beverage and sent on your way.
 
wturber said:
Ah. So you don't want to encapsulate them. I infered from your comment that the diode you are using was too big to encapsulate that you would have preferred to do so.

Yeah, sorry, no. I was just entertaining that line of thought. Keeping them accessible makes more sense for my application.

billvon said:
solarEbike said:
Fast forward to today and we have "active diode" or "smart diode" devices like this Texas Instruments SM74611.
Note that that one works by turning on and off rapidly, so when they are active you will see a ~30KHz voltage signal coming out of the array. It will be about .6 volts riding on top of the voltage at the output. This _probably_ won't affect the Genasun controller, but it's possible if that signal frequency lines up with the sampling period of the MPPT algorithm.

Ooh. Good catch. That could have been a very confusing troubleshooting session. I'll keep it in mind if I'm seeing unexpected results. If they don't work out, I can always swap them out for traditional diodes later.

fechter said:
Those diodes are pretty cool. I haven't seen that particular one before. A higher voltage version would be nice.

They were designed for PV where you typically have 3 bypass diodes for 60-72 cells or 1 diode for every 20-24 cells so I'm not sure if there would be a market for a higher voltage version? There's a 2012 article on DigiKey that does a nice job explaining their application in PV but I have never seen a residential or commercial PV module datasheet that touted having these as a selling point. I suppose if I was a PV module manufacturer who offered 25 year power warranties, I would be wary of new-fangled technology without a proven track record. Accelerated degradation tests can only tell you so much.
 
solarEbike said:
wturber said:
I'd consider adding a bit more distance between the solar trailer and bike as well.
For the final build... I'll move it a little further back. I'll still get the same shadow but it will be later/earlier in the day.

Previously, I had posted anecdotal evidence along the lines of "When I'm riding toward the sun early and late in the day, my head and torso cast a shadow on the solar array. This makes me sad."

Sure, moving the array further back decreases the shading problem but how much is enough? If I double the distance to height ratio of the shading obstruction, will that cut the shading in half? Most of my solar energy capture happens 3-4 hours before and after solar noon when the head shading is not a problem so how much are those early and late hours really worth? Moving the array further back creates a longer trailer which affects handling and maneuverability so ideally I want to make it as long as it needs to be and not an inch longer.

So, I ran a 3D solar shading simulation. I don't know why I didn't think to do this earlier but now that I've done it, the results are glorious.





This shows every travel direction at once so it may take a little work to parse but I think it's worth the effort.

The white cylinders in the center represent the head and torso of the rider on a recumbent bike, simplified to present the same shading profile in every direction. The little yellow/orange squares are 5" (125 mm) SunPower solar cells. The analysis takes into account local weather using a statistically-derived typical meteorological year and calculates the position of the sun for every hour of every day of the year. The numbers on the cells are annual shading derates for a non-tilting array located in Oakland, California. "100" means no shading on that cell. "90" means that cell gets 90% of the energy of an unshaded cell at the same tilt and azimuth.

If you're in the southern hemisphere, you'll need to swap north and south. If you live close to 38° latitude anywhere in the world, your numbers should look very similar to this. If you're much closer or further from the equator, your numbers will look quite different. If you're optimizing only for summer travel rather than year-round, your numbers will also be different.

So what does this tell us?

If you're traveling due north all day in California, all your cells are in rectangle (1). The sun rises to your right and sets to your left and is pretty much behind you all day so your head never casts a shadow on the array. You can put the array as close to the bike as you want. If you look closely, you'll notice some 98's and 97's for the cells closest to the rider. That's because on the summer solstice, the sun rises and sets 31° north of east and west at this latitude meaning the sun is in front of you at sunrise/sunset so you lose a few rays in the summer.

If you're traveling due south, all your cells are in rectangle (2). The sun rises to your left and sets to your right and is pretty much in your eyes all day. Your head casts no shadow on the array early and late in the day but it does cause shading problems in the middle part of the day. Since the sun's elevation is lower in winter, the problem is much worse in the winter because you cast a longer shadow at noon. If you maintain a 1:1 ratio of distance to height from the array to your head, you can limit your year-round losses to 5% (100-95 on the worst cells). Move the array just a little further back and your losses drop to 1-2%. Much better.

Things get much worse when you're traveling east (3) or west (4). You're now traveling directly toward/away from the sun early/late in the day. At a 1:1 ratio, your losses are 10%. At 2:1 they're only 2%. Not linear at all.

If we assume all travel directions are equally likely, the average losses of the closest cell at 1:1 is (0+5+10+10)/4 = 6.2%. If we assume the worst case scenario that a couple of shaded cells can knock out the whole module and we have four modules then the overall self-shading loss is only 1.6% which sounds pretty good to me.

Incidentally, a single-axis sun-tracking system with the axis parallel to the direction of travel provides the most benefit when traveling north/south and the least when traveling earth/west but that's a topic for a future post.
 
Are you thinking about adding a solar panel to your ebike and want to know how many Wh per day you can expect? Maybe you already have a fixed tilt panel and are wondering how many more Wh you can get if you add a tilting mechanism?

I've published a free tool which estimates these values for any location in the world.


Screen Shot 2018-09-03 at 12.44.22 PM.jpg


You can model a solar setup with fixed 0° tilt or single-axis tracking. The tracking can be manual or automatic. For single-axis tracking, the direction of travel is an important factor so you'll need to estimate how much of your travel is along a north-south direction vs. an east-west direction.

You enter your values in the yellow cells in the spreadsheet and read the outputs below. I've pre-loaded the spreadsheet with weather data for Oakland, California so you will need to import data for your own city. I will add instructions for doing this in a follow-up post. It's kind of a pain but you only need to do this once for each city. For longer trips, you'll want to create a separate sheet for each major city along your route.

I've published this as a View-only shared Google Docs spreadsheet. You can look at it without having to log in but if you want to try your own values you will need to log in with a Google account and make your own copy of the sheet under "File > Make a copy...". Please do not "Request Edit Access" to my original sheet. If you spot an error or have ideas for improvements, you can post them here.

[strike]Here's the link to the sheet: [/strike]
UPDATED version here: https://docs.google.com/spreadsheets/d/1Za-3a4wOQHfdXC8ih5EQkNvv1A90bd7S-pbzUT162c4/edit?usp=sharing

My spreadsheet is just a simple summary table using some formulas to combine hourly PV production data in ways that are meaningful for a solar ebike. The underlying PV modeling algorithm and weather data comes from PVWatts, a free tool maintained by the National Renewable Energy Laboratory (NREL) of the U.S. Department of Energy. I've been using this tool professionally for over a decade for residential PV energy modeling. It's not the only tool for this and it's not the most granular or accurate but it's free and the results are better than you might imagine.

The latest version has weather data for most places in the world. Some location have satellite data and some will use ground monitoring stations from the closest major airport. All weather data is based on a statistically derived Typical Meteorological Year (TMY). This means that it will not tell you how many Wh you will get today but it will give a fairly accurate prediction of the average daily Wh you can expect this month, assuming you give it the correct inputs. Remember: garbage in, garbage out.

You can enter your daily distance (in miles or km) and the output will contain your solar Wh/mi or Wh/km. Alternately, you can enter your desired Wh/mi or Wh/km and the output will show your your average daily distance with emphasis on average. Unless the weather is exactly the same every day, your daily Wh will be lower on some days and higher on other days.

Assumptions
  • This tool makes no attempt to account for your elevation gain, vehicle weight, rolling resistance, headwind, etc. so there's no guarantee that your daily Wh and distance are well matched. It is assumed that you will make up the difference by pedaling.
  • Single-axis tracking is assumed to have a full range of motion and no tracking errors. If your tilt is limited to 30° in each direction, your average daily output will be lower. If you are adjusting the angle manually instead of automatically tracking the sun while riding, your output will be lower. You'll just have to estimate how much lower.
  • Stationary charging is assumed to be equivalent to a dual-axis tracker. In other words, it is assumed that you will position your solar array so the sun's rays are hitting it at a 90° angle the entire time you are stopped. Realistically, you may only adjust every 30 or 60 minutes so your output will be a little lower than predicted.
  • I've selected "Module type: Premium" to more accurately reflect the temperature coefficient of power of SunPower cells.
  • PVWatts assumes a traditional glass and aluminum frame solar module with unobstructed airflow above and below the solar module. Semi-flexible modules are not explicitly supported. Reduced airflow will have lower output.
  • The weather data includes wind speed and takes this into account to calculate cell temperature. However, it does not take into account additional airflow when the solar array is moving. In my experience, this can result in as much as 5-10% increase in power on the hottest days, less so on cooler days. [UPDATE: I've used the Koehl formula (8) from this research paper and ran hourly simulations for the whole year using bicycle travel speeds instead of wind speeds and found that at high noon the gains can be 5-7% but averaged out over the whole year it's only 1-3%.]


Uncertainties
  • One of the inputs you will need to provide is "Shading loss." Try not to under-estimate this value. Unless you're riding through a flat, treeless wasteland from sunrise to sunset you will likely experience at least 5-10% shading loss, more if your solar panel is not the highest point on your bike. A lot more if you're riding through a forrest. You will enjoy you trip more if you under-estimate solar Wh.
  • The amount of solar energy reaching your location is not the same every year. June of this year may differ up to 40% from June of last year though +/-10% is much more common.


Glossary

  • Sun-hour: This is a PV modeling term defined as 1000 watts of irradiance falling on 1 square meter for one hour. It is roughly equivalent to an hour of sun at high noon on a cloudless summer day. It should be not confused with the number of daylight hours. A cloudy day will have fewer sun-hours than a cloudless day. A tracking solar panel will have more sun-hours than a fixed tilt panel.
  • Availability: If you are stopped and your battery is fully charged while the sun is shining, there is nowhere for the solar energy to go so this is considered a loss since you're not capturing sunlight which is available at your location. The same is true if your solar charge controller is in CV mode because the battery is nearly full or if your battery is disconnected, etc.

PS: One of the outputs is the maximum solar cell temperature you will get at your location at your selected time of year. This is a useful data point to keep in mind when choosing adhesives, resins or metals for your solar support structure. Thermal expansion coefficients will play a role in the longevity of your design.
 

Attachments

  • california-australia.jpg
    california-australia.jpg
    77.2 KB · Views: 4,416
Back
Top