If you rely on News Feed in Facebook to find my posts, you're missing most of them. On average, only 16% of updates in Facebook make it into News Feeds. Let me suggest that you subscribe to me in Facebook, follow me on Twitter (@ccengct), or use an RSS reader.

Monday, July 6, 2015

Why batteries matter in North Carolina

When you put photovoltaic solar panels on your roof, you care about how much sunlight actually reaches them. Engineers call this "insolation" and it's a function of the time of day, the season of the year, your latitude, the direction your panels face, the tilt angle of the panels on your roof, and how clear the sky is. But in general, the amount of electricity produced by a PV panel in Raleigh on a clear day is proportional to this graph in the winter:

and proportional to this graph in the summer (note that the time of day is not adjusted for daylight savings time):

Installers tweak these curves by adjusting the tilt angles of the panels, but the installers cannot change the laws of physics. Panel output stops a half hour before sunset and doesn't resume until half an hour after sunrise.

One problem in Raleigh and elsewhere in the southern U.S. is that residential consumption of electricity does not match these curves. In the summer down south, the primary use of residential electricity is air conditioning — and those air conditioners are still running flat-out at sunset and beyond. On summer nights the demand for electricity here stays high until 11 pm. See this graph of hourly residential consumption on a summer day specifically in Raleigh-Durham, using data from the U.S. Department of Energy:

This graph reflects the fact that most homes in Raleigh are air-conditioned. If you compare this graph to the second graph above, you'll see that there is a substantial mismatch of timing between the electricity generated by solar panels and the electricity consumed by your home. The peak consumption occurs just as the output from solar is falling off! It doesn't matter how many solar panels you put on your roof — you will still be sucking electricity from Duke Power on summer evenings, and that's when Duke will be burning fossil fuels or increasing output from its nuclear plants to produce the electricity you demand.

In winter, the story might or might not be the same. If your heat is provided by a heat pump, as mine is, your demand for electricity can remain somewhat constant across a 24-hour period because the heat pump is working harder at night than in the day. Looking at my billing data on the Duke website, I see that I use more electricity on cold nights in January than on hot afternoons in July. Of course, the sun is not shining during those cold nights.

With rechargeable batteries we could generate lots of solar electricity in the mornings and early afternoons when the physics make it plentiful, store it, and retrieve it in the late afternoons and evenings. Finding a battery technology that can do that safely and at reasonable cost is not easy. I've done the math and in theory I could meet my needs with 300 car batteries, but that's over $20,000 of batteries that would have to be replaced every 3 years because of the deep discharge cycle daily. Worse, there are concerns about so much sulfuric acid and the possibility of hydrogen gas emissions in a closed space.

The newly hyped Tesla battery sounds better, but I would still need 10 of them at a list price of $3,000 each, and like every other rechargeable the Tesla batteries will eventually wear out. But aside from cost, I've got to consider the implications of so much lithium on site and so much electricity storage. Large lithium batteries are touchy; shipping them falls under hazmat rules. If that much lithium catches fire, my house is ashes. Would you put a hundred-gallon gasoline tank under your house or adjacent to it? Probably not. Ten fully-charged Tesla batteries in one space would worry me. That's a lot of juice.

I'd prefer that Duke use idle land along the edges of Wake County to deploy solar panels on a large scale and then put batteries in secured locations such as substations. My two degrees in electrical engineering assure me that Duke knows more about this than I do. Duke could build these solar systems into their rate equations, and in any event I'd rather see the former tobacco farms in our county filled with solar panels instead of sprawling new residential subdivisions and strip shopping centers.

Or we could build houses that don't need so much air conditioning in the first place... a topic for a blog another day. Or perhaps we could use the insolation peak to cool (or heat) a reservoir of liquid during the day and have our heat pumps draw on that reservoir at night.

Two things to be mindful of:

  • In the context of solar, electric vehicles potentially make the mismatch worse because most EVs will recharge at night when there is no solar output.
  • Be careful when looking at graphs of residential consumption by time of day or season of year. The shape of the graph varies considerably from place to place. Homes in Florida use as much electricity in the summer for air conditioning as we do, but they don't need much heat in the winter. Homes in San Diego don't need much heat or air conditioning. Most homes in northern and central Europe don't have air conditioning; they do need heat, but they don't get it from electricity. Those locales are better situated for solar without need for short-term battery storage. The Internet has many graphs of consumption by time of day, but none of them may match your particular circumstances.