A small experiment with solar

100 watt panel with wood mount
100 watt panel with wood mount

Being carbon-conscious, naturally inclined to tinker, and seeing the falling costs for components, I was curious to know whether I could put together a small solar PV system on my own.  Having experienced a few blackouts this summer and expecting more in the future, I was also curious about providing a small amount of backup power for essential items.  Here’s the story of my first foray into off-grid renewables.

This post at Do The Math (an excellent blog you should read regularly) in which Tom Murphy describes his small, off-grid system really got me started on the whole thing.  I’m not a physics professor, but after reading it and doing some additional google searching, it seemed easy enough for a lay-person armed with a small amount of reading.  A valuable resource (also provided by Tom’s blog) is the Solar Living Sourcebook, available at your local library, which provides all the basics on what solar PV is, how it works, important safety tips, and options for setup.  I also learned a few things from various youtube videos and generally google searching.

The system I put together is 12 volts, which seems very common for small, off-grid installations.  It’s basically only six things: a solar panel, a charge controller, a battery, an inverter, and assorted wires and fuses.  The solar panel provides the electrons, the charge controller controls how those electrons flow to the battery (and makes sure it doesn’t overcharge), the battery stores electrons, and the inverter turns the battery’s 12 volt DC power into 110 volt AC power so it can be used with regular household electronics.  The wires and fuses connect things together and provide safety.

Panel unboxing
Panel unboxing

You can now purchase relatively affordable panels from Amazon or Home Depot in many wattages and sizes.  I chose a 100 watt panel that seemed to receive good reviews and a website that suggested that the company might be around for a while.

Charge controller showing all systems go!
Charge controller showing all systems go!

The other pieces of the system (inverter and charge controller especially) come in a huge range of prices.  After some reading, I decided that it might be better to spend a little more on a charge controller, as many people had complaints about cheap versions, and keeping your battery well-maintained is important (the function the charge controller plays).  I purchased a 30 amp controller from Morningstar, which I think could power up to 300 400 watts of panels if I expand the system in the future.  The battery is rated at 80 amp hours, and is sealed lead acid.  I purchased it from a local battery store, and its a discount version.

A bad photo of the 80 amp hour battery
A bad photo of the 80 amp hour battery

So what can this thing power you ask? That’s a function of how much the panel produces, how much the battery stores, and how much amperage I can draw at one time from the battery and inverter.

According to some assumptions I pulled from NREL’s PVWatts tool, the panel might generate 400 watt hours per day (100 watts X 4 hours equivalent of 100% production) in the peak season and maybe 210 in the low season (November), although I’ve seen higher numbers in other places.  The battery is large enough to store all that daily production and more (80 amp hours X 12 volts = 960 watt hours).  In fact, it would probably take one and a half to two two and a half days of sun to fully charge the battery.

Even in the winter, the daily production of the panel would probably be enough to power a few lights (the LED variety), an efficient laptop, a fan, and a small TV for a few hours.  It won’t run a hotplate, anything but the smallest air conditioner, a heater, or a refrigerator, at least not continuously.  The battery and the inverter could probably handle it (one of these things), but the panel wouldn’t be able to keep up.  As far as a back-up power source, this set up would power my refrigerator for about 12 8 hours, and our 8.8 cubic foot chest freezer for about 24 16 hours.  That’s assuming a fully charged battery, the panel couldn’t keep up with the draw from those appliances for more than a day.  These are just my estimates, I don’t have any real world results yet, but will report back soon.  Right now I’ve got just the chest freezer plugged into the inverter and I’m going to time how long until I get the low battery warning.

Things I’ve learned so far:

  1. It’s all the other stuff that costs money.  At this stage, the panel itself only accounts for 20% of the cost.  If I added two three more panels (which would probably max out the charge controller) to economize, the panels would still only be 40% 51% of the cost.
  2. I need scale to “save” money.  Right now, my costs per watt are about 76% higher than what I have been quoted to put a grid-tied, full size system on my roof.  If I maxed out the charge controller with two more panels and got another battery, I could bring my costs in line with the pros (again, on a per watt basis).  Whether this would continue to scale up I kind of doubt, since batteries get expensive and I would get into more serious electrical work pretty quickly.
  3. I need another battery (or two) for large stuff.  High amperage appliances, like a vaccuum, seem to be within the wattage range of the inverter, but my battery is only 80 Ah.  The internet tells me I should only run things that are 10-12% in amps of that capacity to avoid shortening the batteries life, and indeed I got a low battery warning when trying to run the shopvac.
  4. You should think hard about where to locate a panel before you embark on this kind of project.  I’m still squeamish about getting into roofing for fear I will cause a leak, and others in my household disagree about the aesthetics of a home-built wooden frame.  My goal in the long run is to get this on a roof somewhere.
  5. In the future, our homes should probably run direct current (DC) rather than alternating current (AC).
  6. Solar panels aren’t just for tree-huggers.  If you’ve ever searched youtube for videos about solar back-up systems, you’ll be a lot less surprised about news items like the Atlanta Tea Party teaming with the Sierra Club to promote more solar.  Many of the instructional videos I watched were clearly made by folks of the conservative persuasion who were into solar because they feared the grid would go down or they weren’t comfortable being beholden to utilities/the government.  Maybe there is more common ground here than we thought.

In praise of the prairie


As part of the “Our Fair State” series going on at streets.mn this week, I’ve posted a short photo post praising the western Minnesota prairie.

When most people think summer in Minnesota, they think north: lakes, coniferous forests, large noisy black birds.  But Minnesota actually has four biomes, and I would consider the prairie grassland the most neglected when we engage in communal nostalgia for Minnesota summer.  Luckily, we have some great parks on the prairie where you can see a little of what Minnesota might have been like before settlement.  And it’s beautiful.

The greenhouse gas benefits of autonomous vehicles

stanley side view (2005-023-040)

Autonomous vehicles may bring a myriad of benefits, but I anticipate that one of the largest may be the actual reduction in the total size of the vehicle fleet.  Eventually autonomous vehicles will allow “whistlecar” service, and whether fully autonomous or not would, this service is likely to fundamentally change the ownership model of automobiles.  Like present-day car-sharing services or taxis, a whistlecar subscription would mean one car could serve the needs of many people, instead of remaining parked most of the day waiting for its one owner to return.  Once you’re done with a car, it can drive off and serve someone else in the vicinity, drive to a charging station (if it’s electric), drive to a garage for service, or perhaps even deliver packages.  When you can subscribe to an on-demand travel service available 24-7 (and eventually cheaper than owning a car), many people will choose not to own.

Setting aside all the other benefits of autonomous vehicles for the moment, I’ll explore just this one: the benefits of a reduction in the car fleet.  And in a limited way: the greenhouse gas implications of this reduction in vehicles. Continue reading The greenhouse gas benefits of autonomous vehicles

Water for our future

Over at streets.mn I have a new post on the importance of water supply planning for the next regional plan.

What does all this have to do with Minnesota?  We have tons of water, right?  Well, on the surface yes, but we’re using our groundwater much faster than it’s being replaced, and that’s a problem.  That was one of the main topics at a Thrive MSP 2040 Roundtable discussion I attended a number of weeks ago, and have been meaning to post about since.  The 7-county region now gets70 percent of our water from groundwater sources, up from 15 percent in the 50′s.  In some places this means we’re reducing groundwater levels by over a foot a year.


2012 Nice Ride flows revisited

For the last two years, I’ve mapped the flows of the Nice Ride bikes.  I’ve always been slightly dissatisfied with the results, since bikes were obviously shown taking routes that any sane Nice Rider would never take (Hennepin Avenue between Lake and the bottleneck, for example).  Try as I might, I could never get ArcGIS to prioritize trails, lanes and bike boulevards sufficiently.

Enter the good people at Cyclopath.  Cyclopath is something like a bike route wiki, in that it is constantly updating it’s database of bike routes using ratings from users.  So every street in their database has a rating from bad to awesome (actually 0 to 4).  And this database includes the whole metro and beyond.  Best of all, they were willing to share it!

The latest version of ArcGIS has a new “restriction preference” setting, meaning there are six levels of preference for a link from “Highly Avoid” to “Highly Prefer”.  So I combined cyclopath’s street ratings with these preference settings and got a new and better route analyzer.  Here are the results:

NiceRide2012_cyclopath_routingAs a reminder, here is what the old version looked like:

2012 Nice Ride FlowsA few changes of note:

  • Hennepin is obviously not so popular anymore, save in downtown where there are more Nice Ride Stations.
  • The Cedar Lake Trail got a little more popular, perhaps 500 trips in some locations, since it was a Highly Preferred route.
  • West River Parkway south of the Washington Avenue bridge got a lot less popular (although crossings at Franklin stayed nearly the same).
  • There is generally just a lot less jigging and jogging on small streets as trips tend to condense onto major routes (see the major difference on Summit Avenue in Saint Paul).

Here is a version with a base street map for orientation:


Washington Avenue Traffic Projections

Hennepin County is preparing to reconstruct a portion of Washington Avenue between Hennepin Avenue and 5th Avenue South.  There has been much discussion of this project, in part because the reconstructed road may or may not include some sort of bike facilities.

Today I got an email about an upcoming public meeting for the project, and I noticed the project webpage includes a Traffic Operation Analysis with some traffic projections through 2035.  Hennepin County is projecting a 0.5% annual growth in traffic volumes between 2011 and 2035.

Hennepin County provided traffic volume forecasting information for the Washington
Avenue study area. Several considerations included in the traffic forecasts are:
Minneapolis overall expects to add 36,000 residents and 30,000 employees over
the next 20 years.

  • Closure of Washington Avenue through the U of M, east of the Mississippi River.
  • Construction of the new 4th Street S on-ramp connection to northbound 35W.
  • Reconfiguration of the interchange at Washington Avenue SE/Cedar Avenue.
  • Construction of the Central Corridor LRT line.
  • The impact of continued development in the downtown area including
  • townhomes/condos, office space and retail businesses.

Given the above considerations and through a review of past studies completed within the project area, Hennepin County recommends that the traffic forecasts be based on applying a 0.5 percent per year growth rate (13 percent increase by 2035) to the existing traffic volumes, then adjusting Washington Avenue, 3rd Street S and 4th Street S traffic volumes to account for circulation changes with the future 4th Street S on-ramp connection to northbound 35W.

I don’t feel qualified to speak about hyper-local traffic patterns based on certain street closures and circulation patterns.  That’s traffic engineer stuff.  But here are a few things (and charts) to consider:

  • According to Mark Filipi, who works on regional traffic modeling for the Metropolitan Council, the regional traffic model (based on old comp plan data) projects 0.3% annual growth in total Minneapolis VMT through 2025.  This is lower than 0.5%.
  • Total Minneapolis VMT has basically been falling since 2002, with non-interstate VMT fluctuating around flat growth (all VMT figures from MNDOT).Minneapolis VMT
  • Minnesota total VMT per capita has been falling steadily since 2004 at over half a percent each year, and total VMT has been falling since 2007.  Minnesota VMT and VMT per capita
  • According to the Minneapolis Traffic Count Management System, two of the three traffic count locations on Washington Avenue in the study area show a drop in traffic from their peaks in the late 90’s/early 00’s.  The third shows flat volumes.Washington Traffic Counts Between 3rd Ave & 4th Ave

Does all this mean that 0.5% annual growth rate on Washington Avenue is incorrect?  I’m not sure.  Minneapolis does plan to grow a lot of downtown jobs and housing.  On the other hand, per capita VMT trends have been falling not just in Minnesota, but across the country and world.  In addition, Minneapolis policy makers have stated their goals to shift modes.  It’s troublesome to me that in the “considerations” that Hennepin County used in their traffic forecasts, they didn’t include plans for that mode shift the same way they include plans for development.

Given the severe lack of detail on how the 0.5% growth figure was developed, I don’t think the community should accept any design predicated on that figure without some additional explanation, especially if the capacity needed to accomodate that growth is given as a reason to reject elements that will make this street a livable, vibrant and valuable place, namely, pedestrian and bicycle infrastructure.

Cross-posted at streets.mn

How much should utilities pay for distributed solar power?

Close-up of completed project - Gibbs Dairy goes solar

In the energy and climate circles, there is a lot being written lately about the threat to the traditional utility model from distributed, renewable energy sources.  David Roberts has been running a series describing the problem and looking for solutions.  Chris Nelder also has a good read on the topic.

One of the key issues is the idea that utilities want to avoid “stranded assets”, or infrastructure they still have to pay to maintain with a shrinking pool of customers.  As some customers get more power from solar, sales of electricity shrink, leaving utilities with the same distribution infrastructure to maintain using less revenue.  Some utilities, the latest being a municipal utility in San Antonio profiled by David Roberts, argue they shouldn’t pay customers the “market” rate for electricity their customers generate with rooftop solar, but instead should pay them a wholesale rate, or the same as they pay for other electricity on the grid.

The thinking here is that paying the wholesale price will put renewable energy on an even playing field, and help keep the old utility model more financially whole, since wholesale prices are typically much lower than market prices.  For example, the 5-year average wholesale price for electricity in the grid area that serves Minnesota was $53.62 per MWh for the period ending in 2010, according to FERC.  This is for the “peak” time of day, meaning the afternoon, which is also the time solar is most productive.  That’s equal to roughly 5 cents per kWh, which is the unit at which typical household sales are measured.  Last month I paid about 11 cents per kWh to Xcel before taxes, fees and other charges like WindSource.

At 5 cents/kWh, rooftop solar would take a very long time to pay off.  Many fewer people would likely choose to install it.  However, those in the renewable energy world will tell you that 5 cents/kWh doesn’t pay the owner of a system for some of the benefits solar energy has over wholesale electricity.  We should actually be looking at a “value of solar” that includes not just the wholesale energy price, but reimbursement for other values.  There is movement right now in Minnesota to legislate that a true “value of solar” be computed for future projects.  So what other value does solar energy have that utilities might value?

For one, it can be more efficient.  Whenever you transmit electricity or long distances, you lose some due to resistance (heat).  EIA estimates these loses at 7% nationally and 7.4% in Minnesota.  That means utilities are generating more kWhs than are needed to make up for the losses, and thus the customer is paying more for each kWh.  If you’re generating power very close to where you use it, you minimize these losses and the extra generation.  Distributed solar energy should actually be valued 7% above wholesale prices by a utility if you think it will reduce these line losses.  If you include that 7% bump, 5 cents becomes almost 6 cents per kWh.

The other value is the reduced environmental cost of solar generation.  There is plenty of discussion about what the optimal cost of carbon should be, and it all depends on what you adopt as your discount rate.  Here is a must-read on discount rates, also by David Roberts.  If you think that climate change will have a net drag on the economy in the future, your discount rate is likely low, and the optimal cost of carbon gets up into the $50 to $100/ton range.  Carbon levels per unit of electricity produced vary quite a bit across the county, but in Minnesota and parts of the upper Midwest, they averaged 0.738 metric tons per MWh in 2009 (the latest year for which EPA has data).  At that rate, a high carbon tax might add between 3.5 and 4.5 cents per kwh.

If you add all this up, (an economically optimal price on carbon, savings from transmission losses, and a wholesale price consistent with the 5-year peak average), you get a value of solar energy between 9.5 and 13 cents per kWh.  That’s at or above the market rate I’m paying in Minnesota right now.  Check out my extremely messy spreadsheet if you want to see the math.

Keep in mind there are other values of solar energy I haven’t considered in my calculus.  The Minnesota House legislation includes the savings from delaying capital investments in distribution infrastructure, savings from not having to build more generation, fuel price hedge value savings (not having to bet on fuel costs), and the value of local employment generated by manufacture and installation of solar energy.

“Disruptive challenges”

The investor-owned utilities themselves think that things are about to change dramatically, drawing comparisons to industry disruptions like those faced by regulated airlines, the phone company monopolies and RIM.  Mostly these disruptions will be driven by distributed renewable energy, but also by energy efficiency and market changes.

From a report by the Edison Institute, an association of shareholder-owned utilities:

There are important lessons to be learned from the history of the telephone industry. First, at the onset of the restructuring of the Bell System, there was no vision that the changes to come would be so radical in terms of the services to be provided and the technologies to be deployed. Second, the telephone players acted boldly to consolidate to gain scale and then take action to utilize their market position to expand into new services on a national scale. Finally, and most important, if telephone providers had not pursued new technologies and the transformation of their business model, they would not have been able to survive as viable businesses today. So, while the sector has underperformed the overall market since 2000, and as shown in Exhibit 5, even a leading industry participant like Verizon Communications has not been able to perform in-line with the overall market despite its growth, market share and solid profitability outlook due to the competitive uncertainties inherent in the business. However, those telecom providers that have embraced new technologies and addressed the competitive threats they faced have managed to survive and to protect investors from a “Kodak moment.”

Both David Roberts and Chris Nelder have better and more extensive write-ups of this study.

2012 Nice Ride Flows

2012 Nice Ride FlowsPresented here without scale or legend, are the Nice Ride flows from 2012.  As with the mapping I did for 2011, individual road segments are thickened to represent the volume of Nice Ride traffic that traveled over them during the year. Bike trails and lanes were favored by the routing software, but since it looked for direct routes, some paths may be under or over represented compared with real-life Nice Rider travel (Cedar Lake Trail versus Hennepin Avenue, for example).

St. Paul is much more vibrant in 2012, with the Lake Street bridge seeing a high volume of Nice Riders crossing to our twin city.  Top traffic segments included the Hennepin-Lyndale Bottleneck south of Loring Park, south of the Stone Arch Bridge, West River Parkway, and the Hiawatha trail east of the Metrodome.

Once again, kudos to Nice Ride for releasing all this awesome data.