Monday, December 8, 2014

That follow-up I promised...

A few weeks about a month ago, I "guest-blogged" over at the Smart Grid Library blog about some of the problems with Net Metering, promising to do a follow-up post here in a week offering some ideas for resolving those problems.

The best laid plans of mice and men are approximately equal.

Between increased client work (not comaplining) and a little emergency home-remodeling (I would love to have a few strong words with the guys who put up the bathroom tile in this house when it was built) the blog entry never got published.

So here (better late than never) is the follow-up to that November 3rd blog entry...

To summarize the previous discussion:
I referred to an article appearing in Utility Deep Dive about a report from E3 on the difficulties of renewable generation integration.
  • In that article (and report) it said “The largest integration challenge that emerges,” E3 found, “is overgeneration.” That is when must-run generation (non-dispatchable renewables, combined-heat-and-power, nuclear generation, run-of-river hydro and thermal generation needed for grid stability) is greater than energy demand.”
  • I also discussed the "duck chart" that illustrates this problem.  A "duck chart" is produced when you track the demand for non-renewable generation in an environment where renewable generation is widespread.  It's called a "duck chart" because it looks like the back of a duck, facing right.
  • The biggest challenge is at the “neck” of the “duck” when a significant amount of generation has to come online in a short period of time to meet the demand that is placed on the grid when solar resources drop their output. The lower the “back” of the duck, the more of a problem bringing enough generation online in time becomes (the higher the “ramp rate”.)
For those not in with the technology buzzwords, "dispatchable" is grid-speak for "resources that can be told what to do to help keep the grid operating"

Net metering generation is non-dispatchable, because the utility offering the net metering tariff is required to take the power, whether it's needed to meet load or not. The more non-dispatchable generation you have to deal with, the steeper the neck becomes. One diagram shown in the executive summary of the report shows that at a 40% renewables level, power taken under net metering is more than half of the overgeneration problem. Under those conditions, the “duck diagram” gets positively swaybacked so that, in the study, the ramp rate in one scenario approaches twice what happens currently.  A touchy situation, because this means that the ramp rate may exceed any possible ramp rate with current technologies.

So that's the problem, what are some possible solutions?  At a very high level, more stuff needs to be dispatchable.  Options are:
  • Build more generation, particularly high-ramp-rate generation.
  • Make renewable generation dispatchable.
  • Make something else dispatchable that has not heretofore been dispatchable.
The first option is not seeing much interest because (a) it's expensive, (b) it's time-consuming, and (c) a big part of the whole push for renewable generation is to not have to build more generation. So, let's take that off the table...

The second option, making renewable generation dispatchable, means either an end to net metering, or a significant change to the concept.  There are a few options here:
  • Let the grid operator / utility decide whether to take energy from renewables, based on grid conditions, and pay the consumer either the prevailing rate for power at the time or their net metering rate.  This necessitates; 
    • the addition of significant control systems, because you may need control beyond "on/off", and even if you could work with "on/off" it can be a lot of decisions to keep track of.
    • significant policy changes
    • enough public trust of grid operators that it doesn't squelch the market for renewable generation.
  • Let renewables self-dispatch, based on a price signal.  This provides some significant opportunities, but has its own necessities:
    • Renewables would tend to come online when they are most needed, if, and to the extent, they are available at that time.  
    • In addition, renewable generation sites would be encouraged to add storage, both to take the "free" power when the grid didn't need it and keep it for later, and potentially to provide power to the grid when it needed it, and the location didn't.
    • This also paves the way for microgrids, which have benefits of their own.
    • However, renewable generation could increase somewhat in cost, because of the possible need to add control an communications intelligence.
The third option I discussed, making something dispatchable that hasn't been dispatchable before, is being discussed in some corners, and frankly goes back to Avery Lovins' beloved "negawatts".  Make load disptchable, or responsive.

Making load dispatchable is not a terribly surprising concept in the large Commercial and Industrial realms, where Demand Charges have long been the norm, and customers have become accustomed to making adjustments to operations to minimize energy costs.  In the Residential and Small Commercial world, however, it is limited to the minuscule penetration of Direct Load Control.  I think there is a better way, which I'll talk about next Monday.

No comments:

Post a Comment